daniel543x/pdfcraft
1
0 Fork 0
mirror of https://github.com/storytold/pdfcraft.git synced 2026-11-06 02:21:45 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge pull request #5569 from Piohrun/feat/jbig2-run-allowance

Browse Source 
M2.6: the images of a run, and a page's JBIG2 images, share one allowance of decoding
This commit is contained in:
FredApps authored and GitHub committed 2026-11-02 15:11:32 +01:00
2 parents 1303684851 ec25e7d327
commit e261518727
52 files changed
+2743 -293
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

crates/automation/src/a11y/documents.rs
+1 -1
View File
@@ -15,7 +15,7 @@ impl Automation {
if !export.truncated_pages.is_empty() {
use pdfcraft_engine::compare::{MAX_FORM_DEPTH, MAX_FORM_VISITS};
out["warnings"] = json!([format!(
"forms nested more than {MAX_FORM_DEPTH} deep, or past the first {MAX_FORM_VISITS} forms of a page (forms the page draws itself count too), were left out on the pages in truncated_pages: their text and images are missing from the export, although the viewer may still show them"
"pictures past what one export may hold or decode, or forms nested more than {MAX_FORM_DEPTH} deep, or past the first {MAX_FORM_VISITS} forms of a page (forms the page draws itself count too), were left out on the pages in truncated_pages: their text and images are missing from the export, although the viewer may still show them"
)]);
}
Ok(out)
crates/automation/src/dispatch.rs
+2 -2
View File
@@ -323,13 +323,13 @@ impl Automation {
let id = self.doc(&a)?.id;
let before = self.doc(&a)?.bytes.len();
let path = self.resolve(a.str("path")?, true)?;
let (bytes, merged) = self.session.reduced_bytes(id).map_err(failed)?;
let (bytes, merged, images_refused) = self.session.reduced_bytes(id).map_err(failed)?;
// As in the app, a copy that isn't smaller is not written (#490).
let written = bytes.len() < before;
if written {
write_atomic(&path, &bytes)?;
}
json!({ "path": path.to_string_lossy(), "written": written, "bytes_before": before, "bytes_after": bytes.len(), "merged_objects": merged })
json!({ "path": path.to_string_lossy(), "written": written, "bytes_before": before, "bytes_after": bytes.len(), "merged_objects": merged, "images_refused": images_refused })
}
"doc_optimize" => self.doc_optimize(&a)?,
"doc_initial_view" => self.doc_initial_view(&a)?,
crates/automation/src/documents.rs
+1
View File
@@ -410,6 +410,7 @@ impl Automation {
"images": o.images,
"images_resampled": o.images_resampled,
"images_recompressed": o.images_recompressed,
"images_refused": o.images_refused,
"image_bytes_before": o.image_bytes_before,
"image_bytes_after": o.image_bytes_after,
"thumbnails": o.thumbnails,
crates/automation/src/tools.rs
+4 -4
View File
@@ -1194,7 +1194,7 @@ pub fn tools() -> Vec<ToolDef> {
json!({ "doc": doc(), "field": { "type": "string" }, "event": { "type": "string", "enum": ["keystroke", "format", "validate", "calculate", "mouse_up"] }, "script": { "type": "string" } }),
&["doc", "field", "event"],
)),
t("doc_export_office", "Export to Word, PowerPoint, HTML, RTF or Markdown", "Export a PDF ▸ Word (.docx), PowerPoint (.pptx, one slide per page), HTML (.html, one file with images inline), RTF (.rtf) or Markdown (.md or .markdown: pipe tables, images left out), chosen by path's extension: text, headings, bold and italic. Word uses borderless inferred tables and positioned editable shapes on pages with artwork, rotated or large text; each source page retains its size. Font matching and vector graphics remain limited. Export reads forms 50 deep and up to 50 000 a page, counting forms the page draws itself. Pages that go past either limit are listed in truncated_pages, with a warning; the text and images of the forms left out are missing.")
t("doc_export_office", "Export to Word, PowerPoint, HTML, RTF or Markdown", "Export a PDF ▸ Word (.docx), PowerPoint (.pptx, one slide per page), HTML (.html, one file with images inline), RTF (.rtf) or Markdown (.md or .markdown: pipe tables, images left out), chosen by path's extension: text, headings, bold and italic. Word uses borderless inferred tables and positioned editable shapes on pages with artwork, rotated or large text; each source page retains its size. Font matching and vector graphics remain limited. Export reads forms 50 deep and up to 50 000 a page, counting forms the page draws itself. Pages that go past either limit, or that have pictures past what one export may hold or decode, are listed in truncated_pages, with a warning; what was left out is missing.")
.cmd("export.docx")
.with(schema(json!({ "doc": doc(), "path": { "type": "string" } }), &["doc", "path"])),
t("pdfa_verify", "Verify PDF/A", "Standards ▸ Verify PDF/A compliance: the PDF/A-2b or 3b rules the document breaks (ISO 19005 clause, message, page, whether Save as PDF/A can fix it), plus what it declares.")
@@ -1281,7 +1281,7 @@ pub fn tools() -> Vec<ToolDef> {
.ro()
.cmd("ocr.recognize")
.with(schema(json!({}), &[])),
t("doc_export_all_images", "Export all images", "Write the images that pages use into a folder (`<name>_Page_<n>_Image_<k>.jpg|png`, n zero-padded to the page count's digits), each once: JPEG images unchanged, others as PNG with their soft mask as alpha. min_size skips images with fewer pixels on their shorter side. Images that can't be decoded yet (JPEG 2000, separations) are listed under skipped. Includes unsaved edits.")
t("doc_export_all_images", "Export all images", "Write the images that pages use into a folder (`<name>_Page_<n>_Image_<k>.jpg|png`, n zero-padded to the page count's digits), each once: JPEG images unchanged, others as PNG with their soft mask as alpha. min_size skips images with fewer pixels on their shorter side. Images that can't be decoded yet (JPEG 2000, separations), or past what one export may decode or hold, are listed under skipped, with the reason. Includes unsaved edits.")
.cmd("export.all_images")
.with(schema(
json!({
@@ -1340,7 +1340,7 @@ pub fn tools() -> Vec<ToolDef> {
}),
&["from"],
)),
t("doc_reduce", "Reduce file size", "Write a smaller copy of the document to `path` with Acrobat's Reduce File Size choices: colour and grey images above 225 ppi downsampled to 150 ppi and JPEG-compressed (medium quality), black-and-white images averaged down to 300 ppi above 450 ppi and each stored as CCITT Group 4 (or 1-bit Flate, once resampled or packed from more bits) whenever that is smaller, thumbnails dropped, identical fonts and images merged, unused objects dropped, compressed object streams. If the copy wouldn't be smaller, nothing is written and `written` is false. The open document is unchanged.")
t("doc_reduce", "Reduce file size", "Write a smaller copy of the document to `path` with Acrobat's Reduce File Size choices: colour and grey images above 225 ppi downsampled to 150 ppi and JPEG-compressed (medium quality), black-and-white images averaged down to 300 ppi above 450 ppi and each stored as CCITT Group 4 (or 1-bit Flate, once resampled or packed from more bits) whenever that is smaller, thumbnails dropped, identical fonts and images merged, unused objects dropped, compressed object streams. If the copy wouldn't be smaller, nothing is written and `written` is false. Images past what one run may decode stay as they are, counted in images_refused; streams it only stores with Flate past that stay LZW, uncounted. The open document is unchanged.")
.cmd("optimize.reduce")
.with(schema(json!({ "doc": doc(), "path": { "type": "string" } }), &["doc", "path"])),
t("doc_initial_view", "Initial view", "Read or change how the document opens (Document Properties ▸ Initial View) and its reading options: navigation (page, bookmarks, pages, attachments, layers), layout (default, single, continuous, two_up, two_up_continuous, two_up_cover, two_up_continuous_cover), magnification (default, actual, fit_page, fit_width, fit_height, fit_visible, or a percentage), page, window options (fit_window, center_window, full_screen, display_title), interface options (hide_menubar, hide_toolbar, hide_window_ui), language and binding (left, right). Only the given ones change; returns the result. Undoable.").with(schema(
@@ -1451,7 +1451,7 @@ pub fn tools() -> Vec<ToolDef> {
.with(schema(json!({ "doc": doc() }), &["doc"])),
t("doc_open_revision", "Open a revision", "Open saved revision `revision` (1 = the oldest) of a document as a new, unsaved document, to see the file as it was then.")
.with(schema(json!({ "doc": doc(), "revision": { "type": "integer", "minimum": 1 } }), &["doc", "revision"])),
t("doc_optimize", "Optimize PDF", "Write an optimized copy to `path` (Acrobat's PDF Optimizer). preset: standard (default; Reduce File Size's choices) or mobile (images to 96 ppi above 144, JPEG high quality); the options below change it. compatibility: retain (default; keeps the document's version, or 1.5 when object streams are written), 1.4, 1.5, 1.6, 1.7 or 2.0. merge_fragments (default true) first puts a page picture drawn as many stacked image strips (as some PDF printers write it) back together into one image, so it can be resampled (merged_images, fragments_merged). color / gray: { downsample, ppi, above_ppi, compression: jpeg|zip|retain, quality 1–100 } (defaults: downsample to 150 ppi above 225, JPEG 60). mono: { downsample, ppi, above_ppi, compression: ccitt_g4|jbig2|retain } for black-and-white images and image masks (default: averaged down to 300 ppi above 450 ppi, counted in images_resampled; stored losslessly as CCITT Group 4 or JBIG2 (or 1-bit Flate, once resampled or packed from more bits) when that is smaller, Group 3 fax images, JBIG2 images, black-and-white RGB and indexed images using only palette entries 0 and 1 among them, counted in images_recompressed; retain resamples only, a CCITT image staying CCITT, a JBIG2 image JBIG2 and any other becoming Flate). Images are measured where pages draw them and replaced only when smaller. discard_*: thumbnails (default true), alternate_images (true), tags, print_settings; flate_unencoded (true); object_streams (default true; always off for 1.4); lzw_to_flate (true) decodes LZW streams and stores them with Flate instead, leaving any it cannot decode unchanged (lzw_streams counts the streams converted); remove_invalid_links and remove_unreferenced_dests (true); images and forms the pages list but never draw are always dropped (unused_xobjects). discard: Remove Hidden Information categories (metadata, attachments, comments, form-fields, hidden-text, hidden-layers, bookmarks, links-actions-scripts, private-data). Signed documents are refused. The open document is unchanged.")
t("doc_optimize", "Optimize PDF", "Write an optimized copy to `path` (Acrobat's PDF Optimizer). preset: standard (default; Reduce File Size's choices) or mobile (images to 96 ppi above 144, JPEG high quality); the options below change it. compatibility: retain (default; keeps the document's version, or 1.5 when object streams are written), 1.4, 1.5, 1.6, 1.7 or 2.0. merge_fragments (default true) first puts a page picture drawn as many stacked image strips (as some PDF printers write it) back together into one image, so it can be resampled (merged_images, fragments_merged). color / gray: { downsample, ppi, above_ppi, compression: jpeg|zip|retain, quality 1–100 } (defaults: downsample to 150 ppi above 225, JPEG 60). mono: { downsample, ppi, above_ppi, compression: ccitt_g4|jbig2|retain } for black-and-white images and image masks (default: averaged down to 300 ppi above 450 ppi, counted in images_resampled; stored losslessly as CCITT Group 4 or JBIG2 (or 1-bit Flate, once resampled or packed from more bits) when that is smaller, Group 3 fax images, JBIG2 images, black-and-white RGB and indexed images using only palette entries 0 and 1 among them, counted in images_recompressed; retain resamples only, a CCITT image staying CCITT, a JBIG2 image JBIG2 and any other becoming Flate). Images past what one run may decode stay as they are, counted in images_refused; streams it only stores with Flate past that stay LZW, uncounted. Images are measured where pages draw them and replaced only when smaller. discard_*: thumbnails (default true), alternate_images (true), tags, print_settings; flate_unencoded (true); object_streams (default true; always off for 1.4); lzw_to_flate (true) decodes LZW streams and stores them with Flate instead, leaving any it cannot decode unchanged (lzw_streams counts the streams converted); remove_invalid_links and remove_unreferenced_dests (true); images and forms the pages list but never draw are always dropped (unused_xobjects). discard: Remove Hidden Information categories (metadata, attachments, comments, form-fields, hidden-text, hidden-layers, bookmarks, links-actions-scripts, private-data). Signed documents are refused. The open document is unchanged.")
.cmd("optimize.advanced")
.with(schema(
json!({
crates/automation/tests/it/automation.rs
+2
View File
@@ -6651,10 +6651,12 @@ fn optimizing_through_tools() {
let reduced = ok(&mut a, "doc_reduce", json!({ "doc": doc, "path": "reduced.pdf" }));
assert!(reduced["bytes_after"].as_u64().unwrap() < reduced["bytes_before"].as_u64().unwrap());
assert_eq!(reduced["written"], true);
assert_eq!(reduced["images_refused"], 0);
assert_eq!(std::fs::metadata(dir.join("reduced.pdf")).unwrap().len(), reduced["bytes_after"].as_u64().unwrap());
// Target PDF 1.4 (disables object streams) and PDF 2.0 (with object_streams toggled off).
let v14 = ok(&mut a, "doc_optimize", json!({ "doc": doc, "path": "v14.pdf", "compatibility": "1.4" }));
assert_eq!(v14["images_refused"], 0);
assert_eq!((v14["pdf_version"].as_str(), v14["object_streams"].as_bool()), (Some("1.4"), Some(false)), "{v14}");
let v14_bytes = std::fs::read(dir.join("v14.pdf")).unwrap();
assert!(v14_bytes.starts_with(b"%PDF-1.4\n") && !v14_bytes.windows(7).any(|w| w == b"/ObjStm"));
crates/cos/README.md
+4 -1
View File
@@ -93,5 +93,8 @@ recognisable in a diff and byte-stable when re-serialized.
- Image codecs. `pdfcraft-filters` carries the non-image filters and CCITT fax and JBIG2 decoding
(`Document::decoded_image`, a JBIG2 image's `/JBIG2Globals` within `MAX_JBIG2_GLOBALS`, 32 MiB,
their filters undone once a document for the images that share them, up to 64 MiB kept, while
their segments are decoded for every image);
their segments are decoded for every image, and `ImageBudget` for an operation over many images,
which share one allowance of decoding, filters, decoders and pixels alike, and are refused past
it: `IMAGE_WORK`, 2^34 steps, or 2 to 5 minutes of JBIG2 and about 1,000 pages scanned at
300 dpi, plus `IMAGE_WORK_PER_BYTE`, 2^13 steps for each byte of the file);
DCT and JPX are still decoded through the bootstrap renderer (`core.image-filters`, M2.6).
crates/cos/src/budget.rs
+612
View File
@@ -6,16 +6,218 @@ use pdfcraft_filters::FilterError;
use crate::{CosError, Document, ObjRef, Object, Stream};
/// The steps of decoding an operation over a document's images may take whatever the file's size.
/// A step of JBIG2 decoding ([`hayro_jbig2::Image::decode_within`]) is about a pixel, and takes 6 to
/// 16 ns, so this is 2 to 5 minutes, and about as long again for what an operation does with the
/// pixels. A page scanned at 300 dpi takes about 17 million, 8.4 for its pixels and that again to
/// decode them, so this alone reads about 1,000 of them. A byte into or out of a filter is a step
/// too, and takes less: 3 ns at most, or up to 19 ns through a predictor on samples of a bit.
const IMAGE_WORK: u64 = 1 << 34;
/// The steps each byte of the file adds to [`IMAGE_WORK`], so that a long scan is read whole: such a
/// page pays for itself with about 2 KB of the file, about the smallest a page of text is stored
/// in, and a stream of any filter this reads can be decoded about 8 times at Flate's best, 1,032 to
/// one. So what a hostile file can ask for grows with its length, by about 14 minutes for 10 MB.
const IMAGE_WORK_PER_BYTE: u64 = 1 << 13;
/// Decoded bytes one page's content walk may produce in all: its `/Contents` streams and every
/// form XObject it enters, each occurrence counted again.
pub const MAX_PAGE_DECODED: usize = 256 << 20;
/// Form XObjects one page's content walk may enter, each entry counted again.
pub const MAX_PAGE_FORMS: usize = 50_000;
/// The steps a sample of a JPEG costs: a few bytes can hold a page of one grey, and decoding,
/// resampling and storing it took about 20 to 45 ns a sample.
const JPEG_WORK: u64 = 2;
/// The steps a byte of a JPEG costs to read: its tables took 24 ns a byte before a size was known,
/// and images can share one, so reading it again for each has to count.
const JPEG_READ_WORK: u64 = 4;
/// A stream decode within a byte bound, with the bytes its filter stages decoded: the data or
/// the error, each with that count.
type Decoder = fn(&Stream, usize) -> Result<(Vec<u8>, usize), (FilterError, usize)>;
/// What one operation over a document's images may still spend decoding them: exporting them,
/// optimizing them, and storing LZW streams with Flate as it does, clearing them for redaction.
///
/// An image has its own limit, which doesn't bound a file of thousands of images made to reach it:
/// a few hundred bytes of JBIG2 can ask for seconds of decoding, 4 KB of Flate inside Flate for a
/// gigabyte, and an image that shares `/JBIG2Globals` decodes them again. So images share this
/// allowance, whatever they do with what they decode and whether or not it works, and past it they
/// aren't decoded. A caller then treats them as it does an image it can't read, and
/// [`ImageBudget::refused`] says how many there were.
///
/// An image also pays for its pixels once decoded, as what reads it goes through them one by one
/// however few bytes coded them: blank rows of fax or Flate take almost none. It pays for no more
/// than its data holds, a bit a pixel at least, so an image declaring a size its data doesn't fill
/// costs what a smaller one would, not the run's allowance. A JPEG pays for its samples when
/// it is decoded ([`ImageBudget::pay_jpeg`]), and is otherwise not charged, as JPEG 2000 isn't:
/// their decoders don't run here.
#[derive(Debug)]
pub struct ImageBudget {
work: u64,
refused: usize,
}
impl ImageBudget {
/// The allowance for an operation on `doc`, which a longer file makes larger. The streams made
/// or changed since it was opened count as the file's bytes do, so a document built in memory
/// (combined, created, or given pages of scans) gets what it would once saved and opened again.
pub fn new(doc: &Document) -> Self {
let length = |n: usize| u64::try_from(n).unwrap_or(u64::MAX);
let edited = doc
.scan_objects_subset(doc.modified_objects())
.filter_map(|(_, object)| match &*object.ok()? {
Object::Stream(s) => Some(length(s.raw.len())),
_ => None,
})
.fold(0u64, u64::saturating_add);
let bytes = length(doc.bytes().len()).saturating_add(edited);
Self::with_work(IMAGE_WORK.saturating_add(bytes.saturating_mul(IMAGE_WORK_PER_BYTE)))
}
/// An allowance of `work` steps of decoding.
pub fn with_work(work: u64) -> Self {
Self { work, refused: 0 }
}
/// The samples of image `s`, decoded and charged to the allowance: the one way to decode an
/// image outside the renderer, so that an operation over many images can't decode them without
/// one.
///
/// They go through its filters ([`Stream::decoded`]), or, for a CCITT fax or JBIG2 image (its
/// last filter), through the filters before it and then [`pdfcraft_filters::decode_ccitt_within`]
/// with its [`Document::fax_params`], or [`pdfcraft_filters::decode_jbig2_within`] with its
/// `/JBIG2Globals`, as `CCITTFaxDecode` and `JBIG2Decode` give them (ISO 32000-2 §7.4.6,
/// §7.4.7). Parameters given by reference count, for every filter
/// ([`Document::with_direct_parms`]). A fax image's `/Columns` has to be its `/Width`, and a
/// JBIG2 image's page its `/Width` × `/Height`: PdfCraft's renderer draws the whole page into
/// the image. Up to its `/Height` rows are decoded, whatever a fax image's `/Rows` says; damaged
/// fax data gives fewer, so a caller checks it has the rows it needs. Damaged JBIG2 data, and
/// globals that can't be read (within [`MAX_JBIG2_GLOBALS`](crate::MAX_JBIG2_GLOBALS)), are an
/// error.
///
/// No stage, the filters before a fax or JBIG2 decoder included, may produce more than `max`
/// bytes, nor more than [`MAX_DECODED`](crate::MAX_DECODED). A JBIG2 image's `/JBIG2Globals`
/// are another stream, shared between images, and are decoded within `MAX_JBIG2_GLOBALS`
/// whatever `max` is. Each filter as it runs, the globals when they have to be read, and the fax
/// or JBIG2 decoder spend from the allowance, and the image then pays for its pixels; with
/// nothing left an image is refused before anything is read, its globals included.
///
/// ```
/// # let doc = pdfcraft_cos::Document::new_empty();
/// # let s = pdfcraft_cos::Stream::from_raw(pdfcraft_cos::Dict::new(), Vec::new());
/// let samples = pdfcraft_cos::ImageBudget::new(&doc).decode(&doc, &s, 1 << 20);
/// assert_eq!(samples.unwrap(), b"");
/// ```
///
/// `Document` has no way of its own to decode an image:
///
/// ```compile_fail
/// # let doc = pdfcraft_cos::Document::new_empty();
/// # let s = pdfcraft_cos::Stream::from_raw(pdfcraft_cos::Dict::new(), Vec::new());
/// let _ = doc.decoded_image(&s);
/// ```
///
/// ```compile_fail
/// # let doc = pdfcraft_cos::Document::new_empty();
/// # let s = pdfcraft_cos::Stream::from_raw(pdfcraft_cos::Dict::new(), Vec::new());
/// let _ = doc.decoded_image_within(&s, 1 << 20);
/// ```
pub fn decode(&mut self, doc: &Document, s: &Stream, max: usize) -> Result<Vec<u8>, CosError> {
let decoded = doc.decoded_image_spending(s, max, &mut self.work);
self.count_refused(s, &decoded);
self.pay_size(s, decoded?)
}
/// Image `s` through its filters alone, charged to the allowance: [`Stream::decoded_within`]
/// with its `/DecodeParms` given by reference read too ([`Document::with_direct_parms`]), for an
/// image that has no fax or JBIG2 decoder to go through.
pub fn decode_filters(&mut self, doc: &Document, s: &Stream, max: usize) -> Result<Vec<u8>, CosError> {
let decoded = doc.with_direct_parms(s).decoded_spending(max, &mut self.work);
self.count_refused(s, &decoded);
let decoded = decoded?;
if s.filters().iter().all(|(f, _)| f.is_decodable()) { self.pay_size(s, decoded) } else { Ok(decoded) }
}
/// `data` through `chain`, any corruption being an error ([`pdfcraft_filters::decode_within`]),
/// for a stream a run rewrites rather than an image it reads, charged to the allowance.
pub fn decode_chain(
&mut self,
chain: &[(pdfcraft_filters::Filter, pdfcraft_filters::Params)],
data: &[u8],
max: usize,
) -> Result<Vec<u8>, FilterError> {
pdfcraft_filters::decode_within(chain, data, max, &mut self.work)
}
/// Pays `steps` of other work on decoded data, or spends all there is and says so.
pub fn spend(&mut self, steps: u64) -> bool {
let left = self.work.checked_sub(steps);
self.work = left.unwrap_or(0);
left.is_some()
}
/// Pays for `pixels` a decode goes on to work through, or spends all there is, and the image
/// is refused.
pub fn pay_pixels(&mut self, pixels: u64) -> bool {
let paid = self.spend(pixels);
self.refused += usize::from(!paid);
paid
}
/// Pays for reading JPEG `data` and decoding its `samples`, as [`ImageBudget::pay_pixels`] does.
/// A caller pays before its tables are parsed, which can be most of the file.
pub fn pay_jpeg(&mut self, samples: u64, data: &[u8]) -> bool {
let reading = (data.len() as u64).saturating_mul(JPEG_READ_WORK);
self.pay_pixels(samples.saturating_mul(JPEG_WORK).saturating_add(reading))
}
/// Pays for the pixels of image `s` that its `decoded` data holds, those of its `/Width` and
/// `/Height` but no more than a bit of the data each, unless it is JBIG2, whose decoder pays for
/// its page, or isn't decoded here. What can't be paid for refuses the image.
fn pay_size(&mut self, s: &Stream, decoded: Vec<u8>) -> Result<Vec<u8>, CosError> {
if !charged(s) || matches!(s.filters().last(), Some((pdfcraft_filters::Filter::Jbig2, _))) {
return Ok(decoded);
}
let side = |key: &[u8]| s.dict.int(key).and_then(|v| u64::try_from(v).ok()).unwrap_or(0);
let held = u64::try_from(decoded.len()).unwrap_or(u64::MAX).saturating_mul(8);
if self.pay_pixels(side(b"Width").saturating_mul(side(b"Height")).min(held)) {
Ok(decoded)
} else {
Err(CosError::Filter(FilterError::Spent.to_string()))
}
}
/// An image that failed with nothing left was refused, unless it has a filter that costs
/// nothing here, which fails whatever is left.
fn count_refused(&mut self, s: &Stream, decoded: &Result<Vec<u8>, CosError>) {
if decoded.is_err() && self.work == 0 && charged(s) {
self.refused += 1;
}
}
/// The decodes refused for lack of allowance: an image read twice over is counted twice.
pub fn refused(&self) -> usize {
self.refused
}
/// The steps of decoding still allowed.
pub fn work_left(&self) -> u64 {
self.work
}
}
/// Whether `s` can be decoded here and so charged: general filters, a fax or JBIG2 decoder last at
/// most. Those decoders anywhere else can't run.
fn charged(s: &Stream) -> bool {
use pdfcraft_filters::Filter;
s.readable_filters().is_ok_and(|chain| {
chain.split_last().is_none_or(|((last, _), before)| {
(last.is_decodable() || matches!(last, Filter::Jbig2 | Filter::CcittFax)) && before.iter().all(|(f, _)| f.is_decodable())
})
})
}
/// What one operation may still decode and enter across all the streams it reads.
///
/// [`MAX_DECODED`](crate::MAX_DECODED) bounds one decode. It does not bound a page whose
@@ -143,6 +345,401 @@ impl DecodeBudget {
}
}
#[cfg(test)]
mod image_tests {
use super::*;
use crate::Dict;
fn jbig2(parms: Option<Object>) -> Stream {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
if let Some(parms) = parms {
dict.set(b"DecodeParms".to_vec(), parms);
}
Stream::from_raw(dict, pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap())
}
/// Each image spends what it took from one allowance, and those past it are refused and
/// counted, whatever their filters. What can't be decoded here costs nothing, and fails
/// without being one of these.
#[test]
fn the_images_of_an_operation_share_its_allowance() {
let doc = Document::new_empty();
let image = jbig2(None);
let mut budget = ImageBudget::with_work(u64::MAX);
let samples = budget.decode(&doc, &image, 1 << 20).unwrap();
let cost = u64::MAX - budget.work_left();
assert!(cost > 0);
let mut budget = ImageBudget::with_work(3 * cost);
for left in [2, 1, 0] {
assert_eq!(budget.decode(&doc, &image, 1 << 20).unwrap(), samples);
assert_eq!(budget.work_left(), left * cost);
}
assert_eq!(budget.refused(), 0, "the allowance held exactly three");
for refused in [1, 2] {
assert!(budget.decode(&doc, &image, 1 << 20).is_err());
assert_eq!(budget.refused(), refused);
}
assert!(budget.decode(&doc, &Stream::flate(Dict::new(), b"samples"), 1 << 20).is_err());
assert!(budget.decode(&doc, &Stream::from_raw(Dict::new(), b"samples".to_vec()), 1 << 20).is_err());
assert_eq!(budget.refused(), 4, "other images share it");
for name in ["NoSuchFilter", "DCTDecode", "JPXDecode"] {
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name(name));
assert!(budget.decode(&doc, &Stream::from_raw(dict, b"data".to_vec()), 1 << 20).is_err());
assert_eq!(budget.refused(), 4, "{name} isn't one of these");
}
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::Int(5));
assert!(budget.decode(&doc, &Stream::from_raw(dict, b"data".to_vec()), 1 << 20).is_err());
assert_eq!(budget.refused(), 4, "a /Filter that can't be read isn't either");
}
/// What a file of images each inflating a megabyte asks for: 4 KB of Flate inside Flate can ask
/// for a gigabyte, and these aren't JBIG2, so without the allowance nothing would stop them.
#[test]
fn filter_bombs_stop_at_the_allowance() {
let doc = Document::new_empty();
let samples = vec![0; 1 << 20];
let inner = pdfcraft_filters::encode_flate(&samples);
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::Array(vec![Object::name("FlateDecode"), Object::name("FlateDecode")]));
let bomb = Stream::from_raw(dict, pdfcraft_filters::encode_flate(&inner));
let cost = (bomb.raw.len() + inner.len() + samples.len()) as u64;
let mut budget = ImageBudget::with_work(2 * cost + 1);
for left in [cost + 1, 1] {
assert_eq!(budget.decode(&doc, &bomb, 1 << 20).unwrap(), samples);
assert_eq!(budget.work_left(), left);
}
assert!(budget.decode(&doc, &bomb, 1 << 20).is_err());
assert_eq!((budget.work_left(), budget.refused()), (0, 1), "the first it can't pay for ends it");
}
/// Filters in front of a JBIG2 decoder pay as they run, on top of what the decoder takes, and
/// one that fails has still paid: nothing about the JBIG2 data is known by then, and each image
/// of a file can do this.
#[test]
fn filters_in_front_of_jbig2_pay_whether_or_not_they_work() {
let doc = Document::new_empty();
let alone = jbig2(None);
let mut budget = ImageBudget::with_work(u64::MAX);
let samples = budget.decode(&doc, &alone, 1 << 20).unwrap();
let decoding = u64::MAX - budget.work_left();
let behind = |raw: Vec<u8>| {
let mut dict = alone.dict.clone();
dict.set(b"Filter".to_vec(), Object::Array(vec![Object::name("FlateDecode"), Object::name("JBIG2Decode")]));
Stream::from_raw(dict, raw)
};
let image = behind(pdfcraft_filters::encode_flate(&alone.raw));
let mut budget = ImageBudget::with_work(u64::MAX);
assert_eq!(budget.decode(&doc, &image, 1 << 20).unwrap(), samples);
assert_eq!(u64::MAX - budget.work_left(), decoding + (image.raw.len() + alone.raw.len()) as u64);
// Past this image's limit: an error, with what finding out took paid and the rest left.
let bomb = behind(pdfcraft_filters::encode_flate(&vec![0; 1 << 16]));
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(budget.decode(&doc, &bomb, 1000).is_err());
assert_eq!(u64::MAX - budget.work_left(), bomb.raw.len() as u64 + 1000);
assert_eq!(budget.refused(), 0, "that image's limit isn't the run's");
// A small allowance is a limit too: it stops there, and the image is refused.
let mut budget = ImageBudget::with_work(bomb.raw.len() as u64 + 10);
assert!(budget.decode(&doc, &bomb, 1 << 20).is_err());
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
}
/// Globals that are read again for every image pay every time, and fail having paid. Kept
/// ones have no filters left to undo, so they pay for being read and nothing more.
#[test]
fn globals_pay_for_their_filters_each_time_they_are_decoded() {
let skipped = [0, 0, 0, 0, 0x3E, 0, 0, 0, 0, 0, 4, 0, 0, 0, 1];
let raw = pdfcraft_filters::encode_flate(&skipped);
let stream = |doc: &Document, r: ObjRef| match &*doc.get(r) {
Object::Stream(s) => s.clone(),
_ => panic!("not a stream"),
};
let spent = |doc: &Document, s: &Stream| {
let mut budget = ImageBudget::with_work(u64::MAX);
budget.decode(doc, s, 1 << 20).unwrap();
u64::MAX - budget.work_left()
};
let mut doc = Document::new_empty();
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("FlateDecode"));
let globals = doc.add(Object::Stream(Stream::from_raw(dict, raw.clone())));
let mut parms = Dict::new();
parms.set(b"JBIG2Globals".to_vec(), Object::Ref(globals));
let image = doc.add(Object::Stream(jbig2(Some(Object::Dict(parms)))));
let root = doc.root().unwrap();
doc.update_dict(root, |c| c.set(b"Image".to_vec(), Object::Ref(image))).unwrap();
// Added in memory, so not kept: the filters are undone for each read.
let s = stream(&doc, image);
let first = spent(&doc, &s);
assert_eq!(spent(&doc, &s), first, "not kept, so paid for again");
// Read from a file, so kept: the second read has no filters to undo.
let file = crate::write_full(&doc, &crate::SaveOptions::default()).unwrap();
let doc = Document::open(std::sync::Arc::new(file)).unwrap();
let image = doc.get(doc.root().unwrap()).as_dict().unwrap().reference(b"Image").unwrap();
let s = stream(&doc, image);
let filters = (raw.len() + skipped.len()) as u64;
assert_eq!(spent(&doc, &s), first, "the first read undoes them");
assert_eq!(spent(&doc, &s), first - filters, "kept");
// Globals that fail pay too, with the image not refused: these go past their own limit.
let over = pdfcraft_filters::encode_flate(&vec![0; crate::MAX_JBIG2_GLOBALS + 1]);
let mut doc = Document::new_empty();
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("FlateDecode"));
let globals = doc.add(Object::Stream(Stream::from_raw(dict, over.clone())));
let mut parms = Dict::new();
parms.set(b"JBIG2Globals".to_vec(), Object::Ref(globals));
let image = jbig2(Some(Object::Dict(parms)));
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(budget.decode(&doc, &image, 1 << 20).unwrap_err().to_string().contains("globals"));
assert_eq!(u64::MAX - budget.work_left(), (over.len() + crate::MAX_JBIG2_GLOBALS) as u64);
assert_eq!(budget.refused(), 0);
}
/// A fax image pays for what comes before its decoder and for the rows it asks for, and past
/// the allowance it is refused like any other.
#[test]
fn fax_images_share_the_allowance() {
let doc = Document::new_empty();
let (w, h) = (4096u32, 64u32);
let coded = pdfcraft_filters::encode_g4(&vec![0xFF; 512 * h as usize], w, h).unwrap();
let mut parms = Dict::new();
parms.set(b"K".to_vec(), Object::Int(-1));
parms.set(b"Columns".to_vec(), Object::Int(i64::from(w)));
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(i64::from(w)));
dict.set(b"Height".to_vec(), Object::Int(i64::from(h)));
dict.set(b"Filter".to_vec(), Object::Array(vec![Object::name("FlateDecode"), Object::name("CCITTFaxDecode")]));
dict.set(b"DecodeParms".to_vec(), Object::Array(vec![Object::Null, Object::Dict(parms)]));
let image = Stream::from_raw(dict, pdfcraft_filters::encode_flate(&coded));
let pixels = u64::from(w) * u64::from(h);
let cost = (image.raw.len() + coded.len() + 4 * coded.len() + 512 * h as usize) as u64 + pixels;
let mut budget = ImageBudget::with_work(cost);
assert_eq!(budget.decode(&doc, &image, 1 << 20).unwrap().len(), 512 * h as usize);
assert_eq!(budget.work_left(), 0, "exactly that");
assert!(budget.decode(&doc, &image, 1 << 20).is_err());
assert_eq!(budget.refused(), 1);
}
/// An image pays for its pixels however few bytes hold them, or a blank page of fax or Flate
/// would cost a run a second for a few bytes' worth of allowance; a JBIG2 image doesn't pay
/// twice, and one this doesn't decode pays nothing. Past what is left it isn't read.
#[test]
fn an_image_pays_for_its_pixels() {
let doc = Document::new_empty();
let blank = |filter: Option<&str>, data: Vec<u8>| {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(4096));
dict.set(b"Height".to_vec(), Object::Int(1024));
if let Some(filter) = filter {
dict.set(b"Filter".to_vec(), Object::name(filter));
}
Stream::from_raw(dict, data)
};
let rows = vec![0xFFu8; 512 * 1024];
let flate = blank(Some("FlateDecode"), pdfcraft_filters::encode_flate(&rows));
let cost = (flate.raw.len() + rows.len()) as u64 + 4096 * 1024;
let mut budget = ImageBudget::with_work(cost);
assert_eq!(budget.decode(&doc, &flate, 1 << 24).unwrap(), rows);
assert_eq!((budget.work_left(), budget.refused()), (0, 0), "exactly that");
// Its filters alone pay the same.
let mut budget = ImageBudget::with_work(cost);
assert_eq!(budget.decode_filters(&doc, &flate, 1 << 24).unwrap(), rows);
assert_eq!(budget.work_left(), 0);
// Short of its pixels it is refused, having paid for its filters.
let mut budget = ImageBudget::with_work(4096 * 1024 - 1);
assert!(budget.decode(&doc, &flate, 1 << 24).unwrap_err().to_string().contains("more image decoding than"));
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
let mut budget = ImageBudget::with_work(4096 * 1024 - 1);
assert!(budget.decode_filters(&doc, &flate, 1 << 24).is_err());
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
// What isn't decoded here pays for nothing, and fails as it would.
let mut budget = ImageBudget::with_work(10);
assert!(budget.decode(&doc, &blank(Some("DCTDecode"), b"data".to_vec()), 1 << 24).is_err());
assert_eq!((budget.work_left(), budget.refused()), (10, 0));
// Unfiltered data pays too, and a size that isn't one is no size.
let mut budget = ImageBudget::with_work(u64::MAX);
budget.decode(&doc, &blank(None, rows.clone()), 1 << 24).unwrap();
assert_eq!(u64::MAX - budget.work_left(), rows.len() as u64 + 4096 * 1024);
let mut odd = Dict::new();
odd.set(b"Width".to_vec(), Object::Int(-4));
odd.set(b"Height".to_vec(), Object::Int(9));
let mut budget = ImageBudget::with_work(u64::MAX);
budget.decode(&doc, &Stream::from_raw(odd, vec![1, 2]), 1 << 24).unwrap();
assert_eq!(u64::MAX - budget.work_left(), 2);
// A JPEG pays twice over for each sample, and four times for each byte it is read from.
let mut budget = ImageBudget::with_work(30);
assert!(budget.pay_jpeg(9, b"ab"));
assert_eq!((budget.work_left(), budget.refused()), (4, 0));
assert!(!budget.pay_jpeg(0, b"ab"));
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
}
/// An image pays for the pixels its data holds, not those its size declares: one claiming
/// 131072 by 131072 pixels with a few bytes of Flate costs what those bytes give, and the
/// images after it are decoded as they would be without it.
#[test]
fn an_image_pays_for_no_more_pixels_than_its_data_holds() {
let doc = Document::new_empty();
let sized = |side: i64, data: &[u8]| {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(side));
dict.set(b"Height".to_vec(), Object::Int(side));
Stream::flate(dict, data)
};
let claims = sized(131_072, &[0; 16]);
let image = sized(64, &[7; 64 * 64]);
let cost = |s: &Stream| {
let mut budget = ImageBudget::with_work(u64::MAX);
budget.decode(&doc, s, 1 << 24).unwrap();
u64::MAX - budget.work_left()
};
assert_eq!(cost(&claims), (claims.raw.len() + 16 + 16 * 8) as u64, "its bytes, and a pixel a bit of them");
assert_eq!(cost(&image), (image.raw.len() + 64 * 64) as u64 + 64 * 64, "its bytes, and its pixels");
for decode in [ImageBudget::decode, ImageBudget::decode_filters] {
let mut budget = ImageBudget::with_work(cost(&claims) + cost(&image));
assert_eq!(decode(&mut budget, &doc, &claims, 1 << 24).unwrap(), [0; 16]);
assert_eq!(decode(&mut budget, &doc, &image, 1 << 24).unwrap(), [7; 64 * 64], "the image after it is decoded");
assert_eq!((budget.work_left(), budget.refused()), (0, 0));
}
}
/// A stream a run rewrites is read strictly and charged, and what the caller does with it after
/// comes out of the same allowance: what can't be paid for spends all of it.
#[test]
fn a_rewritten_stream_shares_the_allowance() {
use pdfcraft_filters::{Filter, FilterError, Params};
let samples = vec![7u8; 1000];
let raw = pdfcraft_filters::encode_flate(&samples);
let chain = [(Filter::Flate, Params::default())];
let cost = (raw.len() + samples.len()) as u64;
let mut budget = ImageBudget::with_work(cost + 10);
assert_eq!(budget.decode_chain(&chain, &raw, 1 << 20).unwrap(), samples);
assert_eq!(budget.work_left(), 10);
assert!(budget.spend(4));
assert_eq!(budget.work_left(), 6);
assert!(budget.spend(6), "exactly what is left");
assert_eq!(budget.work_left(), 0);
assert!(budget.spend(0));
let mut budget = ImageBudget::with_work(5);
assert!(!budget.spend(6));
assert_eq!(budget.work_left(), 0, "spends all of it");
assert!(matches!(budget.decode_chain(&chain, &raw, 1 << 20), Err(FilterError::Spent)));
let cut = raw.get(..raw.len() / 2).unwrap();
assert!(matches!(ImageBudget::with_work(u64::MAX).decode_chain(&chain, cut, 1 << 20), Err(FilterError::Corrupt { .. })), "strictly");
}
/// An image read through its filters alone reads parameters given by reference, is charged, and
/// is refused like the others. It isn't given to a decoder.
#[test]
fn filters_alone_are_charged_too() {
let mut doc = Document::new_empty();
let p = pdfcraft_filters::Params { predictor: 12, columns: 4, ..pdfcraft_filters::Params::default() };
let rows = vec![7u8; 400];
let mut parms = Dict::new();
parms.set(b"Predictor".to_vec(), Object::Int(12));
parms.set(b"Columns".to_vec(), Object::Int(4));
let parms = doc.add(Object::Dict(parms));
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("FlateDecode"));
dict.set(b"DecodeParms".to_vec(), Object::Ref(parms));
let raw = pdfcraft_filters::encode(&pdfcraft_filters::Filter::Flate, &p, &rows).unwrap();
let image = Stream::from_raw(dict, raw);
let mut budget = ImageBudget::with_work(u64::MAX);
assert_eq!(budget.decode_filters(&doc, &image, 1 << 20).unwrap(), rows);
let cost = u64::MAX - budget.work_left();
// The tag byte of each of its 100 rows counts, which the predictor takes off.
assert_eq!(cost, (image.raw.len() + rows.len() + 100) as u64);
let mut budget = ImageBudget::with_work(cost - 1);
assert!(budget.decode_filters(&doc, &image, 1 << 20).is_err());
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(budget.decode_filters(&doc, &jbig2(None), 1 << 20).is_err());
assert_eq!(budget.work_left(), u64::MAX, "no decoder here, so nothing was read");
// Nor for a fax image, nor for a decoder that isn't last, however large they say they are.
for chain in [vec!["CCITTFaxDecode"], vec!["CCITTFaxDecode", "FlateDecode"], vec!["JBIG2Decode", "FlateDecode"]] {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(1000));
dict.set(b"Height".to_vec(), Object::Int(1000));
dict.set(b"Filter".to_vec(), Object::Array(chain.iter().map(|f| Object::name(f)).collect()));
let image = Stream::from_raw(dict, b"data".to_vec());
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(budget.decode_filters(&doc, &image, 1 << 20).is_err());
assert_eq!(budget.work_left(), u64::MAX, "{chain:?}");
}
// Through `decode` a decoder that isn't last pays nothing either; one that is does.
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(1000));
dict.set(b"Height".to_vec(), Object::Int(1000));
dict.set(b"Filter".to_vec(), Object::Array(vec![Object::name("JBIG2Decode"), Object::name("FlateDecode")]));
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(budget.decode(&doc, &Stream::from_raw(dict, b"data".to_vec()), 1 << 20).is_err());
assert_eq!(budget.work_left(), u64::MAX);
}
/// The images of a file share its globals, so an image refused doesn't read them: this one's
/// can't be read, which is how to tell.
#[test]
fn an_image_past_the_allowance_does_not_read_its_globals() {
let mut doc = Document::new_empty();
let mut unreadable = Dict::new();
unreadable.set(b"Filter".to_vec(), Object::name("NoSuchFilter"));
let globals = doc.add(Object::Stream(Stream::from_raw(unreadable, b"data".to_vec())));
let mut parms = Dict::new();
parms.set(b"JBIG2Globals".to_vec(), Object::Ref(globals));
let image = jbig2(Some(Object::Dict(parms)));
let read = ImageBudget::with_work(u64::MAX).decode(&doc, &image, 1 << 20).unwrap_err().to_string();
assert!(read.contains("globals"), "{read}");
let refused = ImageBudget::with_work(0).decode(&doc, &image, 1 << 20).unwrap_err().to_string();
assert!(refused.contains("more image decoding than"), "{refused}");
}
/// A longer file gets more, by its length, and an empty one the fixed part.
#[test]
fn a_longer_file_allows_more_decoding() {
assert_eq!(ImageBudget::new(&Document::new_empty()).work_left(), IMAGE_WORK);
let pdf = b"%PDF-1.7\n1 0 obj << /Type /Catalog >> endobj\ntrailer << /Root 1 0 R >>\n%%EOF".to_vec();
let len = pdf.len() as u64;
let doc = Document::open(std::sync::Arc::new(pdf)).unwrap();
assert_eq!(ImageBudget::new(&doc).work_left(), IMAGE_WORK + len * IMAGE_WORK_PER_BYTE);
}
/// A document made in memory, as Combine, Create and inserted scans make one, gets for the
/// streams it was given what it would get for them once saved: what isn't a stream adds nothing.
#[test]
fn streams_made_in_memory_allow_more_decoding_too() {
let mut doc = Document::new_empty();
doc.add(Object::Stream(Stream::from_raw(Dict::new(), vec![0; 3000])));
doc.add(Object::Dict(Dict::new()));
doc.add(Object::Stream(Stream::flate(Dict::new(), &[1; 100])));
let flate = pdfcraft_filters::encode_flate(&[1; 100]).len() as u64;
assert_eq!(ImageBudget::new(&doc).work_left(), IMAGE_WORK + (3000 + flate) * IMAGE_WORK_PER_BYTE);
// A stream of an opened file changed in memory counts with the file.
let pdf = b"%PDF-1.7\n1 0 obj << /Type /Catalog >> endobj\ntrailer << /Root 1 0 R >>\n%%EOF".to_vec();
let len = pdf.len() as u64;
let mut doc = Document::open(std::sync::Arc::new(pdf)).unwrap();
doc.add(Object::Stream(Stream::from_raw(Dict::new(), vec![0; 500])));
assert_eq!(ImageBudget::new(&doc).work_left(), IMAGE_WORK + (len + 500) * IMAGE_WORK_PER_BYTE);
}
// What the constants' comments promise, for a page scanned at 300 dpi: about 1,000 of them
// whatever the file's size, and a page paying for itself with 2 KB of the file.
const PAGE: u64 = 17_000_000;
const _: () = assert!(IMAGE_WORK >= 1_000 * PAGE);
const _: () = assert!(IMAGE_WORK_PER_BYTE * 2_100 >= PAGE);
// And that a stream can be decoded about 8 times at Flate's best, 1,032 to one, and the byte in.
const _: () = assert!(IMAGE_WORK_PER_BYTE >= 7 * 1_033);
}
#[cfg(test)]
mod tests {
use super::*;
@@ -198,6 +795,21 @@ mod tests {
assert_eq!(budget.remaining(), (10 << 20) - partial);
}
/// A stream that fails its reading as zlib, and then as raw deflate after decoding all of it,
/// gives nothing back, and each reference still pays for what it decoded.
#[test]
fn a_decode_that_gives_nothing_back_still_pays() {
let mut data = vec![0x78, 0x01, 0x06, 0xFE, 0xF9];
data.extend(vec![0u8; 0x0601]);
data.push(0x07);
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("FlateDecode"));
let stream = Stream::from_raw(dict, data);
let mut budget = DecodeBudget::new(10 * 0x0601, 1);
assert!(budget.decode(&stream).unwrap().is_empty());
assert_eq!(budget.remaining(), 9 * 0x0601);
}
/// A chain whose later stage shrinks the data pays for every stage's output: a Flate stage
/// that inflates to hex text, which ASCIIHex then halves, charges both.
#[test]
crates/cos/src/document/objects.rs
+28 -27
View File
@@ -274,34 +274,32 @@ impl Document {
.unwrap_or_default()
}
/// The samples of image `s`, decoded: through its filters ([`Stream::decoded`]), or, for a
/// CCITT fax or JBIG2 image (its last filter), through the filters before it and then
/// [`pdfcraft_filters::decode_ccitt`] with its [`Document::fax_params`], or
/// [`pdfcraft_filters::decode_jbig2`] with its `/JBIG2Globals`, as `CCITTFaxDecode` and
/// `JBIG2Decode` give them (ISO 32000-2 §7.4.6, §7.4.7). Parameters given by reference count,
/// for every filter ([`Document::with_direct_parms`]). A fax image's `/Columns` has to be its
/// `/Width`, and a JBIG2 image's page its `/Width` × `/Height`: PdfCraft's renderer draws the
/// whole page into the image. Up to its `/Height` rows are decoded, whatever a fax image's `/Rows` says,
/// within [`MAX_DECODED`] bytes; damaged fax data gives fewer, so a caller checks it has the
/// rows it needs. Damaged JBIG2 data, and globals that can't be read (within
/// [`MAX_JBIG2_GLOBALS`]), are an error.
pub fn decoded_image(&self, s: &Stream) -> Result<Vec<u8>, CosError> {
/// [`Document::decoded_image_spending`] with no limit on the work, for tests of what an image
/// decodes to: everything else decodes images through an [`ImageBudget`](crate::ImageBudget).
#[cfg(test)]
pub(crate) fn decoded_image(&self, s: &Stream) -> Result<Vec<u8>, CosError> {
self.decoded_image_within(s, MAX_DECODED)
}
/// [`Document::decoded_image`] for an image whose decoding has a tighter bound than
/// [`MAX_DECODED`]: no stage, the filters before a fax or JBIG2 decoder included, may produce
/// more than `max` bytes. A JBIG2 image's `/JBIG2Globals` are another stream, shared between
/// images, and are decoded within [`MAX_JBIG2_GLOBALS`] whatever `max` is.
pub fn decoded_image_within(&self, s: &Stream, max: usize) -> Result<Vec<u8>, CosError> {
/// [`Document::decoded_image`] within `max` bytes.
#[cfg(test)]
pub(crate) fn decoded_image_within(&self, s: &Stream, max: usize) -> Result<Vec<u8>, CosError> {
let mut unbounded = u64::MAX;
self.decoded_image_spending(s, max, &mut unbounded)
}
/// The samples of image `s`, decoded within `max` bytes ([`MAX_DECODED`] at most), spending from
/// `work`: see [`ImageBudget::decode`](crate::ImageBudget::decode), the way in from outside this
/// crate, which says what this decodes and how.
pub(crate) fn decoded_image_spending(&self, s: &Stream, max: usize, work: &mut u64) -> Result<Vec<u8>, CosError> {
let max = max.min(MAX_DECODED);
let s = &*self.with_direct_parms(s);
let filter = |e: pdfcraft_filters::FilterError| CosError::Filter(e.to_string());
let chain = s.readable_filters().map_err(filter)?;
// The filters before the image codec, such as ASCII85Decode, are undone first.
let undo = |before: &[(pdfcraft_filters::Filter, pdfcraft_filters::Params)]| match before {
let undo = |before: &[(pdfcraft_filters::Filter, pdfcraft_filters::Params)], work: &mut u64| match before {
[] => Ok(Cow::Borrowed(&s.raw[..])),
_ => pdfcraft_filters::decode_tolerant(before, &s.raw, max).map(|(data, _)| Cow::Owned(data)).map_err(filter),
_ => pdfcraft_filters::decode_tolerant_within(before, &s.raw, max, work).map(Cow::Owned).map_err(filter),
};
// `/Width` and `/Height` are read as given directly, as every caller reads them first: one
// given by reference has the image refused there, before it gets here.
@@ -310,10 +308,12 @@ impl Document {
let (Some(columns), Some(height)) = (s.dict.int(b"Width").and_then(|w| u32::try_from(w).ok()), height) else {
return Err(CosError::Filter("a JBIG2 image's /Width or /Height isn't a size".into()));
};
let globals = self.jbig2_globals(s)?;
return pdfcraft_filters::decode_jbig2(&undo(before)?, globals.as_deref().map(Vec::as_slice), columns, height, max).map_err(filter);
// Nothing to read for what can't be decoded: the images of a file share its globals.
let globals = if *work == 0 { None } else { self.jbig2_globals(s, work)? };
let data = if *work == 0 { Cow::Borrowed(&[][..]) } else { undo(before, work)? };
return pdfcraft_filters::decode_jbig2_within(&data, globals.as_deref().map(Vec::as_slice), columns, height, max, work).map_err(filter);
}
let Some(p) = self.fax_params(s) else { return s.decoded_within(max) };
let Some(p) = self.fax_params(s) else { return s.decoded_spending(max, work) };
// As many rows as the image has: a /Rows below /Height would cut the picture short, and
// rows past /Height aren't drawn.
let p = pdfcraft_filters::CcittParams { rows: 0, ..p };
@@ -322,8 +322,8 @@ impl Document {
}
let stride = usize::try_from(p.columns).unwrap_or(usize::MAX).div_ceil(8).max(1);
let rows = u32::try_from(max / stride).unwrap_or(u32::MAX).min(height.unwrap_or(0));
let data = undo(chain.split_last().map_or(&[][..], |(_, before)| before))?;
pdfcraft_filters::decode_ccitt(&data, &p, rows).map_err(filter)
let data = undo(chain.split_last().map_or(&[][..], |(_, before)| before), work)?;
pdfcraft_filters::decode_ccitt_within(&data, &p, rows, work).map_err(filter)
}
/// The `/JBIG2Globals` stream a JBIG2 image `s` names by reference in its last filter's
@@ -348,8 +348,9 @@ impl Document {
/// it was read (the same in every clone) and with its `/DecodeParms` direct (by reference they
/// are another object, which can change while this one doesn't), has its filters undone once a
/// document and is kept, up to [`JBIG2_GLOBALS_HELD`] in all; others are read each time. That
/// saves the filters only: the segments it holds are decoded again for every image.
fn jbig2_globals(&self, s: &Stream) -> Result<Option<Arc<Vec<u8>>>, CosError> {
/// saves the filters only: the segments it holds are decoded again for every image. Undoing
/// the filters spends from `work`, so globals read for every image pay every time.
fn jbig2_globals(&self, s: &Stream, work: &mut u64) -> Result<Option<Arc<Vec<u8>>>, CosError> {
let Some(entry) = self.last_filter_parms(s).get(b"JBIG2Globals").cloned() else { return Ok(None) };
let held = match entry {
Object::Ref(r) if !self.overlay.contains_key(&r.num) => Some(r.num),
@@ -361,7 +362,7 @@ impl Document {
let Object::Stream(g) = &*self.resolve(&entry) else { return Ok(None) };
let readable = self.with_direct_parms(g);
let held = held.filter(|_| matches!(readable, Cow::Borrowed(_)));
let mut data = readable.decoded_within(MAX_JBIG2_GLOBALS).map_err(|e| {
let mut data = readable.decoded_spending(MAX_JBIG2_GLOBALS, work).map_err(|e| {
let detail = match e {
CosError::Filter(detail) => detail,
e => e.to_string(),
crates/cos/src/lib.rs
+1 -1
View File
@@ -22,7 +22,7 @@ mod parser;
mod security;
mod writer;
pub use budget::{DecodeBudget, MAX_PAGE_DECODED, MAX_PAGE_FORMS};
pub use budget::{DecodeBudget, ImageBudget, MAX_PAGE_DECODED, MAX_PAGE_FORMS};
pub use bytes::Bytes;
pub use document::{Document, ReadState, Revision, SourceIdentity, XrefEntry};
pub use object::{Dict, MAX_DECODED, MAX_JBIG2_GLOBALS, Name, ObjRef, Object, PdfString, Stream};
crates/cos/src/object.rs
+8
View File
@@ -306,6 +306,14 @@ impl Stream {
self.decode_chain(max, decode)
}
/// [`Stream::decoded_within`] spending from `work`, what an operation may still do across a
/// document's images ([`pdfcraft_filters::decode_tolerant_within`]).
pub(crate) fn decoded_spending(&self, max: usize, work: &mut u64) -> Result<Vec<u8>, CosError> {
let filter = |e: pdfcraft_filters::FilterError| CosError::Filter(e.to_string());
let chain = self.readable_filters().map_err(filter)?;
pdfcraft_filters::decode_tolerant_within(&chain, &self.raw, max.min(MAX_DECODED), work).map_err(filter)
}
/// The stream decoded through its filters by `decode`, or copied when it has none, within
/// `max` and [`MAX_DECODED`].
fn decode_chain(
crates/create/README.md
+5 -3
View File
@@ -27,12 +27,14 @@ Layer L4. Create a PDF (Acrobat's Create a PDF tool), execution plan M10.2:
Every function returns a `pdfcraft_cos::Document`; the caller writes it.
Image export and Save Image As apply each lossless image's Decode pair before converting its color components.
JBIG2 images and masks are decoded through `pdfcraft_cos::Document::decoded_image` (`pdfcraft_filters::decode_jbig2`), behind
JBIG2 images and masks are decoded through `pdfcraft_cos::ImageBudget` (`pdfcraft_filters::decode_jbig2_within`), behind
other filters too, with the `/JBIG2Globals` of their own `/DecodeParms` entry, as `JBIG2Decode` gives them: black as 0. An image
whose JBIG2 page isn't its size (PdfCraft's renderer draws the whole page into the image; the page is refused before it is decoded),
or whose globals can't be read or decode to more than 32 MiB (they are decoded before the page's size is known, so they have a
limit of their own), is skipped; such a mask is left out and the image exported opaque, as JPX and CCITT
masks are.
limit of their own), is skipped, and such a mask is left out and the image exported opaque, as JPX and CCITT masks are.
Past what one export may decode, which its images share, filters, decoders and pixels alike (`ImageBudget`; a JPEG when it has to be
decoded, as a mask, to apply one, or to undo a CMYK file's inversion, and not when it is written as it is), an image is skipped, and so is one whose mask is past it, which would otherwise come out showing what the mask hides. An export also stops at 2 GiB of files held, or twice the document if that
is more, since they stay in memory until it ends.
Indexed Decode values select palette entries after rounding and clipping. Color-key masks compare the original samples.
Missing, null or malformed Decode arrays keep the default mapping.
Soft-mask Decode intervals map samples to opacity before the mask is resized, including reversed ranges and indirect entries.
crates/create/src/extract.rs
+86 -22
View File
@@ -6,7 +6,8 @@
use std::collections::HashSet;
use pdfcraft_cos::{Dict, Document, ObjRef, Object, Stream};
use pdfcraft_cos::{Dict, Document, ImageBudget, MAX_DECODED, ObjRef, Object, Stream};
use pdfcraft_filters::FilterError;
/// One image, ready to write.
#[derive(Clone, Debug, PartialEq)]
@@ -31,12 +32,29 @@ pub struct ImageExport {
/// Images larger than this many pixels are skipped (memory).
const MAX_PIXELS: u64 = 100_000_000;
/// What one export may hold of the files it makes: they are all in memory until it ends, and a file
/// of a few KB can make gigabytes of PNG. Images past it are left out, with the reason.
const MAX_EXPORT_BYTES: usize = 2 << 30;
/// What an export of a file of `file_bytes` may hold: twice the file if that is more than
/// [`MAX_EXPORT_BYTES`], so that a file of large images is exported whole.
pub(crate) fn export_bytes(file_bytes: usize) -> usize {
MAX_EXPORT_BYTES.max(file_bytes.saturating_mul(2))
}
/// The images of `pages` (0-based; each image once, on the first page using it), skipping
/// those with fewer than `min_side` pixels on their shorter side.
pub fn extract_images(doc: &Document, pages: &[usize], min_side: u32) -> ImageExport {
extract_images_within(doc, pages, min_side, &mut ImageBudget::new(doc), export_bytes(doc.bytes().len()))
}
/// [`extract_images`] within `budget`, and holding `max_bytes` of files. An image past either is
/// left out, with the reason.
pub(crate) fn extract_images_within(doc: &Document, pages: &[usize], min_side: u32, budget: &mut ImageBudget, max_bytes: usize) -> ImageExport {
let all = pdfcraft_model::pages(doc);
let mut out = ImageExport::default();
let mut seen = HashSet::new();
let mut held = 0usize;
for &p in pages {
let Some(page) = all.get(p) else { continue };
let mut refs = Vec::new();
@@ -53,8 +71,15 @@ pub fn extract_images(doc: &Document, pages: &[usize], min_side: u32) -> ImageEx
if w <= 0 || h <= 0 || (w.min(h) as u64) < min_side as u64 {
continue;
}
match image(doc, s) {
Ok((extension, data)) => out.images.push(ExtractedImage { page: p, object: r, width: w as u32, height: h as u32, extension, data }),
if held >= max_bytes {
out.skipped.push((p, r, "more images than one export can hold".into()));
continue;
}
match image(doc, s, budget) {
Ok((extension, data)) => {
held = held.saturating_add(data.len());
out.images.push(ExtractedImage { page: p, object: r, width: w as u32, height: h as u32, extension, data });
}
Err(why) => out.skipped.push((p, r, why)),
}
}
@@ -64,8 +89,13 @@ pub fn extract_images(doc: &Document, pages: &[usize], min_side: u32) -> ImageEx
/// One image XObject as a file: JPEG as is, else PNG ("Save image as").
pub fn image_file(doc: &Document, image: ObjRef) -> Result<(&'static str, Vec<u8>), String> {
image_file_within(doc, image, &mut ImageBudget::new(doc))
}
/// [`image_file`] within `budget`, for a caller making a file of many images.
pub fn image_file_within(doc: &Document, image: ObjRef, budget: &mut ImageBudget) -> Result<(&'static str, Vec<u8>), String> {
match &*doc.get(image) {
Object::Stream(s) if s.dict.name(b"Subtype") == Some(b"Image") => self::image(doc, s),
Object::Stream(s) if s.dict.name(b"Subtype") == Some(b"Image") => self::image(doc, s, budget),
_ => Err("not an image".into()),
}
}
@@ -100,20 +130,20 @@ fn filters(s: &Stream) -> Vec<Vec<u8>> {
}
}
/// The samples of image `s`, decoded ([`Document::decoded_image`]): through its filters, with a
/// The samples of image `s`, decoded ([`ImageBudget::decode`]): through its filters, with a
/// `/DecodeParms` dictionary given by reference, or as an array's entry, or with numbers inside by
/// reference, read too (a predictor given so would otherwise be lost); a CCITT fax or JBIG2 image
/// through its decoder too, behind other filters as well, a JBIG2 image with its `/JBIG2Globals`.
/// A JBIG2 page has to be the image's size, which the caller checked against the export's limit,
/// before the page is decoded.
fn decoded(doc: &Document, s: &Stream) -> Result<Vec<u8>, String> {
doc.decoded_image(s).map_err(|e| e.to_string())
fn decoded(doc: &Document, s: &Stream, budget: &mut ImageBudget) -> Result<Vec<u8>, String> {
budget.decode(doc, s, MAX_DECODED).map_err(|e| e.to_string())
}
/// The file for one image: its JPEG data as is, or a PNG. An image with transparency (a soft
/// mask, a stencil `/Mask` or a colour-key `/Mask`) is always a PNG with that transparency as
/// alpha, JPEG images included, so cut-out images keep their transparent background.
fn image(doc: &Document, s: &Stream) -> Result<(&'static str, Vec<u8>), String> {
fn image(doc: &Document, s: &Stream, budget: &mut ImageBudget) -> Result<(&'static str, Vec<u8>), String> {
let f = filters(s);
let jpeg = matches!(f.last().map(Vec::as_slice), Some(b"DCTDecode" | b"DCT")) && f.len() == 1;
match f.last().map(Vec::as_slice) {
@@ -132,7 +162,12 @@ fn image(doc: &Document, s: &Stream) -> Result<(&'static str, Vec<u8>), String>
};
let mask = s.dict.get(b"ImageMask").is_some_and(|m| matches!(&*doc.resolve(m), Object::Bool(true)));
// A stencil mask (`ImageMask true`) has no /SMask or /Mask of its own (8.9.6.2).
let mut alpha = if mask { None } else { soft_mask(doc, s, w, h).or_else(|| stencil_mask(doc, s, w, h)) };
let refused = budget.refused();
let mut alpha = if mask { None } else { soft_mask(doc, s, w, h, budget).or_else(|| stencil_mask(doc, s, w, h, budget)) };
// Without its mask the image would come out opaque, showing what the mask hides.
if budget.refused() > refused {
return Err(FilterError::Spent.to_string());
}
let key = if mask { None } else { colour_key(doc, s) };
// Keep the existing JPEG and stencil-mask Decode handling separate from lossless colour images.
let endpoint = |value: Option<&Object>| value.map(|v| doc.resolve(v)).and_then(|v| v.as_f64()).filter(|v| v.is_finite());
@@ -155,13 +190,16 @@ fn image(doc: &Document, s: &Stream) -> Result<(&'static str, Vec<u8>), String>
}
let mut space = if mask { Space::Gray } else { space(doc, s.dict.get(b"ColorSpace"))? };
let (samples, bpc) = if jpeg {
if !budget.pay_jpeg((w as u64).saturating_mul(h as u64).saturating_mul(space.components() as u64), &s.raw) {
return Err(FilterError::Spent.to_string());
}
// Grey, RGB or CMYK samples (YCCK is converted); a JPEG that doesn't decode, or an
// indexed one (not valid), is written as is, without its transparency.
let Some((samples, decoded)) = jpeg_samples(&s.raw, &space, w, h) else { return as_is() };
space = decoded;
(samples, 8)
} else {
let data = decoded(doc, s)?;
let data = decoded(doc, s, budget)?;
let bpc = if mask { 1 } else { s.dict.int(b"BitsPerComponent").unwrap_or(8) };
let bpc = match bpc {
1 => 1,
@@ -252,6 +290,32 @@ fn size(w: i64, h: i64) -> Option<(usize, usize)> {
(w > 0 && h > 0 && pixels <= MAX_PIXELS).then_some((w, h))
}
#[cfg(test)]
thread_local! {
/// JPEGs decoded whole, for the tests that keep one from being decoded for nothing.
pub(crate) static JPEGS_DECODED: std::cell::Cell<usize> = const { std::cell::Cell::new(0) };
}
/// The size a JPEG's header gives, which is what decoding it goes by, whatever its dictionary
/// says: an image of one pixel can hold a header of 500 megapixels.
fn jpeg_size(data: &[u8]) -> Option<(usize, usize)> {
use zune_jpeg::zune_core::{bytestream::ZCursor, options::DecoderOptions};
let options = DecoderOptions::default().set_strict_mode(false).set_max_width(usize::from(u16::MAX)).set_max_height(usize::from(u16::MAX));
let mut dec = zune_jpeg::JpegDecoder::new_with_options(ZCursor::new(data), options);
dec.decode_headers().ok()?;
dec.dimensions()
}
/// JPEG `data`, decoded, if its header says it is `w` × `h`.
fn jpeg_of_size(data: &[u8], w: usize, h: usize) -> Option<image::DynamicImage> {
if jpeg_size(data)? != (w, h) {
return None;
}
#[cfg(test)]
JPEGS_DECODED.with(|n| n.set(n.get() + 1));
image::load_from_memory_with_format(data, image::ImageFormat::Jpeg).ok()
}
/// A JPEG's samples (8 bits): grey for a grey image, CMYK as stored for a CMYK one, else RGB,
/// with the space they are in.
fn jpeg_samples(data: &[u8], space: &Space, w: usize, h: usize) -> Option<(Vec<u8>, Space)> {
@@ -260,10 +324,7 @@ fn jpeg_samples(data: &[u8], space: &Space, w: usize, h: usize) -> Option<(Vec<u
Space::Cmyk => return cmyk_jpeg_samples(data, w, h).map(|px| (px, Space::Cmyk)),
_ => {}
}
let img = image::load_from_memory_with_format(data, image::ImageFormat::Jpeg).ok()?;
if img.width() as usize != w || img.height() as usize != h {
return None;
}
let img = jpeg_of_size(data, w, h)?;
Some(match space {
Space::Gray => (img.into_luma8().into_raw(), Space::Gray),
_ => (img.into_rgb8().into_raw(), Space::Rgb),
@@ -464,32 +525,35 @@ fn cmyk(c: u8, m: u8, y: u8, k: u8) -> [u8; 3] {
}
/// The soft mask (`/SMask`) as 8-bit alpha, scaled to the image's size when it has another.
fn soft_mask(doc: &Document, s: &Stream, w: usize, h: usize) -> Option<Vec<u8>> {
fn soft_mask(doc: &Document, s: &Stream, w: usize, h: usize, budget: &mut ImageBudget) -> Option<Vec<u8>> {
let r = s.dict.get(b"SMask")?.as_ref()?;
let obj = doc.get(r);
let Object::Stream(m) = &*obj else { return None };
gray_plane(doc, m, w, h)
gray_plane(doc, m, w, h, budget)
}
/// A stencil `/Mask` (an image mask stream, 8.9.6.3) as 8-bit alpha: its 1 samples (0 with
/// `/Decode [1 0]`) hide the image.
fn stencil_mask(doc: &Document, s: &Stream, w: usize, h: usize) -> Option<Vec<u8>> {
fn stencil_mask(doc: &Document, s: &Stream, w: usize, h: usize, budget: &mut ImageBudget) -> Option<Vec<u8>> {
let r = s.dict.get(b"Mask")?.as_ref()?;
let obj = doc.get(r);
let Object::Stream(m) = &*obj else { return None };
Some(gray_plane(doc, m, w, h)?.into_iter().map(|v| 255 - v).collect())
Some(gray_plane(doc, m, w, h, budget)?.into_iter().map(|v| 255 - v).collect())
}
/// A one-component mask image as 8-bit values (its `/Decode` interval applied), scaled to
/// `w` × `h`. `None` when it can't be decoded (JPEG 2000 masks).
fn gray_plane(doc: &Document, m: &Stream, w: usize, h: usize) -> Option<Vec<u8>> {
fn gray_plane(doc: &Document, m: &Stream, w: usize, h: usize, budget: &mut ImageBudget) -> Option<Vec<u8>> {
let (mw, mh) = size(m.dict.int(b"Width")?, m.dict.int(b"Height")?)?;
let stencil = m.dict.get(b"ImageMask").is_some_and(|v| matches!(&*doc.resolve(v), Object::Bool(true)));
let f = filters(m);
let mut px = match f.last().map(Vec::as_slice) {
Some(b"DCTDecode" | b"DCT") if f.len() == 1 => {
let img = image::load_from_memory_with_format(&m.raw, image::ImageFormat::Jpeg).ok()?.into_luma8();
(img.width() as usize == mw && img.height() as usize == mh).then(|| img.into_raw())?
// Past the allowance the image goes too: `image` sees this one refused.
if !budget.pay_jpeg((mw as u64).saturating_mul(mh as u64), &m.raw) {
return None;
}
jpeg_of_size(&m.raw, mw, mh)?.into_luma8().into_raw()
}
Some(b"DCTDecode" | b"DCT" | b"JPXDecode") => return None,
_ => {
@@ -503,7 +567,7 @@ fn gray_plane(doc: &Document, m: &Stream, w: usize, h: usize) -> Option<Vec<u8>>
_ => return None,
};
let max = (1u32 << bpc.min(8)) - 1;
unpack(&decoded(doc, m).ok()?, mw, mh, 1, bpc)?.into_iter().map(|v| (v as u32 * 255 / max) as u8).collect()
unpack(&decoded(doc, m, budget).ok()?, mw, mh, 1, bpc)?.into_iter().map(|v| (v as u32 * 255 / max) as u8).collect()
}
};
if let Some([lo, hi]) = decode(doc, m, 1).and_then(|pairs| pairs.first().copied()) {
crates/create/src/lib.rs
+1 -1
View File
@@ -6,7 +6,7 @@ use pdfcraft_cos::{Dict, Document, ObjRef, Object, PdfString, Stream, serialize}
mod extract;
mod upright_image;
pub use extract::{ExtractedImage, ImageExport, extract_images, image_file};
pub use extract::{ExtractedImage, ImageExport, extract_images, image_file, image_file_within};
use pdfcraft_fonts::{literal, win_ansi, wrap};
#[derive(Debug, thiserror::Error, PartialEq)]
crates/create/src/tests.rs
+254
View File
@@ -2234,6 +2234,260 @@ fn jbig2_globals_are_decoded_within_a_limit_of_their_own() {
}
}
/// One allowance for all the images of an export: an image past it is left out, saying so.
#[test]
fn jbig2_images_past_the_exports_allowance_are_left_out_with_the_reason() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(1));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
dict.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
let image = Stream::from_raw(dict, pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap());
let mut cost = ImageBudget::with_work(u64::MAX);
cost.decode(&Document::new_empty(), &image, 1 << 20).unwrap();
let cost = u64::MAX - cost.work_left();
let pdf = b"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 100 100] >> endobj\ntrailer << /Root 1 0 R >>\n%%EOF".to_vec();
let mut doc = Document::open(std::sync::Arc::new(pdf)).unwrap();
let (a, b) = (doc.add(Object::Stream(image.clone())), doc.add(Object::Stream(image)));
let mut xobjects = Dict::new();
xobjects.set(b"A".to_vec(), Object::Ref(a));
xobjects.set(b"B".to_vec(), Object::Ref(b));
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
let page = pdfcraft_model::pages(&doc).swap_remove(0).obj;
doc.update_dict(page, |p| p.set(b"Resources".to_vec(), Object::Dict(resources))).unwrap();
let export = crate::extract::extract_images_within(&doc, &[0], 1, &mut ImageBudget::with_work(cost), usize::MAX);
assert_eq!(export.images.len(), 1);
assert_eq!(export.skipped.len(), 1);
assert!(export.skipped[0].2.contains("more image decoding than"), "{:?}", export.skipped);
// What an export holds is bounded too: the files are all in memory until it ends.
let export = crate::extract::extract_images_within(&doc, &[0], 1, &mut ImageBudget::with_work(u64::MAX), 1);
assert_eq!(export.images.len(), 1, "the first goes past it");
assert_eq!(export.skipped.len(), 1);
assert!(export.skipped[0].2.contains("than one export can hold"), "{:?}", export.skipped);
// A file of many images shares one allowance too, image by image.
let mut budget = ImageBudget::with_work(cost);
assert!(crate::extract::image_file_within(&doc, a, &mut budget).is_ok());
assert!(crate::extract::image_file_within(&doc, b, &mut budget).is_err());
}
/// An image that isn't JBIG2 shares the allowance too: a file of filter bombs would hold an export
/// as long as one of JBIG2. What isn't decoded here, a JPEG, is still exported.
#[test]
fn flate_images_past_the_exports_allowance_are_left_out_with_the_reason() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let image = |filter: &str, raw: Vec<u8>| {
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(2));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
dict.set(b"Filter".to_vec(), Object::name(filter));
Stream::from_raw(dict, raw)
};
let flate = image("FlateDecode", pdfcraft_filters::encode_flate(&[0x80; 4]));
let mut cost = ImageBudget::with_work(u64::MAX);
cost.decode(&Document::new_empty(), &flate, 1 << 20).unwrap();
let cost = u64::MAX - cost.work_left();
let mut doc = Document::new_empty();
let (a, b) = (doc.add(Object::Stream(flate.clone())), doc.add(Object::Stream(flate)));
let mut budget = ImageBudget::with_work(cost);
assert!(crate::extract::image_file_within(&doc, a, &mut budget).is_ok());
let refused = crate::extract::image_file_within(&doc, b, &mut budget).unwrap_err();
assert!(refused.contains("more image decoding than"), "{refused}");
assert_eq!(budget.refused(), 1);
let jpeg = image::RgbImage::from_pixel(2, 2, image::Rgb([9, 9, 9]));
let mut bytes = Vec::new();
jpeg.write_to(&mut std::io::Cursor::new(&mut bytes), image::ImageFormat::Jpeg).unwrap();
let jpeg = doc.add(Object::Stream(image("DCTDecode", bytes)));
assert!(crate::extract::image_file_within(&doc, jpeg, &mut budget).is_ok(), "nothing charged");
assert_eq!(budget.refused(), 1);
}
/// A JPEG is written as it is, at no cost, unless something means decoding it, here a mask: that
/// pays for its samples, and past the allowance the image goes, with the reason, as it would come out without
/// its transparency otherwise.
#[test]
fn a_jpeg_that_has_to_be_decoded_shares_the_allowance() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let mut jpeg = Vec::new();
image::RgbImage::from_pixel(4, 4, image::Rgb([9, 9, 9])).write_to(&mut std::io::Cursor::new(&mut jpeg), image::ImageFormat::Jpeg).unwrap();
let image_dict = |extra: Option<(&[u8], Object)>| {
let mut d = Dict::new();
d.set(b"Subtype".to_vec(), Object::name("Image"));
d.set(b"Width".to_vec(), Object::Int(4));
d.set(b"Height".to_vec(), Object::Int(4));
d.set(b"BitsPerComponent".to_vec(), Object::Int(8));
d.set(b"ColorSpace".to_vec(), Object::name("DeviceRGB"));
d.set(b"Filter".to_vec(), Object::name("DCTDecode"));
if let Some((key, value)) = extra {
d.set(key.to_vec(), value);
}
d
};
let mut doc = Document::new_empty();
let mut mask = Dict::new();
mask.set(b"Subtype".to_vec(), Object::name("Image"));
mask.set(b"Width".to_vec(), Object::Int(4));
mask.set(b"Height".to_vec(), Object::Int(4));
mask.set(b"BitsPerComponent".to_vec(), Object::Int(8));
mask.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
let mask = doc.add(Object::Stream(Stream::from_raw(mask, vec![0x80; 16])));
let doc_jpeg_len = jpeg.len();
let masked = doc.add(Object::Stream(Stream::from_raw(image_dict(Some((b"SMask", Object::Ref(mask)))), jpeg.clone())));
let plain = doc.add(Object::Stream(Stream::from_raw(image_dict(None), jpeg)));
let mut budget = ImageBudget::with_work(0);
assert!(crate::extract::image_file_within(&doc, plain, &mut budget).is_ok(), "written as it is");
assert_eq!(budget.refused(), 0);
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(crate::extract::image_file_within(&doc, masked, &mut budget).is_ok());
let cost = u64::MAX - budget.work_left();
let read = 4 * doc_jpeg_len as u64;
assert_eq!(cost, 16 + 16 + 2 * 4 * 4 * 3 + read, "the mask's bytes and pixels, the JPEG's samples twice and its reading");
let mut budget = ImageBudget::with_work(cost - 1);
let refused = crate::extract::image_file_within(&doc, masked, &mut budget).unwrap_err();
assert!(refused.contains("more image decoding than"), "{refused}");
assert_eq!(budget.refused(), 1);
}
/// A JPEG that is itself a mask pays for its samples, each time an image reads it: images can
/// share one, and only the images are ever seen once.
#[test]
fn a_jpeg_mask_pays_for_its_samples() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let mut jpeg = Vec::new();
image::GrayImage::from_pixel(4, 4, image::Luma([90])).write_to(&mut std::io::Cursor::new(&mut jpeg), image::ImageFormat::Jpeg).unwrap();
let mut mask = Dict::new();
mask.set(b"Subtype".to_vec(), Object::name("Image"));
mask.set(b"Width".to_vec(), Object::Int(4));
mask.set(b"Height".to_vec(), Object::Int(4));
mask.set(b"BitsPerComponent".to_vec(), Object::Int(8));
mask.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
mask.set(b"Filter".to_vec(), Object::name("DCTDecode"));
let mut doc = Document::new_empty();
let read = 4 * jpeg.len() as u64;
let mask = doc.add(Object::Stream(Stream::from_raw(mask, jpeg)));
let mut image = Dict::new();
image.set(b"Subtype".to_vec(), Object::name("Image"));
image.set(b"Width".to_vec(), Object::Int(1));
image.set(b"Height".to_vec(), Object::Int(1));
image.set(b"BitsPerComponent".to_vec(), Object::Int(8));
image.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
image.set(b"SMask".to_vec(), Object::Ref(mask));
let image = doc.add(Object::Stream(Stream::from_raw(image, vec![0x80])));
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(crate::extract::image_file_within(&doc, image, &mut budget).is_ok());
assert_eq!(u64::MAX - budget.work_left(), 2 * 4 * 4 + read + 2, "its samples twice, its reading, and the image's byte and pixel");
// Just short of what the mask costs: it isn't decoded, and takes the image with it.
let mut budget = ImageBudget::with_work(2 * 4 * 4 + read - 1);
let refused = crate::extract::image_file_within(&doc, image, &mut budget).unwrap_err();
assert!(refused.contains("more image decoding than"), "{refused}");
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
}
/// What decoding a JPEG takes goes by its header, so that is what has to agree with the
/// dictionary before it is decoded: one image of a pixel can hold a page of 500 megapixels. The
/// file comes out as it is either way, so what shows is whether it was decoded.
#[test]
fn a_jpeg_larger_than_its_dictionary_says_isnt_decoded() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let mut jpeg = Vec::new();
image::RgbImage::from_pixel(4, 4, image::Rgb([9, 9, 9])).write_to(&mut std::io::Cursor::new(&mut jpeg), image::ImageFormat::Jpeg).unwrap();
// A colour key means decoding it, to find the pixels it hides.
let keyed = |side: i64| {
let mut d = Dict::new();
d.set(b"Subtype".to_vec(), Object::name("Image"));
d.set(b"Width".to_vec(), Object::Int(side));
d.set(b"Height".to_vec(), Object::Int(side));
d.set(b"BitsPerComponent".to_vec(), Object::Int(8));
d.set(b"ColorSpace".to_vec(), Object::name("DeviceRGB"));
d.set(b"Filter".to_vec(), Object::name("DCTDecode"));
d.set(b"Mask".to_vec(), Object::Array(vec![Object::Int(0); 6]));
Stream::from_raw(d, jpeg.clone())
};
let decodes = |doc: &Document, image: pdfcraft_cos::ObjRef| {
crate::extract::JPEGS_DECODED.with(|n| n.set(0));
assert!(crate::extract::image_file_within(doc, image, &mut ImageBudget::with_work(u64::MAX)).is_ok());
crate::extract::JPEGS_DECODED.with(std::cell::Cell::get)
};
let mut doc = Document::new_empty();
let (honest, lying) = (doc.add(Object::Stream(keyed(4))), doc.add(Object::Stream(keyed(1))));
assert_eq!(decodes(&doc, honest), 1);
assert_eq!(decodes(&doc, lying), 0);
// The same as a mask: read as none, without being decoded.
let mut mask = keyed(1);
mask.dict.remove(b"Mask");
let mask = doc.add(Object::Stream(mask));
let mut image = Dict::new();
image.set(b"Subtype".to_vec(), Object::name("Image"));
image.set(b"Width".to_vec(), Object::Int(1));
image.set(b"Height".to_vec(), Object::Int(1));
image.set(b"BitsPerComponent".to_vec(), Object::Int(8));
image.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
image.set(b"SMask".to_vec(), Object::Ref(mask));
let image = doc.add(Object::Stream(Stream::from_raw(image, vec![0x80])));
assert_eq!(decodes(&doc, image), 0);
// Reading the header to find that out is paid for: a mask can be mostly tables, shared by
// every image of a file, and with nothing left it isn't read.
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(crate::extract::image_file_within(&doc, image, &mut budget).is_ok());
assert_eq!(u64::MAX - budget.work_left(), 2 + 4 * jpeg.len() as u64 + 2, "its one declared sample twice, its bytes, the image's byte and pixel");
let mut budget = ImageBudget::with_work(0);
assert!(crate::extract::image_file_within(&doc, image, &mut budget).is_err());
assert_eq!(budget.refused(), 1, "the image goes, with its mask unread");
}
/// A mask past the allowance can't be left out as an unreadable one is, or the image comes out
/// opaque, showing what the mask hides: the image goes, with the reason.
#[test]
fn an_image_whose_jbig2_mask_is_past_the_allowance_is_left_out() {
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
let mut mask = Dict::new();
mask.set(b"Subtype".to_vec(), Object::name("Image"));
mask.set(b"Width".to_vec(), Object::Int(16));
mask.set(b"Height".to_vec(), Object::Int(2));
mask.set(b"BitsPerComponent".to_vec(), Object::Int(1));
mask.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
mask.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
let mut doc = Document::new_empty();
let mask = doc.add(Object::Stream(Stream::from_raw(mask, pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap())));
// A JPEG, read for nothing until its mask is, so only the mask can be what stops it.
let mut jpeg = Vec::new();
image::RgbImage::from_pixel(1, 1, image::Rgb([9, 9, 9])).write_to(&mut std::io::Cursor::new(&mut jpeg), image::ImageFormat::Jpeg).unwrap();
let mut image = Dict::new();
image.set(b"Subtype".to_vec(), Object::name("Image"));
image.set(b"Width".to_vec(), Object::Int(1));
image.set(b"Height".to_vec(), Object::Int(1));
image.set(b"BitsPerComponent".to_vec(), Object::Int(8));
image.set(b"ColorSpace".to_vec(), Object::name("DeviceRGB"));
image.set(b"Filter".to_vec(), Object::name("DCTDecode"));
image.set(b"SMask".to_vec(), Object::Ref(mask));
let image = doc.add(Object::Stream(Stream::from_raw(image, jpeg)));
assert!(crate::extract::image_file_within(&doc, image, &mut ImageBudget::with_work(u64::MAX)).is_ok());
let refused = crate::extract::image_file_within(&doc, image, &mut ImageBudget::with_work(0)).unwrap_err();
assert!(refused.contains("more image decoding than"), "{refused}");
}
/// A large honest file keeps its images: what an export may hold grows with the file past the fixed
/// part, and a file too long to double doesn't wrap round to nothing.
#[test]
fn an_export_may_hold_twice_a_large_file() {
let fixed = crate::extract::export_bytes(0);
assert_eq!(fixed, 2 << 30);
assert_eq!(crate::extract::export_bytes(fixed / 2), fixed);
assert_eq!(crate::extract::export_bytes(fixed / 2 + 1), fixed + 2);
assert_eq!(crate::extract::export_bytes(usize::MAX), usize::MAX);
}
/// `/Decode` says which way round a stencil mask is, and its numbers can be given by reference. Read
/// as absent, they gave the wrong answer, so the file came out a negative of what the page draws.
#[test]
crates/edit/src/background.rs
+7 -2
View File
@@ -98,8 +98,13 @@ pub fn page_background(doc: &Document, page: usize) -> Result<Option<Rgb>, EditE
}
b"sc" | b"scn" => st.fill = colour(st.space, op),
b"gs" => {
if let Some(g) = op.name(0).and_then(|n| gstates.get(n)).map(|g| doc.resolve(g)).and_then(|g| g.as_dict().cloned()) {
apply_gstate(doc, &g, &mut st);
let mut held = None;
let gstate = match op.name(0).and_then(|n| gstates.get(n)) {
Some(g) => crate::text::dict_in_place(doc, g, &mut held),
None => None,
};
if let Some(g) = gstate {
apply_gstate(doc, g, &mut st);
}
}
b"re" => match op.nums::<4>() {
crates/edit/src/images.rs
+10 -10
View File
@@ -379,34 +379,34 @@ pub fn reading_images(doc: &Document, page: usize) -> Result<Vec<PageImage>, Edi
forms: &mut Forms,
out: &mut Vec<PageImage>,
) -> Result<(), EditError> {
let xo = resources.get(b"XObject").map(|x| doc.resolve(x)).and_then(|x| x.as_dict().cloned()).unwrap_or_default();
let mut held = None;
let xo = crate::text::resource_dict(doc, resources, b"XObject", &mut held);
for (i, op) in ops.iter().enumerate() {
st.apply(op)?;
let ctm = st.placement.matrix;
if op.is("Do") {
let Some(name) = op.name(0) else { continue };
let drawn = xo.and_then(|xo| xo.get(name));
if let Some(form) = crate::text::form_call(doc, resources, name, ctm, &forms.path, &mut forms.budget) {
if let Some(r) = form.obj {
forms.path.push(r);
}
let bounds =
xo.get(name).map(|o| doc.resolve(o)).and_then(|o| o.as_dict().and_then(|d| d.get(b"BBox")).map(|b| doc.resolve(b))).and_then(
|b| {
let a = b.as_array()?;
let [x0, y0, x1, y1] = a.as_slice() else { return None };
Some(form.ctm.bbox([x0.as_f64()?, y0.as_f64()?, x1.as_f64()?, y1.as_f64()?]))
},
);
drawn.map(|o| doc.resolve(o)).and_then(|o| o.as_dict().and_then(|d| d.get(b"BBox")).map(|b| doc.resolve(b))).and_then(|b| {
let a = b.as_array()?;
let [x0, y0, x1, y1] = a.as_slice() else { return None };
Some(form.ctm.bbox([x0.as_f64()?, y0.as_f64()?, x1.as_f64()?, y1.as_f64()?]))
});
let clip = bounds.map_or(st.placement.clip, |b| intersection(st.placement.clip, b).unwrap_or([0.0; 4]));
let mut inner = ImageState::new(form.ctm, clip, st.placement.clips.clone());
let piece = stream(op);
walk(doc, &parse(&form.data).ops, &form.resources, &mut inner, &|_| piece, forms, out)?;
walk(doc, &parse(&form.data).ops, form.resources(resources), &mut inner, &|_| piece, forms, out)?;
if form.obj.is_some() {
forms.path.pop();
}
continue;
}
let Some(r) = xo.get(name).and_then(Object::as_ref) else { continue };
let Some(r) = drawn.and_then(Object::as_ref) else { continue };
let obj = doc.get(r);
let Object::Stream(s) = &*obj else { continue };
if s.dict.name(b"Subtype") != Some(b"Image") {
crates/edit/src/tests.rs
+35
View File
@@ -5575,6 +5575,41 @@ fn reading_follows_form_xobjects() {
assert!(own[0].is_form && own[0].object == Some(pdfcraft_cos::ObjRef::new(6, 0)), "{own:?}");
}
/// A page's `/XObject` and a form's `/Resources` are read wherever they are stored, and a form with
/// none of its own, or with something else there, uses the ones it was drawn with (PDF 1.1 files).
/// Each of these pages shows one line of text and one picture through one form.
#[test]
fn reading_follows_forms_however_their_resources_are_stored() {
let body = "BT /F1 12 Tf 72 700 Td (In the form) Tj ET q 20 0 0 10 5 5 cm /Im Do Q";
let (font, direct) = ("/Font << /F1 5 0 R >>", "/XObject << /Fm 6 0 R /Im 7 0 R >>");
for (page, form) in [
(format!("<< {font} {direct} >>"), ""),
(format!("<< {font} /XObject 9 0 R >>"), ""),
(format!("<< /Font 10 0 R {direct} >>"), ""),
(format!("<< {font} {direct} >>"), "/Resources 99 0 R"),
(format!("<< {font} {direct} >>"), "/Resources 12"),
("<< /XObject << /Fm 6 0 R >> >>".to_string(), "/Resources 8 0 R"),
("<< /XObject 9 0 R >>".to_string(), "/Resources << /Font << /F1 5 0 R >> /XObject 9 0 R >>"),
] {
let doc = build(&[
"<< /Type /Catalog /Pages 2 0 R >>",
"<< /Type /Pages /Kids [3 0 R] /Count 1 >>",
&format!("<< /Type /Page /Parent 2 0 R /MediaBox [0 0 600 800] /Resources {page} /Contents 4 0 R >>"),
&stream("", "/Fm Do"),
"<< /Type /Font /Subtype /Type1 /BaseFont /Courier >>",
&stream(&format!("/Type /XObject /Subtype /Form /BBox [0 0 600 800] {form}"), body),
&stream("/Type /XObject /Subtype /Image /Width 1 /Height 1 /ColorSpace /DeviceGray /BitsPerComponent 8", "A"),
"<< /Font << /F1 5 0 R >> /XObject 9 0 R >>",
"<< /Fm 6 0 R /Im 7 0 R >>",
"<< /F1 5 0 R >>",
]);
let blocks = text::reading_blocks(&doc, 0).unwrap();
let texts: Vec<(&str, &str)> = blocks.iter().map(|b| (b.text.as_str(), b.base_font.as_str())).collect();
assert_eq!(texts, [("In the form", "Courier")], "{page} {form}");
assert_eq!(images::reading_images(&doc, 0).unwrap().len(), 1, "{page} {form}");
}
}
/// A chain of forms, each drawing the next twenty times, twelve deep: within the depth cap but
/// 20^11 visits without a budget. Reading stops at the page's visit budget instead of hanging.
#[test]
crates/edit/src/text.rs
+1 -1
View File
@@ -54,7 +54,7 @@ use fallback::{
use fallback::{add_type3_font, is_type3_fallback};
use interp::{Origin, Paint, union};
use lines::page_dict;
pub(crate) use lines::{FormBudget, content_streams, content_streams_within, form_call, rewritten_contents, splice};
pub(crate) use lines::{FormBudget, content_streams, content_streams_within, dict_in_place, form_call, resource_dict, rewritten_contents, splice};
pub use lines::{MAX_FORM_DEPTH, MAX_FORM_VISITS, glyph_text_lines, text_lines};
pub use rewrite::{LineEdit, replace_line, rewrite_block};
use rewrite::{coincident_ops, rewrite_block_from};
crates/edit/src/text/interp.rs
+9 -4
View File
@@ -773,8 +773,9 @@ pub(super) fn interpret(
// `Some(properties)`: read `/ActualText`, looking named property lists up in `properties`.
logical_text: Option<&Dict>,
) -> Result<Vec<Shown>, EditError> {
let fonts_res = resources.get(b"Font").map(|f| doc.resolve(f)).and_then(|f| f.as_dict().cloned()).unwrap_or_default();
let states = resources.get(b"ExtGState").map(|g| doc.resolve(g)).and_then(|g| g.as_dict().cloned()).unwrap_or_default();
let (mut held_fonts, mut held_states, none) = (None, None, Dict::new());
let fonts_res = super::resource_dict(doc, resources, b"Font", &mut held_fonts).unwrap_or(&none);
let states = super::resource_dict(doc, resources, b"ExtGState", &mut held_states).unwrap_or(&none);
let mut out: Vec<Shown> = Vec::new();
let mut ts = carry.ts.clone();
let mut at_bt = (0usize, ts.clone());
@@ -881,8 +882,12 @@ pub(super) fn interpret(
if let Some(name) = op.name(0) {
select_gs(&mut ts.ext_gstate, name);
}
let params = op.name(0).and_then(|name| states.get(name)).map(|p| doc.resolve(p));
if let Some(params) = params.as_deref().and_then(Object::as_dict) {
let mut held = None;
let params = match op.name(0).and_then(|name| states.get(name)) {
Some(p) => super::dict_in_place(doc, p, &mut held),
None => None,
};
if let Some(params) = params {
ts.gs.apply(doc, params);
// `/LW` sets the line width as `w` does (ISO 32000-2 §8.4.5).
let lw = params.get(b"LW").and_then(|w| doc.resolve(w).as_f64()).filter(|w| w.is_finite());
crates/edit/src/text/lines.rs
+46 -11
View File
@@ -1,3 +1,5 @@
use std::sync::Arc;
use super::fallback::{same_font, same_paint};
use super::interp::{Carry, Origin, Pen, Shown, TextState, Ts, direction, interpret, to_frame, union};
use super::{DecodeBudget, Dict, Document, EditError, HashMap, Matrix, Object, Op, Pieces, Stream, TextLine, parse, serialize_ops};
@@ -452,10 +454,41 @@ impl FormBudget {
pub(crate) struct FormCall {
pub(crate) obj: Option<pdfcraft_cos::ObjRef>,
pub(crate) data: Vec<u8>,
pub(crate) resources: Dict,
/// The form, and its `/Resources` when they are an object of their own, which is what
/// [`FormCall::resources`] reads.
form: Arc<Object>,
own: Option<Arc<Object>>,
pub(crate) ctm: Matrix,
}
impl FormCall {
/// The resources its content is read with: its own, or if it has none, `drawn_with`, the ones
/// it was drawn with (PDF 1.1 files). Read in place, as a page can draw a form thousands of times:
/// a copy for each took time in the square of them.
pub(crate) fn resources<'a>(&'a self, drawn_with: &'a Dict) -> &'a Dict {
let direct = || match &*self.form {
Object::Stream(s) => s.dict.get(b"Resources").and_then(Object::as_dict),
_ => None,
};
self.own.as_deref().and_then(Object::as_dict).or_else(direct).unwrap_or(drawn_with)
}
}
/// `o` as a dictionary, read in place for the same reason as [`FormCall::resources`]. `held` keeps it
/// when it is an object of its own. Nothing else is one, and [`Document::resolve`] would copy it.
pub(crate) fn dict_in_place<'a>(doc: &Document, o: &'a Object, held: &'a mut Option<Arc<Object>>) -> Option<&'a Dict> {
match o {
Object::Dict(dict) => Some(dict),
Object::Ref(_) => held.insert(doc.resolve(o)).as_dict(),
_ => None,
}
}
/// The dictionary at `key` of `resources`, `/XObject` or `/Font`, say, as each form reads these.
pub(crate) fn resource_dict<'a>(doc: &Document, resources: &'a Dict, key: &[u8], held: &'a mut Option<Arc<Object>>) -> Option<&'a Dict> {
dict_in_place(doc, resources.get(key)?, held)
}
/// The form XObject `name` in `resources`, drawn with `ctm`. `None` for images, missing names
/// and forms already being read (`path`), and for forms nested too deep or past the page's
/// budget, which `budget` records as truncated; each form entered counts one visit.
@@ -472,13 +505,15 @@ pub(crate) fn form_call(
if limited && budget.truncated {
return None;
}
let xobjects = resources.get(b"XObject").map(|x| doc.resolve(x)).and_then(|x| x.as_dict().cloned())?;
let o = xobjects.get(name)?;
let obj = o.as_ref();
let mut held = None;
let o = resource_dict(doc, resources, b"XObject", &mut held)?.get(name)?;
// Only a reference can name a stream, so nothing else needs reading, or copying.
let obj = Some(o.as_ref()?);
if obj.is_some_and(|r| path.contains(&r)) {
return None;
}
let Object::Stream(s) = &*doc.resolve(o) else { return None };
let form = doc.resolve(o);
let Object::Stream(s) = &*form else { return None };
if s.dict.name(b"Subtype") != Some(b"Form") {
return None;
}
@@ -492,8 +527,7 @@ pub(crate) fn form_call(
let v: Vec<f64> = a.iter().filter_map(|x| doc.resolve(x).as_f64()).collect();
<[f64; 6]>::try_from(v).ok().filter(|v| v.iter().all(|x| x.is_finite()))
});
// A form without its own resources uses the ones it was drawn with (PDF 1.1 files).
let own = s.dict.get(b"Resources").map(|r| doc.resolve(r)).and_then(|r| r.as_dict().cloned());
let own = s.dict.get(b"Resources").filter(|r| matches!(r, Object::Ref(_))).map(|r| doc.resolve(r));
let data = match budget.decode.decode(s) {
Ok(data) => data,
Err(_) if budget.decode.remaining() == 0 => {
@@ -503,7 +537,7 @@ pub(crate) fn form_call(
Err(_) => Vec::new(),
};
budget.visits = budget.visits.saturating_add(1);
Some(FormCall { obj, data, resources: own.unwrap_or_else(|| resources.clone()), ctm: m.map_or(ctm, |m| Matrix(m).then(&ctm)) })
Some(FormCall { obj, data, form, own, ctm: m.map_or(ctm, |m| Matrix(m).then(&ctm)) })
}
/// The lines a page shows as it reads, including text drawn by form XObjects (which
@@ -522,8 +556,9 @@ pub(super) fn reading_lines(doc: &Document, page: usize) -> Result<(Vec<TextLine
fn walk(w: &mut Walk, ops: &[Op], resources: &Dict, carry: &mut Carry, stream: &dyn Fn(usize) -> usize) -> Result<(), EditError> {
let mut cache = HashMap::new();
let mut invoked = Vec::new();
let properties = resources.get(b"Properties").and_then(|o| w.doc.dict(o)).unwrap_or_default();
let shown = interpret(w.doc, ops, resources, &mut cache, carry, Some(&mut invoked), Some(&properties))?;
let (mut held, none) = (None, Dict::new());
let properties = resource_dict(w.doc, resources, b"Properties", &mut held).unwrap_or(&none);
let shown = interpret(w.doc, ops, resources, &mut cache, carry, Some(&mut invoked), Some(properties))?;
let mut last = None;
let mut last_stream = None;
let mut forms = invoked.into_iter().peekable();
@@ -556,7 +591,7 @@ pub(super) fn reading_lines(doc: &Document, page: usize) -> Result<(Vec<TextLine
if let Some(r) = form.obj {
w.path.push(r);
}
walk(w, &ops, &form.resources, &mut carry, &|_| stream)?;
walk(w, &ops, form.resources(resources), &mut carry, &|_| stream)?;
*text_left = carry.text_left;
if form.obj.is_some() {
w.path.pop();
crates/engine/src/actions.rs
+20 -1
View File
@@ -146,6 +146,11 @@ pub struct FileResult {
pub log: Vec<String>,
}
/// What an action's log says of the images a Reduce File Size step left, if there are any.
fn images_refused_note(images: usize) -> Option<String> {
(images > 0).then(|| format!("Reduce file size: {images} image(s) stayed as they were, past what one run may decode"))
}
fn apply_step(s: &mut Session, id: DocId, step: &Step, log: &mut Vec<String>) -> Result<Option<DocId>, EditError> {
let pages = s.get(id).map_or(0, |d| d.info.pages.len());
let all: Vec<usize> = (0..pages).collect();
@@ -156,9 +161,10 @@ fn apply_step(s: &mut Session, id: DocId, step: &Step, log: &mut Vec<String>) ->
return Ok(None);
}
Step::ReduceFileSize => {
let (bytes, _) = s.optimized_bytes(id, &crate::optimize::Settings::default(), &[])?;
let (bytes, report) = s.optimized_bytes(id, &crate::optimize::Settings::default(), &[])?;
let before = s.get(id).map_or(0, |d| d.bytes.len());
log.push(format!("Reduce file size: {} → {} bytes", before, bytes.len()));
log.extend(images_refused_note(report.optimize.images_refused));
let name = s.get(id).map(|d| d.name.clone()).unwrap_or_default();
s.close(id);
let new = s.open(name, None, bytes, None).map_err(|e| EditError::Reopen(e.to_string()))?;
@@ -222,3 +228,16 @@ pub fn run_on(action: &Action, name: &str, bytes: Arc<Vec<u8>>, mut progress: im
let bytes = if d.is_signed() { s.save_bytes(id) } else { s.save_full_bytes(id) }.map_err(|e| e.to_string())?;
Ok(FileResult { name: name.to_string(), bytes, log })
}
#[cfg(test)]
mod tests {
use super::images_refused_note;
/// A log line only for images that stayed, saying how many.
#[test]
fn the_log_says_when_images_stayed_as_they_were() {
assert_eq!(images_refused_note(0), None);
assert!(images_refused_note(1).is_some_and(|note| note.contains("1 image(s)")));
assert!(images_refused_note(3).is_some_and(|note| note.contains("3 image(s)")));
}
}
crates/engine/src/compare.rs
+19 -4
View File
@@ -173,14 +173,15 @@ pub struct OfficeExport {
/// The exported file.
pub bytes: Vec<u8>,
/// Pages (0-based) that draw forms nested deeper than [`MAX_FORM_DEPTH`] or more of them
/// than [`MAX_FORM_VISITS`], or pictures past [`MAX_EXPORT_IMAGE_BYTES`]: export stops
/// reading there to stay bounded on hostile files, so the text and images of those forms, or
/// than [`MAX_FORM_VISITS`], or pictures past [`MAX_EXPORT_IMAGE_BYTES`] or past what one export
/// may decode: export stops reading there to stay bounded on hostile files, so the text and images of those forms, or
/// those pictures, are missing from `bytes`.
pub truncated_pages: Vec<usize>,
}
/// An image object prepared for Office export: its file extension and encoded bytes, or why it is
/// left out (`Err(true)`: past [`MAX_EXPORT_IMAGE_BYTES`]; `Err(false)`: it can't be encoded).
/// left out (`Err(true)`: past [`MAX_EXPORT_IMAGE_BYTES`], or what the export may decode;
/// `Err(false)`: it can't be encoded).
type PreparedImage = Result<(&'static str, Arc<[u8]>), bool>;
impl crate::Document {
@@ -194,6 +195,18 @@ impl crate::Document {
/// pictures went past `image_budget` bytes ([`OfficeExport::truncated_pages`]). Without
/// `images` no picture is prepared.
pub(crate) fn export_pages_checked(&self, images: bool, image_budget: usize) -> (Vec<pdfcraft_export::Page>, Vec<usize>) {
let Some(editor) = self.editor.as_ref() else { return (Vec::new(), Vec::new()) };
self.export_pages_within(images, image_budget, &mut pdfcraft_cos::ImageBudget::new(&editor.cos))
}
/// [`Self::export_pages_checked`], its pictures decoded within `image_work`, which is shared
/// by the whole export: a picture past it is left out like one past `image_budget`.
pub(crate) fn export_pages_within(
&self,
images: bool,
image_budget: usize,
image_work: &mut pdfcraft_cos::ImageBudget,
) -> (Vec<pdfcraft_export::Page>, Vec<usize>) {
let Some(cos) = self.editor.as_ref().map(|e| &e.cos) else { return (Vec::new(), Vec::new()) };
let mut truncated = Vec::new();
// Each image object is encoded once and shared by every place that draws it.
@@ -244,7 +257,9 @@ impl crate::Document {
.filter_map(|im| {
let object = im.object?;
let prepared = encoded.entry(object).or_insert_with(|| {
let (ext, bytes) = pdfcraft_create::image_file(cos, object).map_err(|_| false)?;
let refused = image_work.refused();
let (ext, bytes) =
pdfcraft_create::image_file_within(cos, object, &mut *image_work).map_err(|_| image_work.refused() > refused)?;
image_bytes = image_bytes.checked_sub(bytes.len()).ok_or(true)?;
Ok((if ext == "jpg" { "jpg" } else { "png" }, Arc::from(bytes)))
});
crates/engine/src/session/files.rs
+5 -5
View File
@@ -71,12 +71,12 @@ impl Session {
/// Reduce File Size: Acrobat's defaults (colour and grey images above 225 ppi to 150 ppi,
/// JPEG medium quality; black-and-white averaged down to 300 ppi above 450 ppi, and stored as
/// CCITT Group 4 whenever smaller; thumbnails dropped), identical resources merged, unused
/// objects dropped, objects packed into compressed object streams. Returns the bytes and how
/// many objects were merged. The open document is not changed (Acrobat saves the reduced copy
/// as a new file).
pub fn reduced_bytes(&self, id: DocId) -> Result<(Arc<Vec<u8>>, usize), EditError> {
/// objects dropped, objects packed into compressed object streams. Returns the bytes, how
/// many objects were merged and how many images stayed as they were, past what one run may
/// decode. The open document is not changed (Acrobat saves the reduced copy as a new file).
pub fn reduced_bytes(&self, id: DocId) -> Result<(Arc<Vec<u8>>, usize, usize), EditError> {
let (bytes, report) = self.optimized_bytes(id, &optimize::Settings::default(), &[])?;
Ok((bytes, report.merged))
Ok((bytes, report.merged, report.optimize.images_refused))
}
/// Optimize PDF ▸ Advanced optimization: `settings` for images and objects, plus Remove
crates/engine/src/tests.rs
+82 -2
View File
@@ -3603,7 +3603,7 @@ fn create_and_reduce() {
)
.unwrap();
let full = s2.save_full_bytes(id2).unwrap();
let (reduced, _) = s2.reduced_bytes(id2).unwrap();
let (reduced, ..) = s2.reduced_bytes(id2).unwrap();
assert!(reduced.len() <= full.len(), "{} > {}", reduced.len(), full.len());
let mut s3 = Session::new();
let r = s3.open("r.pdf", None, reduced, None).unwrap();
@@ -8244,10 +8244,35 @@ fn copy_image_reads_the_globals_of_a_jbig2_picture() {
let (w, h, rgba) = d.page_image_rgba(0, 0).unwrap();
assert_eq!((w, h), (16, 2));
assert!(rgba.chunks(4).all(|p| p == [0, 0, 0, 255]), "the page's default pixel, from the globals");
// An Office export shares one allowance across its pictures. One past it is left out, and the
// export says the page is missing something, as for a picture past its bytes.
let (pages, truncated) = d.export_pages_within(true, 1 << 20, &mut pdfcraft_cos::ImageBudget::with_work(u64::MAX));
assert_eq!((pages[0].images.len(), truncated.len()), (1, 0));
let (pages, truncated) = d.export_pages_within(true, 1 << 20, &mut pdfcraft_cos::ImageBudget::with_work(0));
assert!(pages[0].images.is_empty());
assert_eq!(truncated, [0]);
}
/// An Office export leaves out a picture it can't encode, as it did. That isn't a page cut short:
/// only a picture past what the export may do is reported as one.
#[test]
fn an_office_export_doesnt_call_an_unreadable_picture_truncated() {
let pdf = b"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\
2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 /MediaBox [0 0 200 300] >> endobj\n\
3 0 obj << /Type /Page /Parent 2 0 R /Contents 4 0 R /Resources << /XObject << /Im1 5 0 R >> >> >> endobj\n\
4 0 obj << /Length 31 >> stream\nq 160 0 0 20 20 100 cm /Im1 Do Q\nendstream endobj\n\
5 0 obj << /Type /XObject /Subtype /Image /Width 2 /Height 2 /ColorSpace /DeviceGray /BitsPerComponent 8 /Filter /JPXDecode /Length 4 >> stream\nnone\nendstream endobj\n\
trailer << /Root 1 0 R >>\n%%EOF"
.to_vec();
let mut s = Session::new();
let id = s.open("picture.pdf", None, Arc::new(pdf), None).unwrap();
let (pages, truncated) = s.get(id).unwrap().export_pages_within(true, 1 << 20, &mut pdfcraft_cos::ImageBudget::with_work(u64::MAX));
assert!(pages[0].images.is_empty());
assert_eq!(truncated, Vec::<usize>::new());
}
/// Copy Image takes a JBIG2 picture as its pixels and as a PNG file, behind another filter
/// (ASCIIHexDecode) too: Export's decoding, through `Document::decoded_image`.
/// (ASCIIHexDecode) too: Export's decoding, through `ImageBudget::decode`.
#[test]
fn copy_image_copies_a_jbig2_picture_behind_another_filter() {
// 16 × 2: the left half of the top row black, the rest white.
@@ -9215,6 +9240,61 @@ fn a_fields_own_script_may_not_save_as_or_show_a_panel() {
assert_eq!(o.requests, [js::Request::SaveAs]);
}
/// A page of `count` pictures of one flat grey, `side` pixels square, each an object of its own.
fn page_of_pictures(count: usize, side: usize) -> Vec<u8> {
let data = pdfcraft_filters::encode_flate(&vec![0x80; side * side]);
let names: String = (0..count).map(|i| format!("/Im{i} {} 0 R ", 5 + i)).collect();
let content: String = (0..count).map(|i| format!("q 10 0 0 10 0 0 cm /Im{i} Do Q\n")).collect();
let mut pdf = format!(
"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\
2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 /MediaBox [0 0 200 300] >> endobj\n\
3 0 obj << /Type /Page /Parent 2 0 R /Contents 4 0 R /Resources << /XObject << {names}>> >> >> endobj\n\
4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n",
content.len()
)
.into_bytes();
for i in 0..count {
let dict = format!(
"{} 0 obj << /Type /XObject /Subtype /Image /Width {side} /Height {side} /ColorSpace /DeviceGray /BitsPerComponent 8 /Filter /FlateDecode /Length {} >> stream\n",
5 + i,
data.len()
);
pdf.extend_from_slice(dict.as_bytes());
pdf.extend_from_slice(&data);
pdf.extend_from_slice(b"\nendstream endobj\n");
}
pdf.extend_from_slice(b"trailer << /Root 1 0 R >>\n%%EOF");
pdf
}
/// The allowance is the export's, not each picture's: a page of many small pictures stops where it
/// runs out, the rest are left out and counted, and the page is listed, which is what the export's
/// warning goes by. A few large pictures are well within it, and all come through.
#[test]
fn a_page_of_many_pictures_stops_at_the_exports_allowance() {
let export = |pdf: Vec<u8>, work: &mut pdfcraft_cos::ImageBudget| {
let mut s = Session::new();
let id = s.open("pictures.pdf", None, Arc::new(pdf), None).unwrap();
s.get(id).unwrap().export_pages_within(true, 1 << 30, work)
};
let mut one = pdfcraft_cos::ImageBudget::with_work(u64::MAX);
export(page_of_pictures(1, 2), &mut one);
let cost = u64::MAX - one.work_left();
assert!(cost > 0);
let mut work = pdfcraft_cos::ImageBudget::with_work(100 * cost);
let (pages, truncated) = export(page_of_pictures(10_000, 2), &mut work);
assert_eq!(pages[0].images.len(), 100);
assert_eq!(truncated, vec![0]);
assert_eq!(work.refused(), 9_900);
let mut s = Session::new();
let id = s.open("large.pdf", None, Arc::new(page_of_pictures(3, 1_000)), None).unwrap();
let (pages, truncated) = s.get(id).unwrap().export_pages_checked(true, 1 << 30);
assert_eq!(pages[0].images.len(), 3);
assert_eq!(truncated, Vec::<usize>::new());
}
/// A page whose forms go past the reading limit and whose picture goes past the image budget is
/// listed once among the truncated pages.
#[test]
crates/filters/README.md
+4 -1
View File
@@ -15,12 +15,14 @@ Filter::from_name(b"FlateDecode" | b"Fl" | ...) -> Filter // full names and in
Filter::name(&self) -> &str; Filter::is_image_codec(&self) -> bool
pub struct Params { predictor, colors, bits_per_component, columns, early_change } // i64s; Default = 1, 1, 8, 1, 1
pub enum FilterError { Corrupt { filter, detail }, LimitExceeded(usize), Unsupported(String) }
pub enum FilterError { Corrupt { filter, detail }, LimitExceeded(usize), Unsupported(String), Spent }
decode(chain, data, max_output) -> Result<Vec<u8>, FilterError>
decode_counted(chain, data, max_output) -> Result<(Vec<u8>, usize /*decoded*/), (FilterError, usize /*decoded*/)>
decode_tolerant(chain, data, max_output) -> Result<(Vec<u8>, bool /*partial*/), FilterError>
decode_tolerant_counted(chain, data, max_output) -> Result<(Vec<u8>, bool /*partial*/, usize /*decoded*/), (FilterError, usize /*decoded*/)>
decode_tolerant_within(chain, data, max_output, work: &mut u64) -> Result<Vec<u8>, FilterError>
decode_within(chain, data, max_output, work: &mut u64) -> Result<Vec<u8>, FilterError>
encode(filter, params, data) -> Result<Vec<u8>, FilterError>
encode_flate(data) -> Vec<u8> // zlib, level 6
encode_g4(data, columns, rows) -> Result<Vec<u8>, FilterError> // 1-bit rows as CCITT Group 4; columns <= MAX_G4_COLUMNS
@@ -35,6 +37,7 @@ encode_jbig2(data, columns, rows) -> Result<Vec<u8>, FilterError> // 1-bit row
- Image codecs (DCT, JPX, JBIG2, CCITT), `Crypt` and unknown filters anywhere in a chain give `Unsupported`: callers keep such data encoded. So does a chain longer than `MAX_CHAIN` (8) filters, before any stage runs. Decryption belongs to `crypt`.
- `max_output` bounds every intermediate buffer (including the pre-predictor buffer) and gives `LimitExceeded(max_output)`. This also holds when Flate retries raw deflate after a misleading zlib header, in both strict and tolerant decoding.
- `decode_counted` and `decode_tolerant_counted` are for a caller charging a budget (`pdfcraft_cos::DecodeBudget`). `max_output` bounds all stages together (each stage gets what the earlier ones left), and they give the bytes every stage that ran decoded, on success and on failure. A stage's output counts as its decoder produced it, before a predictor or a later stage shrinks it, and a failed decode pays for what it decoded first.
- `decode_tolerant_within` is for a run of images (`pdfcraft_cos::ImageBudget`), as `decode_within` (its strict form), `decode_jbig2_within` and `decode_ccitt_within` are, all spending from one `work`, in the JBIG2 decoder's steps. `max_output` bounds each stage, as in `decode_tolerant`, and no stage decodes past what `work` has left. It pays a step a byte of `data` first and a step a byte each stage decodes, so a chain that fails has paid for what it did: 0.2 to 3 ns a byte in or out, up to 19 ns through a predictor on samples of a bit, against the 6 to 16 ns of a JBIG2 step. What `work` can't pay for spends all of it, and gives `Spent`; with nothing left nothing is read. A chain that can't be decoded at all costs nothing and gives its usual error.
- `decode_tolerant` turns `Corrupt` into "what was decoded so far" plus `partial = true`, then runs the rest of the chain on it (predictors are applied to partial output too). `LimitExceeded` and `Unsupported` are still errors.
## Behaviour
crates/filters/src/ccitt.rs
+73 -18
View File
@@ -50,7 +50,7 @@ pub fn encode_g4(data: &[u8], columns: u32, rows: u32) -> Result<Vec<u8>, Filter
}
}
/// The widest row [`decode_ccitt`] decodes: its row buffer stays within 128 KiB.
/// The widest row [`decode_ccitt_within`] decodes: its row buffer stays within 128 KiB.
pub const MAX_CCITT_COLUMNS: u32 = 1 << 20;
/// A `CCITTFaxDecode` stream's `/DecodeParms` (ISO 32000-2 Table 11). [`Default`] gives the
@@ -74,22 +74,8 @@ impl Default for CcittParams {
}
}
/// CCITT fax `data` decoded as `CCITTFaxDecode` gives it: rows of `columns` 1-bit samples, most
/// significant bit first, each starting on a new byte, with black as 0 (as 1 with `black_is_1`).
///
/// At most `max_rows` rows are decoded (fewer when `rows` says so), so the output is bounded by
/// the caller. Damaged data gives the rows decoded before the damage, as viewers show them: a
/// caller that needs a number of rows checks it got them.
///
/// Group 3 rows on bytes (`k` from 0, `encoded_byte_align`) are read by their end-of-line codes,
/// whether the fill starts each code on a byte or ends it on one, when `data` starts with one (T.4
/// sends one before the first row when it sends them at all), and on bytes otherwise, whatever
/// `end_of_line` says. Group 4 is read as its parameters say. Data with codes but none before the
/// first row is read on bytes, as PdfCraft's renderer draws it; with `k` 0 and the fill ending each
/// code on a byte, that may be other rows than were coded.
///
/// Errors: `columns` that aren't from 1 to [`MAX_CCITT_COLUMNS`] ([`FilterError::Unsupported`]).
pub fn decode_ccitt(data: &[u8], p: &CcittParams, max_rows: u32) -> Result<Vec<u8>, FilterError> {
/// What [`decode_ccitt_within`] decodes once it is paid for.
fn decode_ccitt(data: &[u8], p: &CcittParams, max_rows: u32) -> Result<Vec<u8>, FilterError> {
let columns = u32::try_from(p.columns)
.ok()
.filter(|c| (1..=MAX_CCITT_COLUMNS).contains(c))
@@ -123,6 +109,39 @@ pub fn decode_ccitt(data: &[u8], p: &CcittParams, max_rows: u32) -> Result<Vec<u
Ok(out.out)
}
/// The steps of reading a byte of fax code ([`decode_ccitt_within`]): noise takes 114 ns a byte of
/// rows and about two bytes of code to a byte of rows, so that and a step a byte of rows is about
/// what a step of JBIG2 decoding takes.
const CODE_WORK: u64 = 4;
/// CCITT fax `data` decoded as `CCITTFaxDecode` gives it: rows of `columns` 1-bit samples, most
/// significant bit first, each starting on a new byte, with black as 0 (as 1 with `black_is_1`).
///
/// At most `max_rows` rows are decoded (fewer when `rows` says so), so the output is bounded by
/// the caller. Damaged data gives the rows decoded before the damage, as viewers show them: a
/// caller that needs a number of rows checks it got them.
///
/// Group 3 rows on bytes (`k` from 0, `encoded_byte_align`) are read by their end-of-line codes,
/// whether the fill starts each code on a byte or ends it on one, when `data` starts with one (T.4
/// sends one before the first row when it sends them at all), and on bytes otherwise, whatever
/// `end_of_line` says. Group 4 is read as its parameters say. Data with codes but none before the
/// first row is read on bytes, as PdfCraft's renderer draws it; with `k` 0 and the fill ending each
/// code on a byte, that may be other rows than were coded.
///
/// Errors: `columns` that aren't from 1 to [`MAX_CCITT_COLUMNS`] ([`FilterError::Unsupported`]).
///
/// Decoding spends from `work`, what decoding may still cost across a run of images, and there is
/// no way in without it: a
/// byte of rows a step, whether or not the data holds them, and [`CODE_WORK`] a byte of `data`.
/// A blank row takes a bit of code, so the rows are what a few bytes can ask for. It is all paid
/// before anything is decoded, and what `work` can't pay for spends all of it.
pub fn decode_ccitt_within(data: &[u8], p: &CcittParams, max_rows: u32, work: &mut u64) -> Result<Vec<u8>, FilterError> {
let rows = u32::try_from(p.rows).ok().filter(|r| *r > 0).map_or(max_rows, |r| r.min(max_rows));
let stride = u64::try_from(p.columns).unwrap_or(0).div_ceil(8);
crate::pay(work, stride.saturating_mul(u64::from(rows)).saturating_add((data.len() as u64).saturating_mul(CODE_WORK)))?;
decode_ccitt(data, p, max_rows)
}
/// Whether CCITT fax `data` starts with an end-of-line code: a 1 after 11 zeros or more, fill
/// included. T.4 sends one before the first row when it sends them at all, and no other code starts
/// with more than eight zeros.
@@ -133,7 +152,7 @@ fn starts_with_end_of_line(data: &[u8]) -> bool {
/// Rows of 1-bit samples filled from a decoder's pixels: `columns` a row, each row on a new byte, at
/// most `rows` rows. Pixels past a row's end and rows past the last are dropped. The JBIG2 decoder
/// ([`crate::decode_jbig2`]) fills them too.
/// ([`crate::decode_jbig2_within`]) fills them too.
pub(crate) struct Rows {
pub(crate) out: Vec<u8>,
row: Vec<u8>,
@@ -189,6 +208,42 @@ impl hayro_ccitt::Decoder for Rows {
mod tests {
use super::*;
/// A run pays for the rows an image asks for and the code it reads, before any of it is
/// decoded: rows the data doesn't hold are paid for too, as a few bytes can ask for any number.
#[test]
fn a_run_of_images_pays_for_the_rows_it_asks_for() {
let (w, h) = (4096u32, 64u32);
let stride = (w as usize).div_ceil(8);
let data = encode_g4(&vec![0xFF; stride * h as usize], w, h).unwrap();
let p = CcittParams { k: -1, columns: i64::from(w), ..CcittParams::default() };
let whole = decode_ccitt(&data, &p, h).unwrap();
let cost = (whole.len() + 4 * data.len()) as u64;
let mut work = cost;
assert_eq!(decode_ccitt_within(&data, &p, h, &mut work).unwrap(), whole, "exactly enough");
assert_eq!(work, 0);
let mut work = cost + 7;
decode_ccitt_within(&data, &p, h, &mut work).unwrap();
assert_eq!(work, 7, "only what it took");
let mut work = cost - 1;
let refused = decode_ccitt_within(&data, &p, h, &mut work).unwrap_err();
assert!(matches!(refused, FilterError::Spent), "a step short: {refused}");
assert_eq!(work, 0, "spends all that was left");
// Nothing left: not read at all, so bad columns aren't reported as bad columns.
let bad = CcittParams { columns: 0, ..p };
assert!(decode_ccitt_within(&data, &bad, h, &mut 0).is_err_and(|e| matches!(e, FilterError::Spent)));
// No code at all, for the same rows.
let mut work = u64::MAX;
decode_ccitt_within(&[], &p, h, &mut work).unwrap();
assert_eq!(u64::MAX - work, whole.len() as u64);
// `/Rows` below what is asked for, and no rows.
let mut work = u64::MAX;
decode_ccitt_within(&data, &CcittParams { rows: 3, ..p }, h, &mut work).unwrap();
assert_eq!(u64::MAX - work, (3 * stride + 4 * data.len()) as u64);
let mut work = u64::MAX;
decode_ccitt_within(&data, &p, 0, &mut work).unwrap();
assert_eq!(u64::MAX - work, 4 * data.len() as u64);
}
/// The rows decoded by an independent decoder (`hayro-ccitt`), as 1 for white and 0 for black.
pub(crate) fn decode(data: &[u8], columns: u32, rows: u32) -> Vec<Vec<bool>> {
struct Rows(Vec<Vec<bool>>, Vec<bool>, usize);
crates/filters/src/flate.rs
+3 -2
View File
@@ -28,8 +28,9 @@ pub(crate) fn decode(data: &[u8], max: usize) -> Step {
// A header that only looked valid: retry the whole input as raw deflate.
Err(f) if f.partial.is_empty() && matches!(f.error, crate::FilterError::Corrupt { .. }) => match inflate_raw(data, max) {
// Keep the original corruption diagnostic only if the retry is corrupt
// too. A limit error must remain fatal even for tolerant decoding.
Err(retry) if matches!(retry.error, crate::FilterError::Corrupt { .. }) => Err(f),
// too. A limit error must remain fatal even for tolerant decoding. What the
// retry decoded isn't kept, but a caller charging for work has to hear of it.
Err(retry) if matches!(retry.error, crate::FilterError::Corrupt { .. }) => Err(Failure { discarded: retry.partial.len(), ..f }),
retry => retry,
},
r => r,
crates/filters/src/jbig2.rs
+1 -1
View File
@@ -7,7 +7,7 @@
mod decode;
mod symbols;
pub use decode::{MAX_JBIG2_PIXELS, decode_jbig2};
pub use decode::{MAX_JBIG2_PIXELS, decode_jbig2_within};
use crate::FilterError;
use crate::mq::{Context, MqEncoder};
crates/filters/src/jbig2/decode.rs
+62 -3
View File
@@ -3,7 +3,7 @@
use crate::FilterError;
use crate::ccitt::Rows;
/// The largest JBIG2 page [`decode_jbig2`] decodes, in pixels. The decoder decodes a page whole,
/// The largest JBIG2 page [`decode_jbig2_within`] decodes, in pixels. The decoder decodes a page whole,
/// and takes pages of up to about this many (20,000 × 10,000 decodes).
pub const MAX_JBIG2_PIXELS: u64 = 1 << 28;
@@ -22,10 +22,23 @@ pub const MAX_JBIG2_PIXELS: u64 = 1 << 28;
/// ([`FilterError::Corrupt`]); a page past `max` ([`FilterError::LimitExceeded`]) or
/// [`MAX_JBIG2_PIXELS`] ([`FilterError::Unsupported`], or `Corrupt` from the decoder, which has the
/// same cap and can refuse the page first).
pub fn decode_jbig2(data: &[u8], globals: Option<&[u8]>, columns: u32, rows: u32, max: usize) -> Result<Vec<u8>, FilterError> {
///
/// Decoding spends from `work`: what decoding may still cost across a run of images, in the
/// decoder's count of steps. There is no way in without it, so a run of images can't decode them
/// without sharing an allowance. Reading `data` and `globals` is charged first
/// ([`hayro_jbig2::charge_reading`]), so an image that turns out not to decode has paid for that,
/// then what decoding takes ([`hayro_jbig2::Image::decode_within`]). An image takes at most what
/// one decode may on top of its reading and its page, whatever `work` allows, and one that `work`
/// can't pay for spends all of it. With nothing left an image isn't read at all, so the images after
/// that one cost nothing but this check, however large their globals.
pub fn decode_jbig2_within(data: &[u8], globals: Option<&[u8]>, columns: u32, rows: u32, max: usize, work: &mut u64) -> Result<Vec<u8>, FilterError> {
if rows == 0 {
return Ok(Vec::new());
}
if *work == 0 {
return Err(FilterError::Spent);
}
hayro_jbig2::charge_reading(work, data.len().saturating_add(globals.map_or(0, <[u8]>::len))).map_err(|_| FilterError::Spent)?;
let corrupt = |detail: String| FilterError::Corrupt { filter: "JBIG2Decode", detail };
let image = hayro_jbig2::Image::new_embedded(data, globals).map_err(|e| corrupt(e.to_string()))?;
let (width, height) = (image.width(), image.height());
@@ -41,7 +54,8 @@ pub fn decode_jbig2(data: &[u8], globals: Option<&[u8]>, columns: u32, rows: u32
return Err(FilterError::LimitExceeded(max));
}
let mut out = Rows::new(width, height, false);
image.decode(&mut out).map_err(|e| corrupt(e.to_string()))?;
// Stopped by what was left, it isn't a damaged image, whatever the decoder calls it.
image.decode_within(&mut out, work).map_err(|e| if *work == 0 { FilterError::Spent } else { corrupt(e.to_string()) })?;
Ok(out.out)
}
@@ -63,6 +77,12 @@ mod tests {
use super::*;
use crate::{encode_jbig2, encode_jbig2_generic};
/// [`decode_jbig2_within`] with no limit on the work.
fn decode_jbig2(data: &[u8], globals: Option<&[u8]>, columns: u32, rows: u32, max: usize) -> Result<Vec<u8>, FilterError> {
let mut unbounded = u64::MAX;
decode_jbig2_within(data, globals, columns, rows, max, &mut unbounded)
}
/// `rows` rows of `columns` samples of noise, the bits that pad each row to a byte 0.
fn noise(columns: u32, rows: u32) -> Vec<u8> {
let stride = (columns as usize).div_ceil(8);
@@ -123,6 +143,45 @@ mod tests {
assert_eq!(decode_jbig2(&symbols, None, w, h, usize::MAX).unwrap(), marks);
}
/// What a page costs is the same however much a run has left. Given exactly that it decodes,
/// and less stops it: spending all that's left, and what comes after isn't read. A page that
/// fails still spends what it took.
#[test]
fn a_run_of_images_spends_its_allowance_down_to_the_step() {
let (w, h) = (203u32, 41u32);
let data = encode_jbig2_generic(&noise(w, h), w, h).unwrap();
let whole = decode_jbig2(&data, None, w, h, usize::MAX).unwrap();
let mut work = u64::MAX;
assert_eq!(decode_jbig2_within(&data, None, w, h, usize::MAX, &mut work).unwrap(), whole);
let cost = u64::MAX - work;
assert!(cost >= u64::from(w) * u64::from(h), "a pixel at a time: {cost}");
let mut work = cost;
assert_eq!(decode_jbig2_within(&data, None, w, h, usize::MAX, &mut work).unwrap(), whole, "exactly enough");
assert_eq!(work, 0);
let mut work = cost + 7;
decode_jbig2_within(&data, None, w, h, usize::MAX, &mut work).unwrap();
assert_eq!(work, 7, "only what it took");
let mut work = cost - 1;
assert!(matches!(decode_jbig2_within(&data, None, w, h, usize::MAX, &mut work), Err(FilterError::Spent)), "a step short");
assert_eq!(work, 0, "spends all that was left");
// Nothing left: not read at all, so this garbage isn't reported as garbage.
assert!(matches!(decode_jbig2_within(b"garbage", None, w, h, usize::MAX, &mut work), Err(FilterError::Spent)));
// Too little even for its page.
let mut work = 3;
assert!(decode_jbig2_within(&data, None, w, h, usize::MAX, &mut work).is_err());
assert_eq!(work, 0);
// A region that can't be read after the one that can: the page fails having spent the
// first, and the segment it fails on, and that isn't being out of allowance.
let bad_region = [0, 0, 0, 9, 38, 0, 1, 0, 0, 0, 1, 0xFF];
let mut work = u64::MAX;
assert!(decode_jbig2_within(&[&data[..], &bad_region].concat(), None, w, h, usize::MAX, &mut work).is_err());
assert_eq!(u64::MAX - work, cost + 1 + 4 * bad_region.len() as u64, "the first region, a segment more, and reading it");
// Not a stream at all: it pays for being read, and globals count with it.
let mut work = u64::MAX;
assert!(matches!(decode_jbig2_within(b"garbage", Some(b"globals"), w, h, usize::MAX, &mut work), Err(FilterError::Corrupt { .. })));
assert_eq!(u64::MAX - work, 4 * 14);
}
/// The page has to be the image's size, and is bounded before it is decoded, being decoded
/// whole: within `max` as rows of one bit, and within [`MAX_JBIG2_PIXELS`]. Globals are given
/// to the decoder: ones it skips decode, ones that aren't JBIG2 are an error, as is damaged
crates/filters/src/lib.rs
+272 -26
View File
@@ -7,9 +7,27 @@
//! Image codecs (`DCTDecode`, `JPXDecode`, `JBIG2Decode`, `CCITTFaxDecode`),
//! `Crypt` and unknown filters are reported as [`FilterError::Unsupported`] so
//! callers keep such data encoded. CCITT fax and JBIG2 images decode on their own
//! ([`decode_ccitt`], [`decode_jbig2`]), as `pdfcraft_cos::Document::decoded_image`
//! does after the filters before them. 1-bit samples can be encoded as CCITT Group 4
//! ([`encode_g4`]) or lossless JBIG2 ([`encode_jbig2`]).
//! ([`decode_ccitt_within`], [`decode_jbig2_within`]), each spending from an allowance its caller
//! shares across a run of images, as `pdfcraft_cos::ImageBudget::decode` does after the filters
//! before them. Neither decodes without one:
//!
//! ```compile_fail
//! let _ = pdfcraft_filters::decode_jbig2(&[], None, 1, 1, 1 << 20);
//! ```
//!
//! ```compile_fail
//! let _ = pdfcraft_filters::decode_ccitt(&[], &pdfcraft_filters::CcittParams::default(), 1);
//! ```
//!
//! ```
//! let mut work = 1 << 20;
//! assert!(pdfcraft_filters::decode_jbig2_within(&[], None, 1, 1, 1 << 20, &mut work).is_err());
//! let p = pdfcraft_filters::CcittParams { columns: 8, ..Default::default() };
//! assert!(pdfcraft_filters::decode_ccitt_within(&[], &p, 1, &mut work).is_ok());
//! ```
//!
//! 1-bit samples can be encoded as CCITT Group 4 ([`encode_g4`]) or lossless JBIG2
//! ([`encode_jbig2`]).
//!
//! Every decoder is bounded by a caller-supplied `max_output` (decompression-bomb
//! defence) and never panics on malformed input.
@@ -28,8 +46,8 @@ mod runlength;
use std::borrow::Cow;
pub use ccitt::{CcittParams, MAX_CCITT_COLUMNS, MAX_G4_COLUMNS, decode_ccitt, encode_g4};
pub use jbig2::{MAX_JBIG2_COLUMNS, MAX_JBIG2_PIXELS, decode_jbig2, encode_jbig2, encode_jbig2_generic};
pub use ccitt::{CcittParams, MAX_CCITT_COLUMNS, MAX_G4_COLUMNS, decode_ccitt_within, encode_g4};
pub use jbig2::{MAX_JBIG2_COLUMNS, MAX_JBIG2_PIXELS, decode_jbig2_within, encode_jbig2, encode_jbig2_generic};
/// A PDF stream filter, identified by its `/Filter` name.
#[derive(Clone, Debug, PartialEq, Eq, Hash)]
@@ -91,7 +109,8 @@ impl Filter {
matches!(self, Filter::Dct | Filter::Jpx | Filter::Jbig2 | Filter::CcittFax)
}
fn is_decodable(&self) -> bool {
/// True for the general-purpose filters, which a chain can hold ([`decode`]).
pub fn is_decodable(&self) -> bool {
matches!(self, Filter::Flate | Filter::Lzw | Filter::AsciiHex | Filter::Ascii85 | Filter::RunLength)
}
}
@@ -123,33 +142,53 @@ pub enum FilterError {
LimitExceeded(usize),
#[error("unsupported filter: {0}")]
Unsupported(String),
/// A run of images has nothing left to spend ([`decode_tolerant_within`]).
#[error("more image decoding than one operation may do")]
Spent,
}
/// A decoder failure carrying whatever was decoded before the error.
pub(crate) struct Failure {
pub(crate) error: FilterError,
pub(crate) partial: Vec<u8>,
/// Bytes decoded on the way that `partial` doesn't hold: they took as long.
pub(crate) discarded: usize,
}
impl Failure {
pub(crate) fn corrupt(filter: &'static str, detail: impl Into<String>, partial: Vec<u8>) -> Self {
Failure { error: FilterError::Corrupt { filter, detail: detail.into() }, partial }
Failure { error: FilterError::Corrupt { filter, detail: detail.into() }, partial, discarded: 0 }
}
pub(crate) fn limit(max: usize) -> Self {
Failure { error: FilterError::LimitExceeded(max), partial: Vec::new() }
Failure { error: FilterError::LimitExceeded(max), partial: Vec::new(), discarded: 0 }
}
}
pub(crate) type Step = Result<Vec<u8>, Failure>;
/// Takes `steps` from `work`, what a run of images may still cost. What it can't pay for spends
/// all of it.
pub(crate) fn pay(work: &mut u64, steps: u64) -> Result<(), FilterError> {
match work.checked_sub(steps) {
Some(left) => {
*work = left;
Ok(())
}
None => {
*work = 0;
Err(FilterError::Spent)
}
}
}
/// Decodes `data` through `chain` (first element applied first).
///
/// Image codecs, `Crypt` and unknown filters anywhere in the chain give
/// [`FilterError::Unsupported`] (callers keep such data encoded), as does a chain longer than
/// [`MAX_CHAIN`]. `max_output` bounds every intermediate buffer.
pub fn decode(chain: &[(Filter, Params)], data: &[u8], max_output: usize) -> Result<Vec<u8>, FilterError> {
run(chain, data, max_output, true, false).map(|(v, _, _)| v).map_err(|(e, _)| e)
run(chain, data, max_output, None, true).map(|(v, _, _)| v).map_err(|(e, _)| e)
}
/// Like [`decode`] for a caller charging an allowance: `max_output` bounds the bytes all stages
@@ -157,47 +196,98 @@ pub fn decode(chain: &[(Filter, Params)], data: &[u8], max_output: usize) -> Res
/// bytes, on success and on failure. A stage's output counts even when a later stage or its
/// predictor shrinks it.
pub fn decode_counted(chain: &[(Filter, Params)], data: &[u8], max_output: usize) -> Result<(Vec<u8>, usize), (FilterError, usize)> {
run(chain, data, max_output, true, true).map(|(v, _, spent)| (v, spent))
run(chain, data, max_output, Some(max_output), true).map(|(v, _, spent)| (v, spent))
}
/// Like [`decode`], but on corrupt data returns what was decoded before the error and
/// reports `partial = true` (the behaviour viewers expect). The remaining filters of the
/// chain still run on the partial data. Limit and unsupported errors are still errors.
pub fn decode_tolerant(chain: &[(Filter, Params)], data: &[u8], max_output: usize) -> Result<(Vec<u8>, bool), FilterError> {
run(chain, data, max_output, false, false).map(|(v, partial, _)| (v, partial)).map_err(|(e, _)| e)
run(chain, data, max_output, None, false).map(|(v, partial, _)| (v, partial)).map_err(|(e, _)| e)
}
/// [`decode_tolerant`] bounded and counted as [`decode_counted`] is.
pub fn decode_tolerant_counted(chain: &[(Filter, Params)], data: &[u8], max_output: usize) -> Result<(Vec<u8>, bool, usize), (FilterError, usize)> {
run(chain, data, max_output, false, true)
run(chain, data, max_output, Some(max_output), false)
}
/// [`decode_tolerant`], spending from `work`, what decoding may still cost across a run of images:
/// a step for each byte of `data`, paid before anything is decoded, and one for each byte a stage
/// decodes, so a chain that fails, or a stage that goes past `max_output`, has paid for what it did.
/// No stage decodes past what is left: a chain that `work` can't pay for spends all of it and stops
/// there, so with nothing left nothing is read. A chain that can't be decoded at all costs nothing,
/// whatever is left, so what an image's filters are is known from the error.
///
/// Unlike [`decode_tolerant`], no filters at all gives `data` itself, within `max_output`.
pub fn decode_tolerant_within(chain: &[(Filter, Params)], data: &[u8], max_output: usize, work: &mut u64) -> Result<Vec<u8>, FilterError> {
spending(chain, data, max_output, work, false)
}
/// [`decode`], spending from `work` as [`decode_tolerant_within`] does: corrupt data is an error,
/// having paid for what was decoded before it.
pub fn decode_within(chain: &[(Filter, Params)], data: &[u8], max_output: usize, work: &mut u64) -> Result<Vec<u8>, FilterError> {
spending(chain, data, max_output, work, true)
}
fn spending(chain: &[(Filter, Params)], data: &[u8], max_output: usize, work: &mut u64, strict: bool) -> Result<Vec<u8>, FilterError> {
decodable(chain)?;
pay(work, data.len() as u64)?;
if chain.is_empty() {
return if data.len() > max_output { Err(FilterError::LimitExceeded(max_output)) } else { Ok(data.to_vec()) };
}
let left = usize::try_from(*work).unwrap_or(usize::MAX);
match run(chain, data, max_output, Some(left), strict) {
Ok((out, _, decoded)) => pay(work, decoded as u64).map(|()| out),
// The stage stopped at the lower of the two, which took as much to find out.
Err((FilterError::LimitExceeded(_), decoded)) if left.saturating_sub(decoded) > max_output => {
pay(work, (decoded as u64).saturating_add(max_output as u64))?;
Err(FilterError::LimitExceeded(max_output))
}
Err((FilterError::LimitExceeded(_), _)) => {
*work = 0;
Err(FilterError::Spent)
}
Err((e, decoded)) => pay(work, decoded as u64).and(Err(e)),
}
}
/// Longest filter chain decoded. Real streams use one to three filters; every stage may produce
/// up to the output limit, so the chain length multiplies the work.
pub const MAX_CHAIN: usize = 8;
/// The decoded data, whether it is partial and the bytes all stages decoded, or the error and
/// the bytes decoded before it. `max` bounds each stage, or with `cumulative` all of them
/// together.
fn run(chain: &[(Filter, Params)], data: &[u8], max: usize, strict: bool, cumulative: bool) -> Result<(Vec<u8>, bool, usize), (FilterError, usize)> {
/// Whether a chain is one [`run`] takes, and if not, the error.
fn decodable(chain: &[(Filter, Params)]) -> Result<(), FilterError> {
if chain.len() > MAX_CHAIN {
return Err((FilterError::Unsupported(format!("a chain of {} filters", chain.len())), 0));
return Err(FilterError::Unsupported(format!("a chain of {} filters", chain.len())));
}
if let Some((f, _)) = chain.iter().find(|(f, _)| !f.is_decodable()) {
return Err((FilterError::Unsupported(f.name().to_owned()), 0));
match chain.iter().find(|(f, _)| !f.is_decodable()) {
Some((f, _)) => Err(FilterError::Unsupported(f.name().to_owned())),
None => Ok(()),
}
}
/// The decoded data, whether it is partial and the bytes all stages decoded, or the error and
/// the bytes decoded before it. `max` bounds each stage, and `total` all of them together.
fn run(
chain: &[(Filter, Params)],
data: &[u8],
max: usize,
total: Option<usize>,
strict: bool,
) -> Result<(Vec<u8>, bool, usize), (FilterError, usize)> {
decodable(chain).map_err(|e| (e, 0))?;
let mut cur: Cow<[u8]> = Cow::Borrowed(data);
let mut partial = false;
let mut spent = 0usize;
for (filter, params) in chain {
#[cfg(test)]
STAGE_BYTES.with(|c| c.set(c.get() + cur.len()));
let limit = if cumulative { max.saturating_sub(spent) } else { max };
let limit = total.map_or(max, |total| max.min(total.saturating_sub(spent)));
let (step, work) = decode_one(filter, params, &cur, limit);
spent = spent.saturating_add(work);
cur = Cow::Owned(match step {
Ok(v) => v,
Err(Failure { error: FilterError::Corrupt { .. }, partial: out }) if !strict => {
Err(Failure { error: FilterError::Corrupt { .. }, partial: out, .. }) if !strict => {
partial = true;
out
}
@@ -216,15 +306,15 @@ fn decode_one(filter: &Filter, params: &Params, data: &[u8], max: usize) -> (Ste
Filter::AsciiHex => counted(asciihex::decode(data, max)),
Filter::Ascii85 => counted(ascii85::decode(data, max)),
Filter::RunLength => counted(runlength::decode(data, max)),
other => (Err(Failure { error: FilterError::Unsupported(other.name().to_owned()), partial: Vec::new() }), 0),
other => (Err(Failure { error: FilterError::Unsupported(other.name().to_owned()), partial: Vec::new(), discarded: 0 }), 0),
}
}
/// A step with the bytes it produced, its output or its partial output.
/// A step with the bytes it produced: its output, or its partial output and what it threw away.
fn counted(step: Step) -> (Step, usize) {
let n = match &step {
Ok(v) => v.len(),
Err(f) => f.partial.len(),
Err(f) => f.partial.len().saturating_add(f.discarded),
};
(step, n)
}
@@ -236,12 +326,12 @@ fn with_predictor(step: Step, params: &Params, name: &'static str) -> (Step, usi
let (step, n) = counted(step);
let step = match step {
Ok(v) => predictor::decode(params, v, name),
Err(Failure { error, partial }) => {
Err(Failure { error, partial, discarded }) => {
let partial = match predictor::decode(params, partial, name) {
Ok(v) => v,
Err(f) => f.partial,
};
Err(Failure { error, partial })
Err(Failure { error, partial, discarded })
}
};
(step, n)
@@ -428,6 +518,9 @@ mod tests {
assert_eq!(decode(&chain, &data, max).map(|d| d.len()).ok(), Some(1 << 19));
assert!(matches!(decode_counted(&chain, &data, max), Err((FilterError::LimitExceeded(m), _)) if m == max));
assert_eq!(decode_counted(&chain, &data, 2 << 20).map(|(_, spent)| spent).ok(), Some((1 << 20) + (1 << 19)));
let tolerant = decode_tolerant_counted(&chain, &data, 2 << 20).map(|(out, partial, spent)| (out.len(), partial, spent));
assert_eq!(tolerant.ok(), Some((1 << 19, false, (1 << 20) + (1 << 19))));
assert!(matches!(decode_tolerant_counted(&chain, &data, max), Err((FilterError::LimitExceeded(m), _)) if m == max));
}
/// A PNG predictor drops a tag byte per row, but Flate produced it: the count includes it,
@@ -446,4 +539,157 @@ mod tests {
assert!(matches!(e, FilterError::Corrupt { .. }), "{e}");
assert_eq!(spent, bad.len());
}
fn refused(e: &FilterError) -> bool {
matches!(e, FilterError::Spent)
}
/// A chain pays for what it reads and what each stage decodes, down to the byte: Flate inflates
/// hex that ASCIIHex halves, so a stage counts when a later one shrinks it.
#[test]
fn a_chain_spends_down_to_the_byte() {
let hex = b"ab".repeat(1 << 10);
let data = encode_flate(&hex);
let chain = [(Filter::Flate, Params::default()), (Filter::AsciiHex, Params::default())];
let whole = decode(&chain, &data, 1 << 20).unwrap();
let cost = (data.len() + hex.len() + whole.len()) as u64;
let mut work = u64::MAX;
assert_eq!(decode_tolerant_within(&chain, &data, 1 << 20, &mut work).unwrap(), whole);
assert_eq!(u64::MAX - work, cost);
let mut work = cost;
assert_eq!(decode_tolerant_within(&chain, &data, 1 << 20, &mut work).unwrap(), whole, "exactly enough");
assert_eq!(work, 0);
let mut work = cost + 7;
decode_tolerant_within(&chain, &data, 1 << 20, &mut work).unwrap();
assert_eq!(work, 7, "only what it took");
let mut work = cost - 1;
assert!(refused(&decode_tolerant_within(&chain, &data, 1 << 20, &mut work).unwrap_err()), "a step short");
assert_eq!(work, 0, "spends all that was left");
// Each stage has `max_output` to itself, as in a plain decode: together they are past it.
let mut work = u64::MAX;
assert_eq!(decode_tolerant_within(&chain, &data, hex.len(), &mut work).unwrap(), whole);
}
/// No stage runs past what is left: the first stage of this bomb stops there, so the second
/// isn't given anything to read, and with too little to read it nothing runs at all.
#[test]
fn a_chain_stops_at_what_is_left() {
let inner = encode_flate(&vec![0; 1 << 20]);
let bomb = encode_flate(&inner);
let chain = [(Filter::Flate, Params::default()), (Filter::Flate, Params::default())];
let read = |work: u64| {
super::STAGE_BYTES.with(|c| c.set(0));
let mut work = work;
assert!(refused(&decode_tolerant_within(&chain, &bomb, usize::MAX, &mut work).unwrap_err()));
assert_eq!(work, 0);
super::STAGE_BYTES.with(|c| c.get())
};
assert_eq!(read((bomb.len() + inner.len()) as u64 + 10), bomb.len() + inner.len(), "the second stage stops");
assert_eq!(read(bomb.len() as u64 + 10), bomb.len(), "the first stops, and the second never runs");
assert_eq!(read(bomb.len() as u64 - 1), 0, "it can't pay to be read");
}
/// Past `max_output` the chain has failed, and hasn't run out of anything: it pays what it took
/// to find out, and what is left stays. Whichever limit a stage stops at is the one reported.
#[test]
fn a_stage_past_its_limit_pays_for_finding_out() {
let data = encode_flate(&vec![0; 1 << 16]);
let chain = [(Filter::Flate, Params::default())];
let mut work = u64::MAX;
assert!(matches!(decode_tolerant_within(&chain, &data, 1000, &mut work), Err(FilterError::LimitExceeded(1000))));
assert_eq!(u64::MAX - work, data.len() as u64 + 1000);
// The later stage of two, which pays for the first as well.
let bomb = encode_flate(&data);
let two = [(Filter::Flate, Params::default()), (Filter::Flate, Params::default())];
let mut work = u64::MAX;
assert!(matches!(decode_tolerant_within(&two, &bomb, data.len(), &mut work), Err(FilterError::LimitExceeded(_))));
assert_eq!(u64::MAX - work, (bomb.len() + 2 * data.len()) as u64);
// Both limits at once: a run that can't go on is what a caller needs to hear.
let mut work = data.len() as u64 + 1000;
assert!(refused(&decode_tolerant_within(&chain, &data, 1000, &mut work).unwrap_err()));
assert_eq!(work, 0);
let mut work = data.len() as u64 + 1001;
assert!(matches!(decode_tolerant_within(&chain, &data, 1000, &mut work), Err(FilterError::LimitExceeded(1000))));
assert_eq!(work, 1, "the image's limit, with a step to spare");
}
/// What a damaged stream decoded is paid for, however the decode ends.
#[test]
fn a_chain_that_fails_has_paid_for_what_it_did() {
let data: Vec<u8> = (0..20_000u32).map(|i| (i % 97) as u8 ^ (i / 300) as u8).collect();
let mut cut = encode_flate(&data);
cut.truncate(cut.len() / 2);
let chain = [(Filter::Flate, Params::default())];
let (partial, _) = decode_tolerant(&chain, &cut, 1 << 20).unwrap();
assert!(!partial.is_empty());
let mut work = u64::MAX;
assert_eq!(decode_tolerant_within(&chain, &cut, 1 << 20, &mut work).unwrap(), partial);
assert_eq!(u64::MAX - work, (cut.len() + partial.len()) as u64);
// Read strictly it is an error, for the same work.
let mut work = u64::MAX;
assert!(matches!(decode_within(&chain, &cut, 1 << 20, &mut work), Err(FilterError::Corrupt { .. })));
assert_eq!(u64::MAX - work, (cut.len() + partial.len()) as u64);
let whole = encode_flate(&data);
let mut work = u64::MAX;
assert_eq!(decode_within(&chain, &whole, 1 << 20, &mut work).unwrap(), data);
assert_eq!(u64::MAX - work, (whole.len() + data.len()) as u64);
// A predictor that comes to nothing, after all of this was inflated for it.
let p = Params { predictor: 99, ..Params::default() };
let data = encode_flate(&vec![0; 1 << 16]);
let mut work = u64::MAX;
assert!(decode_tolerant_within(&[(Filter::Flate, p)], &data, 1 << 20, &mut work).unwrap().is_empty());
assert_eq!(u64::MAX - work, data.len() as u64 + (1 << 16));
}
/// A stream that reads as zlib only until it fails, and as raw deflate until that fails too,
/// gives back no data, and has still decoded everything the retry did.
#[test]
fn a_retry_that_fails_pays_for_what_it_decoded() {
// Stored as raw deflate: this many bytes, then a block of a reserved type. As zlib: the
// same reserved type at once, so the first reading fails with nothing.
let mut data = vec![0x78, 0x01, 0x06, 0xFE, 0xF9];
data.extend(vec![0u8; 0x0601]);
data.push(0x07);
let chain = [(Filter::Flate, Params::default())];
let (out, partial) = decode_tolerant(&chain, &data, 1 << 20).unwrap();
assert!(out.is_empty() && partial, "what is read stays as it was");
let (_, _, spent) = decode_tolerant_counted(&chain, &data, 1 << 20).unwrap();
assert_eq!(spent, 0x0601);
let (e, spent) = decode_counted(&chain, &data, 1 << 20).unwrap_err();
assert!(matches!(e, FilterError::Corrupt { .. }), "{e}");
assert_eq!(spent, 0x0601);
let mut work = u64::MAX;
assert!(decode_tolerant_within(&chain, &data, 1 << 20, &mut work).unwrap().is_empty());
assert_eq!(u64::MAX - work, (data.len() + 0x0601) as u64);
// Through a predictor, which would otherwise lose the count.
let p = Params { predictor: 12, columns: 4, ..Params::default() };
let (_, _, spent) = decode_tolerant_counted(&[(Filter::Flate, p)], &data, 1 << 20).unwrap();
assert_eq!(spent, 0x0601);
}
/// With nothing left nothing is read. What can't be decoded at all never costs anything, and
/// says so whatever is left. Without filters, what is read is the data.
#[test]
fn what_a_chain_is_shows_whatever_is_left() {
let flate = [(Filter::Flate, Params::default())];
let mut work = 0;
assert!(refused(&decode_tolerant_within(&flate, b"garbage", 1 << 20, &mut work).unwrap_err()));
let long = vec![(Filter::Flate, Params::default()); MAX_CHAIN + 1];
for chain in [vec![(Filter::Dct, Params::default())], vec![(Filter::Jbig2, Params::default())], long] {
for left in [0, 100] {
let mut work = left;
let e = decode_tolerant_within(&chain, b"data", 1 << 20, &mut work).unwrap_err();
assert!(matches!(e, FilterError::Unsupported(_)) && !refused(&e), "{e}");
assert_eq!(work, left);
}
}
let mut work = 100;
assert_eq!(decode_tolerant_within(&[], b"data", 4, &mut work).unwrap(), b"data");
assert_eq!(work, 96);
assert!(matches!(decode_tolerant_within(&[], b"data", 3, &mut work), Err(FilterError::LimitExceeded(3))));
assert_eq!(work, 92);
let mut work = 3;
assert!(refused(&decode_tolerant_within(&[], b"data", 4, &mut work).unwrap_err()));
assert_eq!(work, 0);
}
}
crates/optimize/README.md
+8 -2
View File
@@ -100,13 +100,19 @@ let report = optimize_with_progress(&mut doc, &settings, &mut |stage| keep_going
- Lossless JBIG2 uses bounded generic-region encoding or exact-copy symbol dictionaries when
smaller. No lossy symbol substitution is used. Export, redaction and this step read JBIG2
images, with bounded shared globals.
images, with bounded shared globals. All of a run's images share one allowance of decoding
(`pdfcraft_cos::ImageBudget`), filters, decoders and pixels alike, JPEG samples and merged
pictures too: about
1,000 scanned pages' worth, plus more for each byte of the file, enough to decode a stream
about 8 times at Flate's best. Images past it stay as they are, and `Report::images_refused`
counts those the image step read. Streams it only stores with Flate (below) aren't counted.
- **Discard objects:** thumbnails, alternate images, document tags, print settings.
- **Clean up:** Flate for streams with no filter; for streams with an outer ASCII85 or ASCIIHex
layer, that layer is removed when the result is smaller (the inner filters keep their encoded
bytes; streams with `/DecodeParms`, images and object streams keep their ASCII layer. LZW-compressed
streams, content and images alike, are decoded and stored with Flate instead (Acrobat's "In
streams that use LZW encoding, use Flate instead"): a predictor carries over to Flate without
streams that use LZW encoding, use Flate instead"), from the run's allowance, past which they
stay LZW: a predictor carries over to Flate without
`EarlyChange`, filters applied before the LZW one go with it and those after it keep their
bytes. A stream whose LZW data doesn't decode cleanly is left as it is. Not yet: inline images.
crates/optimize/src/fragments.rs
+69 -10
View File
@@ -18,7 +18,7 @@
//! drops them.
use pdfcraft_content::{Matrix, Op, parse, serialize_ops};
use pdfcraft_cos::{DecodeBudget, Dict, Document, ObjRef, Object, Stream};
use pdfcraft_cos::{DecodeBudget, Dict, Document, ImageBudget, ObjRef, Object, Stream};
use crate::images::{components, content_bytes, decodable, page_resources, pixels, resample};
use crate::{OptimizeError, Report, Stage};
@@ -152,7 +152,7 @@ fn run_end(ops: &[Op], i: usize, count: usize) -> usize {
}
/// The run's pixels as one image, top row first: `(pixels, width, height)`.
fn merge(doc: &Document, run: &[Fragment], n: usize) -> Option<(Vec<u8>, u32, u32)> {
fn merge(doc: &Document, run: &[Fragment], n: usize, budget: &mut ImageBudget) -> Option<(Vec<u8>, u32, u32)> {
let (top, bottom) = (run.first()?.rect[3], run.last()?.rect[1]);
let width = run.iter().map(|f| f.width).max()?;
// Pixel rows per unit of height: the finest fragment's.
@@ -167,7 +167,12 @@ fn merge(doc: &Document, run: &[Fragment], n: usize) -> Option<(Vec<u8>, u32, u3
return None;
}
let row = (width as usize).checked_mul(n)?;
let mut out = vec![255u8; row.checked_mul(height as usize)?];
// What this builds isn't what its strips cost to read: four small ones can ask for all of it.
let bytes = row.checked_mul(height as usize)?;
if !budget.pay_pixels(u64::try_from(bytes).ok()?) {
return None;
}
let mut out = vec![255u8; bytes];
for f in run {
let r0 = (((top - f.rect[3]) * density).round().max(0.0) as u32).min(height);
let r1 = (((top - f.rect[1]) * density).round().max(0.0) as u32).min(height);
@@ -176,7 +181,7 @@ fn merge(doc: &Document, run: &[Fragment], n: usize) -> Option<(Vec<u8>, u32, u3
}
let obj = doc.get(f.image);
let Object::Stream(s) = &*obj else { return None };
let px = pixels(doc, s, n)?;
let px = pixels(doc, s, n, budget)?;
let px = if (f.width, f.height) == (width, r1 - r0) { px } else { resample(&px, n, f.width, f.height, width, r1 - r0)? };
out.get_mut(r0 as usize * row..r1 as usize * row)?.copy_from_slice(px.get(..(r1 - r0) as usize * row)?);
}
@@ -230,7 +235,13 @@ fn add_xobject(doc: &mut Document, page: ObjRef, name: &[u8], image: ObjRef) ->
}
/// Merge the fragment runs of one page; returns how many fragments were merged.
fn merge_page(doc: &mut Document, page: ObjRef, report: &mut Report, progress: &mut dyn FnMut(Stage) -> bool) -> Result<usize, OptimizeError> {
fn merge_page(
doc: &mut Document,
page: ObjRef,
report: &mut Report,
progress: &mut dyn FnMut(Stage) -> bool,
images: &mut ImageBudget,
) -> Result<usize, OptimizeError> {
let Some(contents) = doc.get(page).as_dict().and_then(|d| d.get(b"Contents").cloned()) else { return Ok(0) };
// A page whose content isn't all known is left alone: rewriting it would drop the rest.
let Some(data) = content_bytes(doc, &contents, &mut DecodeBudget::page()) else { return Ok(0) };
@@ -286,7 +297,7 @@ fn merge_page(doc: &mut Document, page: ObjRef, report: &mut Report, progress: &
if !progress(Stage::Fragments) {
return Err(OptimizeError::Cancelled);
}
let Some((px, w, h)) = merge(doc, &run, n) else {
let Some((px, w, h)) = merge(doc, &run, n, images) else {
out.extend(ops.get(i..j).unwrap_or_default().iter().cloned());
i = j;
continue;
@@ -320,9 +331,15 @@ fn merge_page(doc: &mut Document, page: ObjRef, report: &mut Report, progress: &
Ok(merged)
}
pub(crate) fn run(doc: &mut Document, pages: &[ObjRef], report: &mut Report, progress: &mut dyn FnMut(Stage) -> bool) -> Result<(), OptimizeError> {
pub(crate) fn run(
doc: &mut Document,
pages: &[ObjRef],
report: &mut Report,
progress: &mut dyn FnMut(Stage) -> bool,
budget: &mut ImageBudget,
) -> Result<(), OptimizeError> {
for p in pages {
report.fragments_merged += merge_page(doc, *p, report, progress)?;
report.fragments_merged += merge_page(doc, *p, report, progress, budget)?;
}
Ok(())
}
@@ -343,7 +360,49 @@ mod tests {
Fragment { image, rect: [0.0, edge, 1.0, 1.0], width: 1 << 16, height: 1 },
Fragment { image, rect: [0.0, 0.0, 1.0, edge], width: 1 << 16, height: 1 },
];
assert!(merge(&Document::new_empty(), &run, 3).is_none());
assert!(merge(&Document::new_empty(), &run, 3, &mut ImageBudget::with_work(u64::MAX)).is_none());
}
/// Merging reads the images the pass after it reads, from the same allowance: with nothing
/// left the strips stay as they are, and the budget shows the read it refused.
#[test]
fn merging_spends_the_runs_allowance() {
let mut doc = Document::new_empty();
let mut image = Dict::new();
for (k, v) in
[("Subtype", Object::name("Image")), ("Width", Object::Int(8)), ("Height", Object::Int(1)), ("BitsPerComponent", Object::Int(8))]
{
image.set(k.as_bytes().to_vec(), v);
}
image.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
let (mut xobjects, mut drawn, mut raw) = (Dict::new(), String::new(), 0);
for row in 0..8 {
let strip = Stream::flate(image.clone(), &[row * 30; 8]);
raw += strip.raw.len() as u64;
let r = doc.add(Object::Stream(strip));
xobjects.set(format!("I{row}").into_bytes(), Object::Ref(r));
drawn.push_str(&format!("q 80 0 0 10 0 {} cm /I{row} Do Q\n", 70 - 10 * i32::from(row)));
}
let content = doc.add(Object::Stream(Stream::flate(Dict::new(), drawn.as_bytes())));
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
let mut page = Dict::new();
page.set(b"Type".to_vec(), Object::name("Page"));
page.set(b"Resources".to_vec(), Object::Dict(resources));
page.set(b"Contents".to_vec(), Object::Ref(content));
let page = doc.add(Object::Dict(page));
let mut budget = ImageBudget::with_work(0);
assert_eq!(merge_page(&mut doc.clone(), page, &mut Report::default(), &mut |_| true, &mut budget).unwrap(), 0);
assert_eq!(budget.refused(), 1);
// Reading each strip, and building the picture of 8 rows of 8 they make.
let cost = raw + 8 * (8 + 8) + 8 * 8;
let mut budget = ImageBudget::with_work(u64::MAX);
assert_eq!(merge_page(&mut doc.clone(), page, &mut Report::default(), &mut |_| true, &mut budget).unwrap(), 8);
assert_eq!(u64::MAX - budget.work_left(), cost);
// Enough for the strips and not for the picture: nothing is built.
let mut budget = ImageBudget::with_work(cost - 1);
assert_eq!(merge_page(&mut doc, page, &mut Report::default(), &mut |_| true, &mut budget).unwrap(), 0);
assert_eq!(budget.refused(), 1);
}
/// Many images drawn on top of each other are not one picture; looking for runs among them reads
@@ -370,7 +429,7 @@ mod tests {
page.set(b"Resources".to_vec(), Object::Dict(resources));
let page = doc.add(Object::Dict(page));
GROUP_READS.with(|c| c.set(0));
let merged = merge_page(&mut doc, page, &mut Report::default(), &mut |_| true);
let merged = merge_page(&mut doc, page, &mut Report::default(), &mut |_| true, &mut ImageBudget::with_work(u64::MAX));
assert_eq!(merged.ok(), Some(0));
let reads = GROUP_READS.with(|c| c.get());
assert!(reads <= 40 * 4 * groups, "{reads} group reads");
crates/optimize/src/images.rs
+320 -21
View File
@@ -4,7 +4,7 @@ use std::collections::{HashMap, HashSet};
use std::sync::Arc;
use pdfcraft_content::{Matrix, Pieces};
use pdfcraft_cos::{DecodeBudget, Dict, Document, ObjRef, Object, Stream};
use pdfcraft_cos::{DecodeBudget, Dict, Document, ImageBudget, ObjRef, Object, Stream};
use crate::{Compression, ImageSettings, MonoCompression, MonoSettings, OptimizeError, Report, Settings, Stage};
@@ -479,7 +479,7 @@ fn count_decoded(_n: u64) {
/// every row, and the image would be written back wrong. `None` for a JPEG whose colour transform
/// readers may disagree about ([`ycc_encoded`]), and for data that decodes past what the samples
/// can need ([`allowance`]).
pub(crate) fn pixels(doc: &Document, s: &Stream, n: usize) -> Option<Vec<u8>> {
pub(crate) fn pixels(doc: &Document, s: &Stream, n: usize, budget: &mut ImageBudget) -> Option<Vec<u8>> {
let (w, h) = (u32::try_from(s.dict.int(b"Width")?).ok()?, u32::try_from(s.dict.int(b"Height")?).ok()?);
if w == 0 || h == 0 || u64::from(w) * u64::from(h) > MAX_PIXELS {
return None;
@@ -490,11 +490,14 @@ pub(crate) fn pixels(doc: &Document, s: &Stream, n: usize) -> Option<Vec<u8>> {
// never get here: the colour space check excludes them. The JPEG's own size must be the
// dictionary's: `jpeg_samples` refuses another from its header, before any pixels.
let ycc = n == 3 && ycc_encoded(doc, s)?;
if !budget.pay_jpeg(u64::try_from(bytes).ok()?, &s.raw) {
return None;
}
let v = jpeg_samples(&s.raw, usize::try_from(w).ok()?, usize::try_from(h).ok()?, n, ycc)?;
count_decoded(u64::from(w) * u64::from(h));
(v.len() == bytes).then_some(v)
} else {
let v = doc.with_direct_parms(s).decoded_within(allowance(bytes, h)?).ok()?;
let v = budget.decode_filters(doc, s, allowance(bytes, h)?).ok()?;
Some(v.get(..bytes)?.to_vec())
}
}
@@ -716,7 +719,14 @@ fn image_stream(old: &Dict, px: &[u8], n: usize, w: u32, h: u32, compression: Co
/// One image: resample and/or recompress per `settings`; `ppi` is its effective resolution.
/// Returns the new image and its new soft mask, if the result is smaller.
fn process(doc: &Document, s: &Stream, ppi: f64, settings: &ImageSettings, n: usize) -> Option<(Stream, Option<(ObjRef, Stream)>)> {
fn process(
doc: &Document,
s: &Stream,
ppi: f64,
settings: &ImageSettings,
n: usize,
budget: &mut ImageBudget,
) -> Option<(Stream, Option<(ObjRef, Stream)>)> {
let d = &s.dict;
if d.get(b"Decode").is_some_and(|a| !is_identity_decode(doc, a, n))
|| d.contains(b"Mask")
@@ -748,7 +758,7 @@ fn process(doc: &Document, s: &Stream, ppi: f64, settings: &ImageSettings, n: us
} else {
(w, h)
};
let px = pixels(doc, s, n)?;
let px = pixels(doc, s, n, budget)?;
let px = if (nw, nh) != (w, h) { resample(&px, n, w, h, nw, nh)? } else { px };
let new = image_stream(d, &px, n, nw, nh, settings.compression, was_dct)?;
// A soft mask must keep the image's dimensions: resample it with the image (lossless).
@@ -761,7 +771,7 @@ fn process(doc: &Document, s: &Stream, ppi: f64, settings: &ImageSettings, n: us
return None;
}
let (mw, mh) = (u32::try_from(m.dict.int(b"Width")?).ok()?, u32::try_from(m.dict.int(b"Height")?).ok()?);
let mpx = pixels(doc, m, 1)?;
let mpx = pixels(doc, m, 1, budget)?;
let mpx = resample(&mpx, 1, mw, mh, nw, nh)?;
smask = Some((mr, image_stream(&m.dict, &mpx, 1, nw, nh, Compression::Flate, false)?));
}
@@ -801,7 +811,7 @@ type Pixel = fn(&[u8]) -> Option<bool>;
/// with its width and height, and, when they were packed from more bits, its dictionary as a 1-bit
/// image. These count:
/// - one bit per pixel in one colour component, or an image mask, stored with no filter, with
/// general-purpose ones, or with CCITT fax or JBIG2 last ([`Document::decoded_image_within`], no further
/// general-purpose ones, or with CCITT fax or JBIG2 last ([`ImageBudget::decode`], no further
/// than its rows can need);
/// - with `pack`, and with no colour-key `/Mask` (its ranges are in the image's own values):
/// - eight-bit grey (one component, not indexed) whose every sample is 0 or 255. A decode array
@@ -815,7 +825,7 @@ type Pixel = fn(&[u8]) -> Option<bool>;
/// - an indexed image of 2, 4 or 8 bits that uses only palette entries 0 and 1, which stay 0 and
/// 1 at one bit, whatever their colours, so the colour space stays. A decode array other than
/// the default stops it: in index units it would mean something else at one bit.
fn one_bit(doc: &Document, s: &Stream, pack: bool, defaults: bool) -> Option<(Vec<u8>, u32, u32, Option<Dict>)> {
fn one_bit(doc: &Document, s: &Stream, pack: bool, defaults: bool, budget: &mut ImageBudget) -> Option<(Vec<u8>, u32, u32, Option<Dict>)> {
let d = &s.dict;
// Through u32 first, as everywhere here, so the product can't overflow.
let (w, h) = (u32::try_from(d.int(b"Width")?).ok()?, u32::try_from(d.int(b"Height")?).ok()?);
@@ -836,7 +846,7 @@ fn one_bit(doc: &Document, s: &Stream, pack: bool, defaults: bool) -> Option<(Ve
_ => allowance(len, h)?,
};
count_decoded(u64::from(w) * u64::from(h));
let mut data = doc.decoded_image_within(s, max).ok()?;
let mut data = budget.decode(doc, s, max).ok()?;
if data.len() < len {
return None;
}
@@ -904,7 +914,7 @@ fn one_bit(doc: &Document, s: &Stream, pack: bool, defaults: bool) -> Option<(Ve
let row_len = wu.checked_mul(n)?.checked_mul(bpc)?.div_ceil(8);
// Through its parameters given by reference too: without a PNG predictor's tag bytes taken
// off, rows of 0 and 255 (or palette entries 0 and 1) would still pass below, each shifted.
let samples = doc.with_direct_parms(s).decoded_within(allowance(row_len.checked_mul(hu)?, h)?).ok()?;
let samples = budget.decode_filters(doc, s, allowance(row_len.checked_mul(hu)?, h)?).ok()?;
Some((packed_rows(&samples, row_len, wu, hu, (n, bpc, white))?, w, h, Some(packed)))
}
@@ -1104,7 +1114,7 @@ fn downsample_bits_checked(data: &[u8], w: u32, h: u32, nw: u32, nh: u32, dark_i
/// The samples keep their meaning, so the colour space, decode array and `/Mask` stay, but for
/// black-and-white RGB, which becomes grey with one decode pair, and an indexed image's default
/// decode array, which goes as the default at one bit.
fn mono(doc: &Document, s: &Stream, ppi: f64, defaults: bool, settings: &MonoSettings) -> Option<(Stream, bool)> {
fn mono(doc: &Document, s: &Stream, ppi: f64, defaults: bool, settings: &MonoSettings, budget: &mut ImageBudget) -> Option<(Stream, bool)> {
let d = &s.dict;
let downsample = settings.downsample && ppi.is_finite() && ppi > settings.above_ppi && settings.target_ppi > 0.0 && !d.contains(b"SMask");
let retain = settings.compression == MonoCompression::Retain;
@@ -1118,10 +1128,13 @@ fn mono(doc: &Document, s: &Stream, ppi: f64, defaults: bool, settings: &MonoSet
// JBIG2 (whose symbols may be smaller than its own coding) can be smaller stored again alone.
let alone_to_g4 = s.filters().len() == 1 && settings.compression == MonoCompression::CcittG4;
let only_to_resample = retain || (alone_to_g4 && (jbig2 || fax.is_some_and(|p| p.k < 0)));
if only_to_resample && !downsample {
// What can't be averaged, whatever it holds, comes to the same `None` below, after a decode that
// would spend the run's allowance on it.
let too_wide = d.int(b"Width").is_some_and(|w| w > i64::from(pdfcraft_filters::MAX_CCITT_COLUMNS));
if only_to_resample && (!downsample || defaults || too_wide || settings.target_ppi >= ppi || (jbig2 && dark_is_one(doc, d).is_none())) {
return None;
}
let (data, w, h, packed) = one_bit(doc, s, !retain, defaults)?;
let (data, w, h, packed) = one_bit(doc, s, !retain, defaults, budget)?;
let packed_from_more = packed.is_some();
let k = settings.target_ppi / ppi;
// `downsample_bits` keeps a counter for each column: no more columns than a fax row may have,
@@ -1188,12 +1201,15 @@ fn mono(doc: &Document, s: &Stream, ppi: f64, defaults: bool, settings: &MonoSet
(best.0 < crate::stored_len(s)?).then_some((best.1, resampled))
}
/// One pass over the document's images, within `budget`. An image past it stays as it is, and is
/// counted in the report.
pub(crate) fn run(
doc: &mut Document,
pages: &[ObjRef],
settings: &Settings,
report: &mut Report,
progress: &mut dyn FnMut(Stage) -> bool,
budget: &mut ImageBudget,
) -> Result<(), OptimizeError> {
let (ppi, mut defaults, read) = walk_pages(doc, pages);
// Black-and-white RGB becomes grey, which default colour spaces would draw otherwise, and an
@@ -1222,10 +1238,11 @@ pub(crate) fn run(
let obj = doc.get(r);
let Object::Stream(s) = &*obj else { continue };
report.images += 1;
let refused = budget.refused();
// Packed from more bits a sample, it is lossless at one bit, but the grey or colour
// settings can make it smaller still (below).
let packed_from_more = s.dict.int(b"BitsPerComponent").is_some_and(|b| b > 1);
let (packed, mono) = match mono(doc, s, ppi, defaults.contains(&r.num), &settings.mono) {
let (packed, mono) = match mono(doc, s, ppi, defaults.contains(&r.num), &settings.mono, budget) {
Some(m) if packed_from_more => (Some(m), None),
m => (None, m),
};
@@ -1242,7 +1259,7 @@ pub(crate) fn run(
continue;
}
let n = s.dict.get(b"ColorSpace").and_then(|cs| components(doc, cs));
let processed = n.and_then(|n| process(doc, s, ppi, if n == 1 { &settings.gray } else { &settings.color }, n));
let processed = n.and_then(|n| process(doc, s, ppi, if n == 1 { &settings.gray } else { &settings.color }, n, budget));
// A dithered scan, say, downsampled and JPEG-compressed: the smaller wins, counting a soft
// mask resampled with it.
let smask_len = |smask: Option<&(ObjRef, Stream)>| match smask {
@@ -1268,7 +1285,11 @@ pub(crate) fn run(
doc.set(r, Object::Stream(new));
continue;
}
let Some((new, smask)) = processed else { continue };
let Some((new, smask)) = processed else {
// Read to be packed and again to be resampled, it is still one image, and it stayed.
report.images_refused += usize::from(budget.refused() > refused);
continue;
};
let resized = new.dict.int(b"Width") != s.dict.int(b"Width") || new.dict.int(b"Height") != s.dict.int(b"Height");
report.image_bytes_before += s.raw.len();
report.image_bytes_after += new.raw.len();
@@ -1316,13 +1337,16 @@ mod dimension_tests {
dict.set(b"Width".to_vec(), Object::Int(width));
dict.set(b"Height".to_vec(), Object::Int(height));
let stream = Stream::from_raw(dict, vec![0; 4]);
assert!(pixels(&Document::new_empty(), &stream, channels).is_none(), "{width} x {height} x {channels}");
assert!(
pixels(&Document::new_empty(), &stream, channels, &mut ImageBudget::with_work(u64::MAX)).is_none(),
"{width} x {height} x {channels}"
);
}
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(2));
dict.set(b"Height".to_vec(), Object::Int(2));
let stream = Stream::from_raw(dict, vec![1, 2, 3, 4, 5]);
assert_eq!(pixels(&Document::new_empty(), &stream, 1), Some(vec![1, 2, 3, 4]));
assert_eq!(pixels(&Document::new_empty(), &stream, 1, &mut ImageBudget::with_work(u64::MAX)), Some(vec![1, 2, 3, 4]));
}
/// JBIG2 alone, with Group 4 asked for, is decoded only to be resampled. A cost guard: dropped,
@@ -1344,7 +1368,7 @@ mod dimension_tests {
let settings = MonoSettings { compression: MonoCompression::CcittG4, ..MonoSettings::REDUCE };
let decoded = |s: &Stream, settings: &MonoSettings| {
DECODED_PIXELS.with(|c| c.set(0));
mono(&Document::new_empty(), s, 300.0, false, settings);
mono(&Document::new_empty(), s, 300.0, false, settings, &mut ImageBudget::with_work(u64::MAX));
DECODED_PIXELS.with(|c| c.get())
};
let alone = image(Object::name("JBIG2Decode"), coded.clone());
@@ -1355,6 +1379,281 @@ mod dimension_tests {
assert!(decoded(&alone, &jbig2) > 0, "stored as JBIG2 its symbols may come out smaller");
}
/// A run has one allowance for all its JBIG2 images. Those past it stay as they are, and the
/// report says how many: a file of thousands of them can't hold a run for hours.
#[test]
fn jbig2_images_past_the_runs_allowance_stay_as_they_are() {
let coded = pdfcraft_filters::encode_jbig2_generic(&[0xF0; 4], 16, 2).unwrap();
let mut dict = Dict::new();
dict.set(b"Type".to_vec(), Object::name("XObject"));
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(1));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
dict.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
let image = Stream::from_raw(dict, coded);
let mut cost = ImageBudget::with_work(u64::MAX);
cost.decode(&Document::new_empty(), &image, 1 << 20).unwrap();
let cost = u64::MAX - cost.work_left();
let mut doc = Document::new_empty();
let refs = [doc.add(Object::Stream(image.clone())), doc.add(Object::Stream(image.clone())), doc.add(Object::Stream(image.clone()))];
let mut xobjects = Dict::new();
for (i, r) in refs.iter().enumerate() {
xobjects.set(format!("I{i}").into_bytes(), Object::Ref(*r));
}
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
// Drawn small enough, at 720 ppi, for the settings to resample all three.
let content = doc.add(Object::Stream(Stream::flate(Dict::new(), b"q 1.6 0 0 .2 0 0 cm /I0 Do /I1 Do /I2 Do Q")));
let mut page = Dict::new();
page.set(b"Type".to_vec(), Object::name("Page"));
page.set(b"Resources".to_vec(), Object::Dict(resources));
page.set(b"Contents".to_vec(), Object::Ref(content));
let page = doc.add(Object::Dict(page));
let settings = Settings { mono: MonoSettings { compression: MonoCompression::CcittG4, ..MonoSettings::REDUCE }, ..Settings::default() };
let mut report = Report::default();
let mut budget = ImageBudget::with_work(2 * cost);
run(&mut doc, &[page], &settings, &mut report, &mut |_| true, &mut budget).unwrap();
assert_eq!(report.images_refused, 1);
assert_eq!(budget.work_left(), 0, "two decoded and the third refused");
assert_eq!(report.images_resampled, 2);
let changed = refs.iter().filter(|r| matches!(&*doc.get(**r), Object::Stream(s) if *s != image)).count();
assert_eq!(changed, 2, "the third is as it was");
}
/// Images that aren't JBIG2 share the run's allowance too, both on the way to being resampled
/// and on the way to being packed to a bit a sample: a file of filter bombs would hold a run
/// as long as one of JBIG2, and those past it stay as they are.
#[test]
fn flate_images_past_the_runs_allowance_stay_as_they_are() {
// Noise, resampled, or black and white, packed.
let mut seed = 0x2545_F491_4F6C_DD1Du64;
let noise: Vec<u8> = (0..256 * 256)
.map(|_| {
seed ^= seed << 13;
seed ^= seed >> 7;
seed ^= seed << 17;
seed as u8
})
.collect();
let black_and_white: Vec<u8> = noise.iter().map(|v| if v & 1 == 0 { 0 } else { 255 }).collect();
for (name, samples, settings) in [
("resampled", &noise, Settings::default()),
(
"packed",
&black_and_white,
Settings { mono: MonoSettings { compression: MonoCompression::CcittG4, ..MonoSettings::REDUCE }, ..Settings::default() },
),
] {
let page_of = |images: usize| {
let mut doc = Document::new_empty();
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(256));
dict.set(b"Height".to_vec(), Object::Int(256));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
let image = Stream::flate(dict, samples);
let refs: Vec<ObjRef> = (0..images).map(|_| doc.add(Object::Stream(image.clone()))).collect();
let mut xobjects = Dict::new();
let mut drawn = String::from("q 10 0 0 10 0 0 cm ");
for (i, r) in refs.iter().enumerate() {
xobjects.set(format!("I{i}").into_bytes(), Object::Ref(*r));
drawn.push_str(&format!("/I{i} Do "));
}
drawn.push('Q');
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
let content = doc.add(Object::Stream(Stream::flate(Dict::new(), drawn.as_bytes())));
let mut page = Dict::new();
page.set(b"Type".to_vec(), Object::name("Page"));
page.set(b"Resources".to_vec(), Object::Dict(resources));
page.set(b"Contents".to_vec(), Object::Ref(content));
let page = doc.add(Object::Dict(page));
(doc, page, refs, image)
};
let (mut doc, page, ..) = page_of(1);
let mut budget = ImageBudget::with_work(u64::MAX);
let mut report = Report::default();
run(&mut doc, &[page], &settings, &mut report, &mut |_| true, &mut budget).unwrap();
let cost = u64::MAX - budget.work_left();
assert!(cost > 0, "{name}");
assert_eq!(report.images_refused, 0, "{name}");
let (mut doc, page, refs, image) = page_of(3);
let mut budget = ImageBudget::with_work(2 * cost);
let mut report = Report::default();
run(&mut doc, &[page], &settings, &mut report, &mut |_| true, &mut budget).unwrap();
assert_eq!(report.images_refused, 1, "{name}: {report:?}");
assert_eq!(budget.work_left(), 0, "{name}");
let changed = refs.iter().filter(|r| matches!(&*doc.get(**r), Object::Stream(s) if *s != image)).count();
assert_eq!(changed, 2, "{name}: the third is as it was");
}
}
/// An image the allowance ran out on between being packed and being resampled is rewritten,
/// as the packed one, so it isn't one that stayed as it was.
#[test]
fn a_packed_image_isnt_counted_as_refused() {
let mut seed = 0x2545_F491_4F6C_DD1Du64;
let samples: Vec<u8> = (0..256 * 256)
.map(|_| {
seed ^= seed << 13;
seed ^= seed >> 7;
seed ^= seed << 17;
if seed & 1 == 0 { 0 } else { 255 }
})
.collect();
let mut doc = Document::new_empty();
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(256));
dict.set(b"Height".to_vec(), Object::Int(256));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
let image = Stream::flate(dict, &samples);
let read = (image.raw.len() + samples.len()) as u64 + 256 * 256;
let original = doc.add(Object::Stream(image.clone()));
let mut xobjects = Dict::new();
xobjects.set(b"I0".to_vec(), Object::Ref(original));
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
let content = doc.add(Object::Stream(Stream::flate(Dict::new(), b"q 10 0 0 10 0 0 cm /I0 Do Q")));
let mut page = Dict::new();
page.set(b"Type".to_vec(), Object::name("Page"));
page.set(b"Resources".to_vec(), Object::Dict(resources));
page.set(b"Contents".to_vec(), Object::Ref(content));
let page = doc.add(Object::Dict(page));
let settings = Settings { mono: MonoSettings { compression: MonoCompression::CcittG4, ..MonoSettings::REDUCE }, ..Settings::default() };
let mut report = Report::default();
let mut budget = ImageBudget::with_work(read);
run(&mut doc, &[page], &settings, &mut report, &mut |_| true, &mut budget).unwrap();
assert_eq!(budget.refused(), 1, "its second read was refused");
assert_eq!(report.images_refused, 0, "{report:?}");
assert_eq!(report.images_recompressed + report.images_resampled, 1, "{report:?}");
assert!(matches!(&*doc.get(original), Object::Stream(s) if *s != image));
}
/// An image the run has no use for stayed as it was too, but nothing refused it: this one is
/// too small to compress and drawn at no more than its size.
#[test]
fn an_image_left_alone_isnt_counted_as_refused() {
let mut doc = Document::new_empty();
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(8));
dict.set(b"Height".to_vec(), Object::Int(8));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
let image = doc.add(Object::Stream(Stream::flate(dict, &[0x80; 64])));
let mut xobjects = Dict::new();
xobjects.set(b"I0".to_vec(), Object::Ref(image));
let mut resources = Dict::new();
resources.set(b"XObject".to_vec(), Object::Dict(xobjects));
let content = doc.add(Object::Stream(Stream::flate(Dict::new(), b"q 72 0 0 72 0 0 cm /I0 Do Q")));
let mut page = Dict::new();
page.set(b"Type".to_vec(), Object::name("Page"));
page.set(b"Resources".to_vec(), Object::Dict(resources));
page.set(b"Contents".to_vec(), Object::Ref(content));
let page = doc.add(Object::Dict(page));
let mut report = Report::default();
run(&mut doc, &[page], &Settings::default(), &mut report, &mut |_| true, &mut ImageBudget::with_work(u64::MAX)).unwrap();
assert_eq!((report.images, report.images_refused), (1, 0), "{report:?}");
}
/// A JPEG pays for its samples before it is decoded, however few bytes it takes: past what is
/// left it isn't, and counts as refused.
#[test]
fn a_jpeg_pays_for_its_samples() {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(8));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceRGB"));
let jpeg = image_stream(&dict, &[120; 16 * 8 * 3], 3, 16, 8, Compression::Jpeg(60), false).expect("a JPEG");
let doc = Document::new_empty();
let cost = 2 * 16 * 8 * 3 + 4 * jpeg.raw.len() as u64;
let mut budget = ImageBudget::with_work(cost);
assert_eq!(pixels(&doc, &jpeg, 3, &mut budget).map(|p| p.len()), Some(16 * 8 * 3));
assert_eq!((budget.work_left(), budget.refused()), (0, 0), "exactly that");
let mut budget = ImageBudget::with_work(cost - 1);
assert!(pixels(&doc, &jpeg, 3, &mut budget).is_none());
assert_eq!((budget.work_left(), budget.refused()), (0, 1));
}
/// A soft mask resampled with its image is read from the same allowance as the image.
#[test]
fn a_soft_mask_resampled_with_its_image_spends_the_allowance() {
let mut seed = 0x2545_F491_4F6C_DD1Du64;
let noise: Vec<u8> = (0..256 * 256)
.map(|_| {
seed ^= seed << 13;
seed ^= seed >> 7;
seed ^= seed << 17;
seed as u8
})
.collect();
let image = |space: &str| {
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(256));
dict.set(b"Height".to_vec(), Object::Int(256));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(8));
dict.set(b"ColorSpace".to_vec(), Object::name(space));
Stream::flate(dict, &noise)
};
let mut doc = Document::new_empty();
let mask = image("DeviceGray");
let cost = (mask.raw.len() + noise.len()) as u64 + 256 * 256;
let masked = doc.add(Object::Stream(mask));
let mut with_mask = image("DeviceGray");
with_mask.dict.set(b"SMask".to_vec(), Object::Ref(masked));
let spent = |s: &Stream| {
let mut budget = ImageBudget::with_work(u64::MAX);
process(&doc, s, 720.0, &Settings::default().gray, 1, &mut budget).expect("resampled, and smaller");
u64::MAX - budget.work_left()
};
assert_eq!(spent(&with_mask), spent(&image("DeviceGray")) + cost);
}
/// Decoded only to be resampled, what can't be averaged whatever it holds isn't decoded, so it
/// takes nothing from the images that can be, and isn't counted as refused past the allowance.
#[test]
fn a_jbig2_image_that_cant_be_averaged_spends_nothing() {
let image = |edit: &dyn Fn(&mut Dict)| {
let mut dict = Dict::new();
dict.set(b"Subtype".to_vec(), Object::name("Image"));
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(1));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
dict.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
edit(&mut dict);
Stream::from_raw(dict, pdfcraft_filters::encode_jbig2_generic(&[0xF0; 4], 16, 2).unwrap())
};
let spent_at = |target_ppi: f64, s: &Stream, defaults: bool| {
let settings = MonoSettings { compression: MonoCompression::Retain, above_ppi: 100.0, target_ppi, ..MonoSettings::REDUCE };
let mut budget = ImageBudget::with_work(u64::MAX);
let _ = mono(&Document::new_empty(), s, 720.0, defaults, &settings, &mut budget);
u64::MAX - budget.work_left()
};
let spent = |s: &Stream, defaults: bool| spent_at(300.0, s, defaults);
let plain = image(&|_| {});
assert_eq!(spent_at(720.0, &plain, false), 0, "already at the target");
assert!(spent_at(719.0, &plain, false) > 0, "just above it");
assert!(spent(&plain, false) > 0, "one that can be averaged is decoded");
assert_eq!(spent(&plain, true), 0, "drawn where a device colour space is remapped");
let width = |w: i64| image(&move |d| d.set(b"Width".to_vec(), Object::Int(w)));
let widest = i64::from(pdfcraft_filters::MAX_CCITT_COLUMNS);
assert!(spent(&width(widest), false) > 0, "the widest that can be averaged is decoded");
assert_eq!(spent(&width(widest + 1), false), 0, "too wide to average");
let tinted = image(&|d| d.set(b"ColorSpace".to_vec(), Object::Array(vec![Object::name("CalGray"), Object::Dict(Dict::new())])));
assert_eq!(spent(&tinted, false), 0, "no way to tell which sample is dark");
}
/// A JPEG larger than its dictionary says is refused from its header, before its pixels are decoded.
#[test]
fn a_jpeg_of_another_size_than_its_dictionary_is_not_decoded() {
@@ -1367,9 +1666,9 @@ mod dimension_tests {
Stream::from_raw(dict, data.clone())
};
DECODED_PIXELS.with(|c| c.set(0));
assert!(pixels(&Document::new_empty(), &stream(10, 10), 1).is_none());
assert!(pixels(&Document::new_empty(), &stream(10, 10), 1, &mut ImageBudget::with_work(u64::MAX)).is_none());
assert_eq!(DECODED_PIXELS.with(|c| c.get()), 0, "pixels decoded");
assert_eq!(pixels(&Document::new_empty(), &stream(1000, 1000), 1).map(|p| p.len()), Some(1000 * 1000));
assert_eq!(pixels(&Document::new_empty(), &stream(1000, 1000), 1, &mut ImageBudget::with_work(u64::MAX)).map(|p| p.len()), Some(1000 * 1000));
}
/// Resampling goes through no more than one of the two sizes holds, whatever their proportions:
crates/optimize/src/lib.rs
+47 -10
View File
@@ -36,7 +36,7 @@ mod unused;
pub use audit::{SpaceCategory, SpaceUse, audit_space};
use pdfcraft_cos::{Dict, Document, MAX_PAGE_DECODED, ObjRef, Object, Stream, serialize};
use pdfcraft_cos::{Dict, Document, ImageBudget, MAX_PAGE_DECODED, ObjRef, Object, Stream, serialize};
use pdfcraft_filters::{Filter, Params};
pub use images::effective_resolutions;
@@ -279,6 +279,8 @@ pub struct Report {
pub images: usize,
pub images_resampled: usize,
pub images_recompressed: usize,
/// Images left as they are: decoding them all would take more than one run may spend.
pub images_refused: usize,
/// Encoded image bytes before and after (images that changed).
pub image_bytes_before: usize,
pub image_bytes_after: usize,
@@ -329,10 +331,12 @@ pub fn optimize_with_progress(doc: &mut Document, settings: &Settings, progress:
}
let mut report = Report::default();
let pages = pdfcraft_annot::page_refs(doc).map_err(|_| OptimizeError::NoPages)?;
// Merging fragments, the image pass and the LZW pass read the same images, so they share it.
let mut budget = ImageBudget::new(doc);
if settings.merge_fragments {
fragments::run(doc, &pages, &mut report, progress)?;
fragments::run(doc, &pages, &mut report, progress, &mut budget)?;
}
images::run(doc, &pages, settings, &mut report, progress)?;
images::run(doc, &pages, settings, &mut report, progress, &mut budget)?;
if !progress(Stage::CleanUp) {
return Err(OptimizeError::Cancelled);
}
@@ -428,7 +432,11 @@ pub fn optimize_with_progress(doc: &mut Document, settings: &Settings, progress:
let r = ObjRef { num, generation: doc.generation(num) };
let obj = doc.get(r);
let Object::Stream(s) = &*obj else { continue };
if let Some(replacement) = lzw_to_flate(s, MAX_LZW_DECODED) {
// Each can take a second or more, of a pass that has no other way to stop.
if s.filters().iter().any(|(f, _)| *f == Filter::Lzw) && !progress(Stage::CleanUp) {
return Err(OptimizeError::Cancelled);
}
if let Some(replacement) = lzw_to_flate(s, MAX_LZW_DECODED, &mut budget) {
doc.set(r, Object::Stream(replacement));
report.lzw_streams += 1;
}
@@ -463,14 +471,17 @@ pub fn optimize_with_progress(doc: &mut Document, settings: &Settings, progress:
/// Largest decoded LZW stream converted, as for a page's content: the data and its check are
/// held together, so a few hundred kilobytes of hostile LZW can't ask for gigabytes.
const MAX_LZW_DECODED: usize = MAX_PAGE_DECODED;
/// The steps of a run's allowance a decoded byte costs to store with Flate and read back: up to
/// 78 ns a byte to encode and 4 to check, at the 6 to 16 ns of a step.
const LZW_REWRITE_WORK: u64 = 8;
/// `s` with its LZW compression replaced by Flate (Acrobat's "In streams that use LZW encoding,
/// use Flate instead"). The filters up to the last LZW one are decoded, strictly and with its
/// predictor; the result is Flate-encoded under the same predictor, without `EarlyChange`, and
/// the filters after it keep their encoded bytes. A stream that doesn't decode cleanly, or
/// whose filters or parameters can't be read for certain, or that decodes past `max` bytes, is
/// left as it is.
fn lzw_to_flate(s: &Stream, max: usize) -> Option<Stream> {
/// whose filters or parameters can't be read for certain, or that decodes past `max` bytes or
/// what `budget` has left, is left as it is.
fn lzw_to_flate(s: &Stream, max: usize, budget: &mut ImageBudget) -> Option<Stream> {
// An external file's data isn't in `raw`; cross-reference and object streams are rewritten
// on save, so converting them would change nothing in the file.
if s.dict.contains(b"F") || matches!(s.dict.name(b"Type"), Some(b"XRef" | b"ObjStm")) {
@@ -513,7 +524,10 @@ fn lzw_to_flate(s: &Stream, max: usize) -> Option<Stream> {
if !parms.get(..=lzw)?.iter().all(direct) {
return None;
}
let data = pdfcraft_filters::decode(stages, &s.raw, max).ok()?;
let data = budget.decode_chain(stages, &s.raw, max).ok()?;
if !budget.spend((data.len() as u64).saturating_mul(LZW_REWRITE_WORK)) {
return None;
}
let flate = Params { early_change: 1, ..params.clone() };
let raw = pdfcraft_filters::encode(&Filter::Flate, &flate, &data).ok()?;
// The predictor's encoder must give back exactly what its decoder read (a short last row,
@@ -597,8 +611,31 @@ mod tests {
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("LZWDecode"));
let s = Stream::from_raw(dict, raw);
assert!(lzw_to_flate(&s, data.len() - 1).is_none());
let converted = lzw_to_flate(&s, data.len()).unwrap();
let mut budget = ImageBudget::with_work(u64::MAX);
assert!(lzw_to_flate(&s, data.len() - 1, &mut budget).is_none());
let converted = lzw_to_flate(&s, data.len(), &mut budget).unwrap();
assert_eq!(converted.decoded_strict().unwrap(), data);
}
/// Rewriting a stream spends the run's allowance: reading it, decoding it, and storing and
/// checking what it decoded. One the allowance can't pay for stays LZW, and ends the rewriting.
#[test]
fn lzw_streams_spend_the_runs_allowance() {
let data = b"0 0 m 612 792 l S\n".repeat(64);
let raw = pdfcraft_filters::encode(&Filter::Lzw, &Params::default(), &data).unwrap();
let mut dict = Dict::new();
dict.set(b"Filter".to_vec(), Object::name("LZWDecode"));
let s = Stream::from_raw(dict, raw);
let cost = (s.raw.len() + data.len()) as u64 + data.len() as u64 * LZW_REWRITE_WORK;
let mut budget = ImageBudget::with_work(cost + 7);
assert!(lzw_to_flate(&s, data.len(), &mut budget).is_some());
assert_eq!(budget.work_left(), 7);
let mut budget = ImageBudget::with_work(cost);
assert!(lzw_to_flate(&s, data.len(), &mut budget).is_some(), "exactly enough");
assert_eq!(budget.work_left(), 0);
let mut budget = ImageBudget::with_work(cost - 1);
assert!(lzw_to_flate(&s, data.len(), &mut budget).is_none(), "a step short");
assert_eq!(budget.work_left(), 0, "and the next stream isn't read");
assert!(lzw_to_flate(&s, data.len(), &mut budget).is_none());
}
}
crates/optimize/tests/optimize.rs
+67 -20
View File
@@ -9,6 +9,12 @@ use pdfcraft_optimize::{
effective_resolutions, optimize,
};
/// What image `s` decodes to, its fax or JBIG2 data included, with no limit on the work: the
/// samples an operation would read through its `ImageBudget`.
fn decoded_image(doc: &Document, s: &Stream) -> Result<Vec<u8>, pdfcraft_cos::CosError> {
pdfcraft_cos::ImageBudget::with_work(u64::MAX).decode(doc, s, pdfcraft_cos::MAX_DECODED)
}
fn image(doc: &mut Document, w: u32, h: u32, n: usize, jpeg: bool, smask: Option<ObjRef>) -> ObjRef {
// A smooth gradient with a little texture (photo-like).
let mut px = Vec::with_capacity((w * h) as usize * n);
@@ -910,7 +916,7 @@ fn rgb_and_two_colour_indexed_black_and_white_images_become_one_bit() {
(Some(1), Some(&b"CCITTFaxDecode"[..]), Some(&cs)),
"{i:?}"
);
assert_eq!(doc.decoded_image(&s).unwrap(), text, "{i:?}");
assert_eq!(decoded_image(&doc, &s).unwrap(), text, "{i:?}");
let decode = if i == rgb_inverted { Some(Object::Array(vec![1.into(), 0.into()])) } else { None };
assert_eq!(s.dict.get(b"Decode").cloned(), decode, "{i:?}");
}
@@ -1066,7 +1072,7 @@ fn retain_keeps_a_group_3_halftone_as_it_is() {
let fax = doc.add(Object::Stream(Stream::from_raw(d, g3.clone())));
// 7.68 pt for 64 samples is 600 ppi.
page(&mut doc, &[("A", fax)], "q 7.68 0 0 7.68 100 600 cm /A Do Q");
assert_eq!(doc.decoded_image(&stream(&doc, fax)).unwrap().len(), 64 * 8, "the fixture decodes");
assert_eq!(decoded_image(&doc, &stream(&doc, fax)).unwrap().len(), 64 * 8, "the fixture decodes");
let r = optimize(&mut doc, &Settings { mono: RETAIN, ..Settings::default() }).unwrap();
assert_eq!((r.images_resampled, r.images_recompressed), (0, 0), "{r:?}");
assert_eq!(stream(&doc, fax).raw[..], g3[..]);
@@ -1564,7 +1570,7 @@ fn ambiguous_monochrome_text_is_recompressed_without_lossy_reduction() {
}
assert_eq!(raw(&kept, hex_at_400), raw(&doc, hex_at_400), "and nothing else changes");
for i in [flate, fax, hex_fax] {
assert_eq!((size(&kept, i), kept.decoded_image(&stream(&kept, i)).unwrap()), ((1200, 480), text.clone()), "{i:?}");
assert_eq!((size(&kept, i), decoded_image(&kept, &stream(&kept, i)).unwrap()), ((1200, 480), text.clone()), "{i:?}");
}
let bits = [eight, rgb, rgb_inverted].map(|i| stream(&kept, i).dict.int(b"BitsPerComponent"));
assert_eq!(bits, [Some(8); 3], "Retain leaves eight bits to the grey and colour settings");
@@ -1578,7 +1584,7 @@ fn ambiguous_monochrome_text_is_recompressed_without_lossy_reduction() {
let s = stream(&doc, i);
assert_eq!((size(&doc, i), s.dict.int(b"BitsPerComponent")), ((1200, 480), if i == mask { None } else { Some(1) }));
assert!(matches!(s.filters().as_slice(), [(Filter::CcittFax | Filter::Flate, _)]), "{i:?}");
assert_eq!(doc.decoded_image(&s).unwrap(), text.clone(), "{i:?}");
assert_eq!(decoded_image(&doc, &s).unwrap(), text.clone(), "{i:?}");
}
assert_eq!(stream(&doc, mask).dict.get(b"Decode"), Some(&decode), "the decode array stays");
// Grey now, with grey's decode pair, which makes 1 black: its ties are black too.
@@ -1588,12 +1594,12 @@ fn ambiguous_monochrome_text_is_recompressed_without_lossy_reduction() {
(size(&doc, rgb_inverted), s.dict.get(b"ColorSpace"), s.dict.get(b"Decode")),
((1200, 480), Some(&Object::name("DeviceGray")), Some(&inverting))
);
assert_eq!(doc.decoded_image(&s).unwrap(), text.iter().map(|b| !b).collect::<Vec<u8>>(), "the same picture, ties black");
assert_eq!(decoded_image(&doc, &s).unwrap(), text.iter().map(|b| !b).collect::<Vec<u8>>(), "the same picture, ties black");
for i in [at_400, soft, hex_at_400] {
assert_eq!((size(&doc, i), stream(&doc, i).dict.name(b"Filter")), ((1200, 480), Some(&b"CCITTFaxDecode"[..])), "{i:?}");
}
assert!(raw(&doc, hex_at_400).len() < hex_len, "the ASCIIHexDecode is dropped");
assert_eq!(doc.decoded_image(&stream(&doc, hex_at_400)).unwrap(), text, "and the picture kept");
assert_eq!(decoded_image(&doc, &stream(&doc, hex_at_400)).unwrap(), text, "and the picture kept");
assert_eq!([fax_at_400].map(|r| raw(&doc, r)), untouched);
let after = render(&doc);
assert_eq!(after, before, "conservative retention preserves the rendered page exactly");
@@ -1681,7 +1687,7 @@ fn a_fax_image_behind_a_decompression_bomb_is_left_alone() {
let r = optimize(&mut doc, &Settings::default()).unwrap();
if padding == 0 {
assert_eq!(r.images_resampled, 0, "ambiguous local tones are retained: {r:?}");
assert_eq!(doc.decoded_image(&stream(&doc, img)).unwrap(), bilevel(w, h, false));
assert_eq!(decoded_image(&doc, &stream(&doc, img)).unwrap(), bilevel(w, h, false));
} else {
assert_eq!((r.images_resampled, r.images_recompressed), (0, 0), "{r:?}");
assert_eq!(stream(&doc, img).raw, before);
@@ -1776,7 +1782,7 @@ fn monochrome_images_keep_what_they_should() {
assert_eq!([alone, deep].map(|r| raw(&doc, r)), before, "a Group 4 halftone alone, and sixteen-bit grey");
assert_eq!(filters(&doc, wrapped), 1, "a wrapped Group 4 halftone is stored again alone by default");
assert!(matches!(stream(&doc, scan).filters().as_slice(), [(Filter::CcittFax | Filter::Flate, _)]), "the scan: Group 4 or Flate");
assert_eq!(doc.decoded_image(&stream(&doc, scan)).unwrap(), ragged, "the scan preserves every sample");
assert_eq!(decoded_image(&doc, &stream(&doc, scan)).unwrap(), ragged, "the scan preserves every sample");
assert_eq!((stream(&doc, fax).dict.name(b"Filter"), filters(&doc, fax)), (Some(&b"CCITTFaxDecode"[..]), 1), "the busy fax image");
let (mut doc, [_, wrapped, scan, _, _]) = build();
@@ -1785,7 +1791,7 @@ fn monochrome_images_keep_what_they_should() {
assert_eq!(raw(&doc, wrapped), before, "Retain: a wrapped Group 4 halftone stays as it is");
let s = stream(&doc, scan);
assert_eq!((s.dict.int(b"Width"), s.dict.name(b"Filter")), (Some(1200), Some(&b"FlateDecode"[..])), "Retain: {r:?}");
assert_eq!(doc.decoded_image(&s).unwrap(), ragged, "Retain preserves every sample");
assert_eq!(decoded_image(&doc, &s).unwrap(), ragged, "Retain preserves every sample");
}
/// `rows` coded as CCITT Group 3, one-dimensional (`/K 0`): each row its runs in turn, white first,
@@ -1853,7 +1859,12 @@ fn group_3_fax_images_become_group_4() {
let g3 = group_3(&runs, k > 0, on_bytes);
let (end_of_line, encoded_byte_align) = (k > 0 && on_bytes.is_none(), on_bytes.is_some());
let p = pdfcraft_filters::CcittParams { k, columns: w.into(), end_of_line, encoded_byte_align, ..Default::default() };
assert_eq!(pdfcraft_filters::decode_ccitt(&g3, &p, h).unwrap(), bits, "/K {k}, on bytes {on_bytes:?}: the Group 3 data is the picture");
let mut work = u64::MAX;
assert_eq!(
pdfcraft_filters::decode_ccitt_within(&g3, &p, h, &mut work).unwrap(),
bits,
"/K {k}, on bytes {on_bytes:?}: the Group 3 data is the picture"
);
let mut doc = Document::new_empty();
let mut d = Dict::new();
@@ -1886,7 +1897,7 @@ fn group_3_fax_images_become_group_4() {
let s = stream(&doc, fax);
assert_eq!(doc.fax_params(&s).map(|p| p.k), Some(-1), "/K {k}, on bytes {on_bytes:?}: Group 4");
assert!(s.raw.len() < g3.len(), "{} bytes, from {}", s.raw.len(), g3.len());
assert_eq!(doc.decoded_image(&s).unwrap(), bits, "/K {k}, on bytes {on_bytes:?}: the same picture");
assert_eq!(decoded_image(&doc, &s).unwrap(), bits, "/K {k}, on bytes {on_bytes:?}: the same picture");
assert!(render(&doc) == render(&control), "/K {k}, on bytes {on_bytes:?}: the page looks as the picture does");
}
}
@@ -2102,14 +2113,14 @@ fn jbig2_scans_are_averaged_and_stored_again() {
let im = mono_image(&mut doc, w, h, &squares, None, &grey);
page(&mut doc, &[("Im", im)], "q 144 0 0 57.6 20 600 cm /Im Do Q");
assert_eq!(optimize(&mut doc, &settings).unwrap().images_resampled, 1, "Flate, {compression:?}");
let averaged = doc.decoded_image(&stream(&doc, im)).unwrap();
let averaged = decoded_image(&doc, &stream(&doc, im)).unwrap();
for (filters, globals) in [(&["JBIG2Decode"][..], false), (&["ASCIIHexDecode", "JBIG2Decode"][..], false), (&["JBIG2Decode"][..], true)] {
let (mut doc, im) = jbig2_page(&coded, filters, globals, (w, h), at_600);
let r = optimize(&mut doc, &settings).unwrap();
let s = stream(&doc, im);
let what = format!("{filters:?}, globals {globals}, {compression:?}: {r:?}");
assert_eq!((r.images_resampled, s.dict.int(b"Width"), s.dict.int(b"Height")), (1, Some(600), Some(240)), "{what}");
assert_eq!(doc.decoded_image(&s).unwrap(), averaged, "{what}");
assert_eq!(decoded_image(&doc, &s).unwrap(), averaged, "{what}");
assert!(!s.dict.contains(b"DecodeParms") || s.dict.name(b"Filter") == Some(&b"CCITTFaxDecode"[..]), "{what}");
let filter = s.dict.name(b"Filter").map(<[u8]>::to_vec);
// The smallest of what each compression may store; under Retain, JBIG2 again.
@@ -2140,13 +2151,13 @@ fn jbig2_scans_are_averaged_and_stored_again() {
let s = stream(&doc, im);
assert_eq!((r.images_resampled, r.images_recompressed, s.dict.name(b"Filter")), (0, 1, Some(&b"JBIG2Decode"[..])), "{r:?}");
assert!(s.raw.len() * 2 < generic.len(), "symbols {} bytes, the generic region {}", s.raw.len(), generic.len());
assert_eq!(doc.decoded_image(&s).unwrap(), text);
assert_eq!(decoded_image(&doc, &s).unwrap(), text);
// Behind ASCIIHex it is stored again alone, being smaller so.
let (mut doc, im) = jbig2_page(&generic, &["ASCIIHexDecode", "JBIG2Decode"], false, (tw, th), at_300);
optimize(&mut doc, &jbig2_asked).unwrap();
let s = stream(&doc, im);
assert_eq!((s.filters().len(), s.dict.name(b"Filter")), (1, Some(&b"JBIG2Decode"[..])));
assert_eq!(doc.decoded_image(&s).unwrap(), text);
assert_eq!(decoded_image(&doc, &s).unwrap(), text);
// So it is as Group 4 when that is smaller: here the hex digits are spaced out, as ASCIIHex
// allows, so that the stream is many times the code.
let (mut doc, im) = jbig2_page(&generic, &["ASCIIHexDecode", "JBIG2Decode"], false, (tw, th), at_300);
@@ -2156,7 +2167,7 @@ fn jbig2_scans_are_averaged_and_stored_again() {
optimize(&mut doc, &Settings::default()).unwrap();
let s = stream(&doc, im);
assert_eq!((s.filters().len(), s.dict.name(b"Filter")), (1, Some(&b"CCITTFaxDecode"[..])), "Group 4 alone");
assert_eq!(doc.decoded_image(&s).unwrap(), text);
assert_eq!(decoded_image(&doc, &s).unwrap(), text);
// A JBIG2 page taller than the image: the renderer draws the whole page into it, so it isn't
// read, and stays as it is.
@@ -3399,6 +3410,42 @@ fn lzw_streams_become_flate_and_decode_to_the_same_bytes() {
assert!(!stream(&doc, c_late).dict.contains(b"DecodeParms"), "nothing left to say for the content stream");
}
/// A stream can take a second or more to store again, so a cancel lands between any two of them,
/// and nothing runs after it.
#[test]
fn a_cancel_lands_during_the_lzw_pass() {
use pdfcraft_optimize::{OptimizeError, Stage, optimize_with_progress};
let build = |streams: usize| {
let mut doc = Document::new_empty();
let mut d = Dict::new();
d.set(b"Filter".to_vec(), Object::name("LZWDecode"));
for _ in 0..streams {
doc.add(Object::Stream(Stream::from_raw(d.clone(), lzw(&lzw_ops(), &Params::default()))));
}
page(&mut doc, &[], "0 0 1 1 re f");
doc
};
let asked = |streams: usize| {
let mut asked = 0;
optimize_with_progress(&mut build(streams), &Settings::default(), &mut |s| {
asked += usize::from(s == Stage::CleanUp);
true
})
.unwrap();
asked
};
let none = asked(0);
assert_eq!(asked(3), none + 3, "once a stream");
let mut left = none + 1;
let mut doc = build(3);
let r = optimize_with_progress(&mut doc, &Settings::default(), &mut |s| {
left -= usize::from(s == Stage::CleanUp);
left > 0
});
assert!(matches!(r, Err(OptimizeError::Cancelled)), "{r:?}");
assert_eq!(left, 0, "nothing runs after it");
}
#[test]
fn a_page_drawn_from_lzw_streams_renders_the_same_after_conversion() {
let png = Params { predictor: 12, colors: 3, columns: 64, early_change: 0, ..Params::default() };
@@ -3534,7 +3581,7 @@ fn indexed_thin_rules_survive_averaging_for_both_palette_orders_and_decodes() {
optimize(&mut gray, &Settings::default()).unwrap();
let gray_after = render(&Document::open(Arc::new(write_full(&gray, &SaveOptions::default()).unwrap())).unwrap());
let gray_stream = stream(&gray, gray_image);
let gray_samples = gray.decoded_image(&gray_stream).unwrap();
let gray_samples = decoded_image(&gray, &gray_stream).unwrap();
assert_eq!((gray_stream.dict.int(b"Width"), gray_stream.dict.int(b"Height")), (Some(600), Some(240)));
assert!(gray_samples.iter().any(|b| *b != 0xFF), "the DeviceGray thin rule survives");
for white_first in [false, true] {
@@ -3567,7 +3614,7 @@ fn indexed_thin_rules_survive_averaging_for_both_palette_orders_and_decodes() {
assert_eq!((s.dict.int(b"Width"), s.dict.int(b"Height")), (Some(600), Some(240)), "white-first={white_first}, inverted={inverted}");
// The rows are 600 samples, whole bytes: flipping each byte flips each sample.
let expected: Vec<u8> = gray_samples.iter().map(|b| if white_first != inverted { !b } else { *b }).collect();
assert_eq!(doc.decoded_image(&s).unwrap(), expected, "white-first={white_first}, inverted={inverted}: the rule is dark");
assert_eq!(decoded_image(&doc, &s).unwrap(), expected, "white-first={white_first}, inverted={inverted}: the rule is dark");
let saved = Document::open(Arc::new(write_full(&doc, &SaveOptions::default()).unwrap())).unwrap();
assert!(render(&saved) == gray_after, "palette white-first={white_first}, inverted={inverted}");
}
@@ -3628,7 +3675,7 @@ fn an_indexed_palette_is_read_by_the_luma_of_its_colours() {
page(&mut doc, &[("Im", image)], "q 144 0 0 57.6 20 600 cm /Im Do Q");
optimize(&mut doc, &Settings::default()).unwrap();
let s = stream(&doc, image);
(s.dict.int(b"Width"), doc.decoded_image(&s).unwrap())
(s.dict.int(b"Width"), decoded_image(&doc, &s).unwrap())
};
let (width, samples) = optimized("Indexed");
assert_eq!(width, Some(600));
@@ -3736,7 +3783,7 @@ fn mixed_text_and_small_halftone_regions_keep_their_tones_with_wide_white_margin
optimize(&mut doc, &Settings::default()).unwrap();
let s = stream(&doc, image);
assert_eq!((s.dict.int(b"Width"), s.dict.int(b"Height")), (Some(w.into()), Some(h.into())));
assert_eq!(doc.decoded_image(&s).unwrap(), data, "screened samples survive, not a solid patch");
assert_eq!(decoded_image(&doc, &s).unwrap(), data, "screened samples survive, not a solid patch");
let saved = Document::open(Arc::new(write_full(&doc, &SaveOptions::default()).unwrap())).unwrap();
assert_eq!(render(&saved), before, "local halftone {side}x{side}");
}
crates/redact/README.md
+4
View File
@@ -37,6 +37,10 @@ clear_marks(&mut doc, None)?; // remove marks without app
and a byte per row, plus 4 KiB (ASCIIHex text broken into lines under Flate is a little over
twice; for a fax image's code, about a byte per pixel), are removed whole;
cleared copies drop `/Alternates` so alternate pixels cannot survive;
A run that would decode more of its images than one redaction may (`pdfcraft_cos::ImageBudget`,
filters, decoders and pixels alike) is refused
and changes nothing: such an image can't be cleared, and removing it whole would take what no mark
covers with it.
- **vectors**: covered paths are removed; partly covered paths and shadings are clipped so
nothing paints inside the areas;
- **form XObjects**: rewritten recursively into new objects (pages sharing the original keep it);
crates/redact/src/image.rs
+23 -17
View File
@@ -2,13 +2,13 @@
//! whose footprint intersects a region is set to zero (no ink for image masks), and the result is
//! written as a new image, Flate, or for a one-bit image of one component Group 4, or JBIG2 when it
//! was JBIG2, whichever is smallest; the original object is untouched (other pages may use it).
//! CCITT fax and JBIG2 images decode through `Document::decoded_image_within`, but for Group 3 on
//! CCITT fax and JBIG2 images decode through `ImageBudget::decode`, but for Group 3 on
//! bytes, which PdfCraft's renderer may draw as other rows, and a JBIG2 page of other rows than the
//! image's. Images whose codec PdfCraft can't decode (DCT, JPX), and those, return `None`, and the
//! caller removes the whole image instead; that is fail-closed.
use pdfcraft_content::{Matrix, overlaps};
use pdfcraft_cos::{Dict, Document, Object, Stream};
use pdfcraft_cos::{Dict, Document, ImageBudget, Object, Stream};
/// Most pixels an image may have to be cleared in place (larger ones are removed).
const MAX_PIXELS: u64 = 64 << 20;
@@ -39,11 +39,11 @@ fn components(doc: &Document, s: &Stream) -> Option<usize> {
/// `Ok(Some(copy))` with the covered pixels cleared, `Ok(None)` when no pixel intersects,
/// `Err(())` when the image can't be cleared (the caller removes it).
#[allow(clippy::result_unit_err)]
pub(crate) fn clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]]) -> Result<Option<Stream>, ()> {
decode_and_clear(doc, s, ctm, rects).ok_or(())
pub(crate) fn clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]], budget: &mut ImageBudget) -> Result<Option<Stream>, ()> {
decode_and_clear(doc, s, ctm, rects, budget).ok_or(())
}
fn decode_and_clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]]) -> Option<Option<Stream>> {
fn decode_and_clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]], budget: &mut ImageBudget) -> Option<Option<Stream>> {
// Separate masks can retain the original image data.
if s.dict.contains(b"SMask") || s.dict.contains(b"Mask") {
return None;
@@ -54,19 +54,25 @@ fn decode_and_clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]]
b"FlateDecode" | b"Fl" | b"LZWDecode" | b"LZW" | b"ASCII85Decode" | b"A85" | b"ASCIIHexDecode" | b"AHx" | b"RunLengthDecode" | b"RL"
)
};
// A CCITT fax or JBIG2 image (its last filter) decodes through `Document::decoded_image_within`,
// A CCITT fax or JBIG2 image (its last filter) decodes through `ImageBudget::decode`,
// after the filters before it, so a scan stored as Group 4 or JBIG2 is cleared like any other
// image.
let fax = |f: &Object| f.as_name().is_some_and(|n| matches!(n, b"CCITTFaxDecode" | b"CCF"));
let jbig2 = |f: &Object| f.as_name() == Some(b"JBIG2Decode".as_slice());
let readable = |f: &Object| f.as_name().is_some_and(supported);
// `/Crypt` was undone when the object was loaded (ISO 32000-2 §7.4.10), and decoding leaves it
// out wherever it stands.
let crypt = |f: &Object| f.as_name() == Some(b"Crypt".as_slice());
let (ok, is_fax, is_jbig2) = match s.dict.get(b"Filter") {
None => (true, false, false),
Some(f @ Object::Name(_)) => (readable(f) || fax(f) || jbig2(f), fax(f), jbig2(f)),
Some(Object::Array(a)) => match a.split_last() {
Some((last, before)) if fax(last) || jbig2(last) => (before.iter().all(readable), fax(last), jbig2(last)),
_ => (a.iter().all(readable), false, false),
},
Some(f @ Object::Name(_)) => (readable(f) || crypt(f) || fax(f) || jbig2(f), fax(f), jbig2(f)),
Some(Object::Array(a)) => {
let chain: Vec<&Object> = a.iter().filter(|f| !crypt(f)).collect();
match chain.split_last() {
Some((last, before)) if fax(last) || jbig2(last) => (before.iter().all(|f| readable(f)), fax(last), jbig2(last)),
_ => (chain.iter().all(|f| readable(f)), false, false),
}
}
_ => (false, false, false),
};
if !ok {
@@ -112,7 +118,7 @@ fn decode_and_clear(doc: &Document, s: &Stream, ctm: &Matrix, rects: &[[f64; 4]]
// alone would skip such a predictor, and the cleared bytes would then miss the covered pixels),
// and refuses a JBIG2 page of another size than the image: the renderer draws the whole page
// into the image, so the rows cleared wouldn't be the ones a mark covers on screen.
let mut data = doc.decoded_image_within(s, allowance).ok()?;
let mut data = budget.decode(doc, s, allowance).ok()?;
if data.len() < bytes {
return None;
}
@@ -205,12 +211,12 @@ fn stored_len(s: &Stream) -> usize {
/// JBIG2 image `s` stored again without its `/JBIG2Globals`: decoded through them, as it is drawn,
/// and written as the same samples, Flate, Group 4 or JBIG2 coded alone, whichever is smallest
/// (`smallest`), its other entries kept. `None` when it can't
/// be decoded for certain (too large, data that doesn't decode, a JBIG2 page of another size than
/// the image, or not a one-bit image of one component). Redaction does this to every image that
/// (`smallest`), its other entries kept. `None` when it can't be decoded for certain (too large,
/// past what is left of `budget`, data that doesn't decode, a JBIG2 page of another size than the
/// image, or not a one-bit image of one component). Redaction does this to every image that
/// shares a redacted image's globals, since their symbol dictionary holds the shapes of marks drawn
/// under the area too.
pub(crate) fn without_globals(doc: &Document, s: &Stream) -> Option<Stream> {
pub(crate) fn without_globals(doc: &Document, s: &Stream, budget: &mut ImageBudget) -> Option<Stream> {
let (w, h) = (s.dict.int(b"Width")?, s.dict.int(b"Height")?);
if w <= 0 || h <= 0 || (w as u64) * (h as u64) > MAX_PIXELS {
return None;
@@ -225,7 +231,7 @@ pub(crate) fn without_globals(doc: &Document, s: &Stream) -> Option<Stream> {
// As for `clear`: the filters before the JBIG2 decoder may give up to three bytes for each one
// of the raster and a byte a row.
let allowance = bytes.checked_add(h)?.checked_mul(3)?.checked_add(4096)?;
let mut data = doc.decoded_image_within(s, allowance).ok()?;
let mut data = budget.decode(doc, s, allowance).ok()?;
if data.len() < bytes {
return None;
}
crates/redact/src/interp.rs
+6 -3
View File
@@ -21,7 +21,7 @@ use std::collections::{HashMap, HashSet};
use std::rc::Rc;
use pdfcraft_content::{Matrix, Op, Pieces, contains, num, overlaps, serialize_ops, string};
use pdfcraft_cos::{CosError, DecodeBudget, Dict, Document, MAX_PAGE_FORMS, ObjRef, Object, Stream};
use pdfcraft_cos::{CosError, DecodeBudget, Dict, Document, ImageBudget, MAX_PAGE_FORMS, ObjRef, Object, Stream};
use crate::{RedactError, Report, image, page_streams, tags};
use pdfcraft_fonts::pdf::Metrics;
@@ -105,6 +105,8 @@ pub(crate) struct Scope<'a> {
pub rects: &'a [[f64; 4]],
pub mode: Mode,
pub report: &'a mut Report,
/// What the images it clears may spend, shared by all the pages of a run.
pub images: &'a mut ImageBudget,
/// Sanitize: remove hidden text (invisible render modes, or wholly outside this page box).
pub hidden_text: Option<[f64; 4]>,
/// Sanitize: optional content groups that are off; content marked with them is removed.
@@ -134,11 +136,12 @@ pub(crate) struct Scope<'a> {
impl<'a> Scope<'a> {
/// A scope whose form walk spends `budget`, what is left of the page's allowance after its
/// content streams were decoded.
pub fn new(rects: &'a [[f64; 4]], mode: Mode, report: &'a mut Report, budget: DecodeBudget) -> Self {
pub fn new(rects: &'a [[f64; 4]], mode: Mode, report: &'a mut Report, budget: DecodeBudget, images: &'a mut ImageBudget) -> Self {
Scope {
rects,
mode,
report,
images,
hidden_text: None,
hidden_layers: HashSet::new(),
layers_enabled: false,
@@ -750,7 +753,7 @@ fn xobject(
out.doomed.push(r);
return Some(None);
}
match image::clear(doc, s, &gs.ctm, scope.rects) {
match image::clear(doc, s, &gs.ctm, scope.rects, &mut *scope.images) {
Ok(None) => None,
Ok(Some(new)) => {
let nr = doc.add(Object::Stream(new));
crates/redact/src/lib.rs
+34 -9
View File
@@ -70,6 +70,10 @@ pub enum RedactError {
/// could not be shown complete.
#[error("the document's objects nest too deeply to remove hidden information completely; the operation was refused")]
TooDeep,
/// Past what one redaction may decode an image can't be cleared, and removing it whole would
/// take what no mark covers with it, unreported.
#[error("the images under these marks are more than one redaction can decode, so nothing was changed; redact fewer pages at a time")]
ImagesPastAllowance,
/// A JBIG2 image shares its `/JBIG2Globals` with a redacted one and can't be stored again
/// without them, so the shapes they hold, those of marks under the area included, would stay.
#[error(
@@ -386,6 +390,12 @@ pub fn apply(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, Reda
/// [`apply`] without the copy: an error can leave `doc` part redacted.
fn apply_to(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, RedactError> {
let images = pdfcraft_cos::ImageBudget::new(doc);
apply_within(doc, pages, images)
}
/// [`apply_to`], its images decoded within `images`.
fn apply_within(doc: &mut Document, pages: Option<&[usize]>, images: pdfcraft_cos::ImageBudget) -> Result<Report, RedactError> {
let (all_marks, unreadable) = marks_checked(doc);
// A mark with no readable area is not applied, and applying the others would look complete:
// refuse before anything changes. Only the pages being applied count.
@@ -428,6 +438,7 @@ fn apply_to(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, Redac
retired_properties: Vec::new(),
widget_owner,
removed_everywhere: HashSet::new(),
images,
};
for &(pi, page_marks) in &groups {
let page = pdfcraft_model::pages(doc).swap_remove(pi);
@@ -435,7 +446,10 @@ fn apply_to(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, Redac
redact_page_content(doc, &mut applying, pi, &page, &rects)?;
remove_page_annotations(doc, &mut applying, &page, &rects, page_marks)?;
}
let Applying { mut report, doomed_fields, doomed_xobjects, retired_properties, mut removed_everywhere, .. } = applying;
if applying.images.refused() > 0 {
return Err(RedactError::ImagesPastAllowance);
}
let Applying { mut report, doomed_fields, doomed_xobjects, retired_properties, mut removed_everywhere, mut images, .. } = applying;
// 2b. Replies and pop-ups of a removed annotation can be listed on another page than their
// parent: their `/IRT` or `/Parent` would keep the removed comment reachable.
@@ -518,7 +532,7 @@ fn apply_to(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, Redac
// A removed image or form must not stay reachable through another resource dictionary (an
// inherited or shared /Resources): a full save keeps whatever is still referenced.
interp::unlink_doomed(doc, &doomed_xobjects, &retired_properties)?;
report.images_restored = restore_shared_jbig2(doc, &doomed_xobjects)?;
report.images_restored = restore_shared_jbig2(doc, &doomed_xobjects, &mut images)?;
// 5. Verify: nothing readable may remain under a region.
let mut residue = 0;
@@ -532,7 +546,9 @@ fn apply_to(doc: &mut Document, pages: Option<&[usize]>) -> Result<Report, Redac
list.iter().zip(data).filter(|(o, _)| !o.as_ref().is_some_and(|r| report.overlays.contains(&r))).map(|(_, d)| d).collect();
let resources = page.dict.get(b"Resources").and_then(|r| doc.resolve(r).as_dict().cloned()).unwrap_or_default();
let mut scratch = Report::default();
let mut scope = Scope::new(&rects, Mode::Verify, &mut scratch, budget);
// Verifying doesn't decode images: what clearing left is read as the page draws it.
let mut images = pdfcraft_cos::ImageBudget::with_work(0);
let mut scope = Scope::new(&rects, Mode::Verify, &mut scratch, budget, &mut images);
residue += process(doc, &mut scope, &original, &resources, pdfcraft_content::Matrix::IDENTITY).residue;
}
if residue > 0 {
@@ -554,8 +570,10 @@ const MAX_JBIG2_WALK: usize = 100_000;
/// and stored again in place without them, as a cleared copy is, and a full save then leaves them
/// out. An image that can't be stored again refuses the redaction (nothing changes, as `apply`
/// works on a copy), but for a removed or replaced one, which is either unlinked or still drawn
/// whole on another page. Returns how many images that weren't removed or replaced were stored again.
fn restore_shared_jbig2(doc: &mut Document, doomed: &[ObjRef]) -> Result<usize, RedactError> {
/// whole on another page. They are decoded within what is left of the run's allowance (`images`);
/// past it the redaction is refused as one clearing too many images is. Returns how many images
/// that weren't removed or replaced were stored again.
fn restore_shared_jbig2(doc: &mut Document, doomed: &[ObjRef], images: &mut pdfcraft_cos::ImageBudget) -> Result<usize, RedactError> {
let mut globals = HashSet::new();
let mut seen = HashSet::new();
let mut stack: Vec<ObjRef> = doomed.to_vec();
@@ -587,19 +605,24 @@ fn restore_shared_jbig2(doc: &mut Document, doomed: &[ObjRef]) -> Result<usize,
.map(|(r, _)| r)
.collect();
let doomed: HashSet<ObjRef> = doomed.iter().copied().collect();
let (mut restored, mut stuck) = (0, 0);
let (mut restored, mut stuck, mut past_allowance) = (0, 0, false);
for r in sharing {
let object = doc.get(r);
let Object::Stream(s) = &*object else { continue };
match image::without_globals(doc, s) {
let refused = images.refused();
match image::without_globals(doc, s, images) {
Some(new) => {
doc.set(r, Object::Stream(new));
restored += usize::from(!doomed.contains(&r));
}
None if doomed.contains(&r) => {}
None if images.refused() > refused => past_allowance = true,
None => stuck += 1,
}
}
if past_allowance {
return Err(RedactError::ImagesPastAllowance);
}
if stuck > 0 {
return Err(RedactError::SharedJbig2Globals(stuck));
}
@@ -615,6 +638,8 @@ struct Applying {
retired_properties: Vec<Object>,
widget_owner: Vec<(ObjRef, ObjRef)>,
removed_everywhere: HashSet<ObjRef>,
/// What clearing images may spend, across the run's pages.
images: pdfcraft_cos::ImageBudget,
}
/// Step 1 for one page: rewrites its content so nothing readable stays under a mark, and cleans the tags it no longer shows.
@@ -625,13 +650,13 @@ fn redact_page_content(
page: &pdfcraft_model::Page,
rects: &[[f64; 4]],
) -> Result<(), RedactError> {
let Applying { report, doomed_xobjects, retired_properties, .. } = applying;
let Applying { report, doomed_xobjects, retired_properties, images, .. } = applying;
// 1. Content.
let mut budget = pdfcraft_cos::DecodeBudget::page();
let (list, data) = page_streams_within(doc, &page.dict, pi, &mut budget)?;
let resources = page.dict.get(b"Resources").and_then(|r| doc.resolve(r).as_dict().cloned()).unwrap_or_default();
let out = {
let mut scope = Scope::new(rects, Mode::Apply, report, budget);
let mut scope = Scope::new(rects, Mode::Apply, report, budget, images);
process(doc, &mut scope, &data, &resources, pdfcraft_content::Matrix::IDENTITY)
};
// Fail closed: the interpreter lays vertical-writing text out horizontally, so it cannot
crates/redact/src/sanitize.rs
+2 -1
View File
@@ -369,7 +369,8 @@ fn content_pass(doc: &mut Document, text: bool, layers: bool, write: bool) -> Re
let resources = page.dict.get(b"Resources").and_then(|r| doc.resolve(r).as_dict().cloned()).unwrap_or_default();
let mut report = Report::default();
let mode = if write { Mode::Apply } else { Mode::Verify };
let mut scope = Scope::new(&[], mode, &mut report, budget);
let mut images = pdfcraft_cos::ImageBudget::new(doc);
let mut scope = Scope::new(&[], mode, &mut report, budget, &mut images);
if text {
let c = page.crop(doc);
scope.hidden_text = Some([c[0].min(c[2]), c[1].min(c[3]), c[0].max(c[2]), c[1].max(c[3])]);
crates/redact/src/tests.rs
+86 -7
View File
@@ -52,7 +52,8 @@ fn under(doc: &mut Document, page: usize, rects: &[[f64; 4]]) -> usize {
let (_, data) = page_streams_within(doc, &p.dict, page, &mut budget).unwrap();
let res = p.dict.get(b"Resources").and_then(|r| doc.resolve(r).as_dict().cloned()).unwrap_or_default();
let mut rep = Report::default();
let mut scope = Scope::new(rects, Mode::Verify, &mut rep, budget);
let mut images = pdfcraft_cos::ImageBudget::with_work(0);
let mut scope = Scope::new(rects, Mode::Verify, &mut rep, budget, &mut images);
process(doc, &mut scope, &data, &res, pdfcraft_content::Matrix::IDENTITY).residue
}
@@ -519,10 +520,16 @@ fn two_pages_one_resource_dictionary(inherited: bool, again: bool) -> Document {
])
}
/// What image `s` decodes to, its fax or JBIG2 data included, with no limit on the work: the
/// samples an operation would read through its `ImageBudget`.
fn decoded_image(doc: &Document, s: &pdfcraft_cos::Stream) -> Result<Vec<u8>, pdfcraft_cos::CosError> {
pdfcraft_cos::ImageBudget::with_work(u64::MAX).decode(doc, s, pdfcraft_cos::MAX_DECODED)
}
/// The samples of the image a page's (possibly inherited) resources name `name`, decoded through
/// its image codec too (`Document::decoded_image`).
/// its image codec too ([`decoded_image`]).
fn named_image(doc: &Document, page: usize, name: &[u8]) -> Option<Vec<u8>> {
named_stream(doc, page, name).and_then(|s| doc.decoded_image(&s).ok())
named_stream(doc, page, name).and_then(|s| decoded_image(doc, &s).ok())
}
/// The image a page's (possibly inherited) resources name `name`.
@@ -707,7 +714,8 @@ fn group_3_on_bytes_is_removed_whole_and_other_fax_images_cleared() {
(-1, true, "<< /K -1 /Columns 16 /EncodedByteAlign true >>", vec![0x80, 0x80], vec![], false),
] {
let p = pdfcraft_filters::CcittParams { k, columns: 16, encoded_byte_align: on_bytes, ..Default::default() };
assert_eq!(pdfcraft_filters::decode_ccitt(&data, &p, 2).unwrap(), [0xFF; 4], "{parms}: the picture is white");
let mut work = u64::MAX;
assert_eq!(pdfcraft_filters::decode_ccitt_within(&data, &p, 2, &mut work).unwrap(), [0xFF; 4], "{parms}: the picture is white");
let dict = format!("{image} /DecodeParms {parms}");
let mut objs = vec![stream(&dict, data)];
objs.extend(extra);
@@ -769,7 +777,7 @@ fn no_stream_holds_the_pixels_under_a_mark_on_a_jbig2_scan() {
}
let holds = |doc: &Document, pixels: &[u8]| {
doc.scan_objects().any(|(_, object)| match &*object {
Object::Stream(s) => doc.decoded_image(s).is_ok_and(|d| d.windows(pixels.len()).any(|w| w == pixels)),
Object::Stream(s) => decoded_image(doc, s).is_ok_and(|d| d.windows(pixels.len()).any(|w| w == pixels)),
_ => false,
})
};
@@ -806,6 +814,61 @@ fn jbig2_globals(secret: &[u8]) -> Vec<u8> {
g
}
/// An image the run's allowance can't decode can't be cleared. Removed whole, like an image that
/// can't be read, it would take what no mark covers with it without a word, so the run stops there,
/// and [`apply`] leaves the document as it was.
#[test]
fn a_jbig2_image_past_the_runs_allowance_stops_the_run() {
let mut dict = Dict::new();
dict.set(b"Width".to_vec(), Object::Int(16));
dict.set(b"Height".to_vec(), Object::Int(2));
dict.set(b"BitsPerComponent".to_vec(), Object::Int(1));
dict.set(b"ColorSpace".to_vec(), Object::name("DeviceGray"));
dict.set(b"Filter".to_vec(), Object::name("JBIG2Decode"));
let image = Stream::from_raw(dict, pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap());
let doc = Document::new_empty();
let half = [[0.0, 0.0, 0.5, 1.0]];
let clear = |work| crate::image::clear(&doc, &image, &pdfcraft_content::Matrix::IDENTITY, &half, &mut pdfcraft_cos::ImageBudget::with_work(work));
assert!(matches!(clear(u64::MAX), Ok(Some(_))), "cleared in a copy");
assert!(matches!(clear(0), Err(())), "it can't be cleared");
let coded = pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap();
let dict = "/Type /XObject /Subtype /Image /Width 16 /Height 2 /BitsPerComponent 1 /ColorSpace /DeviceGray /Filter /JBIG2Decode";
let mut doc = one_page(b"q 160 0 0 20 0 0 cm /Im Do Q", "/XObject << /Im 7 0 R >>", vec![stream(dict, coded)]);
mark(&mut doc, 0, &[[0.0, 12.0, 75.0, 18.0]], "");
let refused = crate::apply_within(&mut doc, None, pdfcraft_cos::ImageBudget::with_work(0));
assert!(matches!(refused, Err(crate::RedactError::ImagesPastAllowance)), "{refused:?}");
let mut doc = one_page(
b"q 160 0 0 20 0 0 cm /Im Do Q",
"/XObject << /Im 7 0 R >>",
vec![stream(dict, pdfcraft_filters::encode_jbig2_generic(&[0xFF; 4], 16, 2).unwrap())],
);
mark(&mut doc, 0, &[[0.0, 12.0, 75.0, 18.0]], "");
assert!(crate::apply_within(&mut doc, None, pdfcraft_cos::ImageBudget::with_work(u64::MAX)).is_ok());
}
/// A Flate image past the run's allowance stops the run as a JBIG2 image does: removed whole, it
/// would take what no mark covers with it, unreported. A JPEG, which clearing doesn't read, is
/// removed whole as it always was, and spends nothing.
#[test]
fn a_flate_image_past_the_runs_allowance_stops_the_run() {
let dict = "/Type /XObject /Subtype /Image /Width 2 /Height 2 /BitsPerComponent 8 /ColorSpace /DeviceGray /Filter /FlateDecode";
let page = || one_page(b"q 160 0 0 20 0 0 cm /Im Do Q", "/XObject << /Im 7 0 R >>", vec![stream(dict, pdfcraft_filters::encode_flate(&[9; 4]))]);
let mut doc = page();
mark(&mut doc, 0, &[[0.0, 12.0, 75.0, 18.0]], "");
let refused = crate::apply_within(&mut doc, None, pdfcraft_cos::ImageBudget::with_work(0));
assert!(matches!(refused, Err(crate::RedactError::ImagesPastAllowance)), "{refused:?}");
let mut doc = page();
mark(&mut doc, 0, &[[0.0, 12.0, 75.0, 18.0]], "");
assert!(crate::apply_within(&mut doc, None, pdfcraft_cos::ImageBudget::with_work(u64::MAX)).is_ok());
let dict = "/Type /XObject /Subtype /Image /Width 2 /Height 2 /BitsPerComponent 8 /ColorSpace /DeviceGray /Filter /DCTDecode";
let mut doc = one_page(b"q 160 0 0 20 0 0 cm /Im Do Q", "/XObject << /Im 7 0 R >>", vec![stream(dict, b"not read")]);
mark(&mut doc, 0, &[[0.0, 12.0, 75.0, 18.0]], "");
let report = crate::apply_within(&mut doc, None, pdfcraft_cos::ImageBudget::with_work(0)).unwrap();
assert_eq!(report.images_removed, 1);
}
/// A JBIG2 image (a scan the optimizer stored as JBIG2) has its covered pixels cleared in a copy,
/// like a fax image, instead of being removed whole: with `/JBIG2Globals`, its `/DecodeParms` by
/// reference, behind another filter with its globals at its own place, or as a mask, its decode
@@ -830,6 +893,13 @@ fn a_jbig2_image_has_its_covered_pixels_cleared_instead_of_being_removed() {
(
format!("{image} /ColorSpace /DeviceGray /Filter [/ASCIIHexDecode /JBIG2Decode] /DecodeParms [null << /JBIG2Globals 8 0 R >>]"),
hex,
vec![globals.clone()],
[0x00, 0xFF, 0xFF, 0xFF],
),
// `/Crypt` was undone on load, and takes a place in the parameters.
(
format!("{image} /ColorSpace /DeviceGray /Filter [/Crypt /JBIG2Decode] /DecodeParms [null << /JBIG2Globals 8 0 R >>]"),
white.clone(),
vec![globals],
[0x00, 0xFF, 0xFF, 0xFF],
),
@@ -1199,7 +1269,7 @@ fn globals_a_redacted_jbig2_image_shares_with_another_page_leave_the_file() {
let res = doc.resolve(p.dict.get(b"Resources")?).as_dict().cloned()?;
let xo = doc.resolve(res.get(b"XObject")?).as_dict().cloned()?;
match &*doc.get(xo.get(name)?.as_ref()?) {
Object::Stream(s) => doc.decoded_image(s).ok(),
Object::Stream(s) => decoded_image(doc, s).ok(),
_ => None,
}
};
@@ -1240,6 +1310,15 @@ fn globals_a_redacted_jbig2_image_shares_with_another_page_leave_the_file() {
let mut doc = build(&image.replace(&format!("/Height {h}"), &format!("/Height {}", h + 1)));
mark(&mut doc, 0, &[[0.25, 113.75, 49.75, 120.0]], "");
assert!(matches!(apply(&mut doc, None), Err(RedactError::SharedJbig2Globals(1))));
// Storing the second page's image again decodes it, within what is left of the run's
// allowance: with nothing left the redaction is refused as one past the allowance, not
// reported as an image that can't be read.
let mut doc = build(&image);
let first_image = [ObjRef::new(7, 0)];
let refused = crate::restore_shared_jbig2(&mut doc, &first_image, &mut pdfcraft_cos::ImageBudget::with_work(0));
assert!(matches!(refused, Err(RedactError::ImagesPastAllowance)), "{refused:?}");
let restored = crate::restore_shared_jbig2(&mut doc, &first_image, &mut pdfcraft_cos::ImageBudget::with_work(u64::MAX));
assert_eq!(restored.ok(), Some(1));
}
/// Apply Redactions gives the page its own pruned /Resources, but the dictionary it drew through
@@ -2884,7 +2963,7 @@ fn assert_no_secret(doc: &Document, secret: &[u8]) {
assert!(!bytes.windows(secret.len()).any(|w| w == secret), "secret remains in object {r:?}");
if let Object::Stream(s) = &*object {
// Image codecs, a cleared scan coded as Group 4 or JBIG2, decode through the document.
let data = s.decoded().or_else(|_| saved.decoded_image(s)).unwrap();
let data = s.decoded().or_else(|_| decoded_image(&saved, s)).unwrap();
assert!(!data.windows(secret.len()).any(|w| w == secret), "secret remains in stream {r:?}");
}
}
crates/render/src/raster.rs
+8 -2
View File
@@ -83,6 +83,10 @@ pub struct RenderConfig {
/// viewer and preview rasters: exports, printing and OCR render without it, so nothing drawn with an
/// installed font is saved or printed (AGENTS.md §1.2, §1.4).
pub system_fonts: bool,
/// Steps of JBIG2 decoding a render may take across the page's images, instead of the usual
/// `hayro_syntax::MAX_PAGE_JBIG2_WORK`. Those past it aren't drawn, like an image that can't
/// be decoded.
pub jbig2_work: Option<u64>,
}
impl RenderConfig {
@@ -100,7 +104,8 @@ impl RenderConfig {
/// Passwords belong to parser construction; request-specific preview overrides are applied
/// by `render_page`, which also preserves the caller's warning and cancellation hooks.
fn settings(&self) -> InterpreterSettings {
let standard = InterpreterSettings::default().font_resolver;
let defaults = InterpreterSettings::default();
let standard = defaults.font_resolver.clone();
let system_fonts = self.system_fonts;
InterpreterSettings {
ocg_overrides: self.layers.clone(),
@@ -108,7 +113,8 @@ impl RenderConfig {
font_resolver: Arc::new(move |query| {
japanese_fallback(query).or_else(|| system_fonts.then(|| system_fallback(query)).flatten()).or_else(|| standard(query))
}),
..InterpreterSettings::default()
jbig2_work: self.jbig2_work.unwrap_or(defaults.jbig2_work),
..defaults
}
}
}
crates/render/src/raster/tests.rs
+215 -1
View File
@@ -4058,7 +4058,10 @@ fn jbig2_image_draws_with(stream: &[u8], space: &str, entries: &str, objects: &[
pdf.extend_from_slice(b"\nendstream endobj\n");
pdf.extend_from_slice(objects);
pdf.extend_from_slice(b"trailer << /Root 1 0 R >>\n%%EOF\n");
let mut r = PageRenderer::new(Arc::new(pdf), RenderConfig::default());
// What one image may take, which the page's allowance, tested on its own, would otherwise reach
// first: it is the same size, and the page itself comes out of it.
let config = RenderConfig { jbig2_work: Some(u64::MAX), ..RenderConfig::default() };
let mut r = PageRenderer::new(Arc::new(pdf), config);
let page = r.render(RenderRequest { page: 0, kind: RequestKind::Pixels, tile: None, scale: 1.0, tag: 0, ..Default::default() });
assert!(page.error.is_none(), "{:?}", page.error);
assert_eq!(&page.rgba[((38 * 40 + 1) * 4)..][..4], &[255, 0, 0, 255], "the rest of the page draws");
@@ -4070,6 +4073,217 @@ fn jbig2_image_draws_with(stream: &[u8], space: &str, entries: &str, objects: &[
}
}
/// The JBIG2 images of a page share one allowance, down to the step, and the next page starts with
/// its own. The images differ by an extension segment: one image decoded again isn't decoded twice.
#[test]
fn the_jbig2_images_of_a_page_share_one_allowance() {
let image = |number: u8, id: u8| {
let mut stream = black_jbig2_page();
jbig2_segment(&mut stream, 1, 62, &[], &[0, 0, 0, 0, id]);
let mut object = format!(
"{number} 0 obj << /Type /XObject /Subtype /Image /Width 16 /Height 2 /BitsPerComponent 1 /ColorSpace /DeviceGray /Filter /JBIG2Decode /Length {} >> stream\n",
stream.len()
)
.into_bytes();
object.extend_from_slice(&stream);
object.extend_from_slice(b"\nendstream endobj\n");
object
};
let page = |number: u8, content: u8, xobjects: &str| {
format!("{number} 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 40 40] /Contents {content} 0 R /Resources << /XObject << {xobjects} >> >> >> endobj\n").into_bytes()
};
let content = |number: u8, text: &str| format!("{number} 0 obj << /Length {} >> stream\n{text}\nendstream endobj\n", text.len()).into_bytes();
let mut pdf =
b"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n2 0 obj << /Type /Pages /Kids [3 0 R 4 0 R] /Count 2 >> endobj\n".to_vec();
pdf.extend(page(3, 5, "/A 7 0 R /B 8 0 R"));
pdf.extend(page(4, 6, "/C 9 0 R"));
pdf.extend(content(5, "q 20 0 0 40 0 0 cm /A Do Q q 20 0 0 40 20 0 cm /B Do Q"));
pdf.extend(content(6, "q 20 0 0 40 0 0 cm /C Do Q"));
pdf.extend(image(7, 1));
pdf.extend(image(8, 2));
pdf.extend(image(9, 3));
pdf.extend_from_slice(b"trailer << /Root 1 0 R >>\n%%EOF\n");
let pdf = Arc::new(pdf);
let black = |page: &RenderedPage, x: usize| match &page.rgba[((20 * 40 + x) * 4)..][..4] {
[0, 0, 0, 255] => true,
[255, 255, 255, 255] => false,
other => panic!("neither an image nor the page: {other:?}"),
};
let pages = |work: u64| {
let mut renderer = PageRenderer::new(pdf.clone(), RenderConfig { jbig2_work: Some(work), ..RenderConfig::default() });
[0, 1].map(|page| renderer.render(RenderRequest { page, kind: RequestKind::Pixels, tile: None, scale: 1.0, tag: 0, ..Default::default() }))
};
let drawn = |work: u64| {
let [first, _] = pages(work);
assert!(first.error.is_none(), "{:?}", first.error);
(black(&first, 10), black(&first, 30))
};
let one = (1..1_000).find(|work| drawn(*work).0).expect("a small image is cheap");
assert_eq!(drawn(one - 1), (false, false), "a step short of the first");
assert_eq!(drawn(one), (true, false), "enough for one");
assert_eq!(drawn(2 * one - 1), (true, false), "a step short of both");
assert_eq!(drawn(2 * one), (true, true), "enough for both");
let [first, second] = pages(one);
assert!(first.warnings.contains(&RenderWarning::ContentTruncated), "the page says it left something out: {:?}", first.warnings);
assert!(pages(2 * one)[0].warnings.is_empty(), "and only then");
assert!(black(&second, 10), "the next page has its own allowance, though this one spent all of it");
let usual = PageRenderer::new(pdf.clone(), RenderConfig::default()).render(RenderRequest {
page: 0,
kind: RequestKind::Pixels,
tile: None,
scale: 1.0,
tag: 0,
..Default::default()
});
assert!(black(&usual, 10) && black(&usual, 30), "the usual allowance is far above a page like this");
}
/// Without its mask a picture would show what the mask hides, so a JBIG2 mask past the page's
/// allowance takes the picture with it, and the page says so. A mask that can't be read for any
/// other reason leaves the picture, as it did.
#[test]
fn a_picture_whose_jbig2_mask_is_past_the_pages_allowance_is_not_drawn() {
let draw = |entry: &str, filter: &str, mask: &[u8], work: u64| {
let content = b"q 20 0 0 20 5 5 cm /Im1 Do Q";
let mut pdf = b"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 40 40] /Contents 4 0 R /Resources << /XObject << /Im1 5 0 R >> >> >> endobj\n".to_vec();
pdf.extend_from_slice(format!("4 0 obj << /Length {} >> stream\n", content.len()).as_bytes());
pdf.extend_from_slice(content);
pdf.extend_from_slice(b"\nendstream endobj\n");
pdf.extend_from_slice(
format!("5 0 obj << /Type /XObject /Subtype /Image /Width 1 /Height 1 /BitsPerComponent 8 /ColorSpace /DeviceGray {entry} 6 0 R /Length 1 >> stream\n\0\nendstream endobj\n").as_bytes(),
);
let kind = if entry == "/Mask" { "/ImageMask true" } else { "/ColorSpace /DeviceGray" };
pdf.extend_from_slice(
format!(
"6 0 obj << /Type /XObject /Subtype /Image /Width 16 /Height 2 /BitsPerComponent 1 {kind} /Filter {filter} /Length {} >> stream\n",
mask.len()
)
.as_bytes(),
);
pdf.extend_from_slice(mask);
pdf.extend_from_slice(b"\nendstream endobj\ntrailer << /Root 1 0 R >>\n%%EOF\n");
let config = RenderConfig { jbig2_work: Some(work), ..RenderConfig::default() };
let page = PageRenderer::new(Arc::new(pdf), config).render(RenderRequest {
page: 0,
kind: RequestKind::Pixels,
tile: None,
scale: 1.0,
tag: 0,
..Default::default()
});
assert!(page.error.is_none(), "{:?}", page.error);
let shown = match &page.rgba[((25 * 40 + 15) * 4)..][..4] {
[0, 0, 0, 255] => true,
[255, 255, 255, 255] => false,
other => panic!("neither the picture nor the page: {other:?}"),
};
(shown, page.warnings.contains(&RenderWarning::ContentTruncated))
};
let mask = black_jbig2_page();
let jbig2 = "/JBIG2Decode";
assert_eq!(draw("/SMask", jbig2, &mask, u64::MAX), (false, false), "the mask hides all of it");
assert_eq!(draw("/SMask", jbig2, &mask, 0), (false, true), "past the allowance a soft mask hides it too");
assert_eq!(draw("/Mask", jbig2, &mask, 0), (false, true), "and so does a stencil mask");
assert_eq!(draw("/SMask", jbig2, b"no JBIG2 here", 1 << 20), (true, false), "any other bad mask is no mask");
let other = draw("/SMask", "/DCTDecode", b"no JPEG here", 0);
assert_eq!(other, (true, false), "and a page with nothing left says nothing of a mask that isn't JBIG2");
}
/// A page's allowance ends with it, so what decodes on its thread afterwards has no page to run out
/// on, and a `Context` used again starts afresh.
#[test]
fn a_pages_jbig2_allowance_ends_with_the_page() {
use hayro::hayro_interpret::font::Glyph;
use hayro::hayro_interpret::hayro_syntax::Pdf;
use hayro::hayro_interpret::{
BlendMode, ClipPath, Context, Device, GlyphDrawMode, Image, InterpreterCache, InterpreterSettings, Paint, PathDrawMode, SoftMask,
TransformExt, interpret_page,
};
use hayro_syntax::page_jbig2_work_spent;
use kurbo::{Affine, BezPath, Rect};
/// Notes at each path whether the page has run out.
struct Spent(Vec<bool>);
impl<'a> Device<'a> for Spent {
fn set_soft_mask(&mut self, _: Option<SoftMask<'a>>) {}
fn set_blend_mode(&mut self, _: BlendMode) {}
fn draw_path(&mut self, _: &BezPath, _: Affine, _: &Paint<'a>, _: &PathDrawMode) {
self.0.push(page_jbig2_work_spent());
}
fn push_clip_path(&mut self, _: &ClipPath) {}
fn push_transparency_group(&mut self, _: f32, _: Option<SoftMask<'a>>, _: BlendMode) {}
fn draw_glyph(&mut self, _: &Glyph<'a>, _: Affine, _: Affine, _: &Paint<'a>, _: &GlyphDrawMode) {}
fn draw_image(&mut self, _: Image<'a, '_>, _: Affine) {}
fn pop_clip_path(&mut self) {}
fn pop_transparency_group(&mut self) {}
}
let content = "0 0 4 4 re f";
let bytes = format!(
"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 40 40] /Contents 4 0 R >> endobj\n4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\ntrailer << /Root 1 0 R >>\n%%EOF\n",
content.len()
);
let pdf = Pdf::new(Arc::new(bytes.into_bytes())).expect("parses");
let page = &pdf.pages()[0];
let cache = InterpreterCache::new();
let settings = InterpreterSettings { jbig2_work: 0, ..InterpreterSettings::default() };
let mut ctx = Context::new(page.initial_transform(true).to_kurbo(), Rect::new(0.0, 0.0, 40.0, 40.0), &cache, page.xref(), settings);
let mut device = Spent(Vec::new());
for _ in 0..2 {
interpret_page(page, &mut ctx, &mut device);
assert!(!page_jbig2_work_spent(), "no page is being drawn");
}
assert_eq!(device.0, [true, true], "each time, the page has what its settings give it");
}
/// An image that can't be decoded has still cost its reading, which a page repeats for each time
/// it draws one, since such an image isn't kept. So what it reads comes out of the page's allowance,
/// and an image after it has less. Allowed exactly what the good one takes, it isn't drawn behind
/// the bad one, and is drawn alone.
#[test]
fn an_image_that_fails_spends_the_pages_allowance_too() {
let image = |number: u8, stream: &[u8]| {
let mut object = format!(
"{number} 0 obj << /Type /XObject /Subtype /Image /Width 16 /Height 2 /BitsPerComponent 1 /ColorSpace /DeviceGray /Filter /JBIG2Decode /Length {} >> stream\n",
stream.len()
)
.into_bytes();
object.extend_from_slice(stream);
object.extend_from_slice(b"\nendstream endobj\n");
object
};
let pdf = |content: &str| {
let mut pdf = format!(
"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\
3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 40 40] /Contents 4 0 R /Resources << /XObject << /Bad 5 0 R /Good 6 0 R >> >> >> endobj\n\
4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n",
content.len()
)
.into_bytes();
pdf.extend(image(5, &[0xFF; 64]));
pdf.extend(image(6, &black_jbig2_page()));
pdf.extend_from_slice(b"trailer << /Root 1 0 R >>\n%%EOF\n");
Arc::new(pdf)
};
let good = "q 40 0 0 40 0 0 cm /Good Do Q";
let drawn = |content: &str, work: u64| {
let config = RenderConfig { jbig2_work: Some(work), ..RenderConfig::default() };
let page = PageRenderer::new(pdf(content), config).render(RenderRequest {
page: 0,
kind: RequestKind::Pixels,
tile: None,
scale: 1.0,
tag: 0,
..Default::default()
});
assert!(page.error.is_none(), "{:?}", page.error);
page.rgba[((20 * 40 + 20) * 4)..][..4] == [0, 0, 0, 255]
};
let cost = (1..2_000).find(|work| drawn(good, *work)).expect("a small image is cheap");
assert!(!drawn(&format!("q 1 0 0 1 0 0 cm /Bad Do Q {good}"), cost), "the bad one spent some of it");
assert!(drawn(&format!("q 1 0 0 1 0 0 cm /Bad Do Q {good}"), cost + 4 * 64), "and what it spent is its reading");
}
/// Bits written from the top of each byte, as JBIG2's Huffman-coded data is.
#[derive(Default)]
struct Bits {
docs/architecture.md
+4 -3
View File
@@ -72,9 +72,10 @@ The PDF object graph *is* the model (`crates/engine/README.md` §The editing mod
- PDF in: any revision chain, xref tables or streams, hybrid files, damaged files (repaired and
logged), all standard-security revisions. Image codecs (DCT, JPX) decode through the bootstrap
renderer; CCITT and JBIG2 decode through `pdfcraft-filters` (`decode_ccitt`, `decode_jbig2`,
behind `Document::decoded_image`) for export, redaction and the optimizer, and through the
renderer for display. Export ▸ Images decodes JPEG itself.
renderer; CCITT and JBIG2 decode through `pdfcraft-filters` (`decode_ccitt_within`,
`decode_jbig2_within`, behind `ImageBudget::decode`, the only way to decode an image outside
the renderer, which shares one allowance across an operation's images) for export, redaction
and the optimizer, and through the renderer for display. Export ▸ Images decodes JPEG itself.
- PDF out: incremental or full saves, object streams, re-encryption, resource de-duplication when
combining.
- Other formats: images and text in (`create`); PNG/JPEG/TIFF, text, RTF, Word and HTML out
parity/acrobat-features.toml
+3 -3
View File
@@ -7214,7 +7214,7 @@ status = "shipped"
commands = ["export.all_images"]
tools = ["doc_export_all_images"]
evidence = ["crates/create/src/tests.rs::ccitt_fax_images_export_as_their_picture", "crates/create/src/tests.rs::group_3_ccitt_masks_and_parameters_by_reference_export_as_their_picture", "crates/create/src/tests.rs::images_export_as_jpeg_unchanged_and_others_as_png", "crates/create/src/tests.rs::indexed_and_one_bit_images_decode", "crates/create/src/tests.rs::jpeg_images_with_a_soft_mask_export_as_png_with_alpha", "crates/create/src/tests.rs::soft_masks_of_another_size_and_stencil_and_colour_key_masks_become_alpha", "crates/create/src/tests.rs::soft_mask_decode_ranges_become_exported_alpha", "crates/create/src/tests.rs::soft_mask_decode_maps_packed_and_jpeg_masks_before_resizing", "crates/automation/tests/it/automation.rs::exporting_all_images_through_tools", "crates/ui-egui/tests/it/export.rs::export_dialogs_write_images_and_text", "crates/create/src/tests.rs::lossless_decode_maps_each_component_before_colour_conversion", "crates/create/src/tests.rs::lossless_indexed_decode_rounds_in_index_units_then_clamps", "crates/create/src/tests.rs::jbig2_images_export_as_their_picture", "crates/create/src/tests.rs::jbig2_globals_are_decoded_within_a_limit_of_their_own", "crates/automation/tests/it/automation.rs::doc_optimize_stores_black_and_white_images_as_jbig2", "crates/create/src/tests.rs::symbol_coded_jbig2_images_export_as_their_picture", "crates/create/src/tests.rs::jbig2_images_behind_another_filter_export_as_their_picture"]
notes = "each image once (page resources and nested forms); JPEG unchanged unless it has transparency, others PNG (gray/RGB/CMYK/ICC/Indexed, 1–16 bpc, stencil masks); transparency as PNG alpha: /SMask (scaled to the image size, /Matte un-blended), stencil /Mask and colour-key /Mask, JPEG images included (#645); JPX masks are left out (image exported opaque); JPX images and separations reported as skipped; minimum size filter. Lossless images apply per-component Decode; Indexed values round and clip before palette lookup. Missing, null or malformed Decode arrays keep the default mapping. Soft-mask Decode intervals preserve opacity before resizing, including indirect arrays and endpoints. Still limited: 16-bit samples use high bytes; general JPEG image Decode, ICC transforms, and calibrated colors remain unsupported. CCITT fax images (Group 3 and 4, with their /DecodeParms, behind other filters too) export as their picture, masks included; byte-aligned Group 3 is read by the end-of-line codes when the data starts with one and on bytes otherwise, and one cut short is reported. Still limited: byte-aligned Group 3 with codes but none before the first row (not as T.4 sends them) is reported, or at /K 0 may export wrong. JBIG2 images (with the /JBIG2Globals of their own /DecodeParms entry, within 32 MiB, behind other filters too) export as their picture, masks included; the JBIG2 page has to be the image's size (the renderer draws the whole page into the image)."
notes = "each image once (page resources and nested forms), up to what one export may decode and hold of files (2 GiB or twice the document), images past that left out with the reason; JPEG unchanged unless it has transparency, others PNG (gray/RGB/CMYK/ICC/Indexed, 1–16 bpc, stencil masks); transparency as PNG alpha: /SMask (scaled to the image size, /Matte un-blended), stencil /Mask and colour-key /Mask, JPEG images included (#645); JPX masks are left out (image exported opaque); JPX images and separations reported as skipped; minimum size filter. Lossless images apply per-component Decode; Indexed values round and clip before palette lookup. Missing, null or malformed Decode arrays keep the default mapping. Soft-mask Decode intervals preserve opacity before resizing, including indirect arrays and endpoints. Still limited: 16-bit samples use high bytes; general JPEG image Decode, ICC transforms, and calibrated colors remain unsupported. CCITT fax images (Group 3 and 4, with their /DecodeParms, behind other filters too) export as their picture, masks included; byte-aligned Group 3 is read by the end-of-line codes when the data starts with one and on bytes otherwise, and one cut short is reported. Still limited: byte-aligned Group 3 with codes but none before the first row (not as T.4 sends them) is reported, or at /K 0 may export wrong. JBIG2 images (with the /JBIG2Globals of their own /DecodeParms entry, within 32 MiB, behind other filters too) export as their picture, masks included; the JBIG2 page has to be the image's size (the renderer draws the whole page into the image)."
[[feature]]
id = "create.export-html"
@@ -7433,8 +7433,8 @@ milestone = "M11"
status = "partial"
commands = ["optimize.advanced", "optimize.reduce"]
tools = ["doc_optimize", "doc_reduce"]
evidence = ["crates/optimize/tests/optimize.rs::eight_bit_black_and_white_images_become_one_bit", "crates/filters/src/ccitt.rs::bitmaps_round_trip", "crates/filters/src/ccitt.rs::a_page_of_text_like_marks_round_trips_and_shrinks", "crates/filters/src/ccitt.rs::what_it_cannot_encode_is_an_error", "crates/optimize/tests/optimize.rs::black_and_white_images_become_ccitt_group_4", "crates/optimize/tests/optimize.rs::group_3_fax_images_become_group_4", "crates/optimize/tests/optimize.rs::rgb_and_two_colour_indexed_black_and_white_images_become_one_bit", "crates/optimize/tests/optimize.rs::black_and_white_rgb_used_beyond_the_page_content_read_stays_rgb", "crates/optimize/tests/optimize.rs::black_and_white_rgb_drawn_only_where_the_optimizer_reads_becomes_grey", "crates/optimize/tests/optimize.rs::black_and_white_rgb_in_resources_its_page_tree_and_a_form_share_becomes_grey", "crates/automation/tests/it/automation.rs::doc_optimize_stores_black_and_white_images_as_ccitt_group_4", "crates/ui-egui/tests/it/create.rs::the_pdf_optimizer_dialog_saves_an_optimized_copy", "crates/filters/src/mq.rs::the_annex_h_test_sequence_encodes_to_its_bytes", "crates/filters/src/jbig2.rs::bitmaps_round_trip", "crates/filters/src/jbig2.rs::a_page_of_text_like_marks_round_trips_and_shrinks", "crates/optimize/tests/optimize.rs::black_and_white_images_become_jbig2", "crates/automation/tests/it/automation.rs::doc_optimize_stores_black_and_white_images_as_jbig2", "crates/filters/src/jbig2/symbols.rs::repeated_marks_are_stored_once", "crates/filters/src/jbig2/symbols.rs::scattered_marks_round_trip", "crates/optimize/tests/optimize.rs::rendered_text_becomes_jbig2_symbols", "crates/optimize/tests/optimize.rs::jbig2_scans_are_averaged_and_stored_again"]
notes = "CCITT Group 4, lossless: PDF Optimizer ▸ Images ▸ Monochrome Images (on by default, so Reduce File Size does it too; doc_optimize mono) stores a 1-bit image of one colour component (grey, indexed, a separation) or an image mask, kept with no filter or general-purpose ones, as Group 4 when that is smaller as stored; the samples decode to the same bits, so colour space, decode array and masks stay. Packed to one bit first, then stored the same way (colour-key masks aside): eight-bit grey of only black and white, eight-bit DeviceRGB of only black and white (as DeviceGray; calibrated RGB, RGB with a profile, RGB drawn or listed where /DefaultGray or /DefaultRGB is in force or referred to from anywhere but the page and form resources read (an annotation's appearance or a pattern drawing it, directly or through resources or a form shared with a page; another image's /Alternates), and RGB whose soft mask has a /Matte aside), and an indexed image of 2, 4 or 8 bits that uses only palette entries 0 and 1 (keeping its palette), unless the grey or colour settings make it smaller still, as they can a dithered scan downsampled to JPEG, its soft mask counted. Group 4 wins on scans; on clean computer-drawn 1-bit art Flate can be smaller, and the image then stays. Group 3 fax images are stored again as Group 4 when that is smaller (damaged Group 3 with end-of-line codes as the decoder recovers it, as it reports no damage there; byte-aligned Group 3 with end-of-line codes but none before the first row, not as T.4 sends them, as PdfCraft draws it, which at /K 0 may be other rows than were coded). CCITT fax images are decoded to be downsampled (optimize.downsample-images), and a resampled image is stored as Group 4 or 1-bit Flate, whichever is smaller. Lossless JBIG2 with bounded exact-copy symbol coding is available too, with image export; lossy JBIG2 remains unsupported. JBIG2 images are read too: averaged like any (under Retain stored as JBIG2 again), and, not resampled, stored again as the compression asks when that is smaller (as JBIG2 symbols where its own coding is a generic region, say); a symbol-coded scan, averaged, loses its exact repeats and is often larger, and then stays as it is; one whose JBIG2 page isn't its size is left alone. Not yet: inline images, soft masks and /Mask images, Group 4 for rows wider than 32,767 samples."
evidence = ["crates/optimize/src/images.rs::flate_images_past_the_runs_allowance_stay_as_they_are", "crates/optimize/tests/optimize.rs::eight_bit_black_and_white_images_become_one_bit", "crates/filters/src/ccitt.rs::bitmaps_round_trip", "crates/filters/src/ccitt.rs::a_page_of_text_like_marks_round_trips_and_shrinks", "crates/filters/src/ccitt.rs::what_it_cannot_encode_is_an_error", "crates/optimize/tests/optimize.rs::black_and_white_images_become_ccitt_group_4", "crates/optimize/tests/optimize.rs::group_3_fax_images_become_group_4", "crates/optimize/tests/optimize.rs::rgb_and_two_colour_indexed_black_and_white_images_become_one_bit", "crates/optimize/tests/optimize.rs::black_and_white_rgb_used_beyond_the_page_content_read_stays_rgb", "crates/optimize/tests/optimize.rs::black_and_white_rgb_drawn_only_where_the_optimizer_reads_becomes_grey", "crates/optimize/tests/optimize.rs::black_and_white_rgb_in_resources_its_page_tree_and_a_form_share_becomes_grey", "crates/automation/tests/it/automation.rs::doc_optimize_stores_black_and_white_images_as_ccitt_group_4", "crates/ui-egui/tests/it/create.rs::the_pdf_optimizer_dialog_saves_an_optimized_copy", "crates/filters/src/mq.rs::the_annex_h_test_sequence_encodes_to_its_bytes", "crates/filters/src/jbig2.rs::bitmaps_round_trip", "crates/filters/src/jbig2.rs::a_page_of_text_like_marks_round_trips_and_shrinks", "crates/optimize/tests/optimize.rs::black_and_white_images_become_jbig2", "crates/automation/tests/it/automation.rs::doc_optimize_stores_black_and_white_images_as_jbig2", "crates/filters/src/jbig2/symbols.rs::repeated_marks_are_stored_once", "crates/filters/src/jbig2/symbols.rs::scattered_marks_round_trip", "crates/optimize/tests/optimize.rs::rendered_text_becomes_jbig2_symbols", "crates/optimize/tests/optimize.rs::jbig2_scans_are_averaged_and_stored_again", "crates/optimize/src/images.rs::jbig2_images_past_the_runs_allowance_stay_as_they_are"]
notes = "All of a run's images share one allowance of decoding, filters, decoders and pixels alike (about 1,000 scanned pages' worth, more for a longer file); those past it stay as they are, those the image pass read being counted in the report, which doc_optimize and doc_reduce return and the panel doesn't show yet. CCITT Group 4, lossless: PDF Optimizer ▸ Images ▸ Monochrome Images (on by default, so Reduce File Size does it too; doc_optimize mono) stores a 1-bit image of one colour component (grey, indexed, a separation) or an image mask, kept with no filter or general-purpose ones, as Group 4 when that is smaller as stored; the samples decode to the same bits, so colour space, decode array and masks stay. Packed to one bit first, then stored the same way (colour-key masks aside): eight-bit grey of only black and white, eight-bit DeviceRGB of only black and white (as DeviceGray; calibrated RGB, RGB with a profile, RGB drawn or listed where /DefaultGray or /DefaultRGB is in force or referred to from anywhere but the page and form resources read (an annotation's appearance or a pattern drawing it, directly or through resources or a form shared with a page; another image's /Alternates), and RGB whose soft mask has a /Matte aside), and an indexed image of 2, 4 or 8 bits that uses only palette entries 0 and 1 (keeping its palette), unless the grey or colour settings make it smaller still, as they can a dithered scan downsampled to JPEG, its soft mask counted. Group 4 wins on scans; on clean computer-drawn 1-bit art Flate can be smaller, and the image then stays. Group 3 fax images are stored again as Group 4 when that is smaller (damaged Group 3 with end-of-line codes as the decoder recovers it, as it reports no damage there; byte-aligned Group 3 with end-of-line codes but none before the first row, not as T.4 sends them, as PdfCraft draws it, which at /K 0 may be other rows than were coded). CCITT fax images are decoded to be downsampled (optimize.downsample-images), and a resampled image is stored as Group 4 or 1-bit Flate, whichever is smaller. Lossless JBIG2 with bounded exact-copy symbol coding is available too, with image export; lossy JBIG2 remains unsupported. JBIG2 images are read too: averaged like any (under Retain stored as JBIG2 again), and, not resampled, stored again as the compression asks when that is smaller (as JBIG2 symbols where its own coding is a generic region, say); a symbol-coded scan, averaged, loses its exact repeats and is often larger, and then stays as it is; one whose JBIG2 page isn't its size is left alone. Not yet: inline images, soft masks and /Mask images, Group 4 for rows wider than 32,767 samples."
[[feature]]
id = "optimize.only-if-smaller"
vendor/README.md
Vendored
+3 -3
View File
@@ -39,11 +39,11 @@ Other resource categories retain their existing parent-chain lookup.
| Crate | Version | Licence | Patches | Test |
|---|---|---|---|---|
| hayro-interpret | 0.7.0 | Apache-2.0 OR MIT | (1) select `/AP /N` appearance states via `/AS` (checkboxes, radios, stateful icons) and honour the NoView flag; (2) Type3 Unicode fallback to `/Encoding` glyph names, then printable ASCII; (3) `Type3Glyph::advance_width`; (4) `InterpreterSettings::ocg_overrides` for viewer layer toggles; (5) soft masks whose group lacks `/CS` were dropped (Chrome gradient text); (6) paint nesting cap `MAX_PAINT_NESTING` for tiling patterns and Type 3 glyphs (self-referencing content overflowed the stack); (7) CID width ranges in `/W` and `/W2` clamped to `MAX_CID` (a `0 4294967295 w` range hung); (8) `InterpreterSettings::hide_comments` skips markup annotations but keeps widgets and links (View ▸ Hide all comments); (9) `image_size_ok`: images and masks over 2^28 pixels are not decoded (a fuzzed `/W 4294967295` stencil mask allocated 4 GiB); (10) `MAX_NESTED_PAINTS`: nested paints (forms, Type 3 glyphs and tiling patterns inside one another) share a per-page budget of 50,000 (a glyph showing eight glyphs of its own font, sixteen deep, hung); (11) depends on skrifa 0.47 instead of 0.42 (upstream read-fonts fixed a Type 1 tokenizer that recursed through runs of numbers: a fuzzed font hung or overflowed the stack), with the one API change (`Charset::glyph_id` returns `Option`); (12) annotation appearances whose `/BBox` or `/Matrix` collapses to a line or a point are skipped (mapping them onto `/Rect` divided by zero and handed the device infinite or NaN matrices); (13) a Highlight annotation without a usable normal appearance is drawn from its `/QuadPoints`, `/C` and `/CA` with Multiply blending, as pdfcraft-annot draws highlights (it was not drawn at all); (14) simple-font `/Widths` (`read_widths`): entries only for codes 0–255 (one entry per code below `/FirstChar` was allocated: `4294967295` asked for 34 GB), an entry that isn't a number leaves its code missing instead of ending the array, and a `/LastChar` below `/FirstChar` gives no widths instead of discarding the font for a standard one; (15) the appearance box is mapped onto `/Rect` corner to corner: the translation now accounts for the scale (an appearance whose box didn't start at the origin and differed in size from `/Rect` was drawn shifted); (16) a clip set by `W`/`W*` takes effect after whichever path-painting operator ends the path (`S`, `f`, `B`, … as well as `n`; ISO 32000-2 §8.5.4): it was applied only on `n`, so `re W* S` clipped nothing and the clip stayed pending for a later, unrelated `n`; (17) `InterpreterSettings::cancelled`: once set, interpretation stops before the next content operator (a render pool sets it for every worker when dropped, after an edit or on closing, and for one worker when the watchdog gives up on its render, so renders whose answers nobody can receive stop instead of running to the end); (18) `MAX_PAGE_CONTENT`: the content a page holds at once (its contents, and the forms, tiling patterns and Type 3 glyph procedures being painted) stays within 512 MiB (`decode_content`; once it runs out, `InterpreterWarning::ContentTruncated` is reported (once per page by pdfcraft-render) so a truncated page is not a clean render; each level gives its bytes back when it finishes): forms nest fifty deep and each level kept its content while the next ran, so a form painting itself around a stream inflating to 256 MiB would have held over 12 GB; and Type 0 sampled functions take at most `MAX_SAMPLE_ENTRIES` (2^21) table points and `MAX_SAMPLE_VALUES` (2^23) samples, read no further than `/Size` asks (256 MiB of 1-bit samples asked for over 8 GB); (19) mesh shadings (types 4–7) are sampled only inside the area being drawn (`ShadingPattern::encode_within`, which the vendored hayro renderer calls with that area plus a pixel), within 8 visits per pixel of it plus 4 per triangle (`TriangleSampler`), and patches are triangulated one at a time: points far off the page were sampled over their whole bounding box, up to 65536 × 65536 entries (a fuzzed 147-byte patch stream; a synthetic page passed 6 GB in six seconds), ten thousand overlapping triangles ran for minutes, and every patch became 722 triangles before any limit; meshes hold at most `MAX_MESH_PATCHES` (2^16) patches and `MAX_MESH_TRIANGLES` (2^20) triangles, a lattice needs two to 2^19 vertices per row (`/VerticesPerRow 0` pushed empty rows for ever; one huge row read the whole stream), and a page samples at most `MAX_PAGE_MESH_TRIANGLES` (2^25) mesh triangles in all (a mesh painted a thousand times was triangulated a thousand times); each cut logs a warning; (20) PDF function construction is limited to 64 levels and 10,000 total nodes with cycle detection; calculator parsing and evaluation share procedure depth/work budgets, integer division/bitshift/rotation/index are panic-free, and fixed-size stack errors propagate; (21) a Type 0 font drawn with a substitute that isn't CID-keyed (it isn't embedded, or its program doesn't load) draws `.nLine truncated
| hayro-interpret | 0.7.0 | Apache-2.0 OR MIT | (1) select `/AP /N` appearance states via `/AS` (checkboxes, radios, stateful icons) and honour the NoView flag; (2) Type3 Unicode fallback to `/Encoding` glyph names, then printable ASCII; (3) `Type3Glyph::advance_width`; (4) `InterpreterSettings::ocg_overrides` for viewer layer toggles; (5) soft masks whose group lacks `/CS` were dropped (Chrome gradient text); (6) paint nesting cap `MAX_PAINT_NESTING` for tiling patterns and Type 3 glyphs (self-referencing content overflowed the stack); (7) CID width ranges in `/W` and `/W2` clamped to `MAX_CID` (a `0 4294967295 w` range hung); (8) `InterpreterSettings::hide_comments` skips markup annotations but keeps widgets and links (View ▸ Hide all comments); (9) `image_size_ok`: images and masks over 2^28 pixels are not decoded (a fuzzed `/W 4294967295` stencil mask allocated 4 GiB); (10) `MAX_NESTED_PAINTS`: nested paints (forms, Type 3 glyphs and tiling patterns inside one another) share a per-page budget of 50,000 (a glyph showing eight glyphs of its own font, sixteen deep, hung); (11) depends on skrifa 0.47 instead of 0.42 (upstream read-fonts fixed a Type 1 tokenizer that recursed through runs of numbers: a fuzzed font hung or overflowed the stack), with the one API change (`Charset::glyph_id` returns `Option`); (12) annotation appearances whose `/BBox` or `/Matrix` collapses to a line or a point are skipped (mapping them onto `/Rect` divided by zero and handed the device infinite or NaN matrices); (13) a Highlight annotation without a usable normal appearance is drawn from its `/QuadPoints`, `/C` and `/CA` with Multiply blending, as pdfcraft-annot draws highlights (it was not drawn at all); (14) simple-font `/Widths` (`read_widths`): entries only for codes 0–255 (one entry per code below `/FirstChar` was allocated: `4294967295` asked for 34 GB), an entry that isn't a number leaves its code missing instead of ending the array, and a `/LastChar` below `/FirstChar` gives no widths instead of discarding the font for a standard one; (15) the appearance box is mapped onto `/Rect` corner to corner: the translation now accounts for the scale (an appearance whose box didn't start at the origin and differed in size from `/Rect` was drawn shifted); (16) a clip set by `W`/`W*` takes effect after whichever path-painting operator ends the path (`S`, `f`, `B`, … as well as `n`; ISO 32000-2 §8.5.4): it was applied only on `n`, so `re W* S` clipped nothing and the clip stayed pending for a later, unrelated `n`; (17) `InterpreterSettings::cancelled`: once set, interpretation stops before the next content operator (a render pool sets it for every worker when dropped, after an edit or on closing, and for one worker when the watchdog gives up on its render, so renders whose answers nobody can receive stop instead of running to the end); (18) `MAX_PAGE_CONTENT`: the content a page holds at once (its contents, and the forms, tiling patterns and Type 3 glyph procedures being painted) stays within 512 MiB (`decode_content`; once it runs out, `InterpreterWarning::ContentTruncated` is reported (once per page by pdfcraft-render) so a truncated page is not a clean render; each level gives its bytes back when it finishes): forms nest fifty deep and each level kept its content while the next ran, so a form painting itself around a stream inflating to 256 MiB would have held over 12 GB; and Type 0 sampled functions take at most `MAX_SAMPLE_ENTRIES` (2^21) table points and `MAX_SAMPLE_VALUES` (2^23) samples, read no further than `/Size` asks (256 MiB of 1-bit samples asked for over 8 GB); (19) mesh shadings (types 4–7) are sampled only inside the area being drawn (`ShadingPattern::encode_within`, which the vendored hayro renderer calls with that area plus a pixel), within 8 visits per pixel of it plus 4 per triangle (`TriangleSampler`), and patches are triangulated one at a time: points far off the page were sampled over their whole bounding box, up to 65536 × 65536 entries (a fuzzed 147-byte patch stream; a synthetic page passed 6 GB in six seconds), ten thousand overlapping triangles ran for minutes, and every patch became 722 triangles before any limit; meshes hold at most `MAX_MESH_PATCHES` (2^16) patches and `MAX_MESH_TRIANGLES` (2^20) triangles, a lattice needs two to 2^19 vertices per row (`/VerticesPerRow 0` pushed empty rows for ever; one huge row read the whole stream), and a page samples at most `MAX_PAGE_MESH_TRIANGLES` (2^25) mesh triangles in all (a mesh painted a thousand times was triangulated a thousand times); each cut logs a warning; (20) PDF function construction is limited to 64 levels and 10,000 total nodes with cycle detection; calculator parsing and evaluation share procedure depth/work budgets, integer division/bitshift/rotation/index are panic-free, and fixed-size stack errors propagate; (21) a Type 0 font drawn with a substitute that isn't CID-keyed (it isn't embedded, or its program doesn't load) draws `.nLine truncated
| hayro | 0.7.1 | Apache-2.0 OR MIT | (1) `RenderSettings::{x_offset, y_offset}` to render one tile of a large page; (2) images over `MAX_IMAGE_PIXELS` (2^28) are skipped (a fuzzed `/W 4294967295` hung in resampling); (3) `tiling_cell_scale` keeps a tiling pattern's cell pixmap within 3000 px a side whatever its `/XStep`/`/YStep` (a fuzzed step far beyond `/BBox` allocated 2 GB); (4) stroke widths clamped to 4 × (canvas width + height) device pixels (a fuzzed `/LW 9223372036854775807` allocated 10 GB in stroke expansion); (5) `usable_dash_pattern`: a stroke whose dash pattern could cut it into more than `MAX_DASHES_PER_STROKE` (250,000) pieces, counting kurbo's restart of the pattern at every subpath and `ClosePath`, is drawn solid (a fuzzed `/D [[11] 0]` on a line from x = 92234775807 allocated over 5 GB, `[0.000001] 0 d` on a 20 pt line over 10 GB); (6) a dash pattern with a negative or non-finite entry, or a period that isn't positive, is drawn solid (`[-1 -1] 0 d` looped for ever in kurbo); (7) `render_size` and `render_into`: render into a buffer the caller allocates (PdfCraft renders into pixel-aligned memory that the GUI takes over as texture data); `render` is `render_into` on a new pixmap; (8) checked image source/target/RGBA, alpha-premultiplication scratch and intermediate sizes before resampling, finite transforms and exact source/mask buffers; sources wider than u16 may shrink before backend limits, with a 2^20 source-axis cap and conservative 128 MiB filter-planning estimate; an over-limit or failed resize keeps the original image resolution and geometry when the source fits the backend, otherwise skips it, with a logged warning; staging, resampling output/scratch, RGBA conversion and padding buffers use fallible reserves; (9) `may_paint_canvas`: a fill or stroke whose control-point box, grown by how far its paint reaches past the path (half the stroke width times the larger of the miter limit and √2) plus 0.5 user units and 2 pixels, misses the canvas is not drawn, so a tile no longer strokes and fills every path of its page (item 5 of #623); `set_skip_offscreen_paths` turns this off for the regression test; (10) `/Interpolate` governs the pre-resize too (#624): a minified image that does not request interpolation is pre-resized with nearest-neighbour, planned without the Catmull-Rom filter limits (`image_resampling_size` takes the flag); a Type 3 glyph image is sampled bicubically only when interpolated; a minified image with a matching soft mask is nearest-sampled straight into premultiplied RGBA; (11) `may_paint_rect`: an image whose stream-size box, mapped to the device, misses the canvas is skipped before decoding (decoded data can only shrink against that box), so a tile of an image-heavy page no longer decodes, resamples and rasterizes every image on it; (12) preview renders (`InterpreterSettings::preview`: backdrops, thumbnails, grid cells, signature-drag backgrounds) skip images under `PREVIEW_MIN_IMAGE_AREA` device pixels, while final renders always draw them; (13) reading colours (#897): `RenderSettings::color_replacement` takes a `ColorReplacement` (text colour, background, `only_black`, `line_art`) that remaps solid text and line-art paint while rendering, glyph outlines counting as text and stencil-mask ink as line art; images, image alpha masks, soft masks and pattern cells keep their colours; `render_pdf` gains a `color_replacement: Option<ColorReplacement>` argument (pass `None` for true colours); the upstream-only `[dev-dependencies.log]` entry is removed from `Cargo.toml` (log is a normal dependency since the logging patch), so `cargo test -p hayro` runs this crate's unit tests; (14) a shrunk, interpolated image whose axes lie along the device's (upright, mirrored or a quarter turn, as on a page under `/Rotate`) is snapped to whole device pixels: shrunk once to its rounded device rectangle and drawn 1:1 from the rounded corner, instead of a ceil-sized raster resampled again at a fractional scale and offset (`snap_minified_image`; scans looked soft next to Acrobat, and until #5530 a turned one still did, up to 200 grey levels off the same page drawn upright). An image counts as axis-aligned when the terms the snap drops move its far edges by at most a hundredth of a device pixel (`SNAP_SKEW`), not by a tolerance on the coefficients, which let a slight skew of a very wide image through; and the snapped 1:1 transform is used only when the shrink produced the snapped size, so a failed resize keeps the full transform instead of drawing the source raster at its source size; (15) a soft mask of another size than its image is filtered when shrunk, whatever its `/Interpolate` (#1649): such a mask carries the shape on its own (Word saves text as a tiny black image with a glyph-bitmap mask), and nearest-neighbour dropped every stroke thinner than the shrink factor; (16) under a soft mask or a blend mode, axial and radial shadings are sampled instead of drawn as native gradients (#Line truncated
| hayro-jbig2 | 0.3.0 | Apache-2.0 OR MIT | (1) at most 2^28 pixels per bitmap and page bitmap, and 2^25 cells per halftone grid and grayscale image (each side could be 65535: 4.3 billion pixels decoded one by one, and a 4-byte value per cell); (2) at most 2^26 symbol pixels per symbol dictionary (a fuzzed dictionary decoded for minutes); (3) a symbol dictionary's height classes are capped at one per new symbol and its export runs at `2 × symbols + 1`, each plus `EXTRA_EMPTY_STEPS` (1024): an empty height class or a zero-length export run doesn't advance, and past the end of its data the arithmetic decoder can decode them for ever (one changed byte hung the renderer); (4) a segment header's long-form referred-to segment count is capped at `u16::MAX` and a larger count is refused before its retention or referred-to bytes are read: the 29-bit count went straight to `Vec::with_capacity`, a reservation of about 2 GiB whose failure aborts the process (from storytold/craft-libs#176); (5) what a page's segments hold for later ones (intermediate regions, pattern and symbol dictionaries with their headers and retained contexts, and Huffman tables) is at most 64 MiB, two square bitmaps of 16,384 a side (`MAX_RETAINED_WORDS`), and a dictionary's new and exported symbols are counted against it as they are made: each was bounded on its own, but a few dozen bytes an intermediate region held 33 MB each, and 8 MiB symbols exported again and again killed the process, so a few kilobytes could exhaust memory; (6) a segment may refer to at most 2^20 symbols, checked before any are collected: references may repeat, and 65,535 of them to one dictionary, in a 78 KB stream, asked for 17 GB; (7) a pattern dictionary keeps at most 16 MiB of shifted copies, which only speed up drawing, and none that wouldn't fit what the page has left (`MAX_SHIFTED_WORDS`); (8) a custom Huffman table has at most 2^16 lines, and a stream with its globals at most 2^20 segments and 2^22 references, checked as they are read: both grew before anything counted them, and PDFs of a kilobyte or so reached 5 GB and more behind nested Flate; (9) a page past the bounds of (1) is refused when the stream is opened: a caller sizes its output from the header before the page bitmap is made, and an Indexed image in a 743-byte PDF asked for 137 GB, which aborts; (10) a stream may ask for at most 2^30 units of decoding work, four of the largest pages, about 6 s to 17 s (`MAX_WORK`): decoding a bitmap costs a pixel each and 4 a row, copying one a word each and 1 a row, and each integer an item reads costs 38, what one takes at worst, so one with no pixels pays too, before it is decoded: each bitmap was bounded, but nothing bounded how many there were, and a 795-byte PDF took 265 s; (12) adding to a location, a reference offset, a symbol's corner or a Huffman range line near the end of an `i32`, and negating a decoded integer of 2^31, overflowed: a panic wherever overflow is checked. The refinement context, a text region's corners, the integer decoder and the Huffman table wrap, as a release build did, and `Bitmap::combine` saturates, which comes to the same, as a location that overflows is past the end of any page. What is drawn doesn't change | (1) `raster::tests::jbig2_regions_of_billions_of_pixels_are_refused`, `raster::tests::jbig2_halftone_grids_past_their_bound_are_refused`; (2) the nightly `cargo xtask fuzz` finding, which now finishes in about a second; (3) `raster::tests::jbig2_symbol_dictionaries_that_never_advance_terminate`; (4) `segment::tests::test_referred_to_count_is_capped` (`cargo test --manifest-path vendor/hayro-jbig2/Cargo.toml`); (5) `raster::tests::jbig2_regions_held_past_a_pages_bound_are_refused`, `jbig2_symbols_without_pixels_count_towards_a_pages_bound`, `jbig2_retained_contexts_count_towards_a_pages_bound`, `jbig2_huffman_tables_count_towards_a_pages_bound`, `jbig2_new_symbols_count_towards_a_pages_bound`, `jbig2_new_symbols_are_held_to_what_a_page_has_left`, `jbig2_pattern_dictionaries_count_towards_a_pages_bound`, and in the crate (`cargo test --manifest-path vendor/hayro-jbig2/Cargo.toml`) `decode::symbol::tests::exported_symbols_are_counted_before_they_are_copied`, `tests::the_room_left_shrinks_as_segments_are_stored`; (6) `raster::tests::jbig2_segments_referring_to_too_many_symbols_are_refused`; (7) `raster::tests::jbig2_pattern_dictionaries_too_large_to_shift_still_draw`, `jbig2_shifted_patterns_never_refuse_an_image`; (8) `raster::tests::jbig2_huffman_tables_of_too_many_lines_are_refused`, `jbig2_streams_of_too_many_segments_are_refused`, `jbig2_streams_of_too_many_references_are_refused`, and in the crate `tests::the_segments_of_a_stream_and_its_globals_count_together`; (9) `raster::tests::jbig2_pages_past_the_bounds_are_refused_before_anything_is_sized_from_them`, `tests::a_page_past_the_bounds_is_refused_when_opened`; (10) `raster::tests::jbig2_decoding_work_is_counted_item_by_item`; (12) `raster::tests::jbig2_placements_at_the_endLine truncated
| hayro-jbig2 | 0.3.0 | Apache-2.0 OR MIT | (1) at most 2^28 pixels per bitmap and page bitmap, and 2^25 cells per halftone grid and grayscale image (each side could be 65535: 4.3 billion pixels decoded one by one, and a 4-byte value per cell); (2) at most 2^26 symbol pixels per symbol dictionary (a fuzzed dictionary decoded for minutes); (3) a symbol dictionary's height classes are capped at one per new symbol and its export runs at `2 × symbols + 1`, each plus `EXTRA_EMPTY_STEPS` (1024): an empty height class or a zero-length export run doesn't advance, and past the end of its data the arithmetic decoder can decode them for ever (one changed byte hung the renderer); (4) a segment header's long-form referred-to segment count is capped at `u16::MAX` and a larger count is refused before its retention or referred-to bytes are read: the 29-bit count went straight to `Vec::with_capacity`, a reservation of about 2 GiB whose failure aborts the process (from storytold/craft-libs#176); (5) what a page's segments hold for later ones (intermediate regions, pattern and symbol dictionaries with their headers and retained contexts, and Huffman tables) is at most 64 MiB, two square bitmaps of 16,384 a side (`MAX_RETAINED_WORDS`), and a dictionary's new and exported symbols are counted against it as they are made: each was bounded on its own, but a few dozen bytes an intermediate region held 33 MB each, and 8 MiB symbols exported again and again killed the process, so a few kilobytes could exhaust memory; (6) a segment may refer to at most 2^20 symbols, checked before any are collected: references may repeat, and 65,535 of them to one dictionary, in a 78 KB stream, asked for 17 GB; (7) a pattern dictionary keeps at most 16 MiB of shifted copies, which only speed up drawing, and none that wouldn't fit what the page has left (`MAX_SHIFTED_WORDS`); (8) a custom Huffman table has at most 2^16 lines, and a stream with its globals at most 2^20 segments and 2^22 references, checked as they are read: both grew before anything counted them, and PDFs of a kilobyte or so reached 5 GB and more behind nested Flate; (9) a page past the bounds of (1) is refused when the stream is opened: a caller sizes its output from the header before the page bitmap is made, and an Indexed image in a 743-byte PDF asked for 137 GB, which aborts; (10) a stream may ask for at most 2^30 units of decoding work, four of the largest pages, about 6 s to 17 s (`MAX_WORK`): decoding a bitmap costs a pixel each and 4 a row, copying one a word each and 1 a row, and each integer an item reads costs 38, what one takes at worst, so one with no pixels pays too, before it is decoded: each bitmap was bounded, but nothing bounded how many there were, and a 795-byte PDF took 265 s; (11) `Image::decode_within` spends from what a run of images may still do, starting with its page's pixels, and `charge_reading` charges reading a stream and its globals before anything is parsed, so an operation over a document can bound what `MAX_WORK` bounds for one image, and images that fail or decode nothing add up too: a file of thousands of images made to reach that limit took as many times as long; (12) adding to a location, a reference offset, a symbol's corner or a Huffman range line near the end of an `i32`, and negating a decoded integer of 2^31, overflowed: a panic wherever overflow is checked. The refinement context, a text region's corners, the integer decoder and the Huffman table wrap, as a release build did, and `Bitmap::combine` saturates, which comes to the same, as a location that overflows is past the end of any page. What is drawn doesn't change | (1) `raster::tests::jbig2_regions_of_billions_of_pixels_are_refused`, `raster::tests::jbig2_halftone_grids_past_their_bound_are_refused`; (2) the nightly `cargo xtask fuzz` finding, which now finishes in about a second; (3) `raster::tests::jbig2_symbol_dictionaries_that_never_advance_terminate`; (4) `segment::tests::test_referred_to_count_is_capped` (`cargo test --manifest-path vendor/hayro-jbig2/Cargo.toml`); (5) `raster::tests::jbig2_regions_held_past_a_pages_bound_are_refused`, `jbig2_symbols_without_pixels_count_towards_a_pages_bound`, `jbig2_retained_contexts_count_towards_a_pages_bound`, `jbig2_huffman_tables_count_towards_a_pages_bound`, `jbig2_new_symbols_count_towards_a_pages_bound`, `jbig2_new_symbols_are_held_to_what_a_page_has_left`, `jbig2_pattern_dictionaries_count_towards_a_pages_bound`, and in the crate (`cargo test --manifest-path vendor/hayro-jbig2/Cargo.toml`) `decode::symbol::tests::exported_symbols_are_counted_before_they_are_copied`, `tests::the_room_left_shrinks_as_segments_are_stored`; (6) `raster::tests::jbig2_segments_referring_to_too_many_symbols_are_refused`; (7) `raster::tests::jbig2_pattern_dictionaries_too_large_to_shift_still_draw`, `jbig2_shifted_patterns_never_refuse_an_image`; (8) `raster::tests::jbig2_huffman_tables_of_too_many_lines_are_refused`, `jbig2_streams_of_too_many_segments_are_reLine truncated
| lopdf | 0.45.0 | MIT | (1) PNG predictor rows longer than the data are refused before their buffers are allocated (a fuzzed `/Columns 4294967295` allocated two 4 GiB rows), and empty frames return without row buffers; (2) predictor row widths use saturating arithmetic or reject overflow; (3) sub-byte TIFF Predictor 2 scratch is bounded by the samples present in the input instead of the declared `/Colors`, aligned with [lopdf#589](https://github.com/J-F-Liu/lopdf/pull/589); (4) the object lookup keeps its cycle history through compressed objects, requires an ordinary in-file container, and caps the chain of indirect lookups at 16 (a chain of new objects, each a `/Length` compressed in a container whose own `/Length` names the next, or a stream whose `/Length` names the next stream, overflowed the stack: a dev build's 2 MiB thread within 800 hops); (5) object-stream members are parsed within their own extent (up to the next distinct offset), an entry whose offset repeats an earlier entry's is skipped (so a repeated object number can resolve to its earlier entry), and offsets are processed in index order; (6) `find_latest_trailer`: a rebuild takes a `/Type /XRef` stream dictionary as the trailer when the file has no `trailer` keyword (the latest in the file wins, as a later `trailer` would), so a concatenated or truncated file keeps its /Info and /ID instead of failing to load; a candidate whose /Root is the catalog beats one whose /Root is another dictionary, and an in-file /Root must match its generation (#2365) | (1) `inspect::tests::png_predictor_rows_longer_than_the_data_are_refused_before_allocating`, `inspect::tests::empty_png_predictor_frames_do_not_allocate_rows`; (2) `inspect::tests::tiff_subbyte_predictor_row_width_overflow_is_refused`; (3) `inspect::tests::tiff_subbyte_predictor_scratch_is_bounded_by_available_samples`, `inspect::tests::tiff_subbyte_predictors_keep_components_rows_and_padding`; (4) `inspect::tests::compressed_object_self_container_does_not_recurse`, `inspect::tests::compressed_object_mutual_containers_do_not_recurse`, `inspect::tests::object_stream_length_cycle_does_not_recurse`, `inspect::tests::compressed_object_container_chain_does_not_recurse`, `inspect::tests::compressed_length_chain_does_not_recurse`, `inspect::tests::plain_stream_length_chain_does_not_recurse`, `inspect::tests::compressed_objects_and_indirect_stream_length_still_load`, `inspect::tests::compressed_length_in_a_container_with_an_indirect_length_still_loads`; (5) `inspect::tests::object_stream_member_extents_bound_owned_strings`, `inspect::tests::object_stream_member_extents_preserve_index_order_and_whitespace`; (6) `inspect::tests::rebuilt_file_keeps_the_info_of_its_cross_reference_stream`, `inspect::tests::rebuilt_file_prefers_a_trailer_whose_root_is_the_catalog` |
| hayro-syntax | 0.7.2 | Apache-2.0 OR MIT | (1) Page-tree cycle guard in `resolve_pages` (visited set + depth cap): a `/Kids` loop among compressed objects overflowed the stack and aborted the process; (2) `ccitt_size_ok`: CCITT images wider than 2^20 columns or over 2^28 pixels are not decoded (a fuzzed `/Columns 4294967295` allocated 4 GiB); (3) Flate/LZW predictor row widths are saturating `usize`, a zero width is refused and a row longer than the data is not allocated (a fuzzed `/Columns 9223372036854775807` wrapped to a 2^61-byte allocation and aborted); (4) the inline-image end-of-data check (`EI`) doesn't nest inside its own look-ahead and examines at most 256 candidates and steps (50 images with "EI" in their data ran for over ten minutes); an EI followed by a delimiter (EI%..., EI[...) ends the image like one followed by white space; only a regular character after EI skips the candidate; (5) a stream whose `/Length` is wrong is recovered only when `stream` is the next token — scanning ahead to a later object's `stream` made an indirect appearance-state dictionary (Quartz checkboxes: `/AP /N` → `<< /Yes … /Off … >>`) render as a non-form, so the widget was skipped; (6) JPX rescaling uses an exact integer sample range for 1–32-bit components (32 shifted past u32, and floating-point maxima at 25–31 bits wrapped to zero), and refuses zero-sized rows; (7) `decode_limit`: Flate (through flate2 and the fallback inflater), LZW, RunLength and ASCII85 never write past `MAX_DECODED_STREAM` (256 MiB), or, for the last filter of an image of known size, past twice its pixel bytes (a predictor tag byte per byte), at most 1 GiB; `Stream::decoded_within` for budgeted callers; a page's `/Contents` array shares one `MAX_DECODED_STREAM` (a 2 MB page whose content stream or 8 × 8 image inflated to 2 GB took the renderer past 3 GB in under a second; naming the stream twenty times in `/Contents`, 5 GB); (8) high-precision numbers (jsPDF MediaBox coordinates with over 19 digits) parse without overflowing the u64 mantissa: fractional digits past it are dropped, integer digits past it scale the value (a real, kept finite); (9) large dictionary indexes are built on first use and shared across repeated reads within each PDF (#307); the validated-byte cache is bounded, reader contexts remain separate, repairs and inline/content dictionaries bypass sharing, and small dictionaries stay eager; (10) decoded object-stream slots publish once under contention, cold loading rejects recursive dependencies, and xref repair serializes without holding recursive read guards; this permits one parser per native render pool with private worker caches; (11) `sync.rs`: a poisoned lock is recovered instead of unwrapped (a worker panic no longer cascades into every reader of the same PDF), and xref repair takes a blocking write lock instead of asserting a non-blocking one; (12) `SegmentList` (`util.rs`) returns `None` for an out-of-range index instead of panicking; (13) a page's `/UserUnit` scales `render_dimensions` and `initial_transform`, so the page and everything on it are sized in points, as Acrobat shows them; it is read as Acrobat reads it (the page's own entry, not inherited; a number from 1 up, at most 75,000; anything else counts as 1), and a page without it keeps exactly the transform it had; (14) revision 2–4 passwords are tried as given and then in PDFDocEncoding (`legacy_passwords` in `crypto/mod.rs`, which resolves the upstream `TODO: Convert to PDFDocEncoding`): passwords arrive as UTF-8 text, so a password with characters outside ASCII (`Çok güzel`, `ılık€`) written in PDFDocEncoding, as ISO 32000-2 §7.6.4.3.2 asks and qpdf and Acrobat write it, was always wrong; (15) page content is read a piece at a time (`page::ContentReader`, `Page::content_reader`): plain streams and Flate streams without a predictor are decoded as they are read, with a budget of 64 times their compressed size and never less than `MAX_DECODED_STREAM` (a stream that expands past it is cut and reported); other streams are decoded whole, as before, and share `MAX_DECODED_STREAM`. The content tokenizer reads a window at a time (`UntypedIter::new_window`, `TypedIter::needs_more`, `TypedIter::committed`): an instruction that runs past the window is not decided until more data follows, which `Reader` records by flagging reads and `at_end` past the end of the data, including through the clones the parser makes. The Flate decoder keeps its own 32 KiB history, so it pauses at any output size. `ContentReader::new` reuses raw/decrypted stream bytes for budget accounting and piece creation, and `decoded_raw_within` applies the existing Whole-stream filters and limits to those bytes without decrypting twice; (16) revision 5–6 (AES-256) passwords are tried in their SASLprep form (RFC 4013, then UTF-8, at most 127 bytes, as ISO 32000-2 §7.6.4.3.3 Algorithm 2.A asks and pdfcraft-crypt writes them) and then, if that differs, as given (`aes256_passwLine truncated
| hayro-syntax | 0.7.2 | Apache-2.0 OR MIT | (1) Page-tree cycle guard in `resolve_pages` (visited set + depth cap): a `/Kids` loop among compressed objects overflowed the stack and aborted the process; (2) `ccitt_size_ok`: CCITT images wider than 2^20 columns or over 2^28 pixels are not decoded (a fuzzed `/Columns 4294967295` allocated 4 GiB); (3) Flate/LZW predictor row widths are saturating `usize`, a zero width is refused and a row longer than the data is not allocated (a fuzzed `/Columns 9223372036854775807` wrapped to a 2^61-byte allocation and aborted); (4) the inline-image end-of-data check (`EI`) doesn't nest inside its own look-ahead and examines at most 256 candidates and steps (50 images with "EI" in their data ran for over ten minutes); an EI followed by a delimiter (EI%..., EI[...) ends the image like one followed by white space; only a regular character after EI skips the candidate; (5) a stream whose `/Length` is wrong is recovered only when `stream` is the next token — scanning ahead to a later object's `stream` made an indirect appearance-state dictionary (Quartz checkboxes: `/AP /N` → `<< /Yes … /Off … >>`) render as a non-form, so the widget was skipped; (6) JPX rescaling uses an exact integer sample range for 1–32-bit components (32 shifted past u32, and floating-point maxima at 25–31 bits wrapped to zero), and refuses zero-sized rows; (7) `decode_limit`: Flate (through flate2 and the fallback inflater), LZW, RunLength and ASCII85 never write past `MAX_DECODED_STREAM` (256 MiB), or, for the last filter of an image of known size, past twice its pixel bytes (a predictor tag byte per byte), at most 1 GiB; `Stream::decoded_within` for budgeted callers; a page's `/Contents` array shares one `MAX_DECODED_STREAM` (a 2 MB page whose content stream or 8 × 8 image inflated to 2 GB took the renderer past 3 GB in under a second; naming the stream twenty times in `/Contents`, 5 GB); (8) high-precision numbers (jsPDF MediaBox coordinates with over 19 digits) parse without overflowing the u64 mantissa: fractional digits past it are dropped, integer digits past it scale the value (a real, kept finite); (9) large dictionary indexes are built on first use and shared across repeated reads within each PDF (#307); the validated-byte cache is bounded, reader contexts remain separate, repairs and inline/content dictionaries bypass sharing, and small dictionaries stay eager; (10) decoded object-stream slots publish once under contention, cold loading rejects recursive dependencies, and xref repair serializes without holding recursive read guards; this permits one parser per native render pool with private worker caches; (11) `sync.rs`: a poisoned lock is recovered instead of unwrapped (a worker panic no longer cascades into every reader of the same PDF), and xref repair takes a blocking write lock instead of asserting a non-blocking one; (12) `SegmentList` (`util.rs`) returns `None` for an out-of-range index instead of panicking; (13) a page's `/UserUnit` scales `render_dimensions` and `initial_transform`, so the page and everything on it are sized in points, as Acrobat shows them; it is read as Acrobat reads it (the page's own entry, not inherited; a number from 1 up, at most 75,000; anything else counts as 1), and a page without it keeps exactly the transform it had; (14) revision 2–4 passwords are tried as given and then in PDFDocEncoding (`legacy_passwords` in `crypto/mod.rs`, which resolves the upstream `TODO: Convert to PDFDocEncoding`): passwords arrive as UTF-8 text, so a password with characters outside ASCII (`Çok güzel`, `ılık€`) written in PDFDocEncoding, as ISO 32000-2 §7.6.4.3.2 asks and qpdf and Acrobat write it, was always wrong; (15) page content is read a piece at a time (`page::ContentReader`, `Page::content_reader`): plain streams and Flate streams without a predictor are decoded as they are read, with a budget of 64 times their compressed size and never less than `MAX_DECODED_STREAM` (a stream that expands past it is cut and reported); other streams are decoded whole, as before, and share `MAX_DECODED_STREAM`. The content tokenizer reads a window at a time (`UntypedIter::new_window`, `TypedIter::needs_more`, `TypedIter::committed`): an instruction that runs past the window is not decided until more data follows, which `Reader` records by flagging reads and `at_end` past the end of the data, including through the clones the parser makes. The Flate decoder keeps its own 32 KiB history, so it pauses at any output size. `ContentReader::new` reuses raw/decrypted stream bytes for budget accounting and piece creation, and `decoded_raw_within` applies the existing Whole-stream filters and limits to those bytes without decrypting twice; (16) revision 5–6 (AES-256) passwords are tried in their SASLprep form (RFC 4013, then UTF-8, at most 127 bytes, as ISO 32000-2 §7.6.4.3.3 Algorithm 2.A asks and pdfcraft-crypt writes them) and then, if that differs, as given (`aes256_passwLine truncated
| winit | 0.30.13 | Apache-2.0 | (1) on Windows 11 (build 22000 and later), `WM_DPICHANGED` applies the rect Windows suggests. It used to nudge the window onto the monitor `MonitorFromWindow` names, which during a drag is still the monitor being left; the nudge sent another `WM_DPICHANGED`, so a window dragged to a monitor with another scale factor grew and stayed on the wrong monitor at the wrong scale (#324). Backport of winit master 488c036a ([winit#4341](https://github.com/rust-windowing/winit/pull/4341)), which no 0.30 release has; Windows 10 keeps the old adjustment, as on winit master. (2) on Wayland, a seat's text input is destroyed when the seat goes away, not whenever it loses a capability: losing any capability (a touch screen or tablet unplugged, a remote-desktop session ending) destroyed it, and only an added capability recreated it, so IME stopped working until restart. Port of [winit#4747](https://github.com/rust-windowing/winit/pull/4747); (3) on Windows, an IME composition reports its cursor from `GCS_CURSORPOS` only when `WM_IME_COMPOSITION` says it is set, and at the end of the composition otherwise: the Microsoft Korean IME never sets it, so the cursor read as 0 (egui then drew no caret while composing; it now draws it after the syllable). Port of [winit#4746](https://github.com/rust-windowing/winit/pull/4746); (4) on Windows, a hidden window is not maximized until it is shown, and a hidden window shown while maximized gets one `ShowWindow(SW_MAXIMIZE)`: `SW_MAXIMIZE` shows a hidden window, so PdfCraft, which eframe keeps hidden until its first frame is drawn, appeared unpainted (white) when it restored a maximized window, and could lose its focus. Port of [winit#4587](https://github.com/rust-windowing/winit/pull/4587) (the later, focus-safe form of [winit#4575](https://github.com/rust-windowing/winit/pull/4575)); (5) on X11, when allowing or disallowing IME replaces the focused window's input context (egui does it each time a text field gains or loses focus), the new context is focused: no focus event follows, so it stayed unfocused until the window lost and regained focus. Port of [winit#4727](https://github.com/rust-windowing/winit/pull/4727); (6) on macOS, `insertText` commits through IME when a composition was started (`pending_commit`, set by `setMarkedText`) rather than when marked text is left: input methods that end a composition with `unmarkText` before `insertText` lost the text, and Apple Korean dropped the key typed after a syllable (`한` then `5` or `;`); other `insertText` calls stay `KeyboardInput`; a commit outside a key press (a candidate clicked with the mouse, dictation) clears the marked text and leaves no key to swallow, as in [winit#4748](https://github.com/rust-windowing/winit/pull/4748) (without it the next key, such as Backspace, was lost). `validAttributesForMarkedText` lists the underline and marked-clause attributes, without which some input methods don't compose. Port of [winit#4650](https://github.com/rust-windowing/winit/pull/4650), the 0.30 backport of [winit#4651](https://github.com/rust-windowing/winit/pull/4651) (merged on winit master), with the underline attribute spelled as AppKit's `NSUnderlineStyleAttributeName` (`NSUnderline`); (7) on macOS, disallowing IME with a composition in progress (focus leaving a text field) tells the input method to discard its marked text (`discardMarkedText`): it kept composing it once IME was allowed again, so the abandoned syllable came back (`하`, then `가` typed in the next field, became `학`, then `하` + `가`; Japanese `か` then `さ` became `傘`). Port of [winit#4745](https://github.com/rust-windowing/winit/pull/4745); (8) on macOS, text inserted outside a key press without a composition (the emoji picker) is reported as `Ime::Commit`; it was dropped. Port of the emoji-picker part of [winit#4749](https://github.com/rust-windowing/winit/pull/4749) (0.30 backport of [winit#4748](https://github.com/rust-windowing/winit/pull/4748)); its `selectedRange` change for dictation is left out because it switches on press-and-hold for every held key (an accent menu instead of key repeat, and the chosen accent added after the repeated letters); (9) on macOS, a key whose text was committed through IME isn't also sent as `KeyboardInput` unless AppKit forwards it as a command (Enter): Apple Korean answers Space with no syllable in progress with `setMarkedText(" ")`, `insertText(" ")`, `setMarkedText("")`, and the last call reset winit's state, so the space arrived twice. Same goal as [winit#4478](https://github.com/rust-windowing/winit/pull/4478), whose changes don't reach this call order (macOS 14); to be proposed upstream there; (10) on macOS, when the first key in a fresh app reaches Apple Korean's `insertText` as a raw compatibility jamo instead of starting a composition, that key is interpreted once more, synchronously, before it is sent as `KeyboardInput` (g k s typed ㅎㅏㄴ instead of 한); only an unmodified, Line truncated
| egui-wgpu | 0.36.2 | MIT OR Apache-2.0 | (1) `winit::surface_fit`: a window's surface is configured within the device's `max_texture_dimension_2d`, both sides shrunk by one factor that egui then draws at (`pixels_per_point × render_scale`), so the whole window is drawn into the surface (on DX12 it is stretched over the window and lines up with the pointer; wgpu's GL backend copies it unscaled into a corner); the longer side becomes exactly the limit; the depth and MSAA textures follow, and a recreated surface starts again from the window's own size. egui-wgpu configured the window's size as is, and `Surface::configure` panics beyond the limit: a window restored at another monitor's scale (opening an Outlook attachment on a mixed-DPI laptop asked an 8192-pixel device for 8600 × 3423, #577) or stretched across monitors closed PdfCraft, before its window appeared when it happened at start-up ([emilk/egui#8361](https://github.com/emilk/egui/issues/8361)). The published crate leaves out egui's licence files; `LICENSE-APACHE` and `LICENSE-MIT` are from the 0.36.2 tag; (2) `winit::catch_errors`: a new window's surface is configured inside wgpu error scopes (validation, out of memory, internal), and a failure is returned as `WgpuError::ConfigureSurface` (for the root window at start-up eframe returns it as `Error::Wgpu`, which PdfCraft retries with OpenGL); nothing is kept for the window. wgpu's default handler panicked instead, and eframe configures the root surface before the app is created, so a driver that can't configure one closed PdfCraft at launch (a 2015 Intel GPU through wgpu's GL backend timed out waiting for the GPU to go idle, `GpuWaitTimeout`, #519). Unchanged, so still a panic: an error in a later reconfigure (resize, a lost surface), a device lost during the first configure (wgpu reports it only to a callback), and a window shown at zero size (first configured on resize) | (1) `pdfcraft` `tests::a_window_larger_than_the_gpu_limit_gets_a_surface_within_it`, `tests::surface_fit_never_empties_or_overflows_a_surface`, and `tests::vendor_provenance_egui_wgpu_surface_fit_identity`, which keeps the patch in the build and in the paths that size a surface; (2) `pdfcraft` `tests::a_gpu_error_while_a_window_is_set_up_is_returned_not_a_panic` (a real device: an error inside `catch_errors` is returned, not a panic) and `tests::vendor_provenance_egui_wgpu_catch_errors_identity`, which keeps it around `add_surface`'s configure |
| vello_common | 0.0.8 | Apache-2.0 OR MIT | (1) `clip::intersect` (the clip step every clipped path passes through) starts its row walk at the first row both paths have strips in, not the earlier of the two first rows. The walk visited every row from the top of the clip for each path (on the A1 sample's full-page canvas, about half its strip rows per stroke), and rows where one path has no strips produce no output, so the strips are unchanged (vello_common 0.1 starts at the later row too). Its cost grew with the page, and a full-page render of the A1 sample was about 30% faster; the output is byte-identical | (1) `clip::tests::intersect_walks_only_rows_both_paths_have_strips` counts the rows the walk visits (`cargo test --manifest-path vendor/vello_common/Cargo.toml`) and fails without the patch; the render crate's pixel tests run through the patched crate |
vendor/hayro-interpret/src/context.rs
Vendored
+2
View File
@@ -218,6 +218,8 @@ impl<'a> Context<'a> {
NESTED_PAINTS.with(|n| n.set(0));
CONTENT_LEFT.with(|c| c.set(MAX_PAGE_CONTENT));
CONTENT_TRUNCATED.with(|truncated| truncated.set(false));
// PdfCraft patch (73): and fresh JBIG2 decoding, which every image of it shares.
hayro_syntax::set_page_jbig2_work(settings.jbig2_work);
// PdfCraft patch (55): a fallback left over from another page is not this page's.
crate::color::take_color_fallback();
crate::encode::reset_mesh_budget();
vendor/hayro-interpret/src/interpret/mod.rs
Vendored
+18
View File
@@ -120,6 +120,10 @@ pub struct InterpreterSettings {
/// costs a decode and a resample each for pixels nobody can see. Never set it for a final
/// render (tiles, whole pages, print, export).
pub preview: bool,
/// PdfCraft patch (73): the steps of JBIG2 decoding a page's images may take between them
/// ([`hayro_syntax::MAX_PAGE_JBIG2_WORK`] by default). Images past it aren't drawn, nor images
/// whose masks are past it, each with [`InterpreterWarning::ContentTruncated`].
pub jbig2_work: u64,
}
impl Default for InterpreterSettings {
@@ -142,6 +146,7 @@ impl Default for InterpreterSettings {
ocg_overrides: Arc::new(Vec::new()),
cancelled: None,
preview: false,
jbig2_work: hayro_syntax::MAX_PAGE_JBIG2_WORK,
}
}
}
@@ -163,12 +168,25 @@ pub enum InterpreterWarning {
ColorFallback,
}
/// PdfCraft patch (73): ends a page's JBIG2 allowance with the page, so that what decodes on its
/// thread afterwards has no page to run out on.
struct PageJbig2Work;
impl Drop for PageJbig2Work {
fn drop(&mut self) {
hayro_syntax::set_page_jbig2_work(u64::MAX);
}
}
/// interpret the contents of the page and render them into the device.
pub fn interpret_page<'a>(
page: &Page<'a>,
context: &mut Context<'a>,
device: &mut impl Device<'a>,
) {
// PdfCraft patch (73): a context used again is another page.
hayro_syntax::set_page_jbig2_work(context.settings.jbig2_work);
let _allowance = PageJbig2Work;
let resources = page.resources();
let mut content = page.content_reader();
// PdfCraft patch: the page's own contents count against its content budget too, where they
vendor/hayro-interpret/src/x_object.rs
Vendored
+31 -4
View File
@@ -681,13 +681,23 @@ fn decoded_mask_size(mask: &DecodedMask) -> usize {
/// `/Filter` or its inline abbreviation `/F`. Errs towards `true`: a name the decoder would
/// ignore still counts, which only keeps the target in the key as before.
fn uses_jpx(stream: &Stream<'_>) -> bool {
uses_filter(stream, JPX_DECODE)
}
/// PdfCraft patch (73): whether `stream` is JBIG2, so that what the page has left, which all of it
/// shares, says something about why this stream failed.
fn uses_jbig2(stream: &Stream<'_>) -> bool {
uses_filter(stream, JBIG2_DECODE)
}
fn uses_filter(stream: &Stream<'_>, filter: &[u8]) -> bool {
let dict = stream.dict();
let is_jpx = |name: &Name<'_>| name.deref() == JPX_DECODE;
let is_filter = |name: &Name<'_>| name.deref() == filter;
[F, FILTER].into_iter().any(|key| {
dict.get::<Name<'_>>(key).is_some_and(|name| is_jpx(&name))
dict.get::<Name<'_>>(key).is_some_and(|name| is_filter(&name))
|| dict
.get::<Array<'_>>(key)
.is_some_and(|filters| filters.iter::<Name<'_>>().any(|name| is_jpx(&name)))
.is_some_and(|filters| filters.iter::<Name<'_>>().any(|name| is_filter(&name)))
})
}
@@ -755,7 +765,15 @@ fn decode_context<'a>(
let decoded = obj
.stream
.decoded_image(&decode_params)
.map_err(|_| (obj.warning_sink)(InterpreterWarning::ImageDecodeFailure))
.map_err(|_| {
// PdfCraft patch (73): left out for the page's allowance, it is content the page's
// budget skipped, which is reported, and no failure of this image.
(obj.warning_sink)(if hayro_syntax::page_jbig2_work_spent() && uses_jbig2(&obj.stream) {
InterpreterWarning::ContentTruncated
} else {
InterpreterWarning::ImageDecodeFailure
});
})
.ok()?;
let (mut scale_x, mut scale_y) = (1.0, 1.0);
@@ -1008,6 +1026,15 @@ fn decode_raster(
.flatten()
};
// PdfCraft patch (73): a bad mask leaves the image as it is, but one left out for the page's
// allowance takes the image with it. Both are reported, and only here is the file not to blame
// for what the mask would have hidden being shown.
let dict = obj.stream.dict();
let mask = dict.get::<Stream<'_>>(SMASK).or_else(|| dict.get::<Stream<'_>>(MASK));
if alpha.is_none() && hayro_syntax::page_jbig2_work_spent() && mask.is_some_and(|m| uses_jbig2(&m)) {
return None;
}
Some(DecodedRaster { image, alpha })
}
vendor/hayro-jbig2/src/lib.rs
Vendored
+64 -22
View File
@@ -195,11 +195,37 @@ impl<'a> Image<'a> {
/// This is useful in case you want to convert multiple JBIG2 images,
/// as it allows `hayro-jbig2` to reuse allocations during decoding.
pub fn decode_with<D: Decoder>(&self, decoder: &mut D, ctx: &mut DecoderContext) -> Result<()> {
decode_segments(&self.segments, self.height_from_stripes, ctx)?;
let mut work_left = bitmap::MAX_WORK;
decode_segments(&self.segments, self.height_from_stripes, ctx, &mut work_left)?;
emit_bitmap(&ctx.page_bitmap, decoder);
Ok(())
}
/// PdfCraft patch: [`Image::decode`], spending from `work`, what decoding may still cost
/// across a run of images. An image spends what it took, failing or not, starting with its
/// page's pixels, which whoever asked for them will go through, and its segments, so images
/// that decode nothing add up too. On top of that, it takes at most [`bitmap::MAX_WORK`]
/// whatever `work` says. One that `work` can't pay for spends all of it, so a run ends rather
/// than trying every image it has left. Reading the stream comes before this and is charged
/// by [`charge_reading`].
pub fn decode_within<D: Decoder>(&self, decoder: &mut D, work: &mut u64) -> Result<()> {
let start = bitmap::decode_work(self.width, self.height).saturating_add(self.segments.len() as u64);
if charge_work(work, start).is_err() {
*work = 0;
bail!(OverflowError::DecodingWork);
}
let allowed = bitmap::MAX_WORK.min(*work);
let mut left = allowed;
let mut ctx = DecoderContext::default();
let decoded = decode_segments(&self.segments, self.height_from_stripes, &mut ctx, &mut left);
let refused = decoded == Err(OverflowError::DecodingWork.into());
*work = if refused && allowed < bitmap::MAX_WORK { 0 } else { *work - (allowed - left) };
decoded?;
emit_bitmap(&ctx.page_bitmap, &mut *decoder);
Ok(())
}
}
fn emit_bitmap<D: Decoder>(bitmap: &Bitmap, decoder: &mut D) {
@@ -259,13 +285,29 @@ fn emit_bitmap<D: Decoder>(bitmap: &Bitmap, decoder: &mut D) {
}
}
/// PdfCraft patch: count `work` more decoding work, refusing past [`bitmap::MAX_WORK`] before what
/// asks for it is decoded.
/// PdfCraft patch: count `work` more decoding work, refusing past what is `left` of
/// [`bitmap::MAX_WORK`] or a run's allowance before what asks for it is decoded.
fn charge_work(left: &mut u64, work: u64) -> Result<()> {
*left = left.checked_sub(work).ok_or(OverflowError::DecodingWork)?;
Ok(())
}
/// PdfCraft patch: the steps of reading a byte of a stream or its globals: parsing and sorting
/// segments, and building their tables, measured at about 3 a byte at the worst.
const READ_WORK: u64 = 4;
/// PdfCraft patch: charge `work`, what a run of images may still cost, for reading `bytes` of a stream
/// and its globals. This comes before anything is parsed, as it counts whether or not they turn out
/// to decode: an image that fails has still cost its reading, and it is read again for each use. What
/// `work` can't pay for spends all of it.
pub fn charge_reading(work: &mut u64, bytes: usize) -> Result<()> {
if charge_work(work, (bytes as u64).saturating_mul(READ_WORK)).is_err() {
*work = 0;
bail!(OverflowError::DecodingWork);
}
Ok(())
}
/// PdfCraft patch: the work of a region decoded pixel by pixel.
fn region_work(region: &decode::RegionSegmentInfo) -> u64 {
bitmap::decode_work(region.width, region.height)
@@ -275,6 +317,7 @@ fn decode_segments(
segments: &[segment::Segment<'_>],
height_from_stripes: Option<u32>,
decoder_ctx: &mut DecoderContext,
work_left: &mut u64,
) -> Result<()> {
// Find and parse page information segment first.
if let Some(page_info) = segments
@@ -295,7 +338,6 @@ fn decode_segments(
let page_bitmap = &mut decoder_ctx.page_bitmap;
let page_state = &mut decoder_ctx.page_state;
let scratch_buffers = &mut decoder_ctx.scratch_buffers;
let mut work_left = bitmap::MAX_WORK;
// Process all segments.
for seg in segments {
@@ -308,7 +350,7 @@ fn decode_segments(
SegmentType::ImmediateGenericRegion | SegmentType::ImmediateLosslessGenericRegion => {
let had_unknown_length = seg.header.data_length.is_none();
let header = generic::parse(&mut reader, had_unknown_length)?;
charge_work(&mut work_left, region_work(&header.region_info))?;
charge_work(work_left, region_work(&header.region_info))?;
if page_state.can_decode_directly(page_bitmap, &header.region_info, false) {
generic::decode_into(&header, page_bitmap, scratch_buffers)?;
@@ -326,13 +368,13 @@ fn decode_segments(
SegmentType::IntermediateGenericRegion => {
// Intermediate segments cannot have unknown length.
let header = generic::parse(&mut reader, false)?;
charge_work(&mut work_left, region_work(&header.region_info))?;
charge_work(work_left, region_work(&header.region_info))?;
let region = generic::decode(&header, scratch_buffers)?;
page_state.store_region(seg.header.segment_number, region.bitmap)?;
}
SegmentType::PatternDictionary => {
let header = pattern::parse(&mut reader)?;
charge_work(&mut work_left, header.work())?;
charge_work(work_left, header.work())?;
let dictionary = pattern::decode(&header, scratch_buffers, page_state.retained_room())?;
page_state.store_pattern_dictionary(seg.header.segment_number, dictionary)?;
}
@@ -341,7 +383,7 @@ fn decode_segments(
// (6.5.5, step 1)
// Collect references to avoid cloning; symbols are only cloned if re-exported.
let input_symbols = page_state.referred_symbols(&seg.header.referred_to_segments)?;
charge_work(&mut work_left, input_symbols.len() as u64)?;
charge_work(work_left, input_symbols.len() as u64)?;
// Collect Huffman tables from referred table segments.
let referred_tables: Vec<HuffmanTable> = seg
@@ -368,14 +410,14 @@ fn decode_segments(
&page_state.standard_tables,
retained_contexts,
page_state.retained_room(),
&mut work_left,
work_left,
)?;
page_state.store_symbol_dictionary(seg.header.segment_number, dictionary)?;
}
SegmentType::ImmediateTextRegion | SegmentType::ImmediateLosslessTextRegion => {
// Collect symbols from referred symbol dictionaries (SBSYMS).
let symbols = page_state.referred_symbols(&seg.header.referred_to_segments)?;
charge_work(&mut work_left, symbols.len() as u64)?;
charge_work(work_left, symbols.len() as u64)?;
// Collect Huffman tables from referred table segments.
// "These user-supplied Huffman decoding tables may be supplied either
@@ -389,10 +431,10 @@ fn decode_segments(
.collect();
let header = text::parse(&mut reader, symbols.len() as u32)?;
charge_work(&mut work_left, bitmap::copy_work(header.region_info.width, header.region_info.height))?;
charge_work(work_left, bitmap::copy_work(header.region_info.width, header.region_info.height))?;
if header.flags.use_huffman {
// Its symbol ID table has a line for each symbol.
charge_work(&mut work_left, (symbols.len() as u64).saturating_mul(bitmap::HUFFMAN_LINE_WORK))?;
charge_work(work_left, (symbols.len() as u64).saturating_mul(bitmap::HUFFMAN_LINE_WORK))?;
}
if page_state.can_decode_directly(
@@ -407,7 +449,7 @@ fn decode_segments(
&page_state.standard_tables,
page_bitmap,
scratch_buffers,
&mut work_left,
work_left,
)?;
} else {
let region = text::decode(
@@ -416,7 +458,7 @@ fn decode_segments(
&referred_tables,
&page_state.standard_tables,
scratch_buffers,
&mut work_left,
work_left,
)?;
page_bitmap.combine(
&region.bitmap,
@@ -430,7 +472,7 @@ fn decode_segments(
SegmentType::IntermediateTextRegion => {
// Collect symbols from referred symbol dictionaries (SBSYMS).
let symbols = page_state.referred_symbols(&seg.header.referred_to_segments)?;
charge_work(&mut work_left, symbols.len() as u64)?;
charge_work(work_left, symbols.len() as u64)?;
// Collect Huffman tables from referred table segments.
let referred_tables: Vec<HuffmanTable> = seg
@@ -442,10 +484,10 @@ fn decode_segments(
.collect();
let header = text::parse(&mut reader, symbols.len() as u32)?;
charge_work(&mut work_left, bitmap::copy_work(header.region_info.width, header.region_info.height))?;
charge_work(work_left, bitmap::copy_work(header.region_info.width, header.region_info.height))?;
if header.flags.use_huffman {
// Its symbol ID table has a line for each symbol.
charge_work(&mut work_left, (symbols.len() as u64).saturating_mul(bitmap::HUFFMAN_LINE_WORK))?;
charge_work(work_left, (symbols.len() as u64).saturating_mul(bitmap::HUFFMAN_LINE_WORK))?;
}
let region = text::decode(
&header,
@@ -453,7 +495,7 @@ fn decode_segments(
&referred_tables,
&page_state.standard_tables,
scratch_buffers,
&mut work_left,
work_left,
)?;
page_state.store_region(seg.header.segment_number, region.bitmap)?;
}
@@ -466,7 +508,7 @@ fn decode_segments(
.ok_or(SegmentError::MissingPatternDictionary)?;
let header = halftone::parse(&mut reader)?;
charge_work(&mut work_left, halftone::work(&header, pattern_dict))?;
charge_work(work_left, halftone::work(&header, pattern_dict))?;
if page_state.can_decode_directly(
page_bitmap,
@@ -494,7 +536,7 @@ fn decode_segments(
.ok_or(SegmentError::MissingPatternDictionary)?;
let header = halftone::parse(&mut reader)?;
charge_work(&mut work_left, halftone::work(&header, pattern_dict))?;
charge_work(work_left, halftone::work(&header, pattern_dict))?;
let region = halftone::decode(&header, pattern_dict, scratch_buffers)?;
page_state.store_region(seg.header.segment_number, region.bitmap)?;
}
@@ -508,7 +550,7 @@ fn decode_segments(
.unwrap_or(page_bitmap);
let header = generic_refinement::parse(&mut reader)?;
charge_work(&mut work_left, region_work(&header.region_info))?;
charge_work(work_left, region_work(&header.region_info))?;
let region = generic_refinement::decode(&header, reference, scratch_buffers)?;
page_state.store_region(seg.header.segment_number, region.bitmap)?;
}
@@ -526,7 +568,7 @@ fn decode_segments(
.and_then(|&num| page_state.get_referred_segment(num));
let header = generic_refinement::parse(&mut reader)?;
charge_work(&mut work_left, region_work(&header.region_info))?;
charge_work(work_left, region_work(&header.region_info))?;
if let Some(referred_segment) = referred_segment
&& page_state.can_decode_directly(page_bitmap, &header.region_info, false)
vendor/hayro-syntax/src/filter/jbig2.rs
Vendored
+67 -2
View File
@@ -9,6 +9,65 @@ use alloc::vec;
use alloc::vec::Vec;
use core::iter;
/// PdfCraft patch (31): the steps of decoding a page's JBIG2 images may take between them, as many
/// as one may. Each has its own limit, which doesn't bound a page of many made to reach it. That
/// is about 6 s to 17 s, inside a render's watchdog; an image of the largest size costs a quarter of
/// it to decode and a quarter for its page, and ordinary ones are far below that.
pub const MAX_PAGE_JBIG2_WORK: u64 = 1 << 30;
#[cfg(feature = "std")]
std::thread_local! {
// Unbounded until a page says otherwise: what isn't drawing a page has no page to run out on.
static PAGE_WORK_LEFT: core::cell::Cell<u64> = const { core::cell::Cell::new(u64::MAX) };
}
/// PdfCraft patch (31): what the JBIG2 images decoded next on this thread may take, for a new page.
#[cfg(feature = "std")]
pub fn set_page_jbig2_work(work: u64) {
PAGE_WORK_LEFT.with(|left| left.set(work));
}
/// Pays for reading `bytes` of a stream and its globals out of what the page has left, or `None`.
/// An image that doesn't decode isn't kept, so a page drawing it again and again reads it again.
#[cfg(feature = "std")]
fn charge_page_reading(bytes: usize) -> Option<()> {
let mut left = PAGE_WORK_LEFT.with(core::cell::Cell::get);
let charged = hayro_jbig2::charge_reading(&mut left, bytes);
PAGE_WORK_LEFT.with(|cell| cell.set(left));
charged.ok()
}
#[cfg(not(feature = "std"))]
fn charge_page_reading(_: usize) -> Option<()> {
Some(())
}
/// `image` decoded through `decoder`, spending what the page has left, if there is anything.
#[cfg(feature = "std")]
fn decode_for_page<D: hayro_jbig2::Decoder>(image: &hayro_jbig2::Image<'_>, decoder: &mut D) -> Option<()> {
let mut left = PAGE_WORK_LEFT.with(core::cell::Cell::get);
let decoded = image.decode_within(decoder, &mut left);
PAGE_WORK_LEFT.with(|cell| cell.set(left));
decoded.ok()
}
#[cfg(not(feature = "std"))]
fn decode_for_page<D: hayro_jbig2::Decoder>(image: &hayro_jbig2::Image<'_>, decoder: &mut D) -> Option<()> {
image.decode(decoder).ok()
}
/// PdfCraft patch (31): whether the page has nothing left for another image, which then isn't read
/// at all.
#[cfg(feature = "std")]
pub fn page_jbig2_work_spent() -> bool {
PAGE_WORK_LEFT.with(core::cell::Cell::get) == 0
}
#[cfg(not(feature = "std"))]
pub fn page_jbig2_work_spent() -> bool {
false
}
/// Decode JBIG2 data from a PDF stream.
///
/// The `params` dictionary may contain a `JBIG2Globals` entry pointing to
@@ -18,6 +77,10 @@ pub(crate) fn decode(
params: &Dict<'_>,
image_params: &ImageDecodeParams,
) -> Option<FilterResult<'static>> {
if page_jbig2_work_spent() {
return None;
}
// PdfCraft patch: globals are raw segments, so JBIG2 inside them means nothing, and decoding
// it reads its own globals in turn: a chain of 16,000 such streams overflowed the stack, which
// aborts the process. Such a stream reads as globals that can't be decoded.
@@ -26,6 +89,8 @@ pub(crate) fn decode(
.filter(|g| !g.filters().iter().any(|f| matches!(f, Filter::Jbig2Decode)))
.and_then(|g| g.decoded().ok());
charge_page_reading(data.len().saturating_add(globals.as_deref().map_or(0, <[u8]>::len)))?;
let image = hayro_jbig2::Image::new_embedded(data, globals.as_deref()).ok()?;
// Whenever possible (if we don't have an indexed color space), we convert
@@ -60,7 +125,7 @@ pub(crate) fn decode(
let writer = BitWriter::new(&mut packed, 1)?;
let mut decoder = BitWriterDecoder { writer };
image.decode(&mut decoder).ok()?;
decode_for_page(&image, &mut decoder)?;
(packed, 1)
} else {
@@ -83,7 +148,7 @@ pub(crate) fn decode(
}
let mut decoder = Luma8Decoder { output: Vec::new() };
image.decode(&mut decoder).ok()?;
decode_for_page(&image, &mut decoder)?;
(decoder.output, 8)
};
vendor/hayro-syntax/src/filter/mod.rs
Vendored
+4
View File
@@ -9,6 +9,10 @@ mod dct;
#[cfg(feature = "images")]
mod jbig2;
#[cfg(feature = "images")]
pub use jbig2::{MAX_PAGE_JBIG2_WORK, page_jbig2_work_spent};
#[cfg(all(feature = "images", feature = "std"))]
pub use jbig2::set_page_jbig2_work;
#[cfg(feature = "images")]
mod jpx;
mod lzw_flate;
mod run_length;