diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml index d8549142f7..9403a1e044 100644 --- a/.github/workflows/release.yml +++ b/.github/workflows/release.yml @@ -42,3 +42,4 @@ jobs: AZURE_SIGNING_ENDPOINT: ${{ secrets.AZURE_SIGNING_ENDPOINT }} AZURE_SIGNING_ACCOUNT: ${{ secrets.AZURE_SIGNING_ACCOUNT }} AZURE_CERT_PROFILE: ${{ secrets.AZURE_CERT_PROFILE }} + LINUX_SIGNING_KEY: ${{ secrets.LINUX_SIGNING_KEY }} diff --git a/ATTRIBUTION.md b/ATTRIBUTION.md index c70d84e624..e8fa4adac6 100644 --- a/ATTRIBUTION.md +++ b/ATTRIBUTION.md @@ -4,7 +4,7 @@ Every asset PdfCraft includes, bundles or uses to build its published material, with its author, source and licence. The policy is in [AGENTS.md](AGENTS.md) §1. The machine-readable list, with the SHA-256 of every asset, is [ATTRIBUTION.toml](ATTRIBUTION.toml). Licence texts are kept beside the assets and summarised in [NOTICE](NOTICE). -## In this repository (256) +## In this repository (257) | Asset | Title | Author | Licence | Source | Used for | |---|---|---|---|---|---| @@ -263,6 +263,7 @@ Every asset PdfCraft includes, bundles or uses to build its published material, | `assets/icons/superscript.svg` | Lucide icon "superscript" | Lucide Contributors (portions Cole Bemis, Feather) | ISC AND MIT | https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/superscript.svg) | UI icon, embedded via crates/ui-egui/src/icon_data.rs | | `assets/icons/subscript.svg` | Lucide icon "subscript" | Lucide Contributors (portions Cole Bemis, Feather) | ISC AND MIT | https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/subscript.svg) | UI icon, embedded via crates/ui-egui/src/icon_data.rs | | `assets/icons/pilcrow.svg` | Lucide icon "pilcrow" | Lucide Contributors (portions Cole Bemis, Feather) | ISC AND MIT | https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/pilcrow.svg) | UI icon, embedded via crates/ui-egui/src/icon_data.rs | +| `assets/icons/a-large-small.svg` | Lucide icon "a-large-small" | Lucide Contributors (portions Cole Bemis, Feather) | ISC AND MIT | https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/a-large-small.svg) | UI icon, embedded via crates/ui-egui/src/widgets/icon_data.rs | | `assets/icons/move-horizontal.svg` | Lucide icon "move-horizontal" | Lucide Contributors (portions Cole Bemis, Feather) | ISC AND MIT | https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/move-horizontal.svg) | UI icon, embedded via crates/ui-egui/src/icon_data.rs | ## Compiled in through dependencies (6) diff --git a/ATTRIBUTION.toml b/ATTRIBUTION.toml index 9638749929..f998b6fc3d 100644 --- a/ATTRIBUTION.toml +++ b/ATTRIBUTION.toml @@ -3200,6 +3200,17 @@ kind = "icon" usage = "UI icon, embedded via crates/ui-egui/src/icon_data.rs" sha256 = "d344d9f9222f6c1b698c0c82a753a2212d9bd7785f06c00cd6e4910fe2fee8f8" +[[asset]] +path = "assets/icons/a-large-small.svg" +title = "Lucide icon \"a-large-small\"" +author = "Lucide Contributors (portions Cole Bemis, Feather)" +source = "https://github.com/lucide-icons/lucide (npm lucide-static 1.49.0, icons/a-large-small.svg)" +licence = "ISC AND MIT" +licence_file = "assets/icons/LICENSE-lucide.txt" +kind = "icon" +usage = "UI icon, embedded via crates/ui-egui/src/widgets/icon_data.rs" +sha256 = "d4985fc71174d108e322c3b25cb6abe2837a115fe3f239a1918645db436743ec" + [[asset]] path = "assets/icons/move-horizontal.svg" title = "Lucide icon \"move-horizontal\"" diff --git a/README.md b/README.md index 3c1b67936a..f9a41cb3ca 100644 --- a/README.md +++ b/README.md @@ -94,7 +94,7 @@ Real-world typography: world scripts, vertical Japanese, colour emoji, gradients ### Fearless -Every save appends your changes and leaves the original bytes untouched. Writes are atomic, undo runs deep, and nothing is lost if you close by mistake. +Ordinary saves append your changes and leave the original bytes untouched. Edits to damaged objects whose xref and header generations disagree use a full rewrite, keeping both accepted references connected to the same repaired object. Writes are atomic, undo runs deep, and nothing is lost if you close by mistake. diff --git a/ROADMAP.md b/ROADMAP.md index 00dd8fdee0..1f753c500f 100644 --- a/ROADMAP.md +++ b/ROADMAP.md @@ -21,9 +21,14 @@ > **Last reviewed:** 2026-10-23 · **Last updated:** 2026-10-23 · **Change:** minor (M2, PR #4172: extend on-screen CJK fallback to Combine thumbnails and figure previews; recognise FangSong and subset-prefixed Founder names; estimates unchanged) > **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (typed small Edit font sizes and Fill & Sign text/date colour, #2444/#2496; estimates unchanged) +> **Last reviewed:** 2026-10-20 · **Last updated:** 2026-10-20 · **Change:** patch (M1, #1143: repaired aliases survive preflight, snapshots and normalized saves; retained materialized revisions report generation disagreements; estimates unchanged) + +> **Last reviewed:** 2026-10-18 · **Last updated:** 2026-10-18 · **Change:** minor (M1, #1143: explicit live xref/header generation repair policy and bounded diagnostics; estimates unchanged) + +> **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (M1: refresh object-generation reference identity against current upstream; estimates unchanged) **Stage: alpha** · next: beta, ~25% (ready for real work 50% → 75%) and ~400–700 h away -> **Last reviewed:** 2026-10-29 | **Last updated:** 2026-10-29 | **Change:** minor (M7 #2517: a narrowed then widened paragraph keeps the typed spaces the wrap consumed, after save and reopen; estimates unchanged) +> **Last reviewed:** 2026-10-29 | **Last updated:** 2026-10-29 | **Change:** minor (M2: scans keep their strokes at screen size and are smoothed when zoomed; M7 #2517: a narrowed then widened paragraph keeps the typed spaces the wrap consumed, after save and reopen; estimates unchanged) > **Last reviewed:** 2026-10-27 | **Last updated:** 2026-10-27 | **Change:** minor (M10: OCR model downloads use verified ONNX pairs with legacy discovery fallback; estimates unchanged) > **Last reviewed:** 2026-10-25 | **Last updated:** 2026-10-25 | **Change:** minor (M14 localization current state: Arabic logical catalog, explicit RTL label direction and mirrored stock chrome; native/browser RTL workflows unqualified. Native release Hans face via LXGW WenKai Lite; estimates unchanged) > **Last reviewed:** 2026-10-23 · **Last updated:** 2026-10-23 · **Change:** minor (M7, #2636 partial: pasting added text boxes and images in the engine and through `content_copy`; estimates unchanged) @@ -235,6 +240,8 @@ Ranked; detail and the full milestone table in [docs/roadmap.md](docs/roadmap.md - **2026-11-02 (M10, #5743):** macOS print jobs name their title and owner (`com.apple.print.JobInfo.PMJobName`, `PMJobOwner`), as the system print dialog does, because a Kyocera filter crashes on a job without them. Checked by a test and through `doc_print` against a stand-in `lp`; not yet on a physical printer. +- **2026-11-02 (M14, #5737, #5460):** Preferences can turn off the update check at start. It stays on by default, and Help ▸ Check for updates still asks. + - **2026-11-01 (M7, #5657):** Header/footer, watermark and background additions and replacements stage the complete call before committing. Synthetic later-page failures preserve exact incremental bytes, object count, dirty state and allocation cursor; success controls retain other marks and shared/inherited resources. Estimates unchanged. - **2026-11-01 (M14, #5572):** Preferences ▸ Documents and view sets how many recent files to keep, from 0 to 50 with 12 by default. At 0 Home has no Recent section and File ▸ Open Recent is disabled. The list is cut when Preferences closes, so the number can pass through 0 on its way to another. @@ -261,6 +268,7 @@ Ranked; detail and the full milestone table in [docs/roadmap.md](docs/roadmap.md - **2026-10-26 (M2, PR #4172):** Reconciled current main, preserving its Japanese serif/weight selection and both branches' progress records. Installed CJK viewer and preview fallback remains display-only; estimates unchanged. - **2026-10-29 (M7, #2517):** Widening a paragraph rejoins a wrap with the number of spaces that wrap consumed, including a typed run and none for a Chinese or Japanese word split. The count survives an incremental save and a full rewrite, and a width-only `text_edit` uses it. Estimates unchanged. +- **2026-10-29 (M2):** Scanned pages look like the paper again. An image that doesn't ask for /Interpolate was shrunk by nearest-neighbour (since #635), which kept one sample in N per axis, so the text of a 300 dpi scan at 100% lost most of its strokes; it is now area-averaged to whole device pixels (on a 13-page 300 dpi scan every sample at 96–192 dpi is within 1 of the exact area average; it was up to 255 off). Enlarged, an image of page-scan size is smoothed bilinearly, so a zoomed scan has no staircases; icons, QR codes and pixel art keep hard pixels. Estimates unchanged. - **2026-10-26 (M11, JBIG2 symbols):** Lossless JBIG2 now stores each distinct mark of a black-and-white image once, as a symbol, and places its exact copies, when that is smaller than the generic region: on six pages of rendered 300 dpi text 28% of the generic region's size and 17% of Group 4's, and the same bits back through two independent decoders (hayro-jbig2 and Poppler). Scans, whose letters are seldom identical, gain little until refinement coding; a dictionary shared between pages (`/JBIG2Globals`) is next. Estimates unchanged. - **2026-10-25 (M5, comment replies):** One reply can be deleted from the Comments panel without its thread. On the selected card each reply has its own "…" menu (Edit, Copy text, Delete); Delete takes that reply with its pop-up and the replies to it, as `comment_delete` already did headlessly, and the comment and its other replies stay. The card stays selected, and the panel's editor stays open, on what they were on though the delete moved it in `/Annots`. Undo and Redo close the editor, since either can move what it was open on. A locked reply can't be deleted. Estimates unchanged. - **2026-10-24 (M10, #4158 partial):** Word, HTML, RTF and PowerPoint export keep a block that continues another on its line (an italic case name, a link, a superscript) as a run in its own style inside that paragraph, with the source's word space, instead of a paragraph of its own. Style changes that wrap onto the next line still split paragraphs (the reading blocks keep style boundaries). Estimates unchanged. @@ -336,6 +344,8 @@ Ranked; detail and the full milestone table in [docs/roadmap.md](docs/roadmap.md - **2026-10-20 (M5, arrow keys move comments):** With a comment selected, the arrow keys move it a screen pixel and ⌘/Ctrl or Shift and an arrow ten, instead of scrolling the view or turning the page; up is up on screen at any zoom and rotation, and each press is one undo step. Text markup and locked comments stay put. Estimates unchanged. +- **2026-10-20 (M1, PR #1143 review):** Repair-aware dictionary/preflight edits and snapshot restoration preserve both accepted generations of a damaged live slot. Such edits use the existing full-save fallback; classic and packed rewrites coalesce aliases so page-tree parent links retain one object identity. Explicit generation replacement remains single-generation. Materialized xref diagnostics inspect retained primary and hybrid rows before folding, retaining the 256-object cap plus summary and discarding abandoned-index notes. Synthetic red/green controls cover preflight, undo restoration, equal values, explicit narrowing, encryption packing and four save variants. Estimates unchanged. + - **2026-10-19 (M14, #3373):** The main menu and its long submenus scroll in short windows, keeping their commands reachable. Estimates unchanged. - **2026-10-19 (M3, #3387):** Save-prompt shortcuts yield to later consent and utility dialogs; a synthetic UI regression checks that an edited document stays open. Estimates unchanged. @@ -399,6 +409,8 @@ Ranked; detail and the full milestone table in [docs/roadmap.md](docs/roadmap.md - **2026-10-15 (community, M6, #2608):** In Prepare a form the selected field moves with the arrow keys: a point per press, ten with Shift, as the object nudge and dragging it do. Each press is one undo step ("Change field properties"), and the keys no longer turn the page while a field is selected. Estimates unchanged. - **2026-10-15 (M3, closable community card):** The ArtCraft community card on Home has a close button that hides it for the rest of the run; it returns on the next launch, and its links stay in the top bar's Discord button and the Help menu. A UI regression covers closing and the fresh-launch default. Estimates unchanged. - **2026-10-13 (community, M3, #1569 review):** Pixel-only Wayland wheel, wheel-tilt and continuous input use standalone moved events, so Ctrl zoom does not remain latched after the modifier is released (#991); vendored winit patch (13), with a policy test. Estimates unchanged. +- **2026-10-18 (M1, PR #1143 repair policy):** A live in-file reference resolves when its generation matches the xref row or a same-number header at that indexed offset. Neither-match and free slots remain null, including when old object bytes remain. Opening records bounded xref/header disagreements without parsing object bodies or expanding compact xref rows. Synthetic controls cover read order, clones, overlays, dictionary edits, full classic/packed saves and an independent reader of incremental output. **Correction to the 2026-10-15 note below:** its async-opening-test sentence was not evidenced by this COS change; no fresh async-opening or GUI validation is claimed. The full-reference stream-length recursion key is defensive hardening, not an independently isolated regression fix. Estimates unchanged. +- **2026-10-15 (M1, PR #1143 refresh):** COS reference lookups and full saves distinguish complete object references, so stale generations cannot expose replacement content. Snapshot restoration also keeps edits to objects with nonzero generations before incremental saves. Synthetic regressions cover caches, overlays, snapshots, stream lengths and classic/packed saves. Current upstream async-opening test corrections are included. Estimates unchanged. - **2026-10-15 (M14, #2383 candidate):** Device-limit requests now clamp unsupported maxima and minimum alignments before enabling the adapter's full 2D texture extent. This addresses the reported V3D request of 16 shader variables on an adapter supporting 15. Synthetic regressions fail before the clamp and pass afterward, along with the existing HiDPI case and Windows GPU-required application tests. Raspberry Pi startup remains unverified; milestone and overall estimates are unchanged. - **2026-10-12 (M6):** XFDF field import and data-file extraction join all text nodes in a value, preserving text after XML comments and processing instructions. Synthetic regressions cover Unicode, CDATA, empty values and PDF save/reopen. Estimates unchanged. - **2026-10-12 (M7, picture right-click menu):** The Edit text & images menu of a picture gains Cut Image, Redact Image (a redaction mark over the picture, applied from Redact a PDF), Create Link… and Recognize Text… (for that page), in all 21 catalogs. The menu was drawn only while the pointer was over the picture, so its items outside the picture (most of a logo's menu, and Delete below a page) could not be clicked; an open menu now stays drawn. Still missing from Acrobat's menu: Arrange (bring to front, send backward), Remove crop, Draw freehand and Edit image using… (opening an external editor). @@ -777,6 +789,7 @@ Newest first: the date, what moved, and the new overall estimate. | 2026-10-20 | minor | M6, #3502: form detection preserves comments across content streams; estimates unchanged. | | 2026-10-20 | minor | M10, #3532 partial: vector printing retains existing appearances with absent Subtype; estimates unchanged. | | 2026-10-20 | minor | M11 #3507: image merging preserves comments and state across content streams. Estimates unchanged. | +| 2026-10-18 | minor | M1, #1143: a live reference may match the xref generation or a same-number header at its indexed offset; neither-match and free slots remain null. Bounded repair diagnostics and synthetic cache, free-slot, edit and classic/packed save regressions; estimates unchanged. | | 2026-10-18 | minor | M5, PR #1971: nested source `/Page` and `/Pages` dictionaries export as null without traversing page content or the page tree. Synthetic saved/reopened three-page regression retains unrelated metadata and aliases. FDF import semantics, copied-graph limits and estimates unchanged. | | 2026-10-15 | minor | M5: budget retained FDF entries before cloning; discovery keeps lazy array owners and annotation positions, with alias/cache-eviction stress fixtures. Limits cover the copied graph; source parsing/page discovery retain existing COS/parser limits, with no total source-memory bound. FDF import semantics and estimates unchanged. | | 2026-10-15 | minor | M7: the Edit text caret is as tall as the row it is on in a region with several sizes; estimates unchanged. | @@ -787,6 +800,7 @@ Newest first: the date, what moved, and the new overall estimate. | 2026-10-14 | minor | M1: raw-deflate fallback preserves decoding-limit errors in strict and tolerant decoding; synthetic filter and COS stream regressions. Estimates unchanged. | | 2026-10-16 | minor | M14: Czech catalog completed (2,674 entries) and "tag" terminology unified. Estimates unchanged. | | 2026-10-13 | minor | M9: separate page validation overlays preserve signature artwork, and snapshot tabs reuse instead of multiplying; baked captions remain unchanged (#79 open); estimates unchanged | +| 2026-10-15 | minor | M1: refresh object-generation reference identity, retaining current upstream progress and regression tests. Estimates unchanged. | | 2026-10-16 | minor | M10 #1201 partial: lossless image export applies component Decode pairs and Indexed mapping. Estimates unchanged. | | 2026-10-15 | minor | M6: CSV export and spreadsheet merging quote tabs, and import distinguishes quoted tabs from TSV delimiters without changing field identities. Estimates unchanged. | | 2026-10-13 | minor | M4: rotation queries and edits resolve indirect page-tree values; inherited, negative and large rotations verified through both saves. Estimates unchanged. | diff --git a/assets/icons/a-large-small.svg b/assets/icons/a-large-small.svg new file mode 100644 index 0000000000..7716db3cf2 --- /dev/null +++ b/assets/icons/a-large-small.svg @@ -0,0 +1,18 @@ + + + + + + + diff --git a/crates/cos/README.md b/crates/cos/README.md index 7b576f293f..398b161625 100644 --- a/crates/cos/README.md +++ b/crates/cos/README.md @@ -53,7 +53,18 @@ another. Edits never touch the original bytes. - **Every cross-reference form:** classic tables with their quirky subsections, cross-reference streams, hybrid-reference files (`/XRefStm`), and object streams. - **Repair.** A file whose xref is wrong or missing is reconstructed by scanning for `obj` - headers. Repairs are *recorded*, never silent, so the UI can say the file was damaged. + headers. Repairs are *recorded*, never silent, so the UI can say the file was damaged. A live + in-file reference may match either its xref generation or a same-number object header at + that indexed offset; a missing, free, or neither-matching slot still resolves to null. + Header probes inspect at most 128 bytes at the indexed and header-relative offsets, without + parsing object bodies. On open, disagreements in retained xref revisions are recorded once + per object, up to 256 objects plus an omitted-details summary. This diagnostic scan does + not expand compact xref ranges; the winning xref entry remains authoritative for free + slots. Compressed objects require generation zero. Explicit `set` overlays accept only the + supplied generation; `set_existing` and `update_dict` retain both accepted generations of a + damaged live slot, including through snapshots, and require full rewrite. Headers beyond + this bounded probe are not additionally recovered. Materialized primary and hybrid rows are + diagnosed before folding, so a newer row does not hide a retained disagreement. - **Tolerant, never fatal.** The lexer works on a byte slice and a position; it allocates nothing for skipped data, bounds-checks every index, and caps nesting depth. Malformed input yields `CosError`, never a panic — this crate is the first thing a hostile file meets. Exhausted xref entries are skipped with a repair log; exhausted /Size hints allocate unused numbers without overflowing or replacing existing objects. @@ -66,14 +77,17 @@ another. Edits never touch the original bytes. ## Writing -`write_incremental` appends only the edited objects, a cross-reference section **in the same -style as the file's last one**, and a trailer with `/Prev`. The original bytes are never modified, -which is what keeps a signature over an earlier revision valid. +`write_incremental` ordinarily appends only the edited objects, a cross-reference section **in +the same style as the file's last one**, and a trailer with `/Prev`, preserving the original +bytes and signatures over earlier revisions. It delegates to `write_full` when reconstruction, +encryption changes or `full_save_required()` demand a rewrite. Repair-aware edits to a live +xref/header generation disagreement use this fallback rather than append a contradictory row. `write_full` writes only what the trailer can reach, renumbered densely — the Save As and garbage-collecting path. By default (`SaveOptions::object_streams`) it packs non-stream objects into compressed object streams with a cross-reference stream (PDF 1.5+); turning that off writes a -classic table any reader accepts. +classic table any reader accepts. Both forms map accepted xref/header aliases to one output +object, retaining shared identity such as page-tree parent links; unrelated generations stay null. Dictionaries keep their key order and streams keep their *encoded* bytes, so rewritten objects stay recognisable in a diff and byte-stable when re-serialized. diff --git a/crates/cos/src/document/objects.rs b/crates/cos/src/document/objects.rs index 62ef1e74cc..8d9917d749 100644 --- a/crates/cos/src/document/objects.rs +++ b/crates/cos/src/document/objects.rs @@ -1,6 +1,8 @@ +use std::collections::HashSet; + use super::reader::find; use super::state::{LENGTHS, LoadGuard, ObjStm, ObjectReader, ReadState, Slot, SourceIdentity, held_bytes, objstm_member}; -use super::xref::{Revision, XrefEntry}; +use super::xref::{Revision, XrefEntry, XrefIndex}; use super::{ Arc, CosError, Cow, Dict, Document, Lexer, MAX_DECODED, MAX_JBIG2_GLOBALS, ObjRef, Object, Stream, is_whitespace, parse_indirect_shared, }; @@ -129,8 +131,8 @@ impl Document { pub fn xref_entry_is_exact(&self, num: u32) -> bool { match self.xref_entry(num) { None | Some(XrefEntry::Free { .. }) => true, - Some(XrefEntry::InFile { offset, .. }) => { - let at = |off: usize| matches!(parse_indirect_shared(&self.data, off, &|_| None, &self.endstreams), Ok((id, _)) if id.num == num); + Some(XrefEntry::InFile { offset, generation }) => { + let at = |off: usize| matches!(parse_indirect_shared(&self.data, off, &|_| None, &self.endstreams), Ok((id, _)) if id == ObjRef::new(num, generation)); // Offsets may count from a header that does not start the file. usize::try_from(offset).is_ok_and(|off| at(off) || (self.header_offset != 0 && at(off.saturating_add(self.header_offset)))) } @@ -151,9 +153,93 @@ impl Document { // ── object access ─────────────────────────────────────────────────────────────────────── - /// Fetch an object by reference. Missing objects are `Null` (ISO 32000-2 §7.3.10). + /// Fetch an object by reference. Undefined references are `Null` (ISO 32000-2 §7.3.10). + /// A live file slot also accepts its actual header generation when a damaged xref row + /// disagrees; the disagreement is recorded in [`Self::repair_log`]. Free slots never do. pub fn get(&self, r: ObjRef) -> Arc { - self.try_get(r.num).unwrap_or_else(|_| Arc::new(Object::Null)) + self.try_get_reference(r).unwrap_or_else(|_| Arc::new(Object::Null)) + } + + pub(super) fn reference_matches(&self, r: ObjRef) -> bool { + match self.overlay.get(&r.num) { + Some(Slot::Set(generation, _)) => r.generation == *generation, + Some(Slot::RepairedSet(generation, alias, _)) => r.generation == *generation || r.generation == *alias, + Some(Slot::Freed(_)) => false, + None => match self.entries.lookup(r.num, &self.data) { + Some(XrefEntry::InFile { offset, generation }) => { + r.generation == generation || self.indexed_header_reference(r.num, offset) == Some(r) + } + Some(XrefEntry::InStream { .. }) => r.generation == 0, + None | Some(XrefEntry::Free { .. }) => false, + }, + } + } + + pub(super) fn try_get_reference(&self, r: ObjRef) -> Result, CosError> { + if !self.reference_matches(r) { + return Ok(Arc::new(Object::Null)); + } + self.try_get(r.num) + } + + /// A repair probe, not a whole-file search: only the indexed header (or the same offset + /// relative to a shifted PDF header) can establish this additional generation. A small + /// window bounds hostile whitespace/tokens; ordinary xref-matching reads are unchanged. + pub(super) fn indexed_header_reference(&self, num: u32, offset: u64) -> Option { + let at = |off: usize| { + let bytes = self.data.get(off..off.saturating_add(128).min(self.data.len()))?; + let mut lexer = Lexer::new(bytes, 0); + lexer.skip_ws(); + let number = std::str::from_utf8(lexer.token()).ok()?.parse::().ok()?; + lexer.skip_ws(); + let generation = std::str::from_utf8(lexer.token()).ok()?.parse::().ok()?; + (number == num && lexer.eat_keyword(b"obj")).then_some(ObjRef::new(number, generation)) + }; + let off = usize::try_from(offset).ok()?; + at(off).or_else(|| if self.header_offset == 0 { None } else { off.checked_add(self.header_offset).and_then(at) }) + } + + /// Add at most one bounded diagnostic per object. False stops diagnostic probing only; + /// parsers must still read the remaining xref rows needed to build their index. + pub(super) fn inspect_xref_generation_repair(&self, num: u32, entry: XrefEntry, recorded: &mut HashSet, log: &mut Vec) -> bool { + let XrefEntry::InFile { offset, generation } = entry else { return true }; + if recorded.contains(&num) { + return true; + } + let Some(header) = self.indexed_header_reference(num, offset) else { return true }; + if header.generation == generation { + return true; + } + if recorded.len() == 256 { + log.push("additional xref/header generation disagreements omitted after 256 objects".into()); + return false; + } + recorded.insert(num); + log.push(format!("object {num}: xref generation {generation} disagrees with object header generation {}", header.generation)); + true + } + + /// Diagnose once at open without expanding compact ranges or decoding an xref twice. + /// Materialized readers retain bounded notes while they read accepted sections, before + /// newer rows hide older ones. Discarded parse attempts/rebuilt indexes carry no notes. + pub(super) fn note_xref_generation_repairs(&self, log: &mut Vec) { + match &*self.entries { + XrefIndex::Materialized(_, notes) => log.extend(notes.iter().cloned()), + XrefIndex::Compact { revisions, .. } => { + let mut recorded = HashSet::new(); + for revision in revisions { + for section in std::iter::once(&revision.primary).chain(revision.hybrid.iter()) { + for num in section.numbers() { + if let Some(entry) = section.lookup(num, &self.data) + && !self.inspect_xref_generation_repair(num, entry, &mut recorded, log) + { + return; + } + } + } + } + } + } } /// Follow a reference if `o` is one; otherwise return `o` itself. An object can be nothing but @@ -388,7 +474,7 @@ impl Document { pub fn try_get(&self, num: u32) -> Result, CosError> { match self.overlay.get(&num) { - Some(Slot::Set(_, o)) => return Ok(o.clone()), + Some(Slot::Set(_, o) | Slot::RepairedSet(_, _, o)) => return Ok(o.clone()), Some(Slot::Freed(_)) => return Ok(Arc::new(Object::Null)), None => {} } @@ -519,11 +605,11 @@ impl Document { None | Some(XrefEntry::Free { .. }) => Ok(Object::Null), Some(XrefEntry::InFile { offset, .. }) => { let resolve = |r: ObjRef| { - if let Some(known) = LENGTHS.with(|m| m.borrow().get(&r.num).copied()) { + if let Some(known) = LENGTHS.with(|m| m.borrow().get(&r).copied()) { return known; } - let length = self.try_get(r.num).ok().and_then(|o| o.as_int()); - LENGTHS.with(|m| m.borrow_mut().insert(r.num, length)); + let length = self.try_get_reference(r).ok().and_then(|o| o.as_int()); + LENGTHS.with(|m| m.borrow_mut().insert(r, length)); length }; let off = offset as usize; @@ -640,6 +726,38 @@ impl Document { } } + /// Replace an accepted existing reference without losing a damaged source slot's aliases. + /// Unlike [`Self::set`], this preserves both xref/header generations when they disagree. + /// Such an edit forces a full rewrite so the output has consistent references, rather than + /// appending a new contradictory xref row. Missing, free and stale references are errors. + pub fn set_existing(&mut self, r: ObjRef, obj: impl Into) -> Result<(), CosError> { + if !self.reference_matches(r) { + return Err(CosError::MissingObject(r.num)); + } + let alias = self.alternate_generation(r); + if let Some(alias) = alias { + let obj = self.adopt(obj.into()); + self.overlay.insert(r.num, Slot::RepairedSet(r.generation, alias, Arc::new(obj))); + } else { + self.set(r, obj); + } + Ok(()) + } + + /// Other accepted generation for a known-live reference to a damaged slot. + pub(super) fn alternate_generation(&self, r: ObjRef) -> Option { + match self.overlay.get(&r.num) { + Some(Slot::RepairedSet(generation, alias, _)) => Some(if r.generation == *generation { *alias } else { *generation }), + Some(_) => None, + None => match self.entries.lookup(r.num, &self.data) { + Some(XrefEntry::InFile { offset, generation }) => self.indexed_header_reference(r.num, offset).and_then(|header| { + (header.generation != generation).then_some(if r.generation == generation { header.generation } else { generation }) + }), + _ => None, + }, + } + } + /// Add a new indirect object and return its reference. pub fn add(&mut self, obj: impl Into) -> ObjRef { let r = ObjRef::new(self.allocate_num(), 0); @@ -699,7 +817,8 @@ impl Document { let generation = current.generation(num); let same = saved_nums.binary_search(&num).is_ok() && self.generation(num) == generation - && saved_reader.try_get(ObjRef::new(num, 0))? == current_reader.try_get(ObjRef::new(num, 0))?; + && self.alternate_generation(ObjRef::new(num, generation)) == current.alternate_generation(ObjRef::new(num, generation)) + && saved_reader.try_get(ObjRef::new(num, generation))? == current_reader.try_get(ObjRef::new(num, generation))?; if !same { changed.push(num); } @@ -728,8 +847,14 @@ impl Document { let mut current_reader = current.object_reader(); for num in changed { let generation = current.generation(num); - let object = current_reader.try_get(ObjRef::new(num, 0))?; - self.set(ObjRef::new(num, generation), (*object).clone()); + let object = current_reader.try_get(ObjRef::new(num, generation))?; + if let Some(alias) = current.alternate_generation(ObjRef::new(num, generation)) { + let object = self.adopt((*object).clone()); + self.overlay.insert(num, Slot::RepairedSet(generation, alias, Arc::new(object))); + self.next_num = self.next_num.max(num.saturating_add(1)); + } else { + self.set(ObjRef::new(num, generation), (*object).clone()); + } } *self.trailer_mut() = current.trailer().clone(); Ok(()) @@ -740,14 +865,13 @@ impl Document { let mut obj = (*self.get(r)).clone(); let d = obj.as_dict_mut().ok_or(CosError::NotADictionary(r))?; f(d); - self.set(r, obj); - Ok(()) + self.set_existing(r, obj) } /// Current generation of an object number. pub fn generation(&self, num: u32) -> u16 { match self.overlay.get(&num) { - Some(Slot::Set(g, _)) => *g, + Some(Slot::Set(g, _) | Slot::RepairedSet(g, _, _)) => *g, Some(Slot::Freed(g)) => *g, None => match self.entries.lookup(num, &self.data) { Some(XrefEntry::InFile { generation, .. }) => generation, @@ -759,7 +883,7 @@ impl Document { pub(crate) fn overlay_entries(&self) -> impl Iterator>)> + '_ { self.overlay.iter().map(|(n, s)| match s { - Slot::Set(g, o) => (*n, *g, Some(o.clone())), + Slot::Set(g, o) | Slot::RepairedSet(g, _, o) => (*n, *g, Some(o.clone())), Slot::Freed(g) => (*n, *g, None), }) } @@ -782,7 +906,7 @@ impl Document { if source.peek().is_some_and(|n| *n == num) { source.next(); } - if matches!(slot, Slot::Set(..)) { + if matches!(slot, Slot::Set(..) | Slot::RepairedSet(..)) { numbers.push(num); } } @@ -800,7 +924,7 @@ impl Document { pub(super) fn root_is_catalog(&self) -> bool { let Some(r) = self.root() else { return false }; - matches!(self.try_get(r.num).as_deref(), Ok(Object::Dict(d)) if d.name(b"Type") == Some(b"Catalog") || d.contains(b"Pages")) + matches!(self.try_get_reference(r).as_deref(), Ok(Object::Dict(d)) if d.name(b"Type") == Some(b"Catalog") || d.contains(b"Pages")) } } diff --git a/crates/cos/src/document/open.rs b/crates/cos/src/document/open.rs index 70bbcc313e..7a4337c8da 100644 --- a/crates/cos/src/document/open.rs +++ b/crates/cos/src/document/open.rs @@ -1,5 +1,5 @@ use super::reader::find; -use super::state::Credentials; +use super::state::{Credentials, Slot}; use super::xref::XrefIndex; use super::{Arc, BTreeMap, CosError, Dict, Document, HashSet, ObjRef, Object}; @@ -156,6 +156,9 @@ impl Document { } doc.reconstruct(&mut log)?; } + // Header generations are plaintext, including in encrypted files. Inspect only the + // bounded headers, not object bodies, and retain the compact xref representation. + doc.note_xref_generation_repairs(&mut log); doc.repair_log = Arc::new(log); // Reconstruction may have found a different trailer: (re)authenticate against it. if (!ok || !authenticated) @@ -288,8 +291,9 @@ impl Document { } /// The next save must rewrite the whole file (see [`Document::require_full_save`]). + /// Repair-aware edits retaining two accepted generations cannot use one incremental slot. pub fn full_save_required(&self) -> bool { - self.full_save + self.full_save || self.overlay.values().any(|slot| matches!(slot, Slot::RepairedSet(..))) } /// Security was set or removed since the document was opened (the next save applies it). diff --git a/crates/cos/src/document/reader.rs b/crates/cos/src/document/reader.rs index d21bf9403f..e1691ace90 100644 --- a/crates/cos/src/document/reader.rs +++ b/crates/cos/src/document/reader.rs @@ -17,8 +17,8 @@ impl Document { return Ok((index, trailer, revisions, Vec::new())); } } - let (entries, trailer, revisions, oddities) = self.read_xref_chain_materialized(log)?; - Ok((XrefIndex::Materialized(Arc::new(entries)), trailer, revisions, oddities)) + let (entries, trailer, revisions, oddities, generation_repairs) = self.read_xref_chain_materialized(log)?; + Ok((XrefIndex::Materialized(Arc::new(entries), Arc::new(generation_repairs)), trailer, revisions, oddities)) } fn read_compact_xref_chain(&self, log: &mut Vec) -> Result)>, CosError> { @@ -323,7 +323,7 @@ impl Document { pub(super) fn read_xref_chain_materialized( &self, log: &mut Vec, - ) -> Result<(BTreeMap, Dict, Vec, Vec<(u64, u32)>), CosError> { + ) -> Result<(BTreeMap, Dict, Vec, Vec<(u64, u32)>, Vec), CosError> { let data = &self.data; let tail_start = data.len().saturating_sub(4096); let sx = rfind(data, b"startxref", tail_start).ok_or(CosError::Syntax { offset: data.len(), detail: "no startxref".into() })?; @@ -343,6 +343,9 @@ impl Document { // At most MAX_XREF_OBJECTS entries are kept in all; a stream past that is read again. let mut hybrids: HashMap>> = HashMap::new(); let (mut hybrid_rows, mut noted) = (0usize, HashSet::new()); + let mut generation_repairs = Vec::new(); + let mut recorded = HashSet::new(); + let mut diagnose = true; let mut next = Some(first); while let Some(off) = next.take() { if !visited.insert(off) || visited.len() > 10_000 { @@ -371,6 +374,12 @@ impl Document { } }; revisions.push(Revision { xref_offset: off, is_stream, header_relative }); + for (&num, &entry) in § { + if !diagnose { + break; + } + diagnose = self.inspect_xref_generation_repair(num, entry, &mut recorded, &mut generation_repairs); + } // Hybrid files (§7.5.8.4): the /XRefStm entries supplement the table and replace the // free entries it gives objects only a cross-reference stream can locate. let hybrid = sect_trailer.int(b"XRefStm").filter(|x| *x >= 0); @@ -390,6 +399,12 @@ impl Document { let mut stm = BTreeMap::new(); let read = match self.read_section(x as usize, &mut stm, &mut odd, log) { Ok(_) => { + for (&num, &entry) in &stm { + if !diagnose { + break; + } + diagnose = self.inspect_xref_generation_repair(num, entry, &mut recorded, &mut generation_repairs); + } for (&num, &e) in &stm { match sect.get(&num) { None | Some(XrefEntry::Free { .. }) => { @@ -427,7 +442,7 @@ impl Document { for k in [&b"Prev"[..], b"XRefStm", b"Type", b"W", b"Index", b"Filter", b"DecodeParms", b"Length"] { trailer.remove(k); } - Ok((entries, trailer, revisions, oddities)) + Ok((entries, trailer, revisions, oddities, generation_repairs)) } /// Read one section at `off`, adding entries not already known (newer sections win). `odd` @@ -606,7 +621,7 @@ impl Document { } if found > 0 { log.push(format!("the cross-reference data leaves out {found} objects the file contains; located them by their object headers")); - self.entries = Arc::new(XrefIndex::Materialized(Arc::new(filled))); + self.entries = Arc::new(XrefIndex::Materialized(Arc::new(filled), Arc::default())); } self.located = Arc::new(located); } @@ -631,7 +646,7 @@ impl Document { if entries.is_empty() { return Err(CosError::Syntax { offset: 0, detail: "no objects found".into() }); } - self.entries = Arc::new(XrefIndex::Materialized(Arc::new(entries.clone()))); + self.entries = Arc::new(XrefIndex::Materialized(Arc::new(entries.clone()), Arc::default())); self.cache.lock().map_err(|_| CosError::Poisoned)?.clear(); // Objects inside object streams. for num in objstms { @@ -648,7 +663,7 @@ impl Document { } } let object_count = entries.len(); - self.entries = Arc::new(XrefIndex::Materialized(Arc::new(entries))); + self.entries = Arc::new(XrefIndex::Materialized(Arc::new(entries), Arc::default())); self.cache.lock().map_err(|_| CosError::Poisoned)?.clear(); // Trailer: the last `trailer` dictionary or cross-reference stream dictionary with a // usable /Root (last in the file, whichever kind), else the catalog itself. One whose @@ -658,7 +673,7 @@ impl Document { // A file whose only trailers are cross-reference streams (and whose `startxref` is gone // or points nowhere) keeps its /Info, /ID and /Encrypt this way. let rank = |at: usize, d: &Dict| { - let root = self.try_get(d.reference(b"Root")?.num).ok()?; + let root = self.try_get_reference(d.reference(b"Root")?).ok()?; let Object::Dict(c) = &*root else { return None }; Some((c.name(b"Type") == Some(b"Catalog") || c.contains(b"Pages"), at)) }; diff --git a/crates/cos/src/document/state.rs b/crates/cos/src/document/state.rs index 96dde01c0f..f68a483290 100644 --- a/crates/cos/src/document/state.rs +++ b/crates/cos/src/document/state.rs @@ -7,7 +7,7 @@ thread_local! { /// Indirect `/Length` values resolved during the current outermost load: a damaged file /// retries a parse several ways, and each retry would resolve the same lengths again /// (exponential in the chain of streams whose lengths point at further streams). - pub(super) static LENGTHS: std::cell::RefCell>> = std::cell::RefCell::new(HashMap::new()); + pub(super) static LENGTHS: std::cell::RefCell>> = std::cell::RefCell::new(HashMap::new()); } /// Marks an object as being loaded on this thread. Loading can re-enter (indirect stream @@ -65,6 +65,9 @@ pub(super) fn objstm_member(num: i64, offset: i64, first: Option, len: us #[derive(Clone, Debug, PartialEq)] pub(super) enum Slot { Set(u16, Arc), + /// Both generations accepted from a damaged live source slot survive a repair-aware edit. + /// A full rewrite is required: a new xref slot cannot encode two generations. + RepairedSet(u16, u16, Arc), Freed(u16), } @@ -103,6 +106,7 @@ impl ReadState { a == b && match (x, y) { (Slot::Set(g, o), Slot::Set(h, p)) => g == h && Arc::ptr_eq(o, p), + (Slot::RepairedSet(g, a, o), Slot::RepairedSet(h, b, p)) => g == h && a == b && Arc::ptr_eq(o, p), (Slot::Freed(g), Slot::Freed(h)) => g == h, _ => false, } @@ -133,7 +137,7 @@ impl ObjectReader { } pub(crate) fn try_get(&mut self, reference: ObjRef) -> Result, CosError> { - let object = self.document.try_get(reference.num); + let object = self.document.try_get_reference(reference); // A missed eviction here only costs a re-parse, so never fail the read on a poisoned // lock. The editing-cache evictors propagate instead; they guard user edits. if let Ok(mut streams) = self.document.objstms.lock() { diff --git a/crates/cos/src/document/tests.rs b/crates/cos/src/document/tests.rs index f12437baaf..ad8f86fdea 100644 --- a/crates/cos/src/document/tests.rs +++ b/crates/cos/src/document/tests.rs @@ -599,7 +599,7 @@ pub(crate) fn build(bodies: &[&str], trailer: &str) -> Vec { fn assert_compact_matches_materialized(doc: &Document) { assert!(matches!(doc.entries.as_ref(), XrefIndex::Compact { .. }), "valid xrefs should not retain a per-entry map"); - let (entries, _, revisions, _) = doc.read_xref_chain_materialized(&mut Vec::new()).unwrap(); + let (entries, _, revisions, _, _) = doc.read_xref_chain_materialized(&mut Vec::new()).unwrap(); assert_eq!(revisions.as_slice(), doc.revisions()); let max = entries.keys().next_back().copied().unwrap_or(0); assert_eq!(doc.entries.max_num(), max); @@ -687,7 +687,7 @@ fn malformed_size_uses_the_materialized_compatibility_path() { let at = bytes.windows(marker.len()).position(|window| window == marker).unwrap(); bytes[at..at + marker.len()].copy_from_slice(b"/Size 2"); let doc = Document::open(Arc::new(bytes)).unwrap(); - assert!(matches!(doc.entries.as_ref(), XrefIndex::Materialized(_))); + assert!(matches!(doc.entries.as_ref(), XrefIndex::Materialized(..))); assert_eq!(doc.next_num, 4); assert_eq!(doc.object_numbers(), vec![1, 2, 3]); } @@ -1181,15 +1181,18 @@ fn reads_classic_revision_and_edits_are_copy_on_write() { fn object_numbers_merge_source_and_overlay_in_order() { let mut doc = Document::new_empty(); doc.overlay.clear(); - doc.entries = Arc::new(XrefIndex::Materialized(Arc::new(BTreeMap::from([ - (0, XrefEntry::Free { next_generation: 65535 }), - (2, XrefEntry::InFile { offset: 10, generation: 7 }), - (4, XrefEntry::Free { next_generation: 9 }), - (6, XrefEntry::InStream { stream: 80, index: 1 }), - (8, XrefEntry::InFile { offset: 20, generation: 5 }), - (10, XrefEntry::InFile { offset: 30, generation: 3 }), - (u32::MAX, XrefEntry::InFile { offset: 40, generation: 1 }), - ])))); + doc.entries = Arc::new(XrefIndex::Materialized( + Arc::new(BTreeMap::from([ + (0, XrefEntry::Free { next_generation: 65535 }), + (2, XrefEntry::InFile { offset: 10, generation: 7 }), + (4, XrefEntry::Free { next_generation: 9 }), + (6, XrefEntry::InStream { stream: 80, index: 1 }), + (8, XrefEntry::InFile { offset: 20, generation: 5 }), + (10, XrefEntry::InFile { offset: 30, generation: 3 }), + (u32::MAX, XrefEntry::InFile { offset: 40, generation: 1 }), + ])), + Arc::default(), + )); doc.set(ObjRef::new(1, 2), Object::Int(1)); // before the first live source slot doc.set(ObjRef::new(2, 8), Object::Int(2)); // override, without duplication doc.set(ObjRef::new(4, 9), Object::Int(4)); // reuse a source free slot diff --git a/crates/cos/src/document/xref.rs b/crates/cos/src/document/xref.rs index 02390ff7da..771df28b80 100644 --- a/crates/cos/src/document/xref.rs +++ b/crates/cos/src/document/xref.rs @@ -41,7 +41,7 @@ pub(super) struct XrefRevisionSection { #[derive(Clone, Debug)] pub(super) enum XrefIndex { Compact { revisions: Vec, max_num: u32, has_entries: bool }, - Materialized(Arc>), + Materialized(Arc>, Arc>), } pub(super) const MAX_COMPACT_XREF_ROWS: usize = 2_000_000; @@ -149,8 +149,8 @@ impl XrefSection { } } - pub(super) fn extend_numbers(&self, numbers: &mut Vec) { - for range in &self.ranges { + pub(super) fn numbers(&self) -> impl Iterator + '_ { + self.ranges.iter().flat_map(|range| { let available = match &self.data { XrefSectionData::Table => range.count as usize, XrefSectionData::Stream { bytes, widths } => { @@ -158,14 +158,12 @@ impl XrefSection { bytes.len().saturating_sub(range.offset).checked_div(row_width).unwrap_or(0) } }; - for i in 0..available.min(range.count as usize) { - if let Ok(i) = u32::try_from(i) - && let Some(num) = range.start.checked_add(i) - { - numbers.push(num); - } - } - } + (0..available.min(range.count as usize)).filter_map(move |i| u32::try_from(i).ok().and_then(|i| range.start.checked_add(i))) + }) + } + + pub(super) fn extend_numbers(&self, numbers: &mut Vec) { + numbers.extend(self.numbers()); } } @@ -189,20 +187,20 @@ impl XrefRevisionSection { impl XrefIndex { pub(super) fn empty() -> Self { - Self::Materialized(Arc::new(BTreeMap::new())) + Self::Materialized(Arc::new(BTreeMap::new()), Arc::default()) } pub(super) fn lookup(&self, num: u32, file: &[u8]) -> Option { match self { Self::Compact { revisions, .. } => revisions.iter().find_map(|revision| revision.lookup(num, file)), - Self::Materialized(entries) => entries.get(&num).copied(), + Self::Materialized(entries, _) => entries.get(&num).copied(), } } pub(super) fn is_empty(&self) -> bool { match self { Self::Compact { has_entries, .. } => !has_entries, - Self::Materialized(entries) => entries.is_empty(), + Self::Materialized(entries, _) => entries.is_empty(), } } @@ -211,14 +209,14 @@ impl XrefIndex { pub(super) fn remove_exhausted(&mut self) -> bool { match self { Self::Compact { .. } => false, - Self::Materialized(entries) => Arc::make_mut(entries).remove(&u32::MAX).is_some(), + Self::Materialized(entries, _) => Arc::make_mut(entries).remove(&u32::MAX).is_some(), } } /// Every entry, newest section first, as the materialized reader keeps them. pub(super) fn to_map(&self, file: &[u8]) -> BTreeMap { match self { - Self::Materialized(entries) => entries.as_ref().clone(), + Self::Materialized(entries, _) => entries.as_ref().clone(), Self::Compact { revisions, .. } => { let mut numbers = Vec::new(); for revision in revisions { @@ -232,13 +230,13 @@ impl XrefIndex { pub(super) fn max_num(&self) -> u32 { match self { Self::Compact { max_num, .. } => *max_num, - Self::Materialized(entries) => entries.keys().next_back().copied().unwrap_or(0), + Self::Materialized(entries, _) => entries.keys().next_back().copied().unwrap_or(0), } } pub(super) fn object_numbers(&self, file: &[u8]) -> Vec { match self { - Self::Materialized(entries) => { + Self::Materialized(entries, _) => { entries.iter().filter_map(|(&num, entry)| (!matches!(entry, XrefEntry::Free { .. })).then_some(num)).collect() } Self::Compact { revisions, .. } => { diff --git a/crates/cos/src/writer.rs b/crates/cos/src/writer.rs index dcfcb97997..6c4c71580a 100644 --- a/crates/cos/src/writer.rs +++ b/crates/cos/src/writer.rs @@ -307,8 +307,17 @@ pub fn write_full(doc: &Document, opts: &SaveOptions) -> Result, CosErro if matches!(*o, Object::Null) { continue; // dangling reference: written as null by renumbering below } - map.insert(r, output_number(order.len())?); - order.push((r, matches!(*o, Object::Stream(_)))); + // Accepted xref/header aliases denote one slot, including page-tree cycles. + // Check for null first: an unrelated or freed generation must remain dangling. + let canonical = ObjRef::new(r.num, reader.document.generation(r.num)); + if let Some(&num) = map.get(&canonical) { + map.insert(r, num); + continue; + } + let num = output_number(order.len())?; + map.insert(r, num); + map.insert(canonical, num); + order.push((canonical, matches!(*o, Object::Stream(_)))); visit_refs(&o, &mut queue); } let version = reader.document.version().chars().take(3).collect::(); @@ -319,13 +328,15 @@ pub fn write_full(doc: &Document, opts: &SaveOptions) -> Result, CosErro out.extend_from_slice(&[b'%', 0xE2, 0xE3, 0xCF, 0xD3, b'\n']); // The /Encrypt dictionary and everything it references (e.g. indirect /CF crypt-filter // dictionaries) are needed before anything can be decrypted: never inside object streams. - let mut encryption_objects: std::collections::HashSet = std::collections::HashSet::new(); + let mut encryption_objects: std::collections::HashSet = std::collections::HashSet::new(); let mut pending: Vec = Vec::new(); if let Some(encrypt) = trailer_in.get(b"Encrypt") { collect_refs(encrypt, &mut |r| pending.push(r)); } while let Some(r) = pending.pop() { - if encryption_objects.insert(r) { + if let Some(&num) = map.get(&r) + && encryption_objects.insert(num) + { collect_refs(&reader.get(r), &mut |x| pending.push(x)); } } @@ -333,14 +344,15 @@ pub fn write_full(doc: &Document, opts: &SaveOptions) -> Result, CosErro // damaged and could not be repaired", before it even asks for the password), while other // readers accept it. qpdf keeps the catalog out of object streams too. Verified with // Acrobat DC on macOS (#774). - let root_outside = if reader.document.output_security().0.is_some() { trailer_in.reference(b"Root") } else { None }; + let root_outside = + if reader.document.output_security().0.is_some() { trailer_in.reference(b"Root").and_then(|r| map.get(&r)).copied() } else { None }; let mut rows: BTreeMap = BTreeMap::new(); // Keep only references, not renamed copies of the entire object graph. Standalone // objects still precede object streams, preserving the established output order. let mut packed: Vec<(u32, ObjRef)> = Vec::new(); for (i, (r, is_stream)) in order.iter().enumerate() { let num = output_number(i)?; - let in_stream = opts.object_streams && !is_stream && !encryption_objects.contains(r) && root_outside != Some(*r); + let in_stream = opts.object_streams && !is_stream && !encryption_objects.contains(&num) && root_outside != Some(num); if in_stream { packed.push((num, *r)); } else { diff --git a/crates/cos/tests/it/generations.rs b/crates/cos/tests/it/generations.rs new file mode 100644 index 0000000000..3d56628c93 --- /dev/null +++ b/crates/cos/tests/it/generations.rs @@ -0,0 +1,475 @@ +//! An indirect reference identifies an object number AND its generation (ISO 32000-2 §7.3.10). + +use std::sync::Arc; + +use hayro_syntax::object::{Dict as HayroDict, ObjectIdentifier}; +use pdfcraft_cos::{Document, ObjRef, Object, SaveOptions, XrefEntry, write_full, write_incremental}; + +/// Object 3 was reused: only 3 1 R refers to the current value, never 3 0 R. +fn reused_object() -> Arc> { + object_file(&[ + (1, 0, "<< /Type /Catalog /Pages 2 0 R /Stale 3 0 R /Current 3 1 R >>"), + (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), + (3, 1, "42"), + ]) +} + +fn object_file(objects: &[(u32, u16, &str)]) -> Arc> { + object_file_with_entry(objects, None) +} + +fn object_file_with_entry(objects: &[(u32, u16, &str)], changed: Option<(u32, u16, bool)>) -> Arc> { + let mut bytes = b"%PDF-1.7\n".to_vec(); + let mut offsets = Vec::new(); + for (num, generation, value) in objects { + offsets.push((*num, bytes.len(), *generation)); + bytes.extend_from_slice(format!("{num} {generation} obj\n{value}\nendobj\n").as_bytes()); + } + let xref = bytes.len(); + let size = objects.len() + 1; + bytes.extend_from_slice(format!("xref\n0 {size}\n0000000000 65535 f \n").as_bytes()); + for (num, offset, generation) in offsets { + let (generation, free) = changed.filter(|(n, _, _)| *n == num).map_or((generation, false), |(_, g, free)| (g, free)); + let (offset, kind) = if free { (0, 'f') } else { (offset, 'n') }; + bytes.extend_from_slice(format!("{offset:010} {generation:05} {kind} \n").as_bytes()); + } + bytes.extend_from_slice(format!("trailer\n<< /Size {size} /Root 1 0 R >>\nstartxref\n{xref}\n%%EOF\n").as_bytes()); + Arc::new(bytes) +} + +fn contradictory_entry(free: bool) -> Arc> { + object_file_with_entry( + &[(1, 0, "<< /Type /Catalog /Pages 2 0 R /Stale 3 0 R /Current 3 1 R >>"), (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), (3, 1, "42")], + Some((3, if free { 2 } else { 0 }, free)), + ) +} + +#[test] +fn contradictory_xref_generation_recovers_the_matching_header_and_records_repair() { + for warm in [false, true] { + let doc = Document::open(contradictory_entry(false)).unwrap(); + if warm { + assert_eq!(*doc.try_get(3).unwrap(), Object::Int(42)); + } + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + assert_eq!(*doc.resolve(&Object::Ref(ObjRef::new(3, 1))), Object::Int(42)); + // Preserve the existing xref-matching behavior, but never accept an unrelated generation. + assert_eq!(*doc.get(ObjRef::new(3, 0)), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(3, 2)), Object::Null); + assert!(!doc.xref_entry_is_exact(3), "a contradictory generation is not an exact xref entry"); + assert!( + doc.repair_log() + .iter() + .any(|line| { line.contains("object 3") && line.contains("xref generation 0") && line.contains("header generation 1") }), + "{:?}", + doc.repair_log() + ); + let repairs = doc.repair_log().to_vec(); + for _ in 0..3 { + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + } + assert_eq!(doc.repair_log(), repairs); + + let mut edited = doc.clone(); + edited.free(ObjRef::new(3, 1)); + for generation in [0, 1, 2] { + assert_eq!(*edited.get(ObjRef::new(3, generation)), Object::Null); + } + edited.set(ObjRef::new(3, 2), Object::Int(99)); + assert_eq!(*edited.get(ObjRef::new(3, 1)), Object::Null); + assert_eq!(*edited.get(ObjRef::new(3, 2)), Object::Int(99)); + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + } +} + +/// A damaged row may name generation zero while the retained object and catalog use one. +fn contradictory_dictionary() -> Arc> { + object_file_with_entry( + &[(1, 0, "<< /Type /Catalog /Pages 2 0 R /Current 3 1 R >>"), (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), (3, 1, "<< /Value 42 >>")], + Some((3, 0, false)), + ) +} + +fn assert_repaired_dictionary(bytes: Vec, value: i32) { + let bytes = Arc::new(bytes); + let oracle = hayro_syntax::Pdf::new(bytes.clone()).unwrap(); + let catalog = oracle.xref().get::>(ObjectIdentifier::new(1, 0)).unwrap(); + assert_eq!(catalog.get::>(b"Current").unwrap().get::(b"Value"), Some(value)); + let back = Document::open(bytes).unwrap(); + let catalog = back.get(back.root().unwrap()); + let current = back.resolve(catalog.as_dict().unwrap().get(b"Current").unwrap()); + assert_eq!(current.as_dict().unwrap().int(b"Value"), Some(i64::from(value))); +} + +#[test] +fn repaired_header_references_survive_full_saves_and_dictionary_edits() { + let original = contradictory_dictionary(); + let doc = Document::open(original.clone()).unwrap(); + for object_streams in [false, true] { + assert_repaired_dictionary(write_full(&doc, &SaveOptions { object_streams, ..SaveOptions::default() }).unwrap(), 42); + } + let mut edited = doc.clone(); + edited.update_dict(ObjRef::new(3, 1), |dict| dict.set(b"Value".to_vec(), Object::Int(77))).unwrap(); + assert_eq!(edited.generation(3), 1); + let incremental = write_incremental(&edited, &SaveOptions::default()).unwrap(); + assert!(edited.full_save_required(), "repair-aware dictionary edits normalize on full rewrite"); + assert_repaired_dictionary(incremental, 77); + for object_streams in [false, true] { + assert_repaired_dictionary(write_full(&edited, &SaveOptions { object_streams, ..SaveOptions::default() }).unwrap(), 77); + } + let mut restored = doc.clone(); + restored.overlay_changes_from(&edited).unwrap(); + assert_repaired_dictionary(write_incremental(&restored, &SaveOptions::default()).unwrap(), 77); +} + +#[test] +fn free_xref_slots_never_revive_retained_object_headers() { + for warm in [false, true] { + let doc = Document::open(contradictory_entry(true)).unwrap(); + assert!(matches!(doc.xref_entry(3), Some(XrefEntry::Free { next_generation: 2 }))); + if warm { + assert_eq!(*doc.try_get(3).unwrap(), Object::Null); + } + for generation in [0, 1, 2] { + assert_eq!(*doc.get(ObjRef::new(3, generation)), Object::Null); + } + assert_eq!(*doc.resolve(&Object::Ref(ObjRef::new(3, 1))), Object::Null); + assert!(doc.repair_log().is_empty(), "{:?}", doc.repair_log()); + } + // Control: the retained bytes really are a readable object when its slot is live. + let live = Document::open(contradictory_entry(false)).unwrap(); + assert_eq!(*live.try_get(3).unwrap(), Object::Int(42)); +} + +#[test] +fn incremental_generation_identity_is_visible_to_an_independent_reader() { + let original = reused_object(); + let mut doc = Document::open(original.clone()).unwrap(); + doc.set(ObjRef::new(3, 1), Object::Int(77)); + let bytes = write_incremental(&doc, &SaveOptions::default()).unwrap(); + assert!(bytes.starts_with(&original)); + let oracle = hayro_syntax::Pdf::new(Arc::new(bytes)).unwrap(); + let catalog = oracle.xref().get::>(ObjectIdentifier::new(1, 0)).unwrap(); + assert_eq!(catalog.get::(b"Current"), Some(77)); + assert_eq!(catalog.get::(b"Stale"), None); +} + +#[test] +fn generation_repair_diagnostics_are_bounded_without_disabling_recovery() { + let mut objects = vec![(1, 0, "<< /Type /Catalog /Pages 2 0 R >>"), (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>")]; + objects.extend((3..303).map(|num| (num, 1, "42"))); + let bytes = object_file(&objects); + let bytes = String::from_utf8(bytes.as_ref().clone()).unwrap().replace("00001 n \n", "00000 n \n"); + for materialized in [false, true] { + let bytes = if materialized { bytes.replace("/Size 303", "/Size 302") } else { bytes.clone() }; + let doc = Document::open(Arc::new(bytes.into_bytes())).unwrap(); + assert_eq!(doc.repair_log().len(), 257); + assert!(doc.repair_log().last().unwrap().contains("omitted after 256 objects")); + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(302, 1)), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(302, 2)), Object::Null); + assert_eq!(doc.repair_log().len(), 257); + } +} + +#[test] +fn references_require_the_current_generation_even_with_a_warm_cache() { + let bytes = reused_object(); + let oracle = hayro_syntax::Pdf::new(bytes.clone()).unwrap(); + assert_eq!(oracle.xref().get::(ObjectIdentifier::new(3, 0)), None); + assert_eq!(oracle.xref().get::(ObjectIdentifier::new(3, 1)), Some(42)); + + let doc = Document::open(bytes).unwrap(); + assert!(doc.repair_log().is_empty()); + assert_eq!(*doc.get(ObjRef::new(3, 0)), Object::Null); + assert_eq!(*doc.try_get(3).unwrap(), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(3, 0)), Object::Null); + assert_eq!(*doc.resolve(&Object::Ref(ObjRef::new(3, 0))), Object::Null); + assert_eq!(*doc.get(ObjRef::new(3, 2)), Object::Null); +} + +#[test] +fn overlay_generation_changes_do_not_retarget_old_references_or_clones() { + let mut doc = Document::open(reused_object()).unwrap(); + let snapshot = doc.clone(); + let old = ObjRef::new(3, 1); + let new = ObjRef::new(3, 2); + doc.free(old); + assert_eq!(*doc.get(old), Object::Null); + doc.set(new, Object::Int(99)); + assert_eq!(*doc.get(old), Object::Null); + assert_eq!(*doc.get(new), Object::Int(99)); + assert_eq!(*snapshot.get(old), Object::Int(42)); + assert_eq!(*snapshot.get(new), Object::Null); + + let saved = write_incremental(&doc, &SaveOptions::default()).unwrap(); + let back = Document::open(Arc::new(saved)).unwrap(); + assert!(back.repair_log().is_empty()); + assert_eq!(*back.get(old), Object::Null); + assert_eq!(*back.get(new), Object::Int(99)); +} + +#[test] +fn restoring_snapshot_preserves_changes_to_nonzero_generation_objects() { + for generation in [1, 2] { + let mut saved = Document::open(reused_object()).unwrap(); + let mut current = saved.clone(); + let reference = ObjRef::new(3, generation); + current.set(reference, Object::Int(99)); + current.update_dict(current.root().unwrap(), |catalog| catalog.set(b"Current".to_vec(), Object::Ref(reference))).unwrap(); + + saved.overlay_changes_from(¤t).unwrap(); + assert_eq!(*saved.get(reference), Object::Int(99), "generation {generation}"); + let back = Document::open(Arc::new(write_incremental(&saved, &SaveOptions::default()).unwrap())).unwrap(); + assert!(back.repair_log().is_empty()); + assert_eq!(*back.get(reference), Object::Int(99), "generation {generation}"); + let catalog = back.get(back.root().unwrap()); + assert_eq!(*back.resolve(catalog.as_dict().unwrap().get(b"Current").unwrap()), Object::Int(99)); + } +} + +#[test] +fn full_save_does_not_turn_stale_references_into_live_ones() { + let doc = Document::open(reused_object()).unwrap(); + for object_streams in [false, true] { + let saved = write_full(&doc, &SaveOptions { object_streams, ..SaveOptions::default() }).unwrap(); + let saved = Arc::new(saved); + let oracle = hayro_syntax::Pdf::new(saved.clone()).unwrap(); + let catalog = oracle.xref().get::>(ObjectIdentifier::new(1, 0)).unwrap(); + assert_eq!(catalog.get::(b"Current"), Some(42)); + assert_eq!(catalog.get::(b"Stale"), None); + let back = Document::open(saved).unwrap(); + assert!(back.repair_log().is_empty()); + let catalog = back.get(back.root().unwrap()); + let catalog = catalog.as_dict().unwrap(); + assert_eq!(catalog.get(b"Stale"), Some(&Object::Null)); + assert_eq!(*back.resolve(catalog.get(b"Current").unwrap()), Object::Int(42)); + } +} + +#[test] +fn compressed_objects_only_accept_generation_zero() { + let doc = Document::open(reused_object()).unwrap(); + let saved = write_full(&doc, &SaveOptions::default()).unwrap(); + let back = Document::open(Arc::new(saved)).unwrap(); + let catalog = back.get(back.root().unwrap()); + let current = catalog.as_dict().unwrap().reference(b"Current").unwrap(); + assert_eq!(current.generation, 0); + assert!(matches!(back.xref_entry(current.num), Some(XrefEntry::InStream { .. }))); + assert_eq!(*back.get(current), Object::Int(42)); + assert_eq!(*back.get(ObjRef::new(current.num, 1)), Object::Null); +} + +#[test] +fn indirect_stream_lengths_require_the_referenced_generation() { + for generation in [0, 1] { + // The embedded endstream is data when /Length resolves, but the existing missing- + // length recovery stops there when the reference is stale. + let stream = format!("<< /Length 4 {generation} R >>\nstream\nA\nendstream\nB\nendstream"); + let doc = Document::open(object_file(&[ + (1, 0, "<< /Type /Catalog /Pages 2 0 R >>"), + (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), + (3, 0, &stream), + (4, 1, "13"), + ])) + .unwrap(); + let object = doc.get(ObjRef::new(3, 0)); + let Object::Stream(stream) = &*object else { panic!("expected stream") }; + assert_eq!(&stream.raw[..], if generation == 1 { b"A\nendstream\nB".as_slice() } else { b"A".as_slice() }); + } +} + +/// A newer valid row must not hide damage in a retained older revision. +fn shadowed_generation_disagreement(materialized: bool, shifted: bool) -> Arc> { + let original = contradictory_dictionary(); + let text = std::str::from_utf8(&original).unwrap(); + let previous: usize = text.rsplit("startxref\n").next().unwrap().lines().next().unwrap().parse().unwrap(); + let mut bytes = (*original).clone(); + let offset = bytes.len(); + bytes.extend_from_slice(b"3 1 obj\n<< /Value 77 >>\nendobj\n"); + let xref = bytes.len(); + let size = if materialized { 3 } else { 4 }; + bytes.extend_from_slice( + format!("xref\n3 1\n{offset:010} 00001 n \ntrailer\n<< /Size {size} /Root 1 0 R /Prev {previous} >>\nstartxref\n{xref}\n%%EOF\n").as_bytes(), + ); + if shifted { + bytes.splice(..0, b"junk before the header\n".iter().copied()); + } + Arc::new(bytes) +} + +#[test] +fn retained_generation_disagreements_are_reported_in_materialized_mode_too() { + for materialized in [false, true] { + for shifted in [false, true] { + let doc = Document::open(shadowed_generation_disagreement(materialized, shifted)).unwrap(); + assert_eq!(doc.revisions().len(), 2, "the xref chain is retained"); + assert!(!doc.reconstructed()); + assert_eq!(doc.get(ObjRef::new(3, 1)).as_dict().unwrap().int(b"Value"), Some(77)); + assert!( + doc.repair_log().iter().any(|line| line.contains("object 3: xref generation 0 disagrees with object header generation 1")), + "materialized={materialized}, shifted={shifted}: {:?}", + doc.repair_log() + ); + } + } +} + +#[test] +fn repair_aware_edits_keep_both_aliases_and_full_save_normalizes_them() { + let mut doc = Document::open(contradictory_entry(false)).unwrap(); + let snapshot = doc.clone(); + doc.set_existing(ObjRef::new(3, 0), Object::Int(77)).unwrap(); + for generation in [0, 1] { + assert_eq!(*doc.get(ObjRef::new(3, generation)), Object::Int(77)); + assert_eq!(*snapshot.get(ObjRef::new(3, generation)), Object::Int(42)); + } + let state = doc.read_state(); + assert!(doc.set_existing(ObjRef::new(3, 2), Object::Int(99)).is_err()); + assert!(state.matches(&doc)); + assert_eq!(*doc.get(ObjRef::new(3, 2)), Object::Null); + let mut restored = snapshot.clone(); + restored.overlay_changes_from(&doc).unwrap(); + for changed in [&doc, &restored] { + assert!(changed.full_save_required()); + for bytes in [ + write_incremental(changed, &SaveOptions::default()).unwrap(), + write_full(changed, &SaveOptions { object_streams: false, ..SaveOptions::default() }).unwrap(), + ] { + let bytes = Arc::new(bytes); + let oracle = hayro_syntax::Pdf::new(bytes.clone()).unwrap(); + let catalog = oracle.xref().get::>(ObjectIdentifier::new(1, 0)).unwrap(); + assert_eq!(catalog.get::(b"Current"), Some(77)); + assert_eq!(catalog.get::(b"Stale"), Some(77), "both source aliases stay live"); + let back = Document::open(bytes).unwrap(); + assert!(back.repair_log().is_empty(), "normalized output has no generation repairs"); + } + } + // Explicit replacement keeps its narrower, caller-supplied generation contract. + doc.set(ObjRef::new(3, 2), Object::Int(99)); + assert_eq!(*doc.get(ObjRef::new(3, 0)), Object::Null); + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Null); + assert_eq!(*doc.get(ObjRef::new(3, 2)), Object::Int(99)); +} + +#[test] +fn hybrid_generation_diagnostics_observe_rows_before_the_table_hides_them() { + let mut bytes = b"%PDF-1.7\n".to_vec(); + let mut offsets = Vec::new(); + for (num, generation, body) in [(1, 0, "<< /Type /Catalog /Pages 2 0 R >>"), (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), (3, 1, "42")] { + offsets.push(bytes.len()); + bytes.extend_from_slice(format!("{num} {generation} obj\n{body}\nendobj\n").as_bytes()); + } + let hybrid = bytes.len(); + offsets.push(hybrid); + bytes.extend_from_slice(b"4 0 obj\n<< /Type /XRef /Size 5 /W [1 4 2] /Index [3 1] /Length 7 >>\nstream\n"); + bytes.push(1); + bytes.extend_from_slice(&(offsets[2] as u32).to_be_bytes()); + bytes.extend_from_slice(&0u16.to_be_bytes()); + bytes.extend_from_slice(b"\nendstream\nendobj\n"); + let xref = bytes.len(); + bytes.extend_from_slice(b"xref\n0 5\n0000000000 65535 f \n"); + for (i, offset) in offsets.into_iter().enumerate() { + let generation = u16::from(i == 2); + bytes.extend_from_slice(format!("{offset:010} {generation:05} n \n").as_bytes()); + } + bytes.extend_from_slice(format!("trailer\n<< /Size 5 /Root 1 0 R /XRefStm {hybrid} >>\nstartxref\n{xref}\n%%EOF\n").as_bytes()); + let doc = Document::open(Arc::new(bytes)).unwrap(); + assert!(!doc.reconstructed()); + assert!(doc.xref_oddities().iter().any(|(_, num)| *num == 3), "the conflicting hybrid row is set aside"); + assert_eq!(*doc.get(ObjRef::new(3, 1)), Object::Int(42)); + assert_eq!(*doc.get(ObjRef::new(3, 0)), Object::Null, "the table still determines accepted references"); + assert!(doc.repair_log().iter().any(|line| line.contains("object 3: xref generation 0 disagrees with object header generation 1"))); +} + +#[test] +fn discarded_xref_repair_notes_do_not_survive_object_header_reconstruction() { + let bytes = + String::from_utf8((*contradictory_dictionary()).clone()).unwrap().replace("/Root 1 0 R", "/Root 99 0 R").replace("/Size 4", "/Size 3"); + let doc = Document::open(Arc::new(bytes.into_bytes())).unwrap(); + assert!(doc.reconstructed()); + assert!(doc.repair_log().iter().all(|line| !line.contains("xref generation")), "{:?}", doc.repair_log()); +} + +#[test] +fn restoring_equal_values_still_restores_the_accepted_generation_aliases() { + for edited in [false, true] { + for value in [42, 77] { + let mut current = Document::open(contradictory_entry(false)).unwrap(); + if edited { + current.set_existing(ObjRef::new(3, 0), Object::Int(42)).unwrap(); + } + let mut saved = Document::open(object_file(&[ + (1, 0, "<< /Type /Catalog /Pages 2 0 R /Stale 3 0 R /Current 3 1 R >>"), + (2, 0, "<< /Type /Pages /Kids [] /Count 0 >>"), + (3, 0, &value.to_string()), + ])) + .unwrap(); + assert_eq!(*saved.get(ObjRef::new(3, 1)), Object::Null); + saved.overlay_changes_from(¤t).unwrap(); + for generation in [0, 1] { + assert_eq!(*saved.get(ObjRef::new(3, generation)), Object::Int(42), "edited={edited}, value={value}"); + } + assert!(saved.full_save_required()); + } + } +} + +#[test] +fn repaired_page_tree_aliases_keep_one_identity_in_full_saves() { + let source = object_file_with_entry( + &[ + (1, 0, "<< /Type /Catalog /Pages 2 1 R /Stale 2 2 R >>"), + (2, 1, "<< /Type /Pages /Kids [3 0 R] /Count 1 >>"), + (3, 0, "<< /Type /Page /Parent 2 0 R /MediaBox [0 0 10 10] >>"), + ], + Some((2, 0, false)), + ); + for edited in [false, true] { + let mut doc = Document::open(source.clone()).unwrap(); + if edited { + doc.update_dict(ObjRef::new(2, 0), |dict| dict.set(b"Edited".to_vec(), Object::Bool(true))).unwrap(); + } + let mut restored = Document::open(source.clone()).unwrap(); + restored.overlay_changes_from(&doc).unwrap(); + for current in [&doc, &restored] { + for object_streams in [false, true] { + let opts = SaveOptions { object_streams, ..SaveOptions::default() }; + let mut saves = vec![write_full(current, &opts).unwrap()]; + if edited { + saves.push(write_incremental(current, &opts).unwrap()); + } + for bytes in saves { + let back = Document::open(Arc::new(bytes)).unwrap(); + let catalog = back.get(back.root().unwrap()); + let catalog = catalog.as_dict().unwrap(); + let pages_ref = catalog.reference(b"Pages").unwrap(); + let pages = back.get(pages_ref); + let child_ref = pages.as_dict().unwrap().get(b"Kids").unwrap().as_array().unwrap()[0].as_ref().unwrap(); + let child = back.get(child_ref); + assert_eq!(child.as_dict().unwrap().reference(b"Parent"), Some(pages_ref), "edited={edited}, packed={object_streams}"); + assert!(matches!(catalog.get(b"Stale"), Some(Object::Null)), "unrelated generations stay null"); + assert!(back.repair_log().is_empty()); + } + } + } + } +} + +#[test] +fn restoring_explicit_sets_removes_old_generation_aliases() { + for value in [42, 77] { + let mut current = Document::open(contradictory_entry(false)).unwrap(); + current.set(ObjRef::new(3, 0), Object::Int(42)); + for repaired_overlay in [false, true] { + let mut saved = Document::open(contradictory_entry(false)).unwrap(); + if repaired_overlay { + saved.set_existing(ObjRef::new(3, 0), Object::Int(value)).unwrap(); + } + saved.overlay_changes_from(¤t).unwrap(); + assert_eq!(*saved.get(ObjRef::new(3, 0)), Object::Int(42)); + assert_eq!(*saved.get(ObjRef::new(3, 1)), Object::Null, "value={value}, overlay={repaired_overlay}"); + } + } +} diff --git a/crates/cos/tests/it/main.rs b/crates/cos/tests/it/main.rs index 52cf05d6c2..1bd7ee1fd6 100644 --- a/crates/cos/tests/it/main.rs +++ b/crates/cos/tests/it/main.rs @@ -3,6 +3,7 @@ mod bounded_writer; mod corpus; mod encryption; +mod generations; mod proptests; mod shared_streams; mod writer; diff --git a/crates/cos/tests/it/writer.rs b/crates/cos/tests/it/writer.rs index eb03523189..2716b9242e 100644 --- a/crates/cos/tests/it/writer.rs +++ b/crates/cos/tests/it/writer.rs @@ -358,11 +358,32 @@ fn objects_referenced_by_the_encrypt_dictionary_stay_outside_object_streams() { let cf = doc.get(enc_ref).as_dict().unwrap().get(b"CF").cloned().unwrap(); let cf_ref = doc.add(cf); doc.update_dict(enc_ref, |d| d.set(b"CF".to_vec(), Object::Ref(cf_ref))).unwrap(); - let packed = write_full(&doc, &SaveOptions::default()).unwrap(); - let back = Document::open(Arc::new(packed.clone())).expect("reopens"); - assert!(back.repair_log().is_empty(), "{:?}", back.repair_log()); - assert_eq!(title(&back), b"Caf\xe9 (draft)"); - assert_eq!(hayro_pages(&packed), 4); + // Visit this alias from Root before the encryption walk reaches /CF's other alias. + doc.update_dict(doc.root().unwrap(), |d| d.set(b"CFProbe".to_vec(), Object::Ref(cf_ref))).unwrap(); + let classic = write_full(&doc, &SaveOptions { object_streams: false, ..SaveOptions::default() }).unwrap(); + let saved = Document::open(Arc::new(classic.clone())).unwrap(); + let cf_num = saved.get(saved.root().unwrap()).as_dict().unwrap().reference(b"CFProbe").unwrap().num; + for damaged in [false, true] { + let mut bytes = classic.clone(); + if damaged { + // Equal-length edits preserve the xref offsets and its generation-zero row. + for (from, to) in [(format!("\n{cf_num} 0 obj"), format!("\n{cf_num} 1 obj")), (format!("/CF {cf_num} 0 R"), format!("/CF {cf_num} 1 R"))] + { + let start = bytes.windows(from.len()).position(|w| w == from.as_bytes()).unwrap(); + bytes[start..start + from.len()].copy_from_slice(to.as_bytes()); + } + } + let doc = Document::open(Arc::new(bytes)).unwrap(); + let packed = write_full(&doc, &SaveOptions::default()).unwrap(); + let back = Document::open(Arc::new(packed.clone())).expect("reopens"); + assert!(back.repair_log().is_empty(), "{:?}", back.repair_log()); + assert_eq!(title(&back), b"Caf\xe9 (draft)"); + assert_eq!(hayro_pages(&packed), 4); + let probe = back.get(back.root().unwrap()).as_dict().unwrap().reference(b"CFProbe").unwrap(); + let encrypt = back.get(back.trailer().reference(b"Encrypt").unwrap()); + assert_eq!(encrypt.as_dict().unwrap().reference(b"CF"), Some(probe)); + assert!(matches!(back.xref_entry(probe.num), Some(pdfcraft_cos::XrefEntry::InFile { .. })), "crypt filters must stay outside object streams"); + } } #[test] diff --git a/crates/preflight/src/lib.rs b/crates/preflight/src/lib.rs index 1df39ebf05..1fb7272b33 100644 --- a/crates/preflight/src/lib.rs +++ b/crates/preflight/src/lib.rs @@ -894,7 +894,7 @@ pub fn convert(doc: &mut Document, level: Level) -> Result Result Result Result Arc> { + let mut bytes = b"%PDF-1.7\n".to_vec(); + let bodies = [ + "<< /Type /Catalog /Pages 2 0 R /HeaderImage 5 1 R /RowImage 5 0 R /StaleImage 5 2 R >>", + "<< /Type /Pages /Kids [3 0 R] /Count 1 >>", + "<< /Type /Page /Parent 2 0 R /MediaBox [0 0 10 10] /Resources << /XObject << /Im 5 1 R >> >> >>", + "null", + "<< /Type /XObject /Subtype /Image /Width 1 /Height 1 /BitsPerComponent 8 /ColorSpace /DeviceGray /Interpolate true /Length 1 >>\nstream\nX\nendstream", + ]; + let mut offsets = Vec::new(); + for (i, body) in bodies.iter().enumerate() { + offsets.push(bytes.len()); + let generation = u16::from(i == 4); + bytes.extend_from_slice(format!("{} {generation} obj\n{body}\nendobj\n", i + 1).as_bytes()); + } + let xref = bytes.len(); + bytes.extend_from_slice(b"xref\n0 6\n0000000000 65535 f \n"); + for (i, offset) in offsets.into_iter().enumerate() { + let generation = u16::from(i == 4 && !damaged); + bytes.extend_from_slice(format!("{offset:010} {generation:05} n \n").as_bytes()); + } + bytes.extend_from_slice(format!("trailer\n<< /Size 6 /Root 1 0 R >>\nstartxref\n{xref}\n%%EOF\n").as_bytes()); + Arc::new(bytes) +} + +fn assert_images(doc: &Document, converted: bool, dual: bool) { + let root = doc.get(doc.root().unwrap()); + let catalog = root.as_dict().unwrap(); + for key in [b"HeaderImage".as_slice(), b"RowImage".as_slice()] { + let value = doc.resolve(catalog.get(key).unwrap()); + if key == b"RowImage" && !dual { + assert_eq!(*value, Object::Null); + } else { + let Object::Stream(image) = &*value else { panic!("{key:?} became {value:?}") }; + assert_eq!(image.dict.get(b"Interpolate"), Some(&Object::Bool(!converted))); + assert_eq!(&image.raw[..], b"X"); + } + } + assert_eq!(*doc.resolve(catalog.get(b"StaleImage").unwrap()), Object::Null); +} + +#[test] +fn preflight_preserves_repaired_image_references_in_snapshots_and_saves() { + for damaged in [false, true] { + let original = image_file(damaged); + let mut doc = Document::open(original.clone()).unwrap(); + assert_images(&doc, false, damaged); + let snapshot = doc.clone(); + let report = convert(&mut doc, Level::A2b).unwrap(); + assert!(report.fixed.iter().any(|line| line.contains("cleared image interpolation"))); + assert_images(&doc, true, damaged); + assert_eq!(doc.full_save_required(), damaged); + assert_images(&snapshot, false, damaged); + let mut restored = snapshot.clone(); + restored.overlay_changes_from(&doc).unwrap(); + assert_images(&restored, true, damaged); + for changed in [&doc, &restored] { + for object_streams in [false, true] { + let opts = SaveOptions { object_streams, ..SaveOptions::default() }; + let incremental = write_incremental(changed, &opts).unwrap(); + if !damaged { + assert!(incremental.starts_with(&original)); + } + assert_images(&Document::open(Arc::new(incremental)).unwrap(), true, damaged); + let full = write_full(changed, &opts).unwrap(); + let reopened = Document::open(Arc::new(full)).unwrap(); + assert_images(&reopened, true, damaged); + if damaged { + let catalog = reopened.get(reopened.root().unwrap()); + let catalog = catalog.as_dict().unwrap(); + assert_eq!(catalog.reference(b"HeaderImage"), catalog.reference(b"RowImage"), "saved aliases must share one object"); + } + } + } + } +} diff --git a/crates/render/src/raster/tests.rs b/crates/render/src/raster/tests.rs index 9101012f66..be232d80eb 100644 --- a/crates/render/src/raster/tests.rs +++ b/crates/render/src/raster/tests.rs @@ -6436,8 +6436,8 @@ fn draw_minified_image( /// making scans, screenshots and rasterized text look soft. #[test] fn non_interpolated_images_are_not_smoothed_when_minified() { - // Top band K W R G, bottom band B Y M C. Nearest 8×4 → 4×2 sampling picks each block's - // pixel, so every device pixel is one palette colour; Catmull-Rom blends across blocks. + // Top band K W R G, bottom band B Y M C. Shrinking 8×4 → 4×2 averages each 2×2 block, + // so every device pixel is one palette colour; Catmull-Rom blends across blocks. let palette: [[u8; 3]; 8] = [[0, 0, 0], [255, 255, 255], [255, 0, 0], [0, 255, 0], [0, 0, 255], [255, 255, 0], [255, 0, 255], [0, 255, 255]]; let rgb: Vec = block_source(palette).concat(); let exact: Vec<[u8; 4]> = palette.iter().map(|c| [c[0], c[1], c[2], 255]).collect(); @@ -6455,38 +6455,76 @@ fn non_interpolated_images_are_not_smoothed_when_minified() { assert!(pixels(&enabled).iter().any(|px| px[..3].iter().any(|&c| (32..=223).contains(&c))), "/Interpolate true made no mid-tone"); } -/// A shrunk, interpolated image whose edges fall between device pixels is snapped to whole -/// pixels (shrunk once to that size, drawn 1:1). Before, the shrunk raster was sampled a second -/// time at the fractional offset, so every edge of a scan was a blended, soft pixel. +/// A shrunk image whose edges fall between device pixels is snapped to whole pixels (shrunk +/// once to that size, drawn 1:1), interpolated or not. Before, the shrunk raster was sampled a +/// second time at the fractional offset, so every edge of a scan was a blended, soft pixel. #[test] -fn interpolated_minified_image_edges_snap_to_whole_pixels() { - let black = vec![0u8; 40 * 40]; - let content = "q 10 0 0 10 0.4 0.4 cm /Im0 Do Q"; - let image = hex(&black); - let pdf = format!( - "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ +fn minified_image_edges_snap_to_whole_pixels() { + for interpolate in ["/Interpolate true ", ""] { + let black = vec![0u8; 40 * 40]; + let content = "q 10 0 0 10 0.4 0.4 cm /Im0 Do Q"; + let image = hex(&black); + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 11 11] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ - 5 0 obj << /Type /XObject /Subtype /Image /Width 40 /Height 40 /ColorSpace /DeviceGray /BitsPerComponent 8 /Interpolate true /Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 40 /Height 40 /ColorSpace /DeviceGray /BitsPerComponent 8 {interpolate}/Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ trailer << /Root 1 0 R >>\n%%EOF\n", - content.len(), - image.len() + 1 - ); - let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); - let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); - assert!(page.error.is_none(), "{:?}", page.error); - assert_eq!((page.width, page.height), (11, 11)); - let gray = |x: usize, y: usize| page.rgba[(y * 11 + x) * 4]; - // Device box 0.4..10.4 × 0.6..10.6 snaps to columns 0..10 and rows 1..11. - for y in 1..11 { - for x in 0..10 { - assert_eq!(gray(x, y), 0, "pixel ({x}, {y}) inside the snapped image"); + content.len(), + image.len() + 1 + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); + assert!(page.error.is_none(), "{:?}", page.error); + assert_eq!((page.width, page.height), (11, 11)); + let gray = |x: usize, y: usize| page.rgba[(y * 11 + x) * 4]; + // Device box 0.4..10.4 × 0.6..10.6 snaps to columns 0..10 and rows 1..11. + for y in 1..11 { + for x in 0..10 { + assert_eq!(gray(x, y), 0, "{interpolate:?}: pixel ({x}, {y}) inside the snapped image"); + } + assert_eq!(gray(10, y), 255, "{interpolate:?}: column 10 is outside the snapped image (row {y})"); + } + for x in 0..11 { + assert_eq!(gray(x, 0), 255, "{interpolate:?}: row 0 is outside the snapped image (column {x})"); } - assert_eq!(gray(10, y), 255, "column 10 is outside the snapped image (row {y})"); } - for x in 0..11 { - assert_eq!(gray(x, 0), 255, "row 0 is outside the snapped image (column {x})"); +} + +/// A tile draws exactly the pixels of the whole page under it, also where a shrunk image's +/// edge falls on a half pixel. The edges were snapped by rounding halves away from zero, so +/// the tile at x = 5, which moves the left edge from 2.5 to -2.5, snapped it one pixel left of +/// where the whole page did and shifted every column of the image. +#[test] +fn a_tile_snaps_half_pixel_image_edges_as_the_whole_page_does() { + for interpolate in ["/Interpolate true ", ""] { + let ramp: Vec = (0..40 * 40).map(|i| (i % 40 * 6) as u8).collect(); + let content = "q 10 0 0 10 2.5 2.5 cm /Im0 Do Q"; + let image = hex(&ramp); + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ + 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ + 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 20 20] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ + 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 40 /Height 40 /ColorSpace /DeviceGray /BitsPerComponent 8 {interpolate}/Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ + trailer << /Root 1 0 R >>\n%%EOF\n", + content.len(), + image.len() + 1 + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let req = RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }; + let full = renderer.render(req); + let tile = Tile { x: 5, y: 5, w: 10, h: 10 }; + let part = renderer.render(RenderRequest { tile: Some(tile), ..req }); + assert!(full.error.is_none() && part.error.is_none(), "{:?} {:?}", full.error, part.error); + assert_eq!((full.width, part.width, part.height), (20, 10, 10)); + for y in 0..10usize { + for x in 0..10usize { + let whole = full.rgba[((y + 5) * 20 + x + 5) * 4]; + assert_eq!(part.rgba[(y * 10 + x) * 4], whole, "{interpolate:?}: pixel ({x}, {y}) of the tile"); + } + } } } @@ -6570,28 +6608,161 @@ fn non_interpolated_gray_images_are_not_smoothed_when_minified() { } #[test] -fn non_integer_minification_samples_the_same_columns() { - // Ten columns drawn six pixels wide: centre sampling reads columns 0, 2, 4, 5, 7 and 9. +fn non_integer_minification_averages_the_columns_each_pixel_covers() { + // Ten columns drawn six pixels wide: each pixel covers 10/6 columns. Pixel 0 is 60% of + // column 0 and 40% of column 1 (0.6 × 10 + 0.4 × 20 = 14), pixel 1 is 20% of column 1, + // all of column 2's 60% and 20% of column 3 (30), and so on. Nearest-neighbour read whole + // columns 0, 2, 4, 5, 7 and 9 instead, dropping the four others. let ramp = [10u8, 20, 30, 40, 50, 60, 70, 80, 90, 100]; let page = draw_minified_image("DeviceGray", (10, 1), None, &ramp, None, (6, 1)); let grays: Vec = page.rgba.as_chunks::<4>().0.iter().map(|px| px[0]).collect(); - assert_eq!(grays, [10, 30, 50, 60, 80, 100]); + assert_eq!(grays, [14, 30, 46, 64, 80, 96]); } #[test] -fn non_integer_minification_samples_colour_and_mask_from_the_same_columns() { - // Ten gray columns with a soft mask, drawn six pixels wide: columns 0, 2, 4, 5, 7 and 9 are - // sampled, so the mask must pick the same columns as the colours. +fn non_integer_minification_averages_colour_and_mask_together() { + // Ten gray columns with a soft mask, drawn six pixels wide. Colours are averaged + // premultiplied, over the same columns as the mask, and composited over the white page. + // Pixel 0: 60% of column 0 (gray 10, opaque) and 40% of column 1 (transparent) is + // premultiplied gray 6 at alpha 153, so 6 + 255 × (1 − 153/255) = 108. Pixels 1 and 4 + // cover only transparent columns and show the page. let ramp = [10u8, 20, 30, 40, 50, 60, 70, 80, 90, 100]; let mask = [255u8, 0, 0, 0, 255, 255, 0, 0, 0, 255]; let page = draw_minified_image("DeviceGray", (10, 1), None, &ramp, Some(&mask), (6, 1)); - // Opaque columns keep their gray; transparent columns show the white page. assert_eq!( page.rgba.as_chunks::<4>().0.to_vec(), - vec![[10, 10, 10, 255], [255, 255, 255, 255], [50, 50, 50, 255], [60, 60, 60, 255], [255, 255, 255, 255], [100, 100, 100, 255]] + vec![[108, 108, 108, 255], [255, 255, 255, 255], [132, 132, 132, 255], [138, 138, 138, 255], [255, 255, 255, 255], [162, 162, 162, 255]] ); } +/// A scan shown at screen size: a stroke one sample wide in an image drawn at a third of its +/// resolution. Nearest-neighbour read columns 1, 4 and 7 and lost the stroke in column 3, which +/// is how the text of 300 dpi scans broke up at 100% zoom. Area averaging keeps it as the +/// third of pixel 1 it covers. +#[test] +fn non_interpolated_minification_keeps_thin_strokes() { + let mut stroke = [255u8; 9]; + stroke[3] = 0; + let page = draw_minified_image("DeviceGray", (9, 1), None, &stroke, None, (3, 1)); + let grays: Vec = page.rgba.as_chunks::<4>().0.iter().map(|px| px[0]).collect(); + assert_eq!(grays, [255, 170, 255]); +} + +/// A zoomed scan: an image of page-scan size drawn three times its resolution without +/// `/Interpolate`. Nearest-neighbour turned each sample into a 3 × 3 block, so the edges of +/// scanned text were staircases; it is smoothed (bilinearly) instead, as browsers draw it. +#[test] +fn large_non_interpolated_images_are_smoothed_when_enlarged() { + let side = 512usize; + let half: Vec = (0..side * side).map(|i| if i % side < side / 2 { 0 } else { 255 }).collect(); + let page = draw_minified_image("DeviceGray", (512, 512), None, &half, None, (1536, 1536)); + let row = 768 * 1536; + let edge: Vec = (760..776).map(|x| page.rgba[(row + x) * 4]).collect(); + assert!(edge.iter().any(|v| (32..=223).contains(v)), "a hard edge, not a smoothed one: {edge:?}"); + assert_eq!((page.rgba[(row + 700) * 4], page.rgba[(row + 836) * 4]), (0, 255), "away from the edge"); +} + +/// Small images (icons, QR codes, pixel art) keep their hard pixels when enlarged. +#[test] +fn small_non_interpolated_images_keep_hard_pixels_when_enlarged() { + let checker: Vec = (0..64).map(|i| if (i / 8 + i % 8) % 2 == 0 { 0 } else { 255 }).collect(); + let page = draw_minified_image("DeviceGray", (8, 8), None, &checker, None, (80, 80)); + assert!(page.rgba.as_chunks::<4>().0.iter().all(|px| px[0] == 0 || px[0] == 255), "a small image was smoothed"); +} + +/// A scan shrunk across and stretched down (a fax shown at screen size, zoomed in): the shrunk +/// side is still averaged, and the stretched side is smoothed as a scan enlarged both ways +/// is. Shrinking one side kept the nearest sampler for the other, so its edges were staircases. +#[test] +fn a_scan_shrunk_one_way_is_smoothed_the_way_it_is_stretched() { + // Top half black; bottom half columns alternating black and white, which averages to gray. + let side = 512usize; + let source: Vec = (0..side * side).map(|i| if i / side < side / 2 || i % 2 == 0 { 0 } else { 255 }).collect(); + let page = draw_minified_image("DeviceGray", (512, 512), None, &source, None, (256, 1536)); + let gray = |x: usize, y: usize| page.rgba[(y * 256 + x) * 4]; + let edge: Vec = (760..776).map(|y| gray(128, y)).collect(); + assert!(edge.iter().any(|v| (16..=112).contains(v)), "a hard edge down the stretched side: {edge:?}"); + assert_eq!(gray(128, 700), 0, "above the edge"); + assert!((0..256).all(|x| gray(x, 836).abs_diff(128) <= 1), "the shrunk side is averaged at 1:1"); +} + +/// An image shrunk across and stretched down (a 204 × 98 dpi fax shown at screen size): the +/// shrunk side is still averaged. The area shrink was asked to enlarge the other side too, +/// refused, and the image fell back to its source resolution and the nearest sampler, which +/// read single columns. +#[test] +fn an_image_shrunk_one_way_and_stretched_the_other_is_still_averaged() { + let ramp: Vec = [10u8, 20, 30, 40, 50, 60, 70, 80].repeat(2); + let page = draw_minified_image("DeviceGray", (8, 2), None, &ramp, None, (4, 4)); + let grays: Vec = page.rgba.as_chunks::<4>().0.iter().map(|px| px[0]).collect(); + assert_eq!(grays, [15, 35, 55, 75].repeat(4), "each pixel is the mean of the two columns it covers"); +} + +/// Averaging straight colours would tint an opaque sample with the colour of a transparent +/// neighbour: white beside transparent black, shrunk to one pixel, came out gray over the white +/// page. Premultiplied, the transparent sample adds nothing but its transparency. +#[test] +fn non_interpolated_minification_does_not_bleed_transparent_colour() { + let page = draw_minified_image("DeviceGray", (2, 1), None, &[255, 0], Some(&[255, 0]), (1, 1)); + assert_eq!(page.rgba.as_chunks::<4>().0.to_vec(), vec![[255, 255, 255, 255]]); +} + +/// The same with a soft mask smaller than its image: red, red, blue, blue under a mask that is +/// opaque, then transparent, shrunk to one pixel, is half-covered red over the white page. +/// Colour and mask were averaged apart and multiplied, so the transparent blue came through +/// as purple (191, 127, 191). +#[test] +fn a_smaller_soft_mask_does_not_bleed_transparent_colour_when_minified() { + let content = "q 1 0 0 1 0 0 cm /Im0 Do Q"; + let (image, mask) = ("FF0000FF00000000FF0000FF", "FF00"); + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ + 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ + 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 1 1] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ + 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 4 /Height 1 /ColorSpace /DeviceRGB /BitsPerComponent 8 /SMask 6 0 R /Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ + 6 0 obj << /Type /XObject /Subtype /Image /Width 2 /Height 1 /ColorSpace /DeviceGray /BitsPerComponent 8 /Filter /ASCIIHexDecode /Length {} >> stream\n{mask}>\nendstream endobj\n\ + trailer << /Root 1 0 R >>\n%%EOF\n", + content.len(), + image.len() + 1, + mask.len() + 1 + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); + assert!(page.error.is_none(), "{:?}", page.error); + let [r, g, b, _] = page.rgba.as_chunks::<4>().0.first().copied().unwrap_or_default(); + assert!(r >= 254 && g.abs_diff(128) <= 1 && b == g, "half-covered red over white, got ({r}, {g}, {b})"); +} + +/// A smaller soft mask that asks for `/Interpolate` stays smooth when it is sampled onto the grid +/// of an area-shrunk image. Here a 64 × 64 black image is drawn at 32 × 32 under a 2 × 2 mask, +/// opaque on the left and clear on the right: the mask is enlarged while the image shrinks, and +/// sampled nearest-neighbour onto the image's grid the edge became a hard black and white step. +#[test] +fn an_interpolated_smaller_soft_mask_stays_smooth_when_its_image_is_minified() { + let content = "q 32 0 0 32 0 0 cm /Im0 Do Q"; + let (image, mask) = ("00".repeat(64 * 64), "FF00FF00"); + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ + 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ + 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 32 32] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ + 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 64 /Height 64 /ColorSpace /DeviceGray /BitsPerComponent 8 /SMask 6 0 R /Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ + 6 0 obj << /Type /XObject /Subtype /Image /Width 2 /Height 2 /ColorSpace /DeviceGray /BitsPerComponent 8 /Interpolate true /Filter /ASCIIHexDecode /Length {} >> stream\n{mask}>\nendstream endobj\n\ + trailer << /Root 1 0 R >>\n%%EOF\n", + content.len(), + image.len() + 1, + mask.len() + 1 + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); + assert!(page.error.is_none(), "{:?}", page.error); + assert_eq!((page.width, page.height), (32, 32)); + let row: Vec = page.rgba.as_chunks::<4>().0.iter().skip(16 * 32).take(32).map(|[r, ..]| *r).collect(); + let ramp = row.iter().filter(|v| (1..255).contains(*v)).count(); + assert!(ramp >= 8 && row.is_sorted() && row.first() == Some(&0) && row.last() == Some(&255), "a smooth black to white edge, got {row:?}"); +} + #[test] fn non_interpolated_alpha_images_keep_their_colours_and_mask_when_minified() { let palette: [[u8; 3]; 8] = [[0, 0, 0], [255, 255, 255], [255, 0, 0], [0, 255, 0], [0, 0, 255], [255, 255, 0], [255, 0, 255], [0, 255, 255]]; @@ -6643,6 +6814,59 @@ fn shrunk_soft_mask_larger_than_its_image_keeps_thin_strokes() { assert_eq!(grays[3], 255, "{grays:?}"); } +/// Issue #4800: scanners save the text of a page as a 1-bit `/ImageMask` stencil, painted in the +/// fill colour, and a stencil shrunk by nearest-neighbour lost the strokes thinner than the shrink +/// factor. Here a one-column red stroke of an 8-column stencil is drawn 4 pixels wide: nearest +/// reads columns 1, 3, 5 and 7 and misses it, while the area average makes it a half-covered red. +#[test] +fn a_shrunk_stencil_mask_keeps_thin_strokes() { + // Sample 0 paints (the default `/Decode [0 1]`), so 0xDF is a stroke in column 2. + let content = "q 1 0 0 rg 4 0 0 1 0 0 cm /Im0 Do Q"; + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ + 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ + 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 4 1] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ + 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 8 /Height 1 /ImageMask true /Filter /ASCIIHexDecode /Length 3 >> stream\nDF>\nendstream endobj\n\ + trailer << /Root 1 0 R >>\n%%EOF\n", + content.len() + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); + assert!(page.error.is_none(), "{:?}", page.error); + assert_eq!((page.width, page.height), (4, 1)); + let px = page.rgba.as_chunks::<4>().0.to_vec(); + let [r, g, b, _] = px.get(1).copied().unwrap_or_default(); + assert!(r >= 254 && g.abs_diff(128) <= 1 && b == g, "half-covered red over white, got {px:?}"); + assert_eq!(px.get(3), Some(&[255, 255, 255, 255]), "{px:?}"); +} + +/// A scan drawn at its own resolution keeps its hard pixels when its scale passes 1 only by +/// rounding: `x_scale` is an `f32`, and a 192 dpi scan on a 2× display at 100% can come out a +/// few ulps above 1. Smoothed bilinearly at a half-pixel offset, one-pixel stripes became gray. +/// Here `600.0001` (an `f32` of 600.00012) draws 600 samples 2·10⁻⁷ wider than 1:1. +#[test] +fn a_scan_drawn_a_rounding_error_above_its_resolution_stays_sharp() { + let content = "q 600.0001 0 0 600 0.5 0 cm /Im0 Do Q"; + let image: String = (0..600 * 600).map(|i| if i % 2 == 0 { "00" } else { "FF" }).collect(); + let pdf = format!( + "%PDF-1.4\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\ + 2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\ + 3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 601 600] /Contents 4 0 R /Resources << /XObject << /Im0 5 0 R >> >> >> endobj\n\ + 4 0 obj << /Length {} >> stream\n{content}\nendstream endobj\n\ + 5 0 obj << /Type /XObject /Subtype /Image /Width 600 /Height 600 /ColorSpace /DeviceGray /BitsPerComponent 8 /Filter /ASCIIHexDecode /Length {} >> stream\n{image}>\nendstream endobj\n\ + trailer << /Root 1 0 R >>\n%%EOF\n", + content.len(), + image.len() + 1 + ); + let mut renderer = PageRenderer::new(Arc::new(pdf.into_bytes()), RenderConfig::default()); + let page = renderer.render(RenderRequest { page: 0, kind: RequestKind::Pixels, scale: 1.0, ..Default::default() }); + assert!(page.error.is_none(), "{:?}", page.error); + let row: Vec = page.rgba.as_chunks::<4>().0.iter().skip(300 * page.width as usize).take(page.width as usize).map(|[r, ..]| *r).collect(); + let gray = row.iter().skip(10).take(580).filter(|v| (16..240).contains(*v)).count(); + assert_eq!(gray, 0, "one-pixel stripes were blurred: {:?}", row.get(10..30)); +} + /// Type 3 glyph images used to be sampled bicubically (`ImageQuality::High`) whatever their /// `/Interpolate`, which blends the nearest pre-resize back in. The glyph is 4.6 × 2.3 pixels /// and its image pre-resizes to 5 × 3, so the final draw samples fractionally and bicubic @@ -7716,7 +7940,7 @@ fn stitched_render_partitions_match_a_whole_render() { /// and square caps, a sharp miter, a rotated and a stretched matrix, a hairline, dashes, a /// curve, text, and a page shown rotated. /// -/// Images are skipped the same way (vendored patch (10)): an image on a tile, one crossing +/// Images are skipped the same way (vendored patch (11)): an image on a tile, one crossing /// tile lines, and one fully off the page. #[test] fn skipping_paths_off_the_canvas_changes_no_pixel() { diff --git a/crates/ui-egui/src/app.rs b/crates/ui-egui/src/app.rs index 1e85ef7b8b..4626c802ce 100644 --- a/crates/ui-egui/src/app.rs +++ b/crates/ui-egui/src/app.rs @@ -108,6 +108,7 @@ impl PdfCraftApp { ui_scale_applied: None, flatten_fill_sign_on_save: false, fill_text_size: fill_sign::TEXT_SIZE, + fill_tool: fill_sign::FillTool::Signature, fill_text_color: [0.0; 3], dialog: None, updates: updates::Updates::default(), @@ -128,6 +129,7 @@ impl PdfCraftApp { last_session: Default::default(), quit_session: None, reload_changed_files: true, + update_check_at_start: true, show_secured_notice: true, always_default_view: false, #[cfg(not(target_arch = "wasm32"))] diff --git a/crates/ui-egui/src/app/control.rs b/crates/ui-egui/src/app/control.rs index 0309042bf1..5d51adfdb4 100644 --- a/crates/ui-egui/src/app/control.rs +++ b/crates/ui-egui/src/app/control.rs @@ -47,6 +47,9 @@ impl PdfCraftApp { ("show_page_edge_shadows", _) => { self.show_page_edge_shadows = value.parse::().map_err(|_| "show_page_edge_shadows must be true or false")?; } + ("update_check_at_start", _) => { + self.update_check_at_start = value.parse::().map_err(|_| "update_check_at_start must be true or false")?; + } ("scroll_speed", _) => { self.scroll_speed = scroll_speed::ScrollSpeed::parse(value).ok_or("scroll_speed must be slow, normal, or fast")?; } @@ -333,9 +336,12 @@ impl PdfCraftApp { field if field.starts_with("field-") => QuickTool::Field( prepare::FieldTool::from_command(&format!("form.add.{}", &field[6..])).ok_or_else(|| format!("unknown tool {field}"))?, ), - fill if fill.starts_with("fill-") => QuickTool::Fill( - fill_sign::FillTool::from_command(&format!("sign.fill.{}", &fill[5..])).ok_or_else(|| format!("unknown tool {fill}"))?, - ), + fill if fill.starts_with("fill-") => { + let tool = + fill_sign::FillTool::from_command(&format!("sign.fill.{}", &fill[5..])).ok_or_else(|| format!("unknown tool {fill}"))?; + self.fill_tool = tool; + QuickTool::Fill(tool) + } other => { let t = comments::CommentTool::from_command(&format!("comment.{other}")).ok_or_else(|| format!("unknown tool {other}"))?; self.comment_prefs.group_tool[t.group()] = t; diff --git a/crates/ui-egui/src/app/prefs.rs b/crates/ui-egui/src/app/prefs.rs index 08a20afb45..0df7d6f30b 100644 --- a/crates/ui-egui/src/app/prefs.rs +++ b/crates/ui-egui/src/app/prefs.rs @@ -147,6 +147,9 @@ impl PdfCraftApp { // Set outside json! above, which is at serde_json's macro recursion limit. v["hide_empty_highlight_hover"] = serde_json::Value::Bool(self.comment_prefs.hide_empty_highlight_hover); v["show_page_edge_shadows"] = self.show_page_edge_shadows.into(); + // Not `check_updates_at_start`: that key was an opt-in of its own, off unless chosen, and + // settings written then would turn this off unasked. + v["update_check_at_start"] = self.update_check_at_start.into(); v["hotbar"] = serde_json::to_value(&self.hotbar).unwrap_or_default(); v["comment_tool_pinned"] = serde_json::Value::Bool(self.comment_prefs.pinned); v["show_secured_notice"] = serde_json::Value::Bool(self.show_secured_notice); @@ -219,6 +222,9 @@ impl PdfCraftApp { if let Some(on) = v["reload_changed_files"].as_bool() { self.reload_changed_files = on; } + if let Some(on) = v["update_check_at_start"].as_bool() { + self.update_check_at_start = on; + } if let Some(on) = v["show_secured_notice"].as_bool() { self.show_secured_notice = on; } diff --git a/crates/ui-egui/src/canvas/bars.rs b/crates/ui-egui/src/canvas/bars.rs index dd3fe8c498..fbd4200217 100644 --- a/crates/ui-egui/src/canvas/bars.rs +++ b/crates/ui-egui/src/canvas/bars.rs @@ -315,7 +315,7 @@ mod tests { #[test] fn a_wider_quick_bar_centres_by_its_own_width_and_needs_a_wider_margin() { - // Fill & Sign's Text tool widens the bar to 70 pt: it centres by that, and 72 pt no longer holds it. + // A 70 pt bar centres by that, and 72 pt no longer holds it. assert_eq!(quick_bar_offset(120.0, 70.0), 15.0); assert_eq!(quick_bar_offset(72.0, 70.0), 14.0); } diff --git a/crates/ui-egui/src/commands.rs b/crates/ui-egui/src/commands.rs index 083ada4253..3549418ae3 100644 --- a/crates/ui-egui/src/commands.rs +++ b/crates/ui-egui/src/commands.rs @@ -697,6 +697,7 @@ impl PdfCraftApp { } fill if crate::fill_sign::FillTool::from_command(fill).is_some() => { let Some(tool) = crate::fill_sign::FillTool::from_command(fill) else { return Some(false) }; + self.fill_tool = tool; self.quick_tool = crate::QuickTool::Fill(tool); let initials = tool == crate::fill_sign::FillTool::Initials; if (tool == crate::fill_sign::FillTool::Signature && self.signature.is_none()) || (initials && self.initials.is_none()) { diff --git a/crates/ui-egui/src/dialogs.rs b/crates/ui-egui/src/dialogs.rs index 84c36f41aa..0e7e2c0ed0 100644 --- a/crates/ui-egui/src/dialogs.rs +++ b/crates/ui-egui/src/dialogs.rs @@ -132,7 +132,7 @@ fn dialogs(app: &mut PdfCraftApp, ctx: &egui::Context) { Dialog::Print => 820.0, Dialog::FieldProps | Dialog::BulkFieldProps => 600.0, Dialog::Preferences => 610.0, - Dialog::About => 780.0, + Dialog::About => 820.0, Dialog::PinTool => 540.0, _ => 520.0, }); @@ -1352,6 +1352,7 @@ fn dialog_signature(ui: &mut egui::Ui, app: &mut PdfCraftApp, req: &mut Pending) app.signature = d.saved(); crate::fill_sign::FillTool::Signature }; + app.fill_tool = tool; app.quick_tool = crate::QuickTool::Fill(tool); app.toast = None; } diff --git a/crates/ui-egui/src/dialogs/credits.rs b/crates/ui-egui/src/dialogs/credits.rs index ad45ec77b6..d69d545b3f 100644 --- a/crates/ui-egui/src/dialogs/credits.rs +++ b/crates/ui-egui/src/dialogs/credits.rs @@ -266,12 +266,17 @@ pub fn contributors_ui(ui: &mut egui::Ui) { let counts = crate::i18n::fmt(crate::i18n::t("{n} contributors · {c} commits"), &[("n", &list.len().to_string()), ("c", &group(TOTAL_COMMITS))]); ui.label(RichText::new(counts).small().color(t.text_muted)); ui.separator(); - egui::ScrollArea::both().id_salt("credits_list").max_height(LIST_HEIGHT).auto_shrink([false, true]).show(ui, |ui| { - if list.is_empty() { - ui.label(crate::i18n::t("No contributor data was built into this copy.")); - } else if v.table { - table(ui, &list, &mut v); - } else { + if list.is_empty() { + ui.label(crate::i18n::t("No contributor data was built into this copy.")); + } else if v.table { + // The table scrolls vertically itself, under a fixed header; it only scrolls sideways + // once its columns are dragged wider than the dialog. + ui.allocate_ui(egui::vec2(ui.available_width(), LIST_HEIGHT), |ui| { + ui.set_max_height(LIST_HEIGHT); + egui::ScrollArea::horizontal().id_salt("credits_list").auto_shrink([false, false]).show(ui, |ui| table(ui, &list, &mut v)); + }); + } else { + egui::ScrollArea::vertical().id_salt("credits_list").max_height(LIST_HEIGHT).auto_shrink([false, true]).show(ui, |ui| { ui.horizontal_wrapped(|ui| { for (i, c) in list.iter().enumerate() { if i > 0 { @@ -280,43 +285,154 @@ pub fn contributors_ui(ui: &mut egui::Ui) { ui.hyperlink_to(c.name(v.names), format!("https://github.com/{}", c.login)).on_hover_text(c.summary()); } }); - } - }); + }); + } ui.data_mut(|d| d.insert_temp(id, v)); } -fn table(ui: &mut egui::Ui, list: &[&Contributor], v: &mut View) { - egui::Grid::new("credits_table").striped(true).num_columns(SortKey::ALL.len()).show(ui, |ui| { - for k in SortKey::ALL { - let arrow = if v.key == k { if v.ascending { " ▲" } else { " ▼" } } else { "" }; - if ui - .button(RichText::new(format!("{}{arrow}", crate::i18n::t(k.label().1))).strong()) - .on_hover_text(crate::i18n::t(k.label().0)) - .clicked() - { - if v.key == k { - v.ascending = !v.ascending; - } else { - v.key = k; - v.ascending = k.default_ascending(); - } +/// How the contributor table's spare width is shared out: all of it to Name. +const CONTRIBUTOR_GROW: [f32; 10] = [1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]; + +/// How the model table's spare width is shared out: to Company and Model, the text columns. +const MODEL_GROW: [f32; 6] = [1.0, 2.0, 0.0, 0.0, 0.0, 0.0]; + +/// The narrowest a column can be dragged. +const MIN_COLUMN: f32 = 28.0; + +/// The width a column needs for its header and its widest cell. +fn natural(ui: &egui::Ui, header: RichText, cells: impl Iterator) -> f32 { + let measure = |text: egui::WidgetText| text.into_galley(ui, Some(egui::TextWrapMode::Extend), f32::INFINITY, egui::TextStyle::Body).size().x; + let widest = cells.map(|c| measure(c.into())).fold(measure(header.into()), f32::max); + widest + 2.0 * ui.spacing().button_padding.x +} + +/// Opening column widths: each column's natural width, with the spare room shared out by `grow`, +/// so the columns fill `room` exactly. When there is too little room, only the growing columns +/// give way, each in proportion to its width above [`MIN_COLUMN`], and none goes below that. +fn fit_widths(natural: &[f32], grow: &[f32], room: f32) -> Vec { + let spare = room - natural.iter().sum::(); + let weight = grow.iter().sum::(); + let slack = |w: f32| (w - MIN_COLUMN).max(0.0); + let flexible: f32 = natural.iter().zip(grow).filter(|(_, g)| **g > 0.0).map(|(w, _)| slack(*w)).sum(); + natural + .iter() + .zip(grow) + .map(|(w, g)| { + if *g > 0.0 && spare >= 0.0 { + w + spare * g / weight + } else if *g > 0.0 && flexible > 0.0 { + w - (-spare).min(flexible) * slack(*w) / flexible + } else { + *w } - } - ui.end_row(); - for c in list { - ui.hyperlink_to(c.name(v.names), format!("https://github.com/{}", c.login)).on_hover_text(c.summary()); - ui.label(group(c.prs)); - ui.label(group(c.commits)); - ui.label(group(c.lines_added)); - ui.label(group(c.lines_deleted)); - ui.label(signed(c.lines_delta())); - ui.label(group(c.binary_added)); - ui.label(group(c.binary_deleted)); - ui.label(day(c.first_commit)); - ui.label(day(c.last_commit)); - ui.end_row(); - } - }); + }) + .collect() +} + +/// A striped table whose columns open at [`fit_widths`], so it fills the width and nothing scrolls +/// sideways. Every column, the last included, can be dragged wider or narrower at its right edge: +/// only a table dragged wider than the window scrolls sideways. Text too long for its column is +/// clipped. When a column's natural width or the window changes, the columns go back to a new fit, +/// dropping any dragged widths. +fn fitted_table<'a>(ui: &'a mut egui::Ui, id: &str, natural: &[f32], grow: &[f32]) -> egui_extras::TableBuilder<'a> { + // Every cell is followed by the item gap, and the vertical scrollbar keeps its margins. + let s = ui.spacing(); + let gaps = s.item_spacing.x * natural.len() as f32; + let bar = s.scroll.bar_width + s.scroll.bar_inner_margin + s.scroll.bar_outer_margin; + let widths = fit_widths(natural, grow, (ui.available_width() - gaps - bar).max(0.0)); + // The table keeps its own column widths, so a change in what they were fitted to has to clear + // them. A longer name can leave the fit itself unchanged, and scrolling can nudge the available + // width, so this compares the natural widths and the clip rect, which scrolling never moves. + let fit = egui::Id::unique(("credits_fit", id)); + let measured = (natural.to_vec(), ui.clip_rect().width()); + let refit = ui.data_mut(|d| d.get_temp::<(Vec, f32)>(fit)).as_ref() != Some(&measured); + ui.data_mut(|d| d.insert_temp(fit, measured)); + let mut t = + egui_extras::TableBuilder::new(ui).id_salt(id).striped(true).resizable(true).auto_shrink([false, false]).max_scroll_height(f32::INFINITY); + if refit { + t.reset(); + } + for w in widths { + t = t.column(egui_extras::Column::initial(w).at_least(MIN_COLUMN).clip(true)); + } + t +} + +/// Add a cell's widget right-aligned (counts, dates) or left-aligned (names), vertically centred. +fn aligned(ui: &mut egui::Ui, right: bool, add: impl FnOnce(&mut egui::Ui) -> R) -> R { + let layout = if right { egui::Layout::right_to_left(egui::Align::Center) } else { egui::Layout::left_to_right(egui::Align::Center) }; + ui.with_layout(layout, add).inner +} + +fn row_height(ui: &egui::Ui) -> f32 { + ui.text_style_height(&egui::TextStyle::Body) + 6.0 +} + +/// One contributor's table cells after the name, in `SortKey::ALL` order. +fn cells(c: &Contributor) -> [String; 9] { + [ + group(c.prs), + group(c.commits), + group(c.lines_added), + group(c.lines_deleted), + signed(c.lines_delta()), + group(c.binary_added), + group(c.binary_deleted), + day(c.first_commit).to_string(), + day(c.last_commit).to_string(), + ] +} + +/// A contributor-table header. The sorted column is highlighted; its direction shows on the +/// toolbar's sort-order button, so no arrow widens the column. +fn header_text(k: SortKey) -> RichText { + RichText::new(crate::i18n::t(k.label().1)).strong() +} + +fn table(ui: &mut egui::Ui, list: &[&Contributor], v: &mut View) { + let h = row_height(ui); + let rows: Vec<[String; 9]> = list.iter().map(|c| cells(c)).collect(); + let widths: Vec = SortKey::ALL + .iter() + .enumerate() + .map(|(i, &k)| match i.checked_sub(1) { + None => natural(ui, header_text(k), list.iter().map(|c| c.name(v.names))), + Some(col) => natural(ui, header_text(k), rows.iter().filter_map(|r| r.get(col).cloned())), + }) + .collect(); + fitted_table(ui, "credits_table", &widths, &CONTRIBUTOR_GROW) + .header(h + 4.0, |mut header| { + for k in SortKey::ALL { + header.col(|ui| { + let r = aligned(ui, k != SortKey::Name, |ui| ui.selectable_label(v.key == k, header_text(k))) + .on_hover_text(crate::i18n::t(k.label().0)); + + if r.clicked() { + if v.key == k { + v.ascending = !v.ascending; + } else { + v.key = k; + v.ascending = k.default_ascending(); + } + } + }); + } + }) + .body(|mut body| { + for (c, cells) in list.iter().zip(rows) { + body.row(h, |mut row| { + row.col(|ui| { + aligned(ui, false, |ui| ui.hyperlink_to(c.name(v.names), format!("https://github.com/{}", c.login))) + .on_hover_text(c.summary()); + }); + for text in cells { + row.col(|ui| { + aligned(ui, true, |ui| ui.label(text)); + }); + } + }); + } + }); } /// About ▸ Models: AI models credited in Co-Authored-By trailers. @@ -326,21 +442,50 @@ pub fn models_ui(ui: &mut egui::Ui) { return; } let assisted: u64 = MODELS.iter().map(|m| m.commits).max().unwrap_or(0).max(1); - egui::ScrollArea::both().id_salt("credits_models").max_height(LIST_HEIGHT).auto_shrink([false, true]).show(ui, |ui| { - egui::Grid::new("credits_models").striped(true).num_columns(6).show(ui, |ui| { - for h in MODEL_COLUMNS { - ui.label(RichText::new(crate::i18n::t(h)).strong()); - } - ui.end_row(); - for m in MODELS { - ui.label(m.company); - ui.label(m.model); - ui.label(m.version); - ui.label(group(m.commits)); - ui.label(format!("{:.0}%", 100.0 * m.commits as f64 / TOTAL_COMMITS.max(assisted) as f64)); - ui.label(format!("+{} / −{}", group(m.lines_added), group(m.lines_deleted))); - ui.end_row(); - } + let h = row_height(ui); + let titles = MODEL_COLUMNS; + // Company, Model and Version read as text; the counts line up on the right. + let right = [false, false, false, true, true, true]; + let rows: Vec<[String; 6]> = MODELS + .iter() + .map(|m| { + [ + m.company.to_string(), + m.model.to_string(), + m.version.to_string(), + group(m.commits), + format!("{:.0}%", 100.0 * m.commits as f64 / TOTAL_COMMITS.max(assisted) as f64), + format!("+{} / −{}", group(m.lines_added), group(m.lines_deleted)), + ] + }) + .collect(); + let widths: Vec = titles + .iter() + .enumerate() + .map(|(i, title)| natural(ui, RichText::new(crate::i18n::t(title)).strong(), rows.iter().filter_map(|r| r.get(i).cloned()))) + .collect(); + ui.allocate_ui(egui::vec2(ui.available_width(), LIST_HEIGHT), |ui| { + ui.set_max_height(LIST_HEIGHT); + egui::ScrollArea::horizontal().id_salt("credits_models").auto_shrink([false, false]).show(ui, |ui| { + fitted_table(ui, "credits_models", &widths, &MODEL_GROW) + .header(h + 4.0, |mut header| { + for (title, r) in titles.into_iter().zip(right) { + header.col(|ui| { + aligned(ui, r, |ui| ui.label(RichText::new(crate::i18n::t(title)).strong())); + }); + } + }) + .body(|mut body| { + for cells in rows { + body.row(h, |mut row| { + for (text, r) in cells.into_iter().zip(right) { + row.col(|ui| { + aligned(ui, r, |ui| ui.label(text)); + }); + } + }); + } + }); }); }); } @@ -421,4 +566,131 @@ mod tests { assert_eq!(day("2026-10-07T18:45:23Z"), "2026-10-07"); assert_eq!(day(""), "—"); } + + #[test] + fn columns_fill_the_room_and_spare_width_goes_to_growing_columns() { + assert_eq!(fit_widths(&[100.0, 40.0, 60.0], &[1.0, 0.0, 0.0], 300.0), [200.0, 40.0, 60.0]); + assert_eq!(fit_widths(&[100.0, 40.0, 60.0], &[1.0, 2.0, 0.0], 230.0), [110.0, 60.0, 60.0]); + let w = fit_widths(&[123.0, 45.5, 67.25], &[1.0, 0.0, 2.0], 512.0); + assert!((w.iter().sum::() - 512.0).abs() < 0.01, "a fitted table neither scrolls nor leaves a gap: {w:?}"); + } + + #[test] + fn too_little_room_shrinks_only_growing_columns() { + assert_eq!(fit_widths(&[100.0, 40.0, 60.0], &[1.0, 0.0, 0.0], 150.0), [50.0, 40.0, 60.0]); + assert_eq!(fit_widths(&[100.0, 40.0, 60.0], &[1.0, 0.0, 0.0], 0.0), [MIN_COLUMN, 40.0, 60.0]); + assert_eq!(fit_widths(&[100.0, 40.0], &[0.0, 0.0], 50.0), [100.0, 40.0]); + } + + #[test] + fn a_column_at_the_minimum_leaves_the_shortfall_to_the_others() { + let w = fit_widths(&[100.0, 300.0, 50.0], &[1.0, 1.0, 0.0], 140.0); + assert!((w.iter().sum::() - 140.0).abs() < 0.01, "the columns still fit the room: {w:?}"); + assert!(w.iter().all(|c| *c >= MIN_COLUMN), "no column goes below the minimum: {w:?}"); + assert_eq!(fit_widths(&[100.0, 300.0, 50.0], &[1.0, 1.0, 0.0], 0.0), [MIN_COLUMN, MIN_COLUMN, 50.0]); + } + + const WINDOW: f32 = 640.5; + const NATURAL: [f32; 3] = [120.0, 80.0, 80.0]; + const GROW: [f32; 3] = [1.0, 0.0, 0.0]; + + /// One frame of a table long enough to scroll; the rects of its header cells. + fn table_frame(ctx: &egui::Context, width: f32, natural: &[f32], events: Vec) -> Vec { + scrolled_frame(ctx, width, natural, 0.0, events) + } + + /// [`table_frame`], scrolled `offset` points sideways. + fn scrolled_frame(ctx: &egui::Context, width: f32, natural: &[f32], offset: f32, events: Vec) -> Vec { + let screen_rect = Some(egui::Rect::from_min_size(egui::Pos2::ZERO, egui::vec2(width, 300.0))); + let raw = egui::RawInput { screen_rect, events, ..Default::default() }; + let mut cells = Vec::new(); + let out = ctx.run_ui(raw, |ui| { + cells.clear(); + egui::ScrollArea::horizontal().auto_shrink([false, false]).horizontal_scroll_offset(offset).show(ui, |ui| { + fitted_table(ui, "test", natural, &GROW) + .header(20.0, |mut header| { + for _ in natural { + header.col(|ui| cells.push(ui.max_rect())); + } + }) + .body(|mut body| { + for _ in 0..40 { + body.row(20.0, |mut row| { + for _ in natural { + row.col(|ui| { + ui.label("x"); + }); + } + }); + } + }); + }); + }); + out.drop_without_applying_deltas(); + cells + } + + /// A new context with the table open and settled, and the rects of its header cells. + fn opened_table() -> (egui::Context, Vec) { + let ctx = egui::Context::default(); + table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + let cells = table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + (ctx, cells) + } + + /// Drags a column's right edge, which runs midway through the gap after `cell`, then lets go. + fn drag_edge(ctx: &egui::Context, cell: egui::Rect, by: f32) { + use egui::{Event, Modifiers, PointerButton}; + let start = egui::pos2(cell.right() + 4.0, cell.center().y); + let halfway = start + egui::vec2(by / 2.0, 0.0); + let end = start + egui::vec2(by, 0.0); + let moved = Event::PointerMoved; + let button = |pos, pressed| Event::PointerButton { pos, button: PointerButton::Primary, pressed, modifiers: Modifiers::NONE }; + for event in [moved(start), button(start, true), moved(halfway), moved(end), button(end, false)] { + table_frame(ctx, WINDOW, &NATURAL, vec![event]); + } + } + + #[test] + fn every_column_can_be_dragged_at_its_right_edge_and_stays_there() { + let (_, opened) = opened_table(); + for (dragged, cell) in opened.iter().enumerate() { + let (ctx, _) = opened_table(); + drag_edge(&ctx, *cell, -30.0); + table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + let after = table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + for (column, (was, now)) in opened.iter().zip(&after).enumerate() { + let want = if column == dragged { was.width() - 30.0 } else { was.width() }; + assert!((now.width() - want).abs() < 1.0, "dragging column {dragged}: column {column} is {now:?}, was {was:?}"); + } + } + } + + #[test] + fn a_changed_fit_replaces_the_column_widths() { + let (ctx, opened) = opened_table(); + drag_edge(&ctx, opened[0], -30.0); + let longer_name = table_frame(&ctx, WINDOW, &[200.0, 80.0, 80.0], Vec::new()); + assert!((longer_name[0].width() - opened[0].width()).abs() < 1.0, "a longer name drops a dragged width: {longer_name:?}"); + let wider = table_frame(&ctx, WINDOW, &[120.0, 200.0, 80.0], Vec::new()); + let widths: Vec = wider.iter().map(|c| c.width().round()).collect(); + assert!(widths.ends_with(&[200.0, 80.0]), "a longer cell widens its column: {widths:?}"); + assert!(wider[2].right() < WINDOW, "the table still fits the window: {wider:?}"); + let narrower = table_frame(&ctx, 400.0, &NATURAL, Vec::new()); + assert!(narrower[2].right() < 400.0, "a narrower window refits the table: {narrower:?}"); + } + + #[test] + fn scrolling_sideways_keeps_dragged_widths() { + let (ctx, opened) = opened_table(); + drag_edge(&ctx, opened[0], 100.0); + table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + let dragged = table_frame(&ctx, WINDOW, &NATURAL, Vec::new()); + assert!((dragged[0].width() - opened[0].width() - 100.0).abs() < 1.0, "the drag took: {dragged:?}"); + for offset in [13.3, 27.77, 41.1, 0.6] { + scrolled_frame(&ctx, WINDOW, &NATURAL, offset, Vec::new()); + let scrolled = scrolled_frame(&ctx, WINDOW, &NATURAL, offset, Vec::new()); + assert!((scrolled[0].width() - dragged[0].width()).abs() < 1.0, "scrolled {offset}: {scrolled:?}"); + } + } } diff --git a/crates/ui-egui/src/dialogs/preferences.rs b/crates/ui-egui/src/dialogs/preferences.rs index 9e3182e416..ea985f3207 100644 --- a/crates/ui-egui/src/dialogs/preferences.rs +++ b/crates/ui-egui/src/dialogs/preferences.rs @@ -659,6 +659,8 @@ fn card_general(ui: &mut egui::Ui, app: &mut PdfCraftApp, t: &Tokens, ctx: &mut ctx.checkbox(ui, &mut app.restore_last_view, tl!("Restore the last view (page, zoom and page display) when a document is opened again"), t); #[cfg(not(target_arch = "wasm32"))] ctx.checkbox(ui, &mut app.reload_changed_files, tl!("Reload documents when their files change on disk"), t); + #[cfg(not(target_arch = "wasm32"))] + ctx.checkbox(ui, &mut app.update_check_at_start, tl!("Check for updates when PdfCraft starts"), t); ctx.checkbox(ui, &mut app.show_secured_notice, tl!("Show a notice when a document is secured"), t); #[cfg(not(target_arch = "wasm32"))] diff --git a/crates/ui-egui/src/hotbar.rs b/crates/ui-egui/src/hotbar.rs index 1fd5dbd75c..269f0a12c2 100644 --- a/crates/ui-egui/src/hotbar.rs +++ b/crates/ui-egui/src/hotbar.rs @@ -239,8 +239,20 @@ fn core_tool_buttons(ui: &mut egui::Ui, app: &mut PdfCraftApp, cell: f32, popup_ QuickTool::Fill(f) => Some(f), _ => None, }; - let shown = current_fill.unwrap_or(crate::fill_sign::FillTool::Text); - let resp = icons::button(ui, if current_fill.is_some() { shown.icon() } else { "pen-line" }, cell, current_fill.is_some(), tl!("Fill & Sign")); + let shown = current_fill.unwrap_or(app.fill_tool); + let tip = match shown { + crate::fill_sign::FillTool::Signature | crate::fill_sign::FillTool::Initials if current_fill.is_none() => { + let initials = shown == crate::fill_sign::FillTool::Initials; + let saved = if initials { app.initials.is_some() } else { app.signature.is_some() }; + crate::i18n::fmt( + tl!(if saved { "Place saved {what}" } else { "Add {what}" }), + &[("what", tl!(if initials { "initials" } else { "signature" }))], + ) + } + _ => tl!(shown.label()).to_string(), + }; + let tip = format!("{}: {tip}", tl!("Fill & Sign")); + let resp = icons::button(ui, shown.icon(), cell, current_fill.is_some(), &tip); let r = resp.rect; let tri = [r.right_bottom() + vec2(-4.0, -4.0), r.right_bottom() + vec2(-9.0, -4.0), r.right_bottom() + vec2(-4.0, -9.0)]; ui.painter().add(egui::Shape::convex_polygon(tri.to_vec(), if current_fill.is_some() { t.accent } else { t.text_faint }, Stroke::NONE)); @@ -287,6 +299,30 @@ fn core_tool_buttons(ui: &mut egui::Ui, app: &mut PdfCraftApp, cell: f32, popup_ }); } +fn controls_divider(ui: &mut egui::Ui, t: &Tokens, vertical: bool) { + let (r, _) = if vertical { + ui.allocate_exact_size(vec2(crate::canvas::bars::QUICK_BAR_WIDTH, 9.0), Sense::hover()) + } else { + ui.allocate_exact_size(vec2(9.0, crate::canvas::bars::QUICK_BAR_WIDTH), Sense::hover()) + }; + if vertical { + ui.painter().hline(r.x_range().shrink(6.0), r.center().y, Stroke::new(1.0, t.divider)); + } else { + ui.painter().vline(r.center().x, r.y_range().shrink(6.0), Stroke::new(1.0, t.divider)); + } +} + +fn selected_fill_text(app: &PdfCraftApp) -> Option<(usize, usize)> { + let (index, id) = app.active_ids()?; + let (page, annotation) = app.views.get(index)?.comments.selected?; + let doc = app.session.get(id).filter(|doc| doc.allows_annotation())?; + doc.info + .annotations + .iter() + .any(|a| a.page == page && a.index == annotation && a.fill_sign && a.subtype == "FreeText" && !a.locked) + .then_some((page, annotation)) +} + /// Extra controls rendered below/beside the buttons when a comment or fill tool is active. fn active_tool_controls(ui: &mut egui::Ui, app: &mut PdfCraftApp, t: &Tokens, vertical: bool) { let current_fill = match app.quick_tool { @@ -294,47 +330,44 @@ fn active_tool_controls(ui: &mut egui::Ui, app: &mut PdfCraftApp, t: &Tokens, ve _ => None, }; if matches!(current_fill, Some(crate::fill_sign::FillTool::Text | crate::fill_sign::FillTool::Date)) { - if vertical { - ui.label(tl!("Font size:")); - } - let size = ui.add( - egui::DragValue::new(&mut app.fill_text_size) - .range(crate::fill_sign::MIN_TEXT_SIZE..=crate::fill_sign::MAX_TEXT_SIZE) - .speed(0.25) - .suffix(" pt"), - ); - if ((size.changed() && !size.dragged()) || size.drag_stopped()) - && let Some((index, id)) = app.active_ids() - && let Some((page, annotation)) = app.views.get(index).and_then(|view| view.comments.selected) - && app.session.get(id).is_some_and(|doc| { - doc.allows_annotation() - && doc - .info - .annotations - .iter() - .any(|a| a.page == page && a.index == annotation && a.fill_sign && a.subtype == "FreeText" && !a.locked) - }) - { - app.apply_edit(Edit::FontSizeAnnotation { page, index: annotation, size: app.fill_text_size }); - } - ui.menu_button(tl!("Colour"), |ui| { - if let Some(color) = comments::swatch_grid(ui, Some(app.fill_text_color)) { - app.fill_text_color = color; - ui.close(); + controls_divider(ui, t, vertical); + if let Some(color) = comments::quick_bar_colour(ui, app.fill_text_color) { + app.fill_text_color = color; + if let Some((page, index)) = selected_fill_text(app) { + app.apply_edit(Edit::StyleAnnotation { page, index, color: Some(color), opacity: None, width: None, endings: None }); } - }); + } + let resp = icons::button(ui, "a-large-small", 32.0, false, tl!("Font size")); + let resized = egui::Popup::menu(&resp) + .align(egui::RectAlign::RIGHT_START) + .close_behavior(egui::PopupCloseBehavior::CloseOnClickOutside) + .show(|ui| { + ui.set_min_width(180.0); + ui.label(egui::RichText::new(tl!("Font size")).font(theme::semibold(12.0))); + let size = ui.add( + egui::Slider::new(&mut app.fill_text_size, crate::fill_sign::MIN_TEXT_SIZE..=crate::fill_sign::MAX_TEXT_SIZE) + .logarithmic(true) + .suffix(" pt"), + ); + // #1274: the size also applies to the Fill & Sign text that is selected, + // once per change (a typed value, or the end of a drag), so a drag adds + // one undo step and not one per frame. + (size.changed() && !size.dragged()) || size.drag_stopped() + }) + .is_some_and(|popup| popup.inner); + let size = app.fill_text_size; + if resized + && let Some((page, index)) = selected_fill_text(app) + && app + .active_ids() + .and_then(|(_, id)| app.session.get(id)?.comment_props(page, index)?.font_size) + .is_none_or(|stored| format!("{stored:.3}") != format!("{size:.3}")) + { + app.apply_edit(Edit::FontSizeAnnotation { page, index, size }); + } } if let QuickTool::Comment(tool) = app.quick_tool { - let (r, _) = if vertical { - ui.allocate_exact_size(vec2(crate::canvas::bars::QUICK_BAR_WIDTH, 9.0), Sense::hover()) - } else { - ui.allocate_exact_size(vec2(9.0, crate::canvas::bars::QUICK_BAR_WIDTH), Sense::hover()) - }; - if vertical { - ui.painter().hline(r.x_range().shrink(6.0), r.center().y, Stroke::new(1.0, t.divider)); - } else { - ui.painter().vline(r.center().x, r.y_range().shrink(6.0), Stroke::new(1.0, t.divider)); - } + controls_divider(ui, t, vertical); if let Some(color) = comments::quick_bar_controls(ui, tool, &mut app.comment_prefs) && let Some((index, id)) = app.active_ids() && let Some((page, annotation)) = app.views.get(index).and_then(|view| view.comments.selected) diff --git a/crates/ui-egui/src/lib.rs b/crates/ui-egui/src/lib.rs index d694f317a1..08e212987a 100644 --- a/crates/ui-egui/src/lib.rs +++ b/crates/ui-egui/src/lib.rs @@ -336,6 +336,7 @@ pub struct PdfCraftApp { pub flatten_fill_sign_on_save: bool, /// Point size for new Fill & Sign text and dates, kept across launches. pub fill_text_size: f64, + pub fill_tool: fill_sign::FillTool, /// Colour for new Fill & Sign text and dates, kept across launches. pub fill_text_color: [f64; 3], pub dialog: Option, @@ -375,6 +376,9 @@ pub struct PdfCraftApp { quit_session: Option, /// Preferences: reload a document when its file changes on disk (#431). pub reload_changed_files: bool, + /// Preferences: ask for a newer release when the desktop app starts (#5737). Off, the app + /// asks only from Help ▸ Check for updates. + pub update_check_at_start: bool, /// Preferences: tell the user, above the pages, that a document is secured (its security /// settings restrict changes). Off only hides that notice (#4035); kept in the settings. pub show_secured_notice: bool, diff --git a/crates/ui-egui/src/platform/updates.rs b/crates/ui-egui/src/platform/updates.rs index 6b7ea291fa..bcf6783cdb 100644 --- a/crates/ui-egui/src/platform/updates.rs +++ b/crates/ui-egui/src/platform/updates.rs @@ -21,7 +21,8 @@ pub(crate) struct Updates { updater: Updater, /// The Updates dialog is showing. pub(crate) open: bool, - /// Ask for a newer release at start; the desktop app turns it on, so tests never reach the network. + /// Ask for a newer release at start, unless Preferences says not to; the desktop app turns + /// it on, so tests never reach the network. #[cfg(not(target_arch = "wasm32"))] startup_check: bool, #[cfg(not(target_arch = "wasm32"))] @@ -73,10 +74,12 @@ impl PdfCraftApp { #[cfg(not(target_arch = "wasm32"))] { let Some(ctx) = self.ctx.clone() else { return }; + // Settled on the first frame: a preference turned on later waits for the next start. + let at_start = std::mem::take(&mut self.updates.startup_check) && self.update_check_at_start; let updates = &mut self.updates; if std::mem::take(&mut updates.check_requested) { updates.updater.check_now(&ctx); - } else if updates.startup_check { + } else if at_start { updates.updater.check_at_start(&ctx); } updates.updater.tick(&ctx); diff --git a/crates/ui-egui/src/tools/comments.rs b/crates/ui-egui/src/tools/comments.rs index 8d118ac3c9..5d945136e7 100644 --- a/crates/ui-egui/src/tools/comments.rs +++ b/crates/ui-egui/src/tools/comments.rs @@ -41,7 +41,7 @@ pub(crate) use input::{anchor, context, context_hit, describe, hit, keys, moving pub use menu::{CanvasAction, fill_picker, recent_custom_colours, remember_custom_colour, status_badge, swatch_grid}; #[cfg(test)] use menu::{RECENT_CUSTOM, same_colour}; -pub(crate) use menu::{attachment_items, context_menu, quick_bar_controls}; +pub(crate) use menu::{attachment_items, context_menu, quick_bar_colour, quick_bar_controls}; pub(crate) use paint::paint_page; #[cfg(test)] use paint::{changing_points, drawn_shape, edited_in_view}; diff --git a/crates/ui-egui/src/tools/comments/menu.rs b/crates/ui-egui/src/tools/comments/menu.rs index b577fbbd29..0e0da5caf3 100644 --- a/crates/ui-egui/src/tools/comments/menu.rs +++ b/crates/ui-egui/src/tools/comments/menu.rs @@ -286,25 +286,32 @@ pub fn fill_picker(ui: &mut egui::Ui, current: Option) -> Option Option { - let style = prefs.style(tool); - let mut selected_color = None; - if icons::button(ui, "pin", 32.0, prefs.pinned, if prefs.pinned { tl!("Keep tool selected: on") } else { tl!("Keep tool selected") }).clicked() { - prefs.pinned = !prefs.pinned; - } +pub(crate) fn quick_bar_colour(ui: &mut egui::Ui, current: Rgb) -> Option { let (r, resp) = ui.allocate_exact_size(vec2(32.0, 32.0), Sense::click()); - ui.painter().circle_filled(r.center(), 9.0, color32(style.color)); + ui.painter().circle_filled(r.center(), 9.0, color32(current)); ui.painter().circle_stroke(r.center(), 9.0, Stroke::new(1.0, Color32::from_black_alpha(50))); resp.widget_info(|| egui::WidgetInfo::labeled(egui::Role::Button, true, tl!("Colour"))); let resp = resp.on_hover_text(tl!("Colour")); + let mut picked = None; egui::Popup::menu(&resp).align(egui::RectAlign::RIGHT_START).close_behavior(egui::PopupCloseBehavior::CloseOnClickOutside).show(|ui| { - if let Some(c) = swatch_grid(ui, Some(style.color)) { - prefs.set_color(tool, c); - selected_color = Some(c); + if let Some(c) = swatch_grid(ui, Some(current)) { + picked = Some(c); ui.close(); } }); + picked +} + +/// The comment tools' extra quick-bar controls: pin, colour, fill, opacity and thickness. +pub(crate) fn quick_bar_controls(ui: &mut egui::Ui, tool: CommentTool, prefs: &mut CommentPrefs) -> Option { + let style = prefs.style(tool); + if icons::button(ui, "pin", 32.0, prefs.pinned, if prefs.pinned { tl!("Keep tool selected: on") } else { tl!("Keep tool selected") }).clicked() { + prefs.pinned = !prefs.pinned; + } + let selected_color = quick_bar_colour(ui, style.color); + if let Some(c) = selected_color { + prefs.set_color(tool, c); + } if tool.has_fill() { let resp = fill_button(ui, style.fill); egui::Popup::menu(&resp).align(egui::RectAlign::RIGHT_START).close_behavior(egui::PopupCloseBehavior::CloseOnClickOutside).show(|ui| { diff --git a/crates/ui-egui/src/tools/fill_sign.rs b/crates/ui-egui/src/tools/fill_sign.rs index 55f0624c08..224ba8b49d 100644 --- a/crates/ui-egui/src/tools/fill_sign.rs +++ b/crates/ui-egui/src/tools/fill_sign.rs @@ -374,20 +374,22 @@ pub(crate) fn signature_entries(ui: &mut egui::Ui, app: &mut crate::PdfCraftApp, } continue; }; - ui.horizontal(|ui| { + let on = app.quick_tool == crate::QuickTool::Fill(if i == 0 { FillTool::Signature } else { FillTool::Initials }); + let wash = ui.painter().add(egui::Shape::Noop); + let row = ui.horizontal(|ui| { let (rect, response) = ui.allocate_exact_size(vec2((ui.available_width() - 68.0).max(140.0), 52.0), Sense::click()); let use_label = label("Use {what}"); - response.widget_info(|| egui::WidgetInfo::labeled(egui::Role::Button, true, use_label.clone())); + response.widget_info(|| egui::WidgetInfo::selected(egui::Role::Button, true, on, use_label.clone())); let painter = ui.painter_at(rect); painter.rect_filled(rect, CornerRadius::same(4), Color32::WHITE); painter.rect_stroke( rect, CornerRadius::same(4), - Stroke::new(1.0, if response.hovered() { t.accent } else { t.border }), + Stroke::new(1.0, if on || response.hovered() { t.accent } else { t.border }), egui::StrokeKind::Inside, ); paint_saved(ui, rect.shrink(8.0), saved, &mut app.saved_signature_previews[i], &format!("saved-{what}")); - if response.on_hover_text(label("Place saved {what}")).clicked() { + if crate::theme::hand(response).on_hover_text(label("Place saved {what}")).clicked() { command = Some(use_id); } if crate::icons::button(ui, "pencil", 26.0, false, &label("Change {what}")).clicked() { @@ -397,6 +399,9 @@ pub(crate) fn signature_entries(ui: &mut egui::Ui, app: &mut crate::PdfCraftApp, command = Some(remove_id); } }); + if on { + ui.painter().set(wash, egui::Shape::rect_filled(row.response.rect.expand(3.0), CornerRadius::same(6), t.accent_soft)); + } ui.add_space(4.0); } command diff --git a/crates/ui-egui/src/widgets/icon_data.rs b/crates/ui-egui/src/widgets/icon_data.rs index 72829fe4b8..d58ddcbab6 100644 --- a/crates/ui-egui/src/widgets/icon_data.rs +++ b/crates/ui-egui/src/widgets/icon_data.rs @@ -187,6 +187,7 @@ pub static ICONS: &[(&str, &[u8])] = &[ ("wrench", include_bytes!("../../../../assets/icons/wrench.svg")), ("move-horizontal", include_bytes!("../../../../assets/icons/move-horizontal.svg")), ("pilcrow", include_bytes!("../../../../assets/icons/pilcrow.svg")), + ("a-large-small", include_bytes!("../../../../assets/icons/a-large-small.svg")), ("subscript", include_bytes!("../../../../assets/icons/subscript.svg")), ("superscript", include_bytes!("../../../../assets/icons/superscript.svg")), ("x", include_bytes!("../../../../assets/icons/x.svg")), diff --git a/crates/ui-egui/src/widgets/icons.rs b/crates/ui-egui/src/widgets/icons.rs index 5ad5425e8d..13f50a0d6c 100644 --- a/crates/ui-egui/src/widgets/icons.rs +++ b/crates/ui-egui/src/widgets/icons.rs @@ -99,7 +99,7 @@ mod tests { #[test] fn shared_catalog_preserves_existing_icons_and_fallback_uri() { - assert_eq!(ICONS.len(), 190); + assert_eq!(ICONS.len(), 191); assert!(exists("minimize-2")); for (name, original) in ICONS { let expected = diff --git a/crates/ui-egui/tests/it/comments.rs b/crates/ui-egui/tests/it/comments.rs index e8dfb1b8b0..129bcbf3e4 100644 --- a/crates/ui-egui/tests/it/comments.rs +++ b/crates/ui-egui/tests/it/comments.rs @@ -3286,7 +3286,7 @@ fn hotbar_docked_left_does_not_override_quick_tools() { }); h.run_steps(5); // Controls in the docked bar share one layout, so none may overlap (#5890). - let labels = ["Drag hotbar", "Select (V)", "Hand (H)", "Add a sticky note", "Highlight", "Draw", "Fill & Sign", "Hotbar settings"]; + let labels = ["Drag hotbar", "Select (V)", "Hand (H)", "Add a sticky note", "Highlight", "Draw", "Fill & Sign: Add signature", "Hotbar settings"]; let rects: Vec<_> = labels.iter().map(|&label| h.get_by_label(label).rect()).collect(); for (i, a) in rects.iter().enumerate() { for (j, b) in rects.iter().enumerate().skip(i + 1) { diff --git a/crates/ui-egui/tests/it/fill_sign.rs b/crates/ui-egui/tests/it/fill_sign.rs index 858e082137..9f565d834a 100644 --- a/crates/ui-egui/tests/it/fill_sign.rs +++ b/crates/ui-egui/tests/it/fill_sign.rs @@ -2,7 +2,7 @@ use egui::{Pos2, pos2}; use egui_kittest::Harness; -use egui_kittest::kittest::Queryable; +use egui_kittest::kittest::{NodeT, Queryable}; use pdfcraft_ui_egui::fill_sign::{FillTool, SavedSig}; use pdfcraft_ui_egui::{Dialog, PdfCraftApp, QuickTool}; @@ -164,7 +164,7 @@ fn fill_text_size_controls_new_text_and_survives_settings_restore() { let mut h = harness(); h.state_mut().execute("sign.fill.text"); h.run_steps(2); - h.get_by_label("Font size:"); + h.get_by_label("Font size"); h.state_mut().fill_text_size = 24.0; click(&mut h, 50.0, 250.0); assert_eq!(h.state().views[0].fill_text.as_ref().map(|t| t.size), Some(24.0)); @@ -201,6 +201,8 @@ fn fill_text_size_resizes_the_selected_text() { }; let before = height(&h); let set_size = |h: &mut Harness<'static, PdfCraftApp>, value: &str| { + h.get_by_role_and_label(egui::accesskit::Role::Button, "Font size").click(); + h.run_steps(2); h.query_all_by_role(egui::accesskit::Role::SpinButton) .find(|n| n.value().is_some_and(|v| v.ends_with(" pt"))) .expect("the Font size control") @@ -211,6 +213,8 @@ fn fill_text_size_resizes_the_selected_text() { h.run_steps(1); h.key_press(egui::Key::Enter); h.run_steps(4); + h.get_by_role_and_label(egui::accesskit::Role::Button, "Font size").click(); + h.run_steps(2); }; // Nothing selected: only the size for new text changes. h.state_mut().views[0].comments.selected = None; @@ -423,9 +427,9 @@ fn changing_saved_signatures_and_initials_preserves_placed_marks() { let original = format!("{:?}", h.state().session.get(h.state().views[0].id).unwrap().info.annotations); // The quick toolbar exposes replacement without changing the normal placement action. - h.get_by_label("Fill & Sign").click(); + h.get_by_label("Fill & Sign: Place saved signature").click(); h.run_steps(2); - h.get_by_label("Fill & Sign").click(); + h.get_by_label("Fill & Sign: Sign").click(); h.run_steps(2); h.get_by_label("Change signature").click(); h.run_steps(2); @@ -445,9 +449,9 @@ fn changing_saved_signatures_and_initials_preserves_placed_marks() { click(&mut h, 40.0, 200.0); assert_eq!(items(&h).len(), 2); - h.get_by_label("Fill & Sign").click(); + h.get_by_label("Fill & Sign: Place saved signature").click(); h.run_steps(2); - h.get_by_label("Fill & Sign").click(); + h.get_by_label("Fill & Sign: Sign").click(); h.run_steps(2); h.get_by_label("Change initials").click(); h.run_steps(2); @@ -1813,6 +1817,269 @@ fn mode_tabs_keep_the_hand_over_a_type_box_scrolled_under_them() { assert_eq!(h.state().mode, pdfcraft_ui_egui::Mode::Convert, "the tab takes the click, not the text box"); } +fn place_selected_text(h: &mut Harness<'static, PdfCraftApp>, text: &str) { + h.state_mut().execute("sign.fill.text"); + click(h, 50.0, 250.0); + h.event(egui::Event::Text(text.into())); + h.run_steps(1); + commit_typed_text(h); + let selected = { + let s = h.state(); + s.session + .get(s.views[0].id) + .unwrap() + .info + .annotations + .iter() + .find(|a| a.contents.as_deref() == Some(text)) + .map(|a| (a.page, a.index)) + .unwrap() + }; + h.state_mut().views[0].comments.selected = Some(selected); + h.run_steps(2); +} + +fn placed_height(h: &Harness<'static, PdfCraftApp>, text: &str) -> f32 { + let s = h.state(); + let a = s.session.get(s.views[0].id).unwrap().info.annotations.iter().find(|a| a.contents.as_deref() == Some(text)).unwrap(); + a.rect[3] - a.rect[1] +} + +fn placed_color(h: &Harness<'static, PdfCraftApp>, text: &str) -> Option<[f32; 3]> { + let s = h.state(); + s.session.get(s.views[0].id).unwrap().info.annotations.iter().find(|a| a.contents.as_deref() == Some(text)).unwrap().color +} + +#[test] +fn the_fill_text_size_flyout_opens_from_an_icon_and_leaves_a_stored_size_alone() { + let mut h = harness(); + h.state_mut().fill_text_size = 9.1; + place_selected_text(&mut h, "Keep me"); + let height = placed_height(&h, "Keep me"); + assert_eq!(h.query_all_by_role(egui::accesskit::Role::SpinButton).filter(|n| n.value().is_some_and(|v| v.ends_with(" pt"))).count(), 0); + h.get_by_label("Font size").click(); + h.run_steps(4); + assert!( + h.query_all_by_role(egui::accesskit::Role::SpinButton).any(|n| n.value().is_some_and(|v| v.ends_with(" pt"))), + "the flyout holds the size" + ); + assert_eq!(h.state().fill_text_size, 9.1, "opening the flyout does not round the size"); + assert_eq!(placed_height(&h, "Keep me"), height, "opening the flyout does not resize the selected text"); +} + +#[test] +fn dragging_the_fill_text_size_slider_is_one_undo_step() { + let mut h = harness(); + place_selected_text(&mut h, "Drag me"); + let height = placed_height(&h, "Drag me"); + h.get_by_label("Font size").click(); + h.run_steps(2); + let slider = h.get_by_role(egui::accesskit::Role::Slider).rect(); + let y = (slider.min.y + slider.max.y) / 2.0; + let from = pos2(slider.min.x + 10.0, y); + let to = pos2(slider.max.x - 10.0, y); + h.hover_at(from); + h.run_steps(1); + h.drag_at(from); + h.run_steps(1); + for k in 1..=4 { + h.hover_at(from + (to - from) * (k as f32 / 4.0)); + h.run_steps(1); + } + assert_eq!(placed_height(&h, "Drag me"), height, "nothing is applied before the release"); + h.drop_at(to); + h.run_steps(3); + assert!(placed_height(&h, "Drag me") > height, "the release resizes the selected text"); + h.state_mut().execute("edit.undo"); + h.run_steps(3); + assert_eq!(placed_height(&h, "Drag me"), height, "one undo takes the whole drag back"); +} + +#[test] +fn pressing_the_fill_text_size_slider_without_moving_it_adds_no_undo_step() { + let mut h = harness(); + place_selected_text(&mut h, "Hold me"); + let height = placed_height(&h, "Hold me"); + h.get_by_label("Font size").click(); + h.run_steps(2); + let slider = h.get_by_role(egui::accesskit::Role::Slider).rect(); + let at = pos2(slider.min.x + (slider.max.x - slider.min.x) * 0.75, (slider.min.y + slider.max.y) / 2.0); + let press = |h: &mut Harness<'static, PdfCraftApp>| { + h.hover_at(at); + h.run_steps(1); + h.drag_at(at); + h.run_steps(1); + h.drop_at(at); + h.run_steps(3); + }; + press(&mut h); + let resized = placed_height(&h, "Hold me"); + assert!(resized > height, "the first press moves the size: {height} -> {resized}"); + press(&mut h); + assert_eq!(placed_height(&h, "Hold me"), resized); + h.state_mut().execute("edit.undo"); + h.run_steps(3); + assert_eq!(placed_height(&h, "Hold me"), height, "one undo takes back the only change; the second press added no step"); +} + +#[test] +fn typing_a_fill_text_size_that_rounds_to_the_stored_one_adds_no_undo_step() { + let mut h = harness(); + place_selected_text(&mut h, "Round me"); + let height = placed_height(&h, "Round me"); + h.get_by_label("Font size").click(); + h.run_steps(2); + let type_size = |h: &mut Harness<'static, PdfCraftApp>, value: &str| { + h.query_all_by_role(egui::accesskit::Role::SpinButton) + .find(|n| n.value().is_some_and(|v| v.ends_with(" pt"))) + .expect("the Font size control") + .click(); + h.run_steps(2); + h.key_press_modifiers(egui::Modifiers::COMMAND, egui::Key::A); + h.event(egui::Event::Text(value.into())); + h.run_steps(1); + h.key_press(egui::Key::Enter); + h.run_steps(4); + }; + type_size(&mut h, "36.1234"); + let resized = placed_height(&h, "Round me"); + assert!(resized > height, "the first size applies: {height} -> {resized}"); + type_size(&mut h, "36.1231"); + assert_eq!(h.state().fill_text_size, 36.1231); + h.state_mut().execute("edit.undo"); + h.run_steps(3); + assert_eq!(placed_height(&h, "Round me"), height, "one undo takes back the only change; the rounded retype added no step"); +} + +#[test] +fn the_fill_colour_dot_recolours_the_selected_text() { + let mut h = harness(); + place_selected_text(&mut h, "Paint me"); + let before = placed_color(&h, "Paint me"); + h.get_by_label("Colour").click(); + h.run_steps(2); + h.get_by_label("Blue").click(); + h.run_steps(3); + let color = h.state().fill_text_color.map(|v| v as f32); + assert!(color[2] > color[0], "blue was selected: {color:?}"); + assert_ne!(before, Some(color)); + assert_eq!(placed_color(&h, "Paint me"), Some(color)); + h.state_mut().execute("edit.undo"); + h.run_steps(3); + assert_eq!(placed_color(&h, "Paint me"), before, "one undo takes the colour back"); +} + +#[test] +fn the_quick_bar_fill_button_shows_the_tool_it_picks() { + let mut h = harness(); + h.get_by_label("Fill & Sign: Add signature").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Signature)); + assert_eq!(h.state().dialog, Some(Dialog::Signature)); + + h.state_mut().dialog = None; + h.state_mut().signature = Some(SavedSig::Typed("Ada Lovelace".into())); + h.state_mut().quick_tool = QuickTool::Select; + h.run_steps(2); + h.get_by_label("Fill & Sign: Place saved signature").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Signature)); + assert_eq!(h.state().dialog, None); + h.get_by_label("Fill & Sign: Sign").click(); + h.run_steps(2); + assert_eq!(h.get_by_label("Use signature").accesskit_node().toggled(), Some(egui::accesskit::Toggled::True), "the active tool's card is marked"); + h.key_press(egui::Key::Escape); + h.run_steps(2); + + h.state_mut().execute("sign.fill.check"); + h.state_mut().quick_tool = QuickTool::Select; + h.run_steps(2); + h.get_by_label("Fill & Sign: Checkmark").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Check), "the button keeps the last Fill & Sign tool"); +} + +#[test] +fn a_restored_fill_tool_shows_without_replacing_the_last_selection() { + let mut h = harness(); + h.state_mut().execute("sign.fill.date"); + h.run_steps(2); + h.state_mut().apply_full_screen(true); + h.run_steps(2); + h.state_mut().execute("sign.fill.check"); + h.run_steps(2); + h.state_mut().apply_full_screen(false); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Date)); + h.get_by_label("Fill & Sign: Date"); + + h.state_mut().quick_tool = QuickTool::Select; + h.run_steps(2); + h.get_by_label("Fill & Sign: Checkmark").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Check)); +} + +#[test] +fn a_fill_tool_selected_while_the_quick_bar_is_hidden_is_remembered() { + let mut h = harness(); + let mode = h.state().mode; + h.state_mut().set_option("mode", "read").unwrap(); + h.run_steps(2); + h.state_mut().set_option("quick", "fill-date").unwrap(); + h.run_steps(2); + h.state_mut().set_option("quick", "select").unwrap(); + h.state_mut().mode = mode; + h.run_steps(2); + h.get_by_label("Fill & Sign: Date"); + + h.state_mut().initials = Some(SavedSig::Typed("AL".into())); + h.state_mut().set_option("mode", "read").unwrap(); + h.run_steps(2); + h.state_mut().execute("sign.fill.initials.change"); + h.run_steps(2); + h.get_by_label("Apply").click(); + h.run_steps(3); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Initials)); + h.state_mut().quick_tool = QuickTool::Select; + h.state_mut().mode = mode; + h.run_steps(2); + h.get_by_label("Fill & Sign: Place saved initials"); +} + +#[test] +fn the_quick_bar_fill_button_keeps_date_after_deselecting_and_across_documents() { + let mut h = harness(); + h.get_by_label("Fill & Sign: Add signature").click_secondary(); + h.run_steps(2); + h.get_by_label("Date").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Date)); + h.get_by_label("Select (V)").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Select); + + h.state_mut().open_bytes("second.pdf", None, FIXTURE.to_vec()).unwrap(); + h.state_mut().active = Some(1); + h.run_steps(2); + h.get_by_label("Fill & Sign: Date").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Date), "a click picks the shown tool, not the default"); + assert_eq!(h.state().dialog, None, "the default tool's dialog stays closed"); + h.get_by_label("Select (V)").click(); + h.run_steps(2); + + h.state_mut().close_tab(1); + h.state_mut().close_tab(0); + h.run_steps(2); + h.state_mut().open_bytes("third.pdf", None, FIXTURE.to_vec()).unwrap(); + h.state_mut().active = Some(0); + h.run_steps(2); + h.get_by_label("Fill & Sign: Date").click(); + h.run_steps(2); + assert_eq!(h.state().quick_tool, QuickTool::Fill(FillTool::Date), "the tool outlives closing every document"); +} + fn commit_typed_text(h: &mut Harness<'static, PdfCraftApp>) { let p = pos2(350.0, 850.0); // Blank canvas outside the page and the type box. h.hover_at(p); diff --git a/crates/ui-egui/tests/it/updates.rs b/crates/ui-egui/tests/it/updates.rs index 3ea55e434e..e6cc09d7c5 100644 --- a/crates/ui-egui/tests/it/updates.rs +++ b/crates/ui-egui/tests/it/updates.rs @@ -100,6 +100,93 @@ fn nothing_is_asked_until_the_user_checks() { h.get_by_label_contains("PdfCraft 99.0.0 is available"); } +/// The desktop app's start, with `settings` restored, and the count of what the source was asked. +fn started_with(settings: &'static str) -> (Harness<'static, PdfCraftApp>, Arc) { + let calls = Arc::new(std::sync::atomic::AtomicUsize::new(0)); + let counted = calls.clone(); + let h = Harness::builder().with_size(egui::vec2(1200.0, 800.0)).build_eframe(move |_cc| { + let mut app = PdfCraftApp::new(); + app.set_option("language", "en").unwrap(); + app.set_startup_update_check(true); + app.restore(settings); + let counted = counted.clone(); + app.set_update_source(Arc::new(move || { + counted.fetch_add(1, std::sync::atomic::Ordering::SeqCst); + Ok(Some(release("v99.0.0"))) + })); + app + }); + (h, calls) +} + +/// Run frames for `time`, or until the source has been asked. The check at start shows nothing +/// while it runs, so `settle` can't wait for it. +fn run_until_asked(h: &mut Harness<'static, PdfCraftApp>, calls: &std::sync::atomic::AtomicUsize, time: std::time::Duration) { + let deadline = std::time::Instant::now() + time; + while calls.load(std::sync::atomic::Ordering::SeqCst) == 0 && std::time::Instant::now() < deadline { + h.run_steps(2); + std::thread::sleep(std::time::Duration::from_millis(5)); + } +} + +#[test] +fn the_preference_keeps_the_check_at_start_from_reaching_the_network() { + use std::sync::atomic::Ordering::SeqCst; + assert!(PdfCraftApp::new().update_check_at_start, "on by default"); + + // The same start with the preference left on does ask, so a count of none below means + // the preference held the check back and not that it hadn't run yet. + let (mut h, calls) = started_with("{}"); + run_until_asked(&mut h, &calls, std::time::Duration::from_secs(30)); + assert_eq!(calls.load(SeqCst), 1, "on, the app asks at start"); + + let (mut h, calls) = started_with(r#"{"update_check_at_start": false}"#); + run_until_asked(&mut h, &calls, std::time::Duration::from_millis(500)); + assert_eq!(calls.load(SeqCst), 0, "nothing is asked at start"); + + let mut restarted = PdfCraftApp::new(); + restarted.restore(&h.state().persist()); + assert!(!restarted.update_check_at_start, "the choice is kept"); + + // Help ▸ Check for updates still asks. + h.state_mut().execute("help.check_updates"); + settle(&mut h); + assert_eq!(calls.load(SeqCst), 1); + h.get_by_label_contains("PdfCraft 99.0.0 is available"); +} + +/// The setting is about starting: turning it on asks nothing until the next start. +#[test] +fn turning_the_check_at_start_on_waits_for_the_next_start() { + use std::sync::atomic::Ordering::SeqCst; + let (mut h, calls) = started_with(r#"{"update_check_at_start": false}"#); + h.run_steps(3); + h.state_mut().set_option("update_check_at_start", "true").unwrap(); + run_until_asked(&mut h, &calls, std::time::Duration::from_millis(500)); + assert_eq!(calls.load(SeqCst), 0, "nothing is asked this session"); +} + +#[test] +fn preferences_turn_the_check_at_start_off() { + let (mut h, _calls) = started_with(r#"{"update_check_at_start": true}"#); + h.state_mut().execute("app.preferences"); + h.run_steps(3); + h.get_by_label("Check for updates when PdfCraft starts").click(); + h.run_steps(3); + assert!(!h.state().update_check_at_start); + // The control channel's option, as an agent sets it. + h.state_mut().set_option("update_check_at_start", "true").unwrap(); + assert!(h.state().update_check_at_start); + assert!(h.state_mut().set_option("update_check_at_start", "maybe").is_err()); +} + +#[test] +fn settings_from_the_old_opt_in_leave_the_check_at_start_on() { + let mut app = PdfCraftApp::new(); + app.restore(r#"{"check_updates_at_start": false}"#); + assert!(app.update_check_at_start, "a key of the removed option decides nothing"); +} + #[test] fn the_check_at_start_offers_the_prompt_and_later_dismisses_it() { let mut h = Harness::builder().with_size(egui::vec2(1200.0, 800.0)).build_eframe(|_cc| { diff --git a/docs/roadmap.md b/docs/roadmap.md index d17893c1aa..25c0da7e00 100644 --- a/docs/roadmap.md +++ b/docs/roadmap.md @@ -22,7 +22,7 @@ > **Last reviewed:** 2026-10-26 · **Last updated:** 2026-10-26 · **Change:** patch (M2, PR #4172: document installed-font helpers and regression contracts; behavior and estimates unchanged) > **Last reviewed:** 2026-10-26 · **Last updated:** 2026-10-26 · **Change:** patch (M2, PR #4172: print-preview font policy matches output, with isolated caches and GBK switching regressions; estimates unchanged) > **Last reviewed:** 2026-10-26 · **Last updated:** 2026-10-26 · **Change:** patch (M2, PR #4172: reconcile current main and retain Japanese serif/weight selection alongside display-only CJK fallback; estimates unchanged) -> **Last reviewed:** 2026-10-29 | **Last updated:** 2026-10-29 | **Change:** minor (M7 #2517: a narrowed then widened paragraph keeps the typed spaces the wrap consumed, after save and reopen; estimates unchanged) +> **Last reviewed:** 2026-10-29 | **Last updated:** 2026-10-29 | **Change:** minor (M2: scans keep their strokes at screen size and are smoothed when zoomed; M7 #2517: a narrowed then widened paragraph keeps the typed spaces the wrap consumed, after save and reopen; estimates unchanged) > **Last reviewed:** 2026-10-27 | **Last updated:** 2026-10-27 | **Change:** minor (M10: OCR model downloads use verified ONNX pairs with legacy discovery fallback; estimates unchanged) > **Last reviewed:** 2026-10-25 | **Last updated:** 2026-10-25 | **Change:** minor (M5: one reply in the Comments panel is edited, copied or deleted from its own "…" menu; estimates unchanged) > **Last reviewed:** 2026-10-26 · **Last updated:** 2026-10-26 · **Change:** minor (M11: lossless JBIG2 stores each distinct mark once, as a symbol; estimates unchanged) @@ -63,6 +63,10 @@ > **Last reviewed:** 2026-10-17 · **Last updated:** 2026-10-17 · **Change:** minor (M8, gap 6 #2662: the redaction audit counts a cover over one word of a long line; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) > **Last reviewed:** 2026-10-13 · **Last updated:** 2026-10-13 · **Change:** minor (M10: dotted-leader contents entries export as complete paragraphs; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) > **Last reviewed:** 2026-10-22 · **Last updated:** 2026-10-22 · **Change:** minor (M6: calculation scripts read their notation only from a comment, and a regular expression after a condition as one; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) + +> **Last reviewed:** 2026-10-20 · **Last updated:** 2026-10-20 · **Change:** patch (M1, #1143: repaired aliases survive preflight, snapshots and normalized saves; retained materialized revisions report generation disagreements; estimates unchanged) + +> **Last reviewed:** 2026-10-18 · **Last updated:** 2026-10-18 · **Change:** minor (M1, #1143: explicit live xref/header generation repair policy and bounded diagnostics; estimates unchanged) > **Last reviewed:** 2026-10-18 · **Last updated:** 2026-10-18 · **Change:** minor (M5, PR #1971: bounded nested FDF export excludes source page/page-tree dictionaries; refreshed against upstream 4db6ac4a; estimates unchanged) > **Last reviewed:** 2026-10-18 · **Last updated:** 2026-10-18 · **Change:** patch (M2 #3086: dotted TOC leaders no longer hide word spaces; estimates unchanged) > @@ -78,6 +82,7 @@ > **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (M6: AcroForm XML text; estimates unchanged) > **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (M6: preserve XML/XFDF attribute whitespace; refreshed upstream; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) > **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (M1: raw-deflate fallback preserves decoding-limit errors; refreshed upstream; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) +> **Last reviewed:** 2026-10-15 · **Last updated:** 2026-10-15 · **Change:** minor (M1: refresh object-generation reference identity against current upstream; estimates unchanged) > **Last reviewed:** 2026-10-14 · **Last updated:** 2026-10-15 · **Change:** minor (M2: objects in one large object stream are read in linear time; estimates unchanged) · **Target:** Adobe Acrobat Pro (Acrobat DC, continuous track 26.002.21931, macOS) > **Last reviewed:** 2026-10-16; **Last updated:** 2026-10-16; **Change:** minor (M5 #2669, item 5: reply groups and page root counts per frame; estimates unchanged) > @@ -204,7 +209,7 @@ total matches the parity estimate. | M | Milestone | Plan size (h) | Done | Remaining (h, 2026-10-10) | Notes | |---|---|---|---|---|---| | M0 | Skeleton: workspace, xtask gates, CI | 15–30 | 98% | 1–2 | GitHub workflow, `deny.toml` (licence audit of every dependency), parity checklist (826 features, `xtask parity`) done. Missing: remaining crate stubs, testkit/oracle crates | -| M1 | COS: filters, crypt, parser, xref, writer | 120–200 | 80% | 20–40 | Done:
- filters and crypt: every standard-security revision R2–R6 (RC4, AES-128/256), SASLprep, permissions, creating encryption;
- cos: parse and repair (including invalid xref object-number ranges), decoded-stream limits including unfiltered data, decrypt on load, re-encrypt on save, incremental and full writing.
Corpus: open/edit/save passes on 958 files, and all 7 password-protected files open.
Full saves now pack objects into compressed object streams. An unchanged incremental save of a classic table writes a zero-length cross-reference subsection instead of a section with no subsections (#1339). A full rewrite keeps unknown trailer entries, including objects reachable only from them (#1338). Fuzzing runs nightly (`xtask fuzz`). Raw-deflate fallback preserves decoding-limit errors, including through tolerant COS stream decoding. Done since: #2658: macOS atomic saves retain destination extended attributes, including Finder tags, and reject partial resource-fork reads after size changes. Metadata failures retain the original PDF and unsaved edit. Missing: ≥ 250 tests, own image codecs | +| M1 | COS: filters, crypt, parser, xref, writer | 120–200 | 80% | 20–40 | Done:
- filters and crypt: every standard-security revision R2–R6 (RC4, AES-128/256), SASLprep, permissions, creating encryption;
- cos: parse and repair (including invalid xref object-number ranges), decoded-stream limits including unfiltered data, decrypt on load, re-encrypt on save, incremental and full writing.
Corpus: open/edit/save passes on 958 files, and all 7 password-protected files open.
Full saves now pack objects into compressed object streams. An unchanged incremental save of a classic table writes a zero-length cross-reference subsection instead of a section with no subsections (#1339). A full rewrite keeps unknown trailer entries, including objects reachable only from them (#1338). Fuzzing runs nightly (`xtask fuzz`). Raw-deflate fallback preserves decoding-limit errors, including through tolerant COS stream decoding. Refreshed against upstream `99941daa` (2026-10-15). COS reference lookups and full saves distinguish complete object references, so stale generations cannot expose replacement content. Snapshot restoration preserves edited nonzero-generation objects through incremental save/reopen. Synthetic regressions cover caches, overlays, snapshots, stream lengths and classic/packed saves. For damaged live rows, either the xref generation or a same-number header at its indexed offset may identify the object; neither-match and free slots remain null. Header probes read at most 128 bytes per candidate offset, and opening records at most 256 object disagreements plus a summary without expanding compact xref rows. Synthetic dictionary-edit and independent save/reopen controls cover this repair policy. Repair-aware preflight edits and restored snapshots retain both accepted aliases, with full rewrite for repaired edits; classic/packed output coalesces them to one object identity. Materialized diagnostics inspect retained primary/hybrid rows before folding, with the same cap. Done since: #2658: macOS atomic saves retain destination extended attributes, including Finder tags, and reject partial resource-fork reads after size changes. Metadata failures retain the original PDF and unsaved edit. Missing: ≥ 250 tests, own image codecs | | M2 | Model, render, text | 200–350 | 15% | 130–230 | hayro bootstrap renderer (vendored patches); inspector predictor scratch is bounded by available input, and JPX sample rescaling handles 1–32 bits safely. Text extraction reaches word-F1 0.98 against pdftotext. Short same-line redraws split before placement rather than interleaving. Missing: model crate, fonts, DisplayList, renderer independent of hayro. Inspector page-label traversal and output are resource-bounded, with physical-number fallback warnings. Packed TIFF-predictor RGB rows retain later rows in decoding and rendering. Page marks, added content and print output retain small page-unit transform coefficients when serialized. Forms without Resources use page Properties for ActualText and layer visibility; own dictionaries stop fallback. Oversized missing inline-image samples, including header and terminator overhead, no longer pin the content window (#3107 / PR #3438); normal large streamed images retain deterministic sample boundaries. Radial luminosity fades have synthetic coverage for the existing #2132 mask guard, including per-pixel references, antialiased clips, previews, scales and tiles; no additional production fix was needed. Dotted TOC leaders no longer inflate letter tracking or hide word spaces (#3086); phrase search and selection are covered. Short tracked ellipses retain their gap samples and remain searchable. Missing spacing glyphs in non-embedded Japanese CID substitutes render blank without changing PDF advances or Unicode (#3601). The desktop viewer can use installed CJK fonts, including Windows GBK Song/Hei/Kai/FangSong and subset-prefixed Founder names; Combine thumbnails and accessibility figure previews use the same display-only fallback, without changing export or print font policy (#234, continuation of #1567, review #4172). Print previews exclude installed faces and isolate their font cache; GBK regressions cover view/print switching for inline and pooled whole-page/tile rendering. Installed-font helpers and their regression contracts have function documentation. GBK regression rasters match explicit-Unicode references with independently selected installed faces; wrong-glyph and wrong-family negative controls fail. Integration with main fa5f9f8c retains viewer fallback across the split engine session modules, renderer tests and Combine thumbnail workers. | | M2 | Model, render, text | 200–350 | 15% | 130–230 | hayro bootstrap renderer (vendored patches); inspector predictor scratch is bounded by available input, and JPX sample rescaling handles 1–32 bits safely. Text extraction reaches word-F1 0.98 against pdftotext. Short same-line redraws split before placement rather than interleaving. Missing: model crate, fonts, DisplayList, renderer independent of hayro. Inspector page-label traversal and output are resource-bounded, with physical-number fallback warnings. Packed TIFF-predictor RGB rows retain later rows in decoding and rendering. The renderer keeps its one-point crop-size floor in physical points after UserUnit scaling, with regression coverage for crop origins, rotations and f32 reciprocal rounding. Existing unit policy and nearly-zero fallback are retained. | | M2 | Model, render, text | 200–350 | 15% | 130–230 | hayro bootstrap renderer (vendored patches); inspector predictor scratch is bounded by available input, and JPX sample rescaling handles 1–32 bits safely. Text extraction reaches word-F1 0.98 against pdftotext. Short same-line redraws split before placement rather than interleaving. Missing: model crate, fonts, DisplayList, renderer independent of hayro. Inspector page-label traversal and output are resource-bounded, with physical-number fallback warnings | @@ -308,6 +313,7 @@ Estimates unchanged. | 2026-11-03 | patch | M5 #5850 review: indirect typewriter intent and composer edits preserve outer whitespace; movement regression commits by outside click; estimates unchanged. | | 2026-11-03 | patch | M5 #5850: Enter adds Fill & Sign line breaks; outside clicks commit, Escape cancels, and saved multiline appearances retain their text and height; estimates unchanged. | | 2026-11-02 | patch | M10 #5743: macOS print jobs carry the job name and owner some vendors' filters need; not yet checked on a physical printer | +| 2026-11-02 | patch | M14 #5737, #5460: Preferences turns off the update check at start, on by default | | 2026-11-01 | patch | M7 #5657: header/footer, watermark and background operations commit all selected pages together; later-page errors preserve marks, content, shared/inherited resources and pending edits; estimates unchanged. | | 2026-10-31 | patch | #1577: reconcile optional page shadows with current viewer/organizer and settings modules; CI validation pending. | | 2026-10-31 | patch | M3 #1354: reconcile the launch override with split app/file modules and the structured open result; preserve headless help/version exits. CI validation pending. | @@ -322,6 +328,7 @@ Estimates unchanged. | 2026-10-23 | minor | M11: black-and-white images (1-bit, and image masks) are stored losslessly as CCITT Group 4 when that is smaller: PDF Optimizer ▸ Images ▸ Monochrome Images (on by default, so Reduce File Size too), `doc_optimize` `mono`; `pdfcraft_filters::encode_g4`. JBIG2 and monochrome downsampling not yet. Estimates unchanged. | | 2026-10-23 | minor | M6: calculation checks: a formula reading a field the form doesn't have, and a script building the names it reads, are flagged by `form_calculation_order` and on the Calculation order page (a red mark, a card naming the broken formulas, a note for built names); `form_calculation_order` gives each formula and `form_fields` a custom Calculate script. Estimates unchanged. | | 2026-10-29 | minor | M7 #2517: widening a paragraph rejoins a wrap with the spaces that wrap consumed, including a typed run and none for a Chinese or Japanese word split, after an incremental save and a full rewrite. Estimates unchanged. | +| 2026-10-29 | minor | M2: non-interpolated images drawn below their resolution are area-averaged to whole device pixels instead of nearest-neighbour (scanned text kept one sample in N per axis); page-scan-size images are smoothed when enlarged. Estimates unchanged. | | 2026-10-27 | minor | M10: OCR stages SHA-256-verified ONNX models after the upstream RTEN URL brownout; discovery prefers a complete ONNX pair per directory, falls back to a complete RTEN pair, and never mixes formats or directories; language/web limits and estimates unchanged. | | 2026-10-24 | minor | M3, #3970 follow-up: the button release given back after a window move or resize goes before the pointer leaving, so hover doesn't stick. Estimates unchanged. | | 2026-10-23 | minor | M8, #3205 follow-up: phone-pattern redaction includes the +39 country code on labelled Italian geographic numbers with a slash; dates, protocol identifiers and whole-run guards remain. Estimates unchanged. | @@ -350,6 +357,8 @@ Estimates unchanged. | 2026-10-20 | minor | M10, #3532 partial: vector printing retains existing appearances with absent Subtype; estimates unchanged. | | 2026-10-19 | minor | M4 #3309: named-destination pruning keeps large documents usable; 60,000-name and deep-tree regressions verify targets through undo/redo and both save modes. Estimates unchanged. | | 2026-10-19 | minor | M6: Text field table (rows and columns of text fields as one edit, `form_add_field_table`); estimates unchanged | +| 2026-10-20 | patch | M1, #1143 review: preserve accepted aliases through preflight/snapshots, normalize them to one output identity with full-save fallback for repaired edits, and report retained materialized/hybrid generation disagreements before folding. Synthetic red/green and four-save controls; estimates unchanged. | +| 2026-10-18 | minor | M1, #1143: a live reference may match the xref generation or a same-number header at its indexed offset; neither-match and free slots remain null. Bounded repair diagnostics and synthetic cache, free-slot, edit and classic/packed save regressions; estimates unchanged. | | 2026-10-18 | minor | M5, PR #1971: nested source `/Page` and `/Pages` dictionaries export as null without traversing page content or the page tree. Synthetic saved/reopened three-page regression retains unrelated metadata and aliases. FDF import semantics, copied-graph limits and estimates unchanged. | | 2026-10-15 | minor | M5: budget retained FDF entries before cloning; discovery keeps lazy array owners and annotation positions, with alias/cache-eviction stress fixtures. Limits cover the copied graph; source parsing/page discovery retain existing COS/parser limits, with no total source-memory bound. FDF import semantics and estimates unchanged. | | 2026-10-15 | minor | M7: Edit text caret height follows its row, with a kittest regression; estimates unchanged | @@ -362,6 +371,7 @@ Estimates unchanged. | 2026-10-14 | minor | M1: raw-deflate fallback preserves decoding-limit errors in strict and tolerant decoding; synthetic filter and COS stream regressions. Estimates unchanged. | | 2026-10-16 | minor | M5 #2669, item 5: the Comments panel groups replies and counts all page roots once per frame. Estimates unchanged. | | 2026-10-15 | minor | M9 #1895: generic appearance rewriting deferred; separate validation overlays preserve original artwork, exports and printing. Baked captions remain unchanged (#79 open). Estimates unchanged. | +| 2026-10-15 | minor | M1: refresh object-generation reference identity, retaining current upstream progress and regression tests. Estimates unchanged. | | 2026-10-16 | minor | M10 #1201 partial: lossless image export applies component Decode pairs and Indexed mapping. Estimates unchanged. | | 2026-10-15 | minor | M6: CSV export and spreadsheet merging quote tabs, and import distinguishes quoted tabs from TSV delimiters without changing field identities. Estimates unchanged. | | 2026-10-13 | minor | M4: rotation queries and edits resolve indirect page-tree values; inherited, negative and large rotations verified through both saves. Estimates unchanged. | diff --git a/parity/acrobat-features.toml b/parity/acrobat-features.toml index 986a3e8c6a..949c1c472b 100644 --- a/parity/acrobat-features.toml +++ b/parity/acrobat-features.toml @@ -70,8 +70,8 @@ milestone = "M1" status = "shipped" commands = [] tools = ["doc_info"] -evidence = ["crates/ui-egui/tests/it/viewer.rs::damaged_files_say_they_were_repaired"] -notes = "a notice says the file was repaired; Document Properties ▸ Advanced lists the repairs; doc_info reports them" +evidence = ["crates/ui-egui/tests/it/viewer.rs::damaged_files_say_they_were_repaired", "crates/cos/tests/it/generations.rs::contradictory_xref_generation_recovers_the_matching_header_and_records_repair", "crates/cos/tests/it/generations.rs::generation_repair_diagnostics_are_bounded_without_disabling_recovery", "crates/cos/tests/it/generations.rs::retained_generation_disagreements_are_reported_in_materialized_mode_too", "crates/cos/tests/it/generations.rs::hybrid_generation_diagnostics_observe_rows_before_the_table_hides_them", "crates/cos/tests/it/generations.rs::discarded_xref_repair_notes_do_not_survive_object_header_reconstruction"] +notes = "a notice says the file was repaired; Document Properties ▸ Advanced lists the repairs; doc_info reports them. COS records bounded live xref/header generation disagreements on open, including retained materialized and hybrid rows before folding; this follow-up validates the COS policy, without a fresh GUI claim." [[feature]] id = "core.xref-streams" @@ -82,8 +82,8 @@ milestone = "M1" status = "shipped" commands = [] tools = [] -evidence = ["crates/cos/tests/it/writer.rs::incremental_saves_stack_on_an_xref_stream_file", "crates/cos/tests/it/writer.rs::page_tree_loops_in_object_streams_do_not_crash_readers"] -notes = "" +evidence = ["crates/cos/tests/it/generations.rs::indirect_stream_lengths_require_the_referenced_generation", "crates/cos/tests/it/generations.rs::compressed_objects_only_accept_generation_zero", "crates/cos/tests/it/writer.rs::incremental_saves_stack_on_an_xref_stream_file", "crates/cos/tests/it/writer.rs::page_tree_loops_in_object_streams_do_not_crash_readers"] +notes = "Compressed objects accept generation zero only; indirect stream lengths are resolved by their complete reference. Synthetic regression coverage." [[feature]] id = "core.hybrid-xref" @@ -298,8 +298,8 @@ milestone = "M1" status = "shipped" commands = ["file.save_as"] tools = ["doc_save"] -evidence = ["crates/organize/src/tests.rs::full_save_drops_unreachable_objects", "crates/engine/src/tests.rs::full_save_is_readable_and_smaller_after_deletes"] -notes = "" +evidence = ["crates/cos/tests/it/generations.rs::full_save_does_not_turn_stale_references_into_live_ones", "crates/cos/tests/it/generations.rs::repaired_page_tree_aliases_keep_one_identity_in_full_saves", "crates/organize/src/tests.rs::full_save_drops_unreachable_objects", "crates/engine/src/tests.rs::full_save_is_readable_and_smaller_after_deletes"] +notes = "Full rewrites coalesce accepted repair aliases to one output object, preserving page-tree parent identity. They preserve reference-generation identity: stale references become null while live replacement objects remain reachable. Synthetic classic and object-stream save tests; this does not establish general rendering fidelity." [[feature]] id = "core.object-stream-compression" @@ -334,8 +334,8 @@ milestone = "M1" status = "shipped" commands = [] tools = [] -evidence = ["crates/cos/src/document/tests.rs::reads_classic_revision_and_edits_are_copy_on_write"] -notes = "" +evidence = ["crates/cos/tests/it/generations.rs::free_xref_slots_never_revive_retained_object_headers", "crates/cos/tests/it/generations.rs::restoring_equal_values_still_restores_the_accepted_generation_aliases", "crates/cos/tests/it/generations.rs::restoring_explicit_sets_removes_old_generation_aliases", "crates/preflight/tests/generation_repairs.rs::preflight_preserves_repaired_image_references_in_snapshots_and_saves", "crates/cos/tests/it/generations.rs::repaired_header_references_survive_full_saves_and_dictionary_edits", "crates/cos/tests/it/generations.rs::incremental_generation_identity_is_visible_to_an_independent_reader", "crates/cos/tests/it/generations.rs::restoring_snapshot_preserves_changes_to_nonzero_generation_objects", "crates/cos/tests/it/generations.rs::overlay_generation_changes_do_not_retarget_old_references_or_clones", "crates/cos/tests/it/generations.rs::references_require_the_current_generation_even_with_a_warm_cache", "crates/cos/src/document/tests.rs::reads_classic_revision_and_edits_are_copy_on_write"] +notes = "Generation-aware references stay distinct across warm caches, edit overlays and snapshots, including restoration and incremental save/reopen of edited nonzero-generation objects. A damaged live xref row may additionally resolve the matching same-number object header; unrelated generations and free slots remain null. Repair-aware edits preserve both aliases and require full rewrite; explicit replacement keeps only its supplied generation. Synthetic preflight/snapshot and classic/packed full/incremental controls include an independent reader. Broader real-file fidelity remains tracked separately." [[feature]] id = "core.write-validity" @@ -734,8 +734,8 @@ milestone = "M2" status = "partial" commands = [] tools = ["page_render"] -evidence = ["crates/render/src/raster/tests.rs::renders_a_blue_rectangle", "crates/automation/tests/it/automation.rs::open_inspect_render_and_find", "crates/render/src/raster/tests.rs::user_unit_scales_the_page_and_everything_on_it"] -notes = "hayro bootstrap renderer; own DisplayList devices in M2.6. Pages with /UserUnit are shown at their size in points, as Acrobat reads the value" +evidence = ["crates/render/src/raster/tests.rs::renders_a_blue_rectangle", "crates/automation/tests/it/automation.rs::open_inspect_render_and_find", "crates/render/src/raster/tests.rs::user_unit_scales_the_page_and_everything_on_it", "crates/render/src/raster/tests.rs::non_interpolated_minification_keeps_thin_strokes", "crates/render/src/raster/tests.rs::a_shrunk_stencil_mask_keeps_thin_strokes", "crates/render/src/raster/tests.rs::minified_image_edges_snap_to_whole_pixels", "crates/render/src/raster/tests.rs::large_non_interpolated_images_are_smoothed_when_enlarged", "crates/render/src/raster/tests.rs::small_non_interpolated_images_keep_hard_pixels_when_enlarged"] +notes = "hayro bootstrap renderer; own DisplayList devices in M2.6. Pages with /UserUnit are shown at their size in points, as Acrobat reads the value. An image drawn below its resolution is shrunk once to whole device pixels: by area averaging unless it asks for /Interpolate (then Catmull-Rom), so the text of a 300 dpi scan at 100% keeps its strokes, a 1-bit /ImageMask text layer included (it was nearest-neighbour, which dropped them). Enlarged by more than 1%, an image of page-scan size (512 samples or more on its shorter side) is smoothed bilinearly whatever its /Interpolate, so a zoomed scan has no staircases; smaller images keep their hard pixels. Not measured against Acrobat" [[feature]] id = "view.render-shadings-patterns" @@ -1778,8 +1778,8 @@ milestone = "M3" status = "partial" commands = ["help.check_updates"] tools = [] -evidence = ["crates/ui-egui/tests/it/updates.rs::a_newer_release_is_offered_for_download", "crates/ui-egui/tests/it/updates.rs::an_up_to_date_or_failed_check_says_so", "crates/ui-egui/tests/it/updates.rs::the_check_at_start_offers_the_prompt_and_later_dismisses_it", "crates/ui-egui/tests/it/updates.rs::nothing_is_asked_until_the_user_checks"] -notes = "asks GitHub Releases through craft-update at start (desktop app only) and on request, offers the release page, and updates and restarts in place where the install type allows it; an install waits while documents are unsaved" +evidence = ["crates/ui-egui/tests/it/updates.rs::a_newer_release_is_offered_for_download", "crates/ui-egui/tests/it/updates.rs::an_up_to_date_or_failed_check_says_so", "crates/ui-egui/tests/it/updates.rs::the_check_at_start_offers_the_prompt_and_later_dismisses_it", "crates/ui-egui/tests/it/updates.rs::nothing_is_asked_until_the_user_checks", "crates/ui-egui/tests/it/updates.rs::turning_the_check_at_start_on_waits_for_the_next_start", "crates/ui-egui/tests/it/updates.rs::the_preference_keeps_the_check_at_start_from_reaching_the_network", "crates/ui-egui/tests/it/updates.rs::settings_from_the_old_opt_in_leave_the_check_at_start_on"] +notes = "asks GitHub Releases through craft-update at start (desktop app only) and on request, offers the release page, and updates and restarts in place where the install type allows it; an install waits while documents are unsaved; Preferences ▸ Check for updates when PdfCraft starts (update_check_at_start, on by default) turns the check at start off, leaving only the request from the Help menu; turning it on takes effect at the next start" [[feature]] id = "view.single-key-accelerators" diff --git a/vendor/README.md b/vendor/README.md index e1a5676254..7ae867cb11 100644 --- a/vendor/README.md +++ b/vendor/README.md @@ -40,7 +40,7 @@ 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 `.notdef` for a CID of an Adobe character collection (Japan1, GB1, CNS1, Korea1) that it can't find by Unicode: hayro took the CID for a glyph index of the substitute, which numbers its glyphs its own way, so the character was drawn as an unrelated one (𠮷 as "Ｑ" in BIZ UDPGothic); (22) a Type 0 font drawn with a substitute finds a character given with a variation selector (`/ToUnicode`, or Adobe-Japan1-UCS2, which maps the JIS X 0208 forms of the kanji JIS X 0213:2004 redrew that way: CID 1247 is 噂 U+E0100) through the substitute's variation sequences, else as the character alone: such a CID found no glyph and was drawn as an unrelated one; (23) backport of hayro 0.8's `gradient` module: `EncodedShadingPattern::as_svg_gradient` turns an axial or radial shading into stops of an sRGB gradient (within 0.01 per channel, a stop at every stitching-function bound via `Function::stitching_bounds`) when a native gradient draws it exactly (extended past the box drawn in; a radial one with no background colour); (24) `DecodedImageCache`: decoded image pixels shared across renders through the interpreter cache within a 64 MiB budget (an image-heavy page decoded its thousands of images again per tile and per scale), keyed by every decode input (`DecodedImageKey`: the stream's place and dictionary bytes, its object, mask/stencil/raster output, interpolation, decode target size, preview or final); images under a transfer function and inline images are not cached, an image over the budget is never kept, and a poisoned lock decodes uncached; (25) `Device::actual_text` hands the device a marked content sequence's `/ActualText` (inline or named in `/Properties`), so text extraction keeps the Thai, Indic and Arabic clusters Chrome marks this way (their shaped tone marks and vowels have no Unicode of their own); (26) non-embedded simple fonts drawn with a standard-font stand-in build the Latin Extended-A letters the stand-in lacks (`/Gbreve`, `/Idotaccent`, `/Scedilla`, `/aogonek`, `/ccaron`…) from the base letter and accent it has (`StandardKind::composite_glyph`, `ACCENTED_LETTERS`): PDFium's Foxit stand-ins hold only the Standard/WinAnsi/Mac Roman glyphs, so iText letters with Times New Roman unembedded showed blanks for Ğ, İ, Ş, ğ and ş; the comma-below letters (`/scommaaccent`, `/Tcommaaccent`, `/kcommaaccent`…, found by character, so `/uni0136` and the glyph list's `/Kcedilla` too) are the letter and the font's comma, smaller and under the baseline, since the stand-ins have no `commaaccent` (Romanian ș drew nothing and ț a cedilla in forms laid out from XFA; ģ, whose comma is turned and above, is not built); (27) an embedded TrueType font without a base encoding reads its codes as WinAnsi for text extraction (`EmbeddedKind::code_to_unicode_name`), matching pdftotext's output, instead of Standard: OCR text layers of scans write WinAnsi codes with no `/Encoding`, and Standard has no ü, ö or ß and reads é as Ø (glyph selection unchanged); (28) an image whose pixels hold at most 16 bits of samples (an 8-bit Separation or DeviceN image, a 1-, 2- or 4-bit gray or indexed one) is converted once per possible pixel value through a lookup table instead of once per pixel (`rgb_via_lookup`): a spot-colour tint transform is often a PostScript calculator program, run for every pixel, so a fillable character sheet's 3304 × 4105 DeviceN background took four seconds and 430 MB to draw a page; each entry is converted on its own (`convert_each`), so a tint transform giving the wrong number of components for a value the image never uses cannot shift the colours of the values it does; (29) `Device::begin_annotations`/`end_annotations` bracket the page's annotations, so a device can tell comment paints from document paints (reading colours recolour the document, never a comment, #1548); (30) the Annex D.2 alias codes read as the ink they draw: WinAnsiEncoding 0xA0 and 0xAD, and MacRomanEncoding 0xCA, carry the glyph names `space` and `hyphen` -- the names those tables also give 0x20 and 0x2D -- not `nbspace` and `sfthyphen`, whose glyph list values U+00A0 and U+00AD are invisible formatting characters. Extraction hid drawn ink behind them, so search and redaction matched neither (`XiaoBiaoSong.pdf` in the pdf.js corpus drew 104 soft hyphens; across the 989 files the soft hyphens we extract fall from 117 to 13 and the non-breaking spaces from 30 to 26, what remains coming from `/ToUnicode` maps, which say so explicitly). The base table only, in the text direction only: a `/Differences` name keeps its glyph list value and glyph lookup is untouched, both as pdftotext does; (31) a Type 0 sampled function keeps its samples in one flat array indexed by strides (`Type0::entry`) instead of a hash map keyed by each grid point: the map was built again, one hashed insert per sample, every time a shading using the function was painted (a 2,097,152-sample ramp painted 30 times took 38 s; `bug1721218` in the pdf.js corpus spent most of a 1.5 s page on it); the interpolation is unchanged, so results are bit for bit the same, a sample missing from a short stream still gives no value, and the maximum sample value is computed without overflowing at 32 bits per sample; (32) the decoded-image cache (24) keys a decode by the size the image is drawn at only when a JPX stream decodes it or its soft mask or mask (`decode_reads_target`): JPX is the only filter that decodes a reduced resolution for a size, so a JPEG or Flate scan was decoded again at every zoom step and kept once per zoom level; one decode now serves every size; `InterpreterCache::decoded_image_decodes` counts the cache's decodes; (33) Coons and tensor-product patches copy their vertices' colours less: each point of a patch's 20 × 20 triangle grid is interpolated once (`generate_patch_triangles`; each cell interpolated its four corners itself, 1,444 interpolations a patch for 400 points), a cell's second triangle takes the vertices its first copied, and the mesh sampler moves a patch's triangles into device space instead of copying each vertex (`TriangleSampler::sample_owned`); the same values, 10–16% faster on mesh-gradient pages; (34) `color::tint_transform_runs`: how many colours the calling thread has run through a Separation or DeviceN tint transform (counted once per batch in `convert_f32`), so a test counts patch (28)'s once-per-sample-value runs instead of timing them (its 10 s deadline failed on a busy FreeBSD VM in CI); (35) `EncodedShadingPattern::extent`: where a mesh shading (types 4–7) can paint at all, the pixels its triangles touched when it was encoded (`None` for the types that paint everywhere, and when the shading has a `/Background`), so the renderer samples a texture there alone and skips a mesh with nothing on the canvas; (36) a mesh shading's pixel whose corner no triangle holds strictly (on an edge two triangles share, or on the mesh's side) takes the triangle that covers its centre, edges included (`Triangle::covers`): testing only the corner left a white seam along every shared edge, and the colour is still taken at the corner; (38) the `/DefaultGray`, `/DefaultRGB` and `/DefaultCMYK` colour spaces of a resource dictionary (ISO 32000-2 §8.6.5.6) replace the device space content selects: with `g`/`G`, `rg`/`RG`, `k`/`K` and `cs`/`CS` (by name, or through a `/ColorSpace` resource that is a device space), and as the `/ColorSpace` of an image XObject, an inline image or a shading painted by `sh` or as a shading pattern (`DefaultColorSpaces`, looked up once per content window in the resources and those they inherit, so a form XObject without a default of its own draws in its page's, as Poppler and MuPDF draw it); a default that is a Lab, Indexed, Pattern or device space, has another number of components or is no colour space is ignored and device colour is drawn as before; device spaces inside other spaces (an Indexed base, a Separation or DeviceN alternate, a Pattern's underlying space), shading dictionaries, masks and transparency group spaces are unchanged, and the decoded-image cache (24) keys a decode by the default it was drawn in. hayro drew device colour as it was, so pages that calibrate it this way (common in PDF/A) came out in the wrong colours; (39) knockout groups (#2971): `Device::set_knockout` right after a `/K true` group is pushed; `Device::knockout_image` for an unmasked image before its group is pushed, when `Device::in_knockout_group`: the device clears its placement rectangle only if it will draw the image (on the canvas, not a preview's tiny image, and it decodes), decoded at its draw size so the draw reuses that decode; and, when `Device::knockout_erase_wanted`, a nested group's content drawn once between `begin_knockout_erase` and `end_knockout_erase` before the group itself, which then always gets a layer of its own; (40) an optional content membership dictionary's visibility expression (`/VE`: `/And`, `/Or`, `/Not`, nested up to 32 deep) decides its visibility ahead of `/OCGs` and `/P` (ISO 32000-2 §8.11.2.2); it was ignored, so content under it stayed visible with its layers hidden; a malformed expression, including one naming an object the document doesn't list in `/OCProperties /OCGs`, falls back to `/OCGs` and `/P` (#2986); (41) a transparency group that is neither isolated nor knockout, entered at full opacity with no soft mask and Normal blending, is drawn straight onto the page instead of into a layer of its own (§11.4.7: its elements composite with the backdrop; the layer made a `/BM /Multiply` highlight appearance an opaque block, #1968); (42) text in a Type 3 font is drawn filled whatever the text rendering mode, except mode 3 (invisible): only that mode has any effect on Type 3 fonts (ISO 32000-1 §9.3.6). Stroking ran the glyph procedure a second time, so `2 Tr` made bitmap fonts look bold, `1 Tr` painted an uncoloured glyph in the stroking colour, and `7 Tr` hid it (`7 Tr` since follows 64); (43) Lab colour spaces honour their own `/WhitePoint`: L\*a\*b\* goes to XYZ relative to that white, is Bradford-adapted to D65 and converted to sRGB (ISO 32000-2 §8.6.5.4) instead of through `LAB.icc` as if every white were D50 (a D65 blue came out 45 levels too red; the profile is removed); components normalised to [0, 1] (`manual_scale`, an `/Indexed` palette over Lab) map onto L\* 0..100 and the a\*/b\* `/Range`; (44) a Type 0 font whose `/ToUnicode` is the name `/Identity-H` or `/Identity-V` (not a CMap stream; WSCAD writes it) reads each code as its Unicode value, as Acrobat shows it (`Type0Font::identity_to_unicode`); the identity CMap the name resolved to maps codes to CIDs only, so such pages had no text to select or find; (46) an image XObject's group applies the fill opacity, so the fill alpha is reset to 1 before a stencil mask takes the fill colour as its paint, and a pattern paint drops the opacity it captured at `scn` (`Pattern::without_opacity`) (it was applied twice: a mask at `ca 0.5` was drawn at a quarter, #2975); (47) `path_as_rect`: a five-element path is a rectangle only when its four points are distinct, each step from one to the next runs along a single axis, and the last returns to the first. The length test admits one MoveTo and four LineTo as well as one MoveTo, three LineTo and a ClosePath, and asking only that every corner of the bounding box was visited says nothing about the order or the steps between, so a five-point polyline whose ends reached the corners was painted as that box -- a fourth side the path does not have -- and a clip of that shape was discarded; the coordinates are compared exactly, since the matrix is applied after this check and a `cm` of 10,000 turns `is_nearly_equal`'s 256th of a unit into tens of device units (#3327); (48) an `/ICCBased` colour space is cached by what its profile is read from (`IccProfileKey`): the stream object it names, looked up without reading the stream, and then the stream's dictionary and its own bytes (decrypted, not decoded), so a byte-for-byte copy of a profile in another object is read once, and a profile stream written directly in its array, which reports the object holding the array as every other stream written there does, is told apart by its bytes. hayro keyed it by the stream dictionary's bytes alone, so two different profiles with the same `/N` and `/Length` shared whichever was read first, and colours in the second were drawn through the first. `InterpreterCache::icc_profile_reads` counts the profiles read; (49) `OpenTypeFontBlob::new` also accepts a `post` table version 1.0, 2.0 or 3.0 written with its 16-bit halves swapped (`00 00 00 03`), as ScanSoft PDF Create! embeds every font; such fonts were refused and their text drawn with a substitute (Hebrew came out garbled or missing, #2946); `glyph_names` reads such a table from a copy with the version written correctly, so a 2.0 table's names serve `/Differences`. Any other unknown version is still refused (pdf.js 9462); (50) `CffFontBlob::glyph_index_by_name` looks a glyph name up in a name to glyph map built on first use (shared between clones) instead of searching the whole charset, with a string lookup per glyph, for every character code of a font; the first glyph with a name still wins; (51) at most 4096 graphics states saved at once, holding at most 64 MiB together (each `q` copies the whole state, dash array and clip list included, so short operators multiplied one large state); a `q` past either limit is counted but not copied, and its `Q` still removes the clips added since (from storytold/craft-libs#239); (52) a `/W2` run whose first CID is past `MAX_CID` stops there, as in `/W` (7), checked before the run's metrics are read: a first CID of 4294967295 inserted one entry and then overflowed, discarding the whole `/W2` table (from storytold/craft-libs#254); (55) a Separation or DeviceN tint transform that cannot be read keeps its colour space, and one that cannot be read or evaluated paints the alternate space's initial colour, as evaluation failures already did; that fallback, and a `cs`/`CS` colour space that cannot be read (drawn in DeviceGray as before), are now reported once per content stream as `InterpreterWarning::ColorFallback` (`RenderWarning::ColourFallback` in `pdfcraft-render`) instead of silently (#4260); (56) an annotation whose Invisible flag (bit 1) is set and whose `/Subtype` is none of the standard types of ISO 32000-2 Table 171 (`is_standard_annotation_type`; no `/Subtype` counts as unknown) is not drawn (Table 167), and a `/Popup` annotation's appearance is never painted on the page (§12.5.6.14: a popup has no appearance stream of its own; the viewer shows it as a window for its parent annotation). Both were drawn from their `/AP`; (57) a simple font drawn with a standard-font stand-in because the document names a font it doesn't embed (`StandardKind` with `fallback`) narrows a glyph to a narrower `/Widths` advance but no longer widens it to a wider one: the advance already places the next glyph (ISO 32000-2 §9.2.4; /Widths in Table 109), and widening drew a non-embedded `Arial,Bold` with 600-unit widths as a heavy fixed-pitch face whose `i` and `l` were solid blocks twice their width; an exact standard-14 font still matches its standard metrics both ways; (58) the fallback request for a Type 0 font that is not embedded (`FallbackFontQuery::with_descriptor`) takes its serif, italic and small-cap flags, `/FontWeight`, `/FontFamily` and `/FontStretch` from the descendant CIDFont's font descriptor (ISO 32000-2 §9.7.4.1, §9.7.6.1; §9.8): it read them from the Type 0 dictionary, which has none, so every such request asked for a regular sans-serif face and no resolver could tell a Song or Mincho font from a Hei or Gothic one, or see `/FontWeight 700`; (59) the NoRotate flag (`/F` bit 5) is honoured: the appearance, fitted to `/Rect`, is turned against the page's `/Rotate` about the rectangle's upper-left corner, which stays where it is, so it reads upright as the page is displayed (ISO 32000-2 §12.5.3). Sticky notes, which are saved NoRotate, were drawn turned with the page (#2583). Form fields and links stay in `/Rect`, where PdfCraft takes their click targets from it; (64) text in a Type 3 font follows ISO 32000-2 §9.3.6 in the clipping modes: under `7 Tr` it is not rendered (reported to the device as clipping text, `GlyphDrawMode::Clip`, so extraction keeps it), and in modes 4 to 7 a Type 3 glyph adds nothing to the clipping path (`clip_glyph`), so text shown only in a Type 3 font leaves no clip at `ET`. Since (42) `7 Tr` painted the glyphs in the fill colour (black by default), so a fill meant to follow showed them wherever it didn't cover them; 4 to 6 still render filled; (66) a Separation colour space whose colourant is the special name `All` paints every colourant at once (ISO 32000-2 §8.6.6.4): on screen and in an RGB print raster its tint t is the gray 1 − t, complemented because the device is additive, whatever its alternate space and tint transform say (hayro drew it through the tint transform, so a `/DeviceCMYK` transform giving cyan drew registration marks light cyan); `/None` still paints nothing, any other name still goes through its alternate, and a DeviceN space naming `All` keeps its alternate, since the clause gives the name that meaning in a Separation space; (63) a luminosity soft mask whose group colour space is `/DeviceCMYK` (by name, not an ICC-based or calibrated space) reads a `/DeviceCMYK` fill or stroke colour drawn in it, and its `/BC`, as their gray `1 - min(1, 0.3 c + 0.59 m + 0.11 y + k)` (ISO 32000-2 §10.4.2.3, §11.5.3: device colour converts to gray with no colour calibration; `Context::luminosity_in_device_cmyk`, `color::device_cmyk_to_gray`): it went through the CMYK display profile, where 100% K is a dark grey, so a 100% K mask let about an eighth of the masked paint through. CMYK images, shadings and tiling patterns inside such a mask, and CMYK on the page, still use the profile; (71) a lookup table in an indirect string reads: it was read as a stream first, which uses up a reference (ISO 32000-2 §8.6.6.3); (72) an uncoloured pattern's underlying space given as a name or a reference is used: it was skipped, so a CMYK cyan drew red (ISO 32000-2 §8.7.3.3); (73) `InterpreterSettings::jbig2_work`, set for each page in `interpret_page` and ended with it, and a JBIG2 image or an image with a JBIG2 mask past it left out with `ContentTruncated`; (74) colour spaces nest at most 32 levels, through every way one names another, so a chain of them ends instead of overflowing the stack; an ICC profile's `/Alternate` is read where it's met and isn't kept with the profile, or a page could build a chain past the limit on spaces already read; (75) a soft hyphen (U+00AD) that a CFF substitute for a CID font has no glyph for draws its hyphen, as Annex D.2 reads WinAnsi 0xAD in (30), instead of the CID taken as a glyph index (#5888) | (1) `raster::tests::widget_appearance_states`; (2, 3) `xtask text-oracle` on corpus/pdfjs issue918; (4) `raster::tests::layer_override_hides_content`; (5) `raster::tests::soft_mask_without_group_colour_space_is_applied`; (6) `raster::tests::self_referencing_tiling_pattern_terminates`, `raster::tests::self_referencing_type3_glyph_terminates`; (7) `raster::tests::huge_cid_width_ranges_terminate`; (8) `raster::tests::hiding_comments_keeps_fields`; (9) `raster::tests::absurd_mask_and_fax_sizes_are_refused_before_decoding`; (10) `raster::tests::type3_glyphs_that_fan_out_into_themselves_finish`; (11) `raster::tests::type1_font_programs_with_long_runs_of_numbers_load_quickly`; (12) `raster::tests::degenerate_annotation_appearances_are_skipped`; (13) `raster::tests::highlights_without_an_appearance_are_drawn`; (14) `raster::tests::malformed_simple_font_widths_keep_later_widths_and_the_font`, `raster::tests::huge_simple_font_first_char_terminates`; (15) `raster::tests::annotation_appearance_box_maps_onto_rect`; (16) `raster::tests::clipping_path_applies_after_any_painting_operator`; (17) `raster::tests::dropping_the_pool_cancels_its_renders`, `raster::tests::the_watchdog_stops_the_render_it_gives_up_on`; (18) `raster::tests::page_contents_and_nested_forms_share_one_budget`, `raster::tests::content_painted_again_and_again_keeps_drawing`, `raster::tests::sampled_functions_with_billions_of_samples_are_refused`; (19) `raster::tests::mesh_shadings_far_off_the_page_render`, `raster::tests::many_overlapping_mesh_triangles_finish`, `raster::tests::mesh_shadings_with_many_patches_or_no_rows_finish`; (20) `raster::tests::function_limits_calculator_arithmetic_and_depth`, `raster::tests::function_limits_public_construction_work_and_depth`, `raster::tests::function_limits_stitching_cycles_are_refused_and_shared_children_work`, `raster::tests::function_limits_invalid_calculator_keeps_the_page_renderable`; (21) `raster::tests::cids_a_substitute_lacks_are_not_drawn_as_other_glyphs`; (22) `raster::tests::variation_sequences_draw_their_kanji_in_a_substitute`; (23) `raster::tests::axial_and_radial_shadings_draw_as_native_gradients`; (24) `raster::tests::tiles_match_full_render`, `raster::tests::skipping_paths_off_the_canvas_changes_no_pixel` (both decode through the cache), `raster::tests::decoded_images_with_identical_dictionaries_keep_their_own_pixels`, `raster::tests::decoded_images_under_a_transfer_function_are_not_shared`; (25) `text::tests::actual_text_replaces_the_glyphs_it_marks_in_every_writing_direction`, `text::tests::actual_text_follows_the_pdf_rules_for_marked_content`; (26) `raster::tests::accented_letters_the_standard_stand_ins_lack_are_built_from_their_parts`, `raster::tests::comma_below_letters_in_standard_fonts_draw_a_comma_under_the_letter`; (27) `text::tests::truetype_without_an_encoding_reads_its_codes_as_win_ansi`; (28) `raster::tests::spot_colour_images_run_their_tint_transform_once_per_sample_value`, `raster::tests::images_converted_through_a_table_keep_every_samples_colour`, `raster::tests::a_malformed_tint_for_an_unused_sample_keeps_the_used_colours`; (29) `raster::tests::reading_colors_keep_highlight_annotations_readable`; (30) `text::tests::the_annex_d2_alias_codes_read_as_an_ordinary_hyphen_and_space`; (31) `raster::tests::shadings_with_large_sampled_functions_paint_quickly`, `raster::tests::sampled_functions_interpolate_their_grid_as_before`; (32) `raster::tests::decoded_images_serve_every_size_they_are_drawn_at`; (33) `raster::tests::patch_meshes_keep_their_corner_colours`; (34) `raster::tests::spot_colour_images_run_their_tint_transform_once_per_sample_value`; (35) `mesh_extent_tests::{a_mesh_paints_its_triangles_alone, a_mesh_keeps_the_graphics_state_alpha, a_mesh_off_the_canvas_paints_nothing}` (`cargo test -p hayro`); (36) `raster::tests::mesh_triangles_sharing_an_edge_leave_no_seam`; (38) `raster::tests::default_colour_spaces::device_rgb_operators_draw_in_default_rgb`, `raster::tests::default_colour_spaces::device_cmyk_operators_draw_in_default_cmyk`, `raster::tests::default_colour_spaces::device_gray_operators_draw_in_default_gray`, `raster::tests::default_colour_spaces::device_images_draw_in_the_default_colour_spaces`, `raster::tests::default_colour_spaces::device_shadings_draw_in_the_default_colour_spaces`, `raster::tests::default_colour_spaces::an_image_drawn_with_and_without_a_default_keeps_both_colours`, `raster::tests::default_colour_spaces::forms_inherit_the_default_colour_spaces_they_do_not_override`, `raster::tests::default_colour_spaces::unusable_defaults_leave_device_colour_alone`; (39) `raster::tests::knockout_groups_replace_what_their_elements_cover`; (40) `raster::tests::visibility_expressions_decide_what_a_membership_dictionary_shows`; (41) `raster::tests::a_multiply_highlight_inside_a_transparency_group_keeps_the_text`; (42) `raster::tests::rendering_modes_but_3_and_7_draw_type3_glyphs_as_filled`; (43) `raster::tests::lab_honours_the_space_white_point`, `raster::tests::indexed_palettes_over_lab_keep_their_colours`; (44) `text::tests::identity_to_unicode_names_give_text`; (46) `raster::tests::stencil_masks_are_drawn_at_the_fill_opacity_once`; (47) `raster::tests::a_five_point_polyline_is_not_painted_as_its_bounding_box`, `raster::tests::a_path_that_visits_every_corner_out_of_order_is_not_a_rectangle`, `raster::tests::a_rectangle_is_still_painted_as_a_rectangle`; (48) `raster::tests::icc_profile_cache::icc_profiles_with_identical_dictionaries_draw_their_own_colours`, `raster::tests::icc_profile_cache::icc_profiles_used_again_are_read_once`; (49) `raster::tests::a_font_whose_post_version_has_swapped_halves_is_still_used`; (50) `font::blob::tests::glyph_names_resolve_to_the_first_glyph_with_that_name`; (51) `context::saved_state_tests::repeated_saves_stay_bounded_and_balanced`; (52) `font::cid::tests::w2_run_past_the_largest_cid_keeps_the_rest_of_the_table` (hayro-interpret); (55) `raster::tests::devicen_inks::a_tint_transform_that_cannot_be_evaluated_draws_the_rest_of_the_page_and_warns`, `raster::tests::devicen_inks::a_tint_transform_that_cannot_be_read_falls_back_and_warns`; (56) `raster::tests::invisible_unknown_annotations_and_popups_paint_nothing`; (57) `substitute_widths::a_substitute_keeps_its_glyph_shapes_at_wider_advances`, `substitute_widths::a_substitute_still_narrows_glyphs_to_narrower_advances` (pdfcraft-render `tests/it`); (58) `cid_fallback_query::a_cid_font_fallback_request_carries_its_descendants_style`, `cid_fallback_query::a_cid_font_without_a_descriptor_asks_for_the_default_style` (pdfcraft-render integration tests); (59) `raster::tests::no_rotate_notes_stay_upright_at_their_corner_on_rotated_pages` (hayro-interpret); (64) `type3_text_clip::{clip_only_type3_text_is_not_rendered_and_leaves_no_clip, invisible_type3_text_is_not_rendered, other_modes_render_type3_glyphs_filled_and_never_clip, only_outline_glyphs_shape_a_clip_shared_with_type3_text, self_nesting_type3_text_in_a_clip_mode_stays_bounded}` (`cargo test -p pdfcraft-render --test it`), `raster::tests::rendering_modes_but_3_and_7_draw_type3_glyphs_as_filled`, `text::tests::clipping_mode_type3_text_is_kept_by_the_ocr_filter`; (66) `separation_all::{separation_all_fills_paint_the_complemented_tint_as_gray, separation_all_strokes_paint_the_complemented_tint_as_gray, separation_all_images_paint_each_sample_as_the_complemented_gray, separation_all_shadings_paint_the_complemented_gray, separation_none_paints_nothing, a_named_separation_still_paints_through_its_alternate, a_device_n_space_naming_all_still_paints_through_its_alternate}` (`cargo test -p pdfcraft-render --test it`); (63) `color::device_cmyk_gray_tests::{each_ink_darkens_by_its_iso_weight, ink_past_full_coverage_is_black_and_bad_components_read_as_zero}` (hayro-interpret), `it::luminosity_group_space::{black_ink_in_a_device_cmyk_group_masks_fully_and_no_ink_shows_fully, each_process_ink_in_a_device_cmyk_group_masks_by_its_weight, a_device_cmyk_backdrop_colour_masks_by_its_gray, gray_and_rgb_groups_mask_as_before, device_cmyk_on_the_page_keeps_its_display_colour}` (pdfcraft-render); (71) `raster::tests::an_indexed_lookup_table_is_read_in_every_legal_form` (pdfcraft-render); (72) `raster::tests::an_uncoloured_pattern_uses_its_underlying_space_in_every_form` (pdfcraft-render); (73) `raster::tests::the_jbig2_images_of_a_page_share_one_allowance`, `a_picture_whose_jbig2_mask_is_past_the_pages_allowance_is_not_drawn` and `a_pages_jbig2_allowance_ends_with_the_page` (pdfcraft-render); (74) `raster::tests::{a_chain_of_colour_spaces_ends, colour_spaces_are_read_to_the_limit_of_their_nesting, a_chain_built_on_spaces_already_read_ends, a_profile_read_at_the_limit_keeps_its_alternate_for_a_read_nearer_the_top}` (pdfcraft-render); (75) `raster::tests::a_soft_hyphen_a_substitute_lacks_draws_its_hyphen` (pdfcraft-render) | -| 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` 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 (#2128): vello_cpu 0.0.8 copies an opaque gradient over fully covered tiles without the mask or blend mode, so PowerPoint's faded gradient fills came out solid; (17) the renderer keeps annotation paints out of the reading-colours mapping between `begin_annotations` and `end_annotations` (#1548); (18) the page renders as one transparency group; under reading colours its background is painted inside it as the initial backdrop, so a Multiply highlight blends with the dark background instead of staying an opaque block, and otherwise the background stays outside, so a blend mode over an empty part of the page keeps its colour instead of blending with the white paper (#2970); (19) `set_paint` samples a mesh shading's texture over its triangles alone (`EncodedShadingPattern::extent`, padded rather than repeated) and skips a mesh with none on the canvas: a BricsCAD plan's 310 one-triangle `sh` each covered the whole tile, 2 s per 1024 px tile at any zoom (the watchdog gave the page up on a busy machine), now about 70 ms; (20) a tiling pattern whose step is too large for a cell pixmap, where what is painted (a fill, or a stroke with its width) can only show the first cell (PowerPoint's `/XStep 2147480100` for a picture fill drawn once), has its `/BBox` drawn at full scale with a transparent border and padded instead of tiled (`shows_one_cell`; a cell whose scale isn't finite and positive, from a `/BBox` too thin to divide by, keeps the old path: `usable_cell_scale`); shrinking the cell to fit the 3000 px limit of (3) left the picture under a pixel, so network diagrams lost their pictures (#3349); (22) knockout groups: while one is innermost, fills, strokes, glyph outlines and shadings are drawn with `Compose::Copy` (weighted by coverage), so each element replaces what earlier elements of the group drew instead of compositing over it; an image or nested group, drawn into a layer of its own, first has what is under its shape cleared (`knockout_image` for an unmasked image's rectangle, only when `draw_image` will draw it, decoded at the size it is drawn; a knockout erase pass that draws a nested group's content with every paint clearing instead of painting, a shading fill's clear clipped to the shading's `/BBox` as its paint is) and then composites over it (#2971); (23) backport of hayro 0.8: shadings that `as_svg_gradient` can represent are painted as a `vello_cpu` linear or two-point radial gradient instead of a texture sampled pixel by pixel through the shading function (InDesign fades are luminosity soft masks of one full-page axial `sh`: 40 on a report page took 15 s at 144 dpi, now under 1 s); without upstream's half-pixel shift, which suits vello_cpu 0.3 but moves 0.0.8's gradients off pixel centres; (25) a tiling cell is drawn at the scale that makes one `/XStep` (`/YStep`) exactly the cell pixmap's whole number of pixels (`tiling_cell_pixels`, at least 1 and at most the cap), so the repeated pixmap puts cells at whole multiples of the step at any device scale; drawing at the unrounded scale repeated the cell up to half a pixel off at every step, and the error grew with distance from the pattern's origin (ISO 32000-2 8.7.3.1) | (1) `raster::tests::tiles_match_full_render`; (2) `raster::tests::absurd_image_dimensions_are_skipped`; (3) `raster::tests::tiling_cells_stay_small_whatever_the_step`; (4) `raster::tests::absurd_predictor_columns_and_line_widths_render`; (5) `raster::tests::dash_patterns_that_cut_a_stroke_into_billions_of_pieces_render`; (6) `raster::tests::invalid_dash_patterns_terminate_and_are_drawn_solid`; (7) `raster::tests::pixels_are_a_plain_renders_bytes_in_a_buffer_a_gui_can_take_over`; (8) `raster::tests::image_resampling_rejects_anisotropic_target_growth`, `raster::tests::image_resampling_bounds_intermediate_before_planning`, `raster::tests::image_resampling_rejects_invalid_sources_and_scales`, `raster::tests::image_resampling_keeps_valid_sizes_and_exact_limits`, `raster::tests::resampling_fallback_preserves_original_geometry_and_pixels`, `raster::tests::image_resampling_wide_sources_shrink_before_backend_side_limits`, `raster::tests::image_resampling_wide_transparent_sources_still_shrink`, `raster::tests::image_resampling_padded_backend_dimensions_stay_checked`, `raster::tests::image_resampling_mismatched_alpha_mask_keeps_pixels_and_placement`; (9) `raster::tests::skipping_paths_off_the_canvas_changes_no_pixel`; (10) `raster::tests::non_interpolated_images_are_not_smoothed_when_minified`, `raster::tests::non_interpolated_gray_images_are_not_smoothed_when_minified`, `raster::tests::non_integer_minification_samples_the_same_columns`, `raster::tests::non_integer_minification_samples_colour_and_mask_from_the_same_columns`, `raster::tests::non_interpolated_alpha_images_keep_their_colours_and_mask_when_minified`, `raster::tests::non_interpolated_type3_glyph_images_are_not_smoothed_when_minified`, `raster::tests::image_resampling_nearest_accepts_wide_sources_the_filter_refuses`, `raster::tests::wide_non_interpolated_sources_draw_after_shrinking`; (11) `raster::tests::skipping_paths_off_the_canvas_changes_no_pixel`; (12) `raster::tests::preview_renders_skip_tiny_images_but_finals_draw_everything`; (13) `color_replacement_tests::{black_maps_to_text_colour, white_maps_to_background, midgrey_interpolates, alpha_copied_verbatim, only_black_threshold, line_art_false_skips_paths_but_not_text}`, `color_replacement_render_tests::{color_replacement_maps_text_and_line_art, color_replacement_maps_stencil_mask_ink}`; (14) `raster::tests::interpolated_minified_image_edges_snap_to_whole_pixels` (`cargo test -p hayro`), `raster::tests::a_turned_or_mirrored_minified_image_snaps_to_whole_pixels`, `rotated_image_snap::a_shrunk_image_on_a_turned_page_is_the_upright_render_turned` (`cargo test -p pdfcraft-render --test it`), `renderer::snap_tests::a_skew_that_moves_a_far_edge_is_not_snapped` (`cargo test -p hayro`); (15) `raster::tests::shrunk_soft_mask_larger_than_its_image_keeps_thin_strokes`; (16) `raster::tests::soft_mask_applies_to_a_shading_pattern_fill`; (17), (18) `raster::tests::reading_colors_keep_highlight_annotations_readable`; (19) `mesh_extent_tests::{a_mesh_paints_its_triangles_alone, a_mesh_keeps_the_graphics_state_alpha, a_mesh_off_the_canvas_paints_nothing}` (`cargo test -p hayro`); (20) `raster::tests::a_tiling_pattern_drawn_once_shows_its_cell`, `raster::tests::a_large_step_still_tiles_where_the_next_cell_shows`, `raster::tests::a_tiling_cell_too_thin_to_scale_keeps_the_old_path`; (22) `raster::tests::knockout_groups_replace_what_their_elements_cover`; (23) `raster::tests::axial_and_radial_shadings_draw_as_native_gradients`; (25) `tiling_cell_placement::{far_cells_of_a_fill_sit_at_whole_steps, far_cells_sit_at_whole_steps_under_a_turned_scaled_matrix, far_cells_of_a_stroke_sit_at_whole_steps, a_cell_pixmap_spans_exactly_one_step}` (`cargo test -p pdfcraft-render --test it`) | +| 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 (since (21), only inside Type 3 glyphs), 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` 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 image that is interpolated (or, since (21), area-averaged) and 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 (halves round up, so a tile snaps an edge where the whole page does) 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 (#2128): vello_cpu 0.0.8 copies an opaque gradient over fully covered tiles without the mask or blend mode, so PowerPoint's faded gradient fills came out solid; (17) the renderer keeps annotation paints out of the reading-colours mapping between `begin_annotations` and `end_annotations` (#1548); (18) the page renders as one transparency group; under reading colours its background is painted inside it as the initial backdrop, so a Multiply highlight blends with the dark background instead of staying an opaque block, and otherwise the background stays outside, so a blend mode over an empty part of the page keeps its colour instead of blending with the white paper (#2970); (19) `set_paint` samples a mesh shading's texture over its triangles alone (`EncodedShadingPattern::extent`, padded rather than repeated) and skips a mesh with none on the canvas: a BricsCAD plan's 310 one-triangle `sh` each covered the whole tile, 2 s per 1024 px tile at any zoom (the watchdog gave the page up on a busy machine), now about 70 ms; (20) a tiling pattern whose step is too large for a cell pixmap, where what is painted (a fill, or a stroke with its width) can only show the first cell (PowerPoint's `/XStep 2147480100` for a picture fill drawn once), has its `/BBox` drawn at full scale with a transparent border and padded instead of tiled (`shows_one_cell`; a cell whose scale isn't finite and positive, from a `/BBox` too thin to divide by, keeps the old path: `usable_cell_scale`); shrinking the cell to fit the 3000 px limit of (3) left the picture under a pixel, so network diagrams lost their pictures (#3349); (21) an image that does not request interpolation and is drawn below its resolution, outside a Type 3 glyph, is shrunk by area averaging (`area_shrink`: each device pixel the coverage-weighted mean of the samples it covers, in 16.16 fixed point whose weights sum to exactly one, premultiplied when the image has a mask; a soft mask no larger than the image is first sampled onto the image's grid by `mask_on_image_grid` (nearest-neighbour, or bilinearly when the mask asks for `/Interpolate`, so its edge stays as smooth as an enlarged mask's), so its colour and mask are averaged together instead of the averaged colour of transparent samples showing under a mask layer at the mask's edges) to its snapped device rectangle, instead of nearest-neighbour: `/Interpolate` asks for smoothing only when an image is enlarged (ISO 32000-2 §8.9.5.3), and nearest-neighbour kept one sample in N per axis, so the text of a 300 dpi scan at 100% lost most of its strokes (sampling them again at a fractional offset dropped more). PicScale's `Box` and `Area` filters were not used: shrinking, both average a window four target pixels wide. On a 13-page 300 dpi scan every rendered sample at 96, 120 and 192 dpi is within 1 of the exact area average (it was up to 255 off, 17-23% of samples by more than 1). A snapped image whose shrink failed keeps the full transform, so it is not drawn at its source size. The shrink never enlarges: a side drawn at or above its resolution (an image shrunk across and stretched down, as a 204 × 98 dpi fax is) keeps its samples for the sampler. Each target row averages the whole source rows it covers (contiguous, cheap) and is then shrunk across with taps worked out once per image, which made the shrink about three times faster (a 300 dpi page at 96 dpi: 25 ms, nearest-neighbour 11 ms); working a row at a time (`area_shrink_rows`) keeps its work memory to a few rows. An image with a soft mask, a stencil's ink included, is premultiplied a source row at a time as it is averaged (`area_shrink_premultiplied`), never as a full-resolution RGBA copy: a 10718 × 13701 stencil shown at page size peaked at 1.16 GB with that copy and peaks at 1.03 GB without it, as with nearest-neighbour. A source with a side over `MAX_RESAMPLING_SOURCE_SIDE` (2^20) keeps (10)'s nearest-neighbour shrink, and `area_shrink` refuses it, so its accumulator row and tap table stay a few times one source row (a 2^27 × 2 image of zeros is a tiny file); (22) knockout groups: while one is innermost, fills, strokes, glyph outlines and shadings are drawn with `Compose::Copy` (weighted by coverage), so each element replaces what earlier elements of the group drew instead of compositing over it; an image or nested group, drawn into a layer of its own, first has what is under its shape cleared (`knockout_image` for an unmasked image's rectangle, only when `draw_image` will draw it, decoded at the size it is drawn; a knockout erase pass that draws a nested group's content with every paint clearing instead of painting, a shading fill's clear clipped to the shading's `/BBox` as its paint is) and then composites over it (#2971); (23) backport of hayro 0.8: shadings that `as_svg_gradient` can represent are painted as a `vello_cpu` linear or two-point radial gradient instead of a texture sampled pixel by pixel through the shading function (InDesign fades are luminosity soft masks of one full-page axial `sh`: 40 on a report page took 15 s at 144 dpi, now under 1 s); without upstream's half-pixel shift, which suits vello_cpu 0.3 but moves 0.0.8's gradients off pixel centres; (25) a tiling cell is drawn at the scale that makes one `/XStep` (`/YStep`) exactly the cell pixmap's whole number of pixels (`tiling_cell_pixels`, at least 1 and at most the cap), so the repeated pixmap puts cells at whole multiples of the step at any device scale; drawing at the unrounded scale repeated the cell up to half a pixel off at every step, and the error grew with distance from the pattern's origin (ISO 32000-2 8.7.3.1); (28) an image of page-scan size (`SMOOTHED_IMAGE_SIDE`: 512 samples or more on its shorter side) that does not request interpolation is drawn with `ImageQuality::Medium` (bilinear, as interpolated images are) when enlarged by more than 1% on either side (`SMOOTHED_IMAGE_SCALE`: a scan drawn at its resolution can come out a few ulps above 1, and blurred by half a pixel at a half-pixel offset), also when the other side is shrunk (the area shrink leaves that side at 1:1), outside a Type 3 glyph: scans rarely set `/Interpolate`, and drawn nearest-neighbour every sample of a zoomed scan became a block, so its text had staircase edges; icons, QR codes and pixel art keep their hard pixels. Bicubic (`High`) looked the same on a zoomed 300 dpi scan and drew four 1024 px tiles at 400% in 267 ms against 50 ms | (1) `raster::tests::tiles_match_full_render`; (2) `raster::tests::absurd_image_dimensions_are_skipped`; (3) `raster::tests::tiling_cells_stay_small_whatever_the_step`; (4) `raster::tests::absurd_predictor_columns_and_line_widths_render`; (5) `raster::tests::dash_patterns_that_cut_a_stroke_into_billions_of_pieces_render`; (6) `raster::tests::invalid_dash_patterns_terminate_and_are_drawn_solid`; (7) `raster::tests::pixels_are_a_plain_renders_bytes_in_a_buffer_a_gui_can_take_over`; (8) `raster::tests::image_resampling_rejects_anisotropic_target_growth`, `raster::tests::image_resampling_bounds_intermediate_before_planning`, `raster::tests::image_resampling_rejects_invalid_sources_and_scales`, `raster::tests::image_resampling_keeps_valid_sizes_and_exact_limits`, `raster::tests::resampling_fallback_preserves_original_geometry_and_pixels`, `raster::tests::image_resampling_wide_sources_shrink_before_backend_side_limits`, `raster::tests::image_resampling_wide_transparent_sources_still_shrink`, `raster::tests::image_resampling_padded_backend_dimensions_stay_checked`, `raster::tests::image_resampling_mismatched_alpha_mask_keeps_pixels_and_placement`; (9) `raster::tests::skipping_paths_off_the_canvas_changes_no_pixel`; (10) `raster::tests::non_interpolated_images_are_not_smoothed_when_minified`, `raster::tests::non_interpolated_gray_images_are_not_smoothed_when_minified`, `raster::tests::non_integer_minification_averages_the_columns_each_pixel_covers`, `raster::tests::non_integer_minification_averages_colour_and_mask_together`, `raster::tests::non_interpolated_alpha_images_keep_their_colours_and_mask_when_minified`, `raster::tests::non_interpolated_type3_glyph_images_are_not_smoothed_when_minified`, `raster::tests::image_resampling_nearest_accepts_wide_sources_the_filter_refuses`, `raster::tests::wide_non_interpolated_sources_draw_after_shrinking`; (11) `raster::tests::skipping_paths_off_the_canvas_changes_no_pixel`; (12) `raster::tests::preview_renders_skip_tiny_images_but_finals_draw_everything`; (13) `color_replacement_tests::{black_maps_to_text_colour, white_maps_to_background, midgrey_interpolates, alpha_copied_verbatim, only_black_threshold, line_art_false_skips_paths_but_not_text}`, `color_replacement_render_tests::{color_replacement_maps_text_and_line_art, color_replacement_maps_stencil_mask_ink}`; (14) `raster::tests::minified_image_edges_snap_to_whole_pixels`, `raster::tests::a_tile_snaps_half_pixel_image_edges_as_the_whole_page_does` (`cargo test -p hayro`), `raster::tests::a_turned_or_mirrored_minified_image_snaps_to_whole_pixels`, `rotated_image_snap::a_shrunk_image_on_a_turned_page_is_the_upright_render_turned` (`cargo test -p pdfcraft-render --test it`), `renderer::snap_tests::a_skew_that_moves_a_far_edge_is_not_snapped` (`cargo test -p hayro`); (15) `raster::tests::shrunk_soft_mask_larger_than_its_image_keeps_thin_strokes`; (16) `raster::tests::soft_mask_applies_to_a_shading_pattern_fill`; (17), (18) `raster::tests::reading_colors_keep_highlight_annotations_readable`; (19) `mesh_extent_tests::{a_mesh_paints_its_triangles_alone, a_mesh_keeps_the_graphics_state_alpha, a_mesh_off_the_canvas_paints_nothing}` (`cargo test -p hayro`); (20) `raster::tests::a_tiling_pattern_drawn_once_shows_its_cell`, `raster::tests::a_large_step_still_tiles_where_the_next_cell_shows`, `raster::tests::a_tiling_cell_too_thin_to_scale_keeps_the_old_path`; (21) `raster::tests::non_interpolated_minification_keeps_thin_strokes`, `raster::tests::non_integer_minification_averages_the_columns_each_pixel_covers`, `raster::tests::non_integer_minification_averages_colour_and_mask_together`, `raster::tests::non_interpolated_minification_does_not_bleed_transparent_colour`, `raster::tests::a_smaller_soft_mask_does_not_bleed_transparent_colour_when_minified`, `raster::tests::an_interpolated_smaller_soft_mask_stays_smooth_when_its_image_is_minified`, `raster::tests::a_shrunk_stencil_mask_keeps_thin_strokes`, `raster::tests::minified_image_edges_snap_to_whole_pixels`, `raster::tests::an_image_shrunk_one_way_and_stretched_the_other_is_still_averaged`, `raster::tests::wide_non_interpolated_sources_draw_after_shrinking`, `renderer::area_shrink_tests` (`cargo test -p hayro`); (22) `raster::tests::knockout_groups_replace_what_their_elements_cover`; (23) `raster::tests::axial_and_radial_shadings_draw_as_native_gradients`; (25) `tiling_cell_placement::{far_cells_of_a_fill_sit_at_whole_steps, far_cells_sit_at_whole_steps_under_a_turned_scaled_matrix, far_cells_of_a_stroke_sit_at_whole_steps, a_cell_pixmap_spans_exactly_one_step}` (`cargo test -p pdfcraft-render --test it`); (28) `raster::tests::large_non_interpolated_images_are_smoothed_when_enlarged`, `raster::tests::small_non_interpolated_images_keep_hard_pixels_when_enlarged`, `raster::tests::a_scan_shrunk_one_way_is_smoothed_the_way_it_is_stretched`, `raster::tests::a_scan_drawn_a_rounding_error_above_its_resolution_stays_sharp` | | 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; (13) a bitmap filled with a default pixel of 1 (a page, or an intermediate text or halftone region) had 1s past its width in the last word of each row, and a refinement, which reads its reference a word at a time, read them as pixels where the standard has 0 (6.3.5.2): refining such a reference not a multiple of 32 wide decoded to other pixels than were coded, release builds too. `Bitmap::clear_padding` clears those bits when a bitmap is filled with 1s, as cutting symbols and patterns from a collective bitmap already did. Ghostscript reads the test's page and text streams as they decode now; (14) a region's location is unsigned (5.4.4), so one past `i32::MAX` is off the page, where the decoder took it for a signed one: a region there was drawn back onto the page from the left or the top (at 0xFFFFFFFC, four pixels left of it), and a refinement there, of the page or of a region at the same place (7.4.7.5), was refused as an overflow, and the page with it. `Bitmap::combine_region` keeps such a location past the page. A refinement of another region refines it where it is, as Table 35 has it (GRREFERENCEDX and GRREFERENCEDY 0), whatever the two locations say; a refinement of the page, see (16); (15) a refinement whose reference is offset right of it (GRREFERENCEDX above 0, which a refined text instance has from a wider bitmap, Table 12, and a refined symbol from an RDX above 0, Table 18) read the reference's first 32 columns as 0: `RefinementContextGatherer::maybe_reload_buffers` emptied its buffers left of the reference and didn't load them again at its first column. Such a refinement decoded to other pixels than were coded, or failed and lost the page. Left of the reference the buffers now keep the words of its first column each row starts with; with an offset of 0 or less nothing changes; (16) a refinement of the page refines "the current contents of the page buffer, restricted to the area of the page buffer specified by this segment's region segment information field" (7.4.7.4), at an offset of 0 (Table 35): `Bitmap::area` copies that area a word at a time, 0 off the page and past its width whatever the padding of its rows holds, and every refinement region reads its reference at offset 0. The page around the area was read too, so a refinement of the page decoded to other pixels than were coded, and one wider or taller than the page was refused. And `RefinementContextGatherer::start_row` loads the reference's words a column left of where a row starts in it, as the reload would, not from column 0: with the reference offset 32 or more left of the refinement (GRREFERENCEDX below −31, as a refined text instance has from a narrower bitmap or an RDX below −31, and a refined symbol from such an RDX), each row's first context read 0 for the reference column it starts at, and from 33 left for the one before it too, and the refinement decoded to other pixels or failed | (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`; (11) `jbig2::decode::tests::a_run_of_images_spends_its_allowance_down_to_the_step` (pdfcraft-filters); (12) `raster::tests::jbig2_placements_at_the_end_of_an_i32_draw_nothing`; (13) `jbig2::decode::tests::a_refinement_reads_nothing_past_its_references_width` (pdfcraft-filters), and in the crate `tests::a_bitmap_of_1s_has_nothing_past_its_width`; (14) `jbig2::decode::tests::a_region_located_past_i32_max_is_off_the_page` (pdfcraft-filters), and in the crate `tests::a_region_located_past_i32_max_is_off_the_page` and `tests::a_refinement_of_a_region_decodes_the_same_wherever_they_are`; (15) `jbig2::decode::tests::a_refinement_reads_a_reference_right_of_it_from_its_first_column` (pdfcraft-filters), and in the crate `tests::a_refinement_reads_a_reference_right_of_it_from_its_first_column`; (16) `jbig2::decode::tests::a_refinement_refines_the_page_in_its_area_and_any_reference_where_it_is` (pdfcraft-filters), and in the crate `tests::a_refinement_refines_the_page_in_its_area_and_any_reference_where_it_is` and `tests::a_page_area_is_the_page_inside_it_and_0_outside` | | 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_passwords` in `crypto/mod.rs`, which resolves the upstream `TODO: Do the above.`; `stringprep`, already in the tree through pdfcraft-crypt, is an optional dependency enabled by `std`, and without `std` passwords are used as given): the renderer used the password as typed, so one that SASLprep changes (a decomposed `café`, a no-break space, the fullwidth `ａｂ` a Japanese IME types in full-width mode) was wrong, although pdfcraft-cos accepted it: protecting a document with such an open password was rolled back ("the password is incorrect"), and a file with such a password did not open with it.; (17) CCITT `/Rows 0` means the height is not predetermined (ISO 32000-2 Table 11), not zero rows: scanned Group 4 pages written that way rendered blank; (18) The streaming Flate decoder (`lzw_flate::flate::fallback`) copies a back-reference in runs of up to its distance (`extend_from_within`), not byte by byte. The copy reads only bytes already in the output, and a copy whose source does not exist still writes nothing, so the bytes are those a byte-by-byte copy writes. Inflating the A1 sample's content stream (278.6 MB) takes about 14% less time (0.66 s to 0.56 s); the output is identical; (19) CCITT `/Rows` below the image's `/Height` decodes to the height: a plotter driver wrote `/Rows` equal to `/Columns`, so only the first rows were decoded and the rest drawn black (#3308) (20) `stream` followed by spaces or tabs and then the line break (`stream \r\n`, as Microsoft Reporting Services writes it) starts the data after the line break; the content stream was taken to start at the space and the page rendered blank.; (21) a JP2 colour specification naming an enumerated colour space the decoder refuses (CIEJab in a veraPDF fixture, YCbCr, PhotoYCC, YPbPr, bilevel…) no longer fails the image: it is decoded again from a copy whose `colr` names the space the decoder assumes without one, greyscale, sRGB or CMYK by component count, as MuPDF and Poppler draw it (`without_refused_enumerated_colour`; the page drew blank, #3306); (22) an object stream's index of `objnum offset` pairs is parsed once per stream and kept beside its decoded data (`ObjectStreamIndexes`, `XRef::with_object_stream_indexes`), and an offset past `usize` is malformed instead of wrapping: `XRef::get_with` parsed the whole index again for every object it fetched from the stream, so reading all of a stream of N objects took N² steps (a Firefox page whose 65,542 objects share one stream took 37 s to draw and 40 s to extract its text); (23) the inline-image end-of-data look-ahead no longer rejects a candidate `EI` because a byte above 0x7F lies between it and the next one: when the candidate is the end, those bytes are the content stream that follows the image, and its strings carry such bytes (WinAnsi `(Grüße) Tj`, the two-byte codes of any Type 0 font), so the operator check that follows decides alone, bounded as in (4) (a scanned page whose OCR text held an accented word and a later capital "EI" — EINGANG, BEI, KLEIN — drew and extracted nothing after the image: 40 of 275 pages across five real documents); (24) `xref::prev_offset`: a trailer `/Prev` written as a reference `N G R` is read as the offset that object holds when the section's own entries place it (iText's `indirectPrev.pdf`; each section's entries are read before its `Prev` is followed), and otherwise as offset N (tiff2pdf; JHOVE's PDF-HUL-157 files, #2366), as `pdfcraft-cos` and other readers take it, so the earlier sections' pages are seen instead of one; (25) direct arrays and dictionaries share a reader-local nesting cap of 128, counted while skipping and while parsing a dictionary (its lenient repair reads, rather than skips, an object found where a key belongs, so `<< << …` recursed through that read); deeper objects return `None` instead of overflowing the stack (including unused trailer values, trailer entries, content operands and marked-content properties); a chain of 65 direct stitching functions now stops at this cap, before the 64-function guard; (26) an xref stream subsection must be backed by its decoded data: all-zero `/W` widths (an entry reads no bytes) and a declared count longer than the data are refused before any entry is inserted, where a tiny stream with a large `/Size` or `/Index` count inserted that many entries (from storytold/craft-libs#239); a subsection running past object number `i32::MAX` is refused instead of overflowing or turning negative; (27) an xref stream's type field is read big-endian at its full `/W[0]` width, where only its first byte was read and wider fields shifted every later entry, and `/W[0]` over 8 bytes is refused like `/W[1]` (from storytold/craft-libs#241); (28) a content instruction keeps up to 256 operands (`MAX_OPERANDS`) instead of 10, and past that drops the oldest instead of ending the content stream: the eleventh operand of a twelve-ink DeviceN `scn` (ISO 32000-2 §8.6.6.5 allows any number of components) blanked the rest of the page (#4260); (29) a `/JBIG2Globals` stream that is itself JBIG2 isn't decoded: globals are raw segments, and a chain of 16,000 such streams overflowed the stack, which aborts the process; (30) a stream's own crypt filter: `/Crypt` in its chain read as the filter it names, `/Identity` without one, instead of the document's filter and then a filter that can't be decoded, which left such a stream unreadable in any spelling (ISO 32000-2 §7.4.10); (31) `set_page_jbig2_work`: the JBIG2 images decoded on a thread spend one allowance, set per page, 2^30 steps (`MAX_PAGE_JBIG2_WORK`), so a page of many images made to reach their own limit stops at one, and what they cost to read counts, whether they decode or not; unbounded until a page sets it; (32) a group of array items read together (a tuple: an image's `/Decode`, `/W`, `/W2`, and a function's `/Domain`, `/Range`, `/Encode` and `/Decode`) can start with a reference: it was taken for the whole group and lost, which ended the array, so an image with `/Decode [7 0 R 0]` didn't draw and CID widths fell back to `/DW`; (33) an object that holds only a reference stands for what that names, stored on its own or in an object stream, up to 32 references: read as a number it was the reference's own object number, so a `/Decode` or `/W` entry naming such an object took a wrong value, and as anything else it was lost. Past 32, as any cycle goes, it reads as a missing object. An object stream isn't reached that way, as its checks are of the object named; reading its index counts against the 64 nested loads and reads its dictionary alone, as a chain of them, each with a member of the next for its `/N`, `/First` or `/Length`, ran out of stack; and a comment inside a reference counts as white space; (34) a table rebuilt by scanning the file (`fallback_xref_map`) finds an object by its number when its header names another generation, including one no generation may be (`3 18446744073709551616 obj` in `poppler-85140-0.pdf`, #5580): the header is read leniently only in that scan (`read_object_header`), the object is also filed under generation 0 unless a real `N 0 obj` holds that key (a later header for the number moves the stand-in), `XRef::get_with` matches by object number only in a repaired table, and `XRef::objects` yields each location once; (35) PNG predictors (`lzw_flate::unfilter_png`) are undone a byte row at a time on slices, as the PNG filters are defined, instead of through the bit-level readers TIFF predictors need: a 993 × 936 RGB image took about 63 ms to unfilter whatever its row filters, and now takes about 3 ms (Up) to 16 ms (Paeth) in a release build (#5430); a row ending part-way through a pixel (three 4-bit samples, an odd width) keeps its last byte, where the bit-level readers dropped it and each later row started a byte early, shearing the picture | (1) `cos/tests/writer.rs::page_tree_loops_in_object_streams_do_not_crash_readers`; (2) `raster::tests::absurd_mask_and_fax_sizes_are_refused_before_decoding`; (3) `raster::tests::absurd_predictor_columns_and_line_widths_render`; (4) `raster::tests::inline_images_with_ei_in_their_data_parse_in_linear_time`, `content::tests::inline_image_ei_may_be_followed_by_a_delimiter`; (5) `raster::tests::indirect_appearance_state_dict_is_not_another_stream`; (6) `raster::tests::jpx_stream_bit_depth_boundaries_do_not_panic`, `raster::tests::jpx_32_bit_image_does_not_prevent_other_page_content_from_rendering`, `raster::tests::jpx_gray_rgb_and_alpha_keep_eight_bit_samples`, `raster::tests::jpx_streams_reject_out_of_range_bit_depths`; (7) `raster::tests::decompression_bombs_stop_at_the_stream_limit`, `raster::tests::page_contents_and_nested_forms_share_one_budget`; (8) `inspect::tests::jspdf_high_precision_mediabox_does_not_overflow_dimensions`, `object::number::tests::high_precision_float_does_not_overflow_mantissa` (hayro-syntax); (9) `object::dict::lazy_index_tests`, `raster::tests::large_resource_indexes_preserve_pixels_and_text`; (10) `data::tests`, `xref::tests::concurrent_cold_object_stream_reads_all_succeed`, `raster::tests::shared_parser_is_lazy_and_initializes_once_under_contention`; (11) `data::tests::panicking_initializer_wakes_waiters_and_prevents_retries`, `xref::tests::concurrent_cold_object_stream_reads_all_succeed`; (12) `util::tests::segment_list_rejects_out_of_range_indices`; (13) `raster::tests::user_unit_scales_the_page_and_everything_on_it`; (14) `crypto::pdfcraft_password_tests`, `pdfcraft-engine` `tests::legacy_passwords_with_turkish_characters_open`; (15) `content::tests::windows_keep_every_instruction_at_every_split`, `content::tests::windows_keep_every_instruction_in_pieces_of_any_size`, `content::tests::windows_match_the_slice_on_random_content`, `content::tests::windows_drop_blanks_and_keep_a_comment_that_may_go_on`, `page::content_reader_tests::content_past_the_cap_is_read_to_the_end`, `page::content_reader_tests::a_stream_that_expands_past_the_budget_is_cut_there`, `page::content_reader_tests::a_small_expansion_under_the_cap_is_not_cut`, `raster::tests::page_content_past_the_old_cap_renders`, `raster::tests::content_reader_returns_the_same_bytes_in_any_pieces`, `raster::tests::content_bombs_stop_at_the_budget_and_say_so`, `page::content_reader_tests::plain_and_whole_pieces_keep_their_bytes`, `page::content_reader_tests::repeated_content_reference_counts_compressed_bytes_once`, `page::content_reader_tests::whole_cap_still_counts_later_unique_streams`; (16) `crypto::pdfcraft_password_tests::aes256_passwords_are_tried_in_their_saslprep_form_then_as_given`, `crypto::pdfcraft_password_tests::aes256_passwords_open_as_typed_and_in_their_saslprep_form`, `pdfcraft-engine` `tests::aes256_open_passwords_that_saslprep_changes_protect_and_reopen_as_typed`, `tests::aes256_files_whose_passwords_saslprep_changes_open_as_typed`; (17) `raster::tests::group4_fax_with_rows_zero_decodes_to_the_image_height`; (18) `filter::lzw_flate::tests::back_reference_longer_than_its_distance_repeats_its_source`, `back_reference_cut_by_the_limit_resumes_in_the_next_call` and `back_reference_reads_bytes_written_by_an_earlier_call`; each fails when the copy returns the wrong remaining count; (19) `raster::tests::group4_fax_with_rows_below_its_height_decodes_to_the_image_height`; (20) `raster::tests::space_before_the_line_break_after_stream_still_renders`; (21) `raster::tests::a_jp2_image_in_a_refused_enumerated_colour_space_is_drawn`, `filter::jpx::tests::refused_enumerated_colour_spaces_become_the_component_counts_default`, `filter::jpx::tests::what_the_decoder_handles_or_cant_be_read_is_left_alone` (`cargo test --manifest-path vendor/hayro-syntax/Cargo.toml`); (22) `raster::tests::objects_in_one_large_object_stream_are_read_in_linear_time`; (23) `text::tests::text_after_an_inline_image_survives_a_non_ascii_byte_before_a_later_ei`, `content::tests::an_inline_image_ends_at_its_own_ei_when_later_text_holds_a_non_ascii_byte` (hayro-syntax); (24) `xref::tests::prev_as_reference_reaches_the_earlier_section`, `xref::tests::prev_as_reference_to_an_object_holding_the_offset_reads_that_offset`, `pdfcraft-cos` `document::tests::a_prev_written_as_a_reference_still_reaches_the_earlier_sections`, `document::tests::a_prev_reference_to_an_object_holding_the_offset_reads_that_offset`; (25) `raster::tests::deeply_nested_direct_objects_do_not_overflow_the_stack`, `raster::tests::dictionaries_nested_without_keys_in_content_do_not_overflow_the_stack`, `raster::tests::dictionaries_nested_without_keys_in_the_trailer_do_not_overflow_the_stack`, `raster::tests::dictionaries_nested_without_keys_as_marked_content_properties_do_not_overflow_the_stack`, `raster::tests::dictionaries_nested_without_keys_are_read_up_to_the_nesting_cap`, `raster::tests::nesting_at_the_cap_fits_a_512_kib_stack` (native only), `raster::tests::function_limits_depth_guard_refuses_a_chain_of_references`; (26) `xref::tests::xref_stream_entries_need_backing_bytes`; (27) `xref::tests::xref_stream_type_field_uses_the_full_w0_width`, `xref::tests::xref_stream_loader_rejects_oversized_w0` (`cargo test --manifest-path vendor/hayro-syntax/Cargo.toml`); (28) `content::ops::tests::scn_with_twelve_components_keeps_them_all`, `content::ops::tests::operands_past_the_cap_keep_the_latest_in_order` (hayro-syntax), `raster::tests::devicen_inks::a_twelve_ink_devicen_colour_paints`, `raster::tests::devicen_inks::a_four_ink_devicen_colour_paints`, `raster::tests::devicen_inks::an_operator_with_hundreds_of_operands_does_not_end_the_content`; (29) `raster::tests::jbig2_globals_that_are_jbig2_themselves_are_not_decoded`, `raster::tests::jbig2_globals_are_read`, `raster::tests::the_jbig2_images_of_a_page_share_one_allowance`, `raster::tests::an_image_that_fails_spends_the_pages_allowance_too`; (30) `raster::tests::a_stream_with_a_crypt_filter_of_its_own_is_drawn`; (32) `raster::tests::a_decode_array_with_references_in_it_is_read`, `raster::tests::a_functions_arrays_can_start_with_a_reference` (pdfcraft-render), `font::cid::width_tests::a_width_group_can_start_with_a_reference` (hayro-interpret); (33) `raster::tests::a_reference_to_a_reference_is_what_it_names`, `raster::tests::a_decode_array_with_references_in_it_is_read` (pdfcraft-render), `font::cid::width_tests::a_width_group_can_start_with_a_reference` (hayro-interpret), `xref::tests::an_object_stream_isnt_reached_through_a_reference`, `xref::tests::a_chain_of_object_streams_ends_at_the_limit_of_nested_loads`, `xref::tests::a_chain_of_object_streams_takes_no_more_work_for_two_ways_on`, `xref::tests::a_reference_held_in_either_storage_is_followed`, `xref::tests::a_page_a_reference_is_held_to_is_found_once` (hayro-syntax); (34) `raster::tests::a_rebuilt_table_finds_an_object_whose_header_names_another_generation`, `raster::tests::a_rebuilt_tables_generation_0_stand_in_is_not_a_second_object` (pdfcraft-render), `xref::tests::an_object_stream_and_its_first_member_are_both_enumerated` (hayro-syntax); (35) `filter::lzw_flate::pdfcraft_png_predictor_tests` (hayro-syntax: every filter type at 1–16 bits and 1–4 colours against the PNG specification's byte-by-byte definition, a row ending part-way through a pixel, Average and Paeth over 6-byte pixels), `raster::tests::png_predicted_images::a_four_bit_rgb_image_of_odd_width_with_up_predicted_rows_is_not_sheared` | diff --git a/vendor/hayro/src/renderer.rs b/vendor/hayro/src/renderer.rs index 0b9cfe779e..1e46e3b79e 100644 --- a/vendor/hayro/src/renderer.rs +++ b/vendor/hayro/src/renderer.rs @@ -193,6 +193,67 @@ fn image_buffer(len: usize) -> Option> { Some(out) } +/// PdfCraft patch: `mask` sampled onto the grid of a `width` × `height` image (no side smaller +/// than the mask's). Each image sample takes the mask sample under its centre or, for a mask that +/// asks for `/Interpolate`, the bilinear blend of the four around it, so the mask's edges stay as +/// smooth as when it was enlarged on its own. `None` for an empty side, an image side smaller than +/// the mask's, or a buffer over the limits. +fn mask_on_image_grid(mask: &LumaData, width: u32, height: u32) -> Option> { + if width == 0 || height == 0 || mask.width == 0 || mask.height == 0 { + return None; + } + let mut out = image_buffer(source_byte_len(width, height, 1, MAX_IMAGE_PIXELS)?)?; + let columns = grid_axis(mask.width, width, mask.interpolate)?; + let rows = grid_axis(mask.height, height, mask.interpolate)?; + let mask_width = mask.width as usize; + let row = |y: usize| { + let start = y.checked_mul(mask_width)?; + mask.data.get(start..start.checked_add(mask_width)?) + }; + let one = u64::from(GRID_ONE); + for &(y0, y1, fy) in &rows { + let (top, bottom, fy) = (row(y0)?, row(y1)?, u64::from(fy)); + for &(x0, x1, fx) in &columns { + let fx = u64::from(fx); + let blend = |r: &[u8]| Some(u64::from(*r.get(x0)?) * (one - fx) + u64::from(*r.get(x1)?) * fx); + // At most 255 × GRID_ONE², so the sum fits in a u64 and the result in a u8. + let v = (blend(top)? * (one - fy) + blend(bottom)? * fy + one * one / 2) / (one * one); + out.push(v.min(255) as u8); + } + } + Some(out) +} + +/// PdfCraft patch: fixed-point unit of [`grid_axis`]'s blend weights. +const GRID_ONE: u32 = 1 << 16; + +/// PdfCraft patch: for each of `dst` samples (`dst >= src`) taken from `src` along one axis, the +/// two source samples it lies between and the weight of the second, in [`GRID_ONE`] units. +/// Centres line up: target `t` lies at `(t + ½)·src/dst − ½` source samples, held to the ends. +/// Without `smooth` both are the nearest sample, so the weight does not matter. +fn grid_axis(src: u32, dst: u32, smooth: bool) -> Option> { + if src == 0 || dst < src { + return None; + } + let mut out = Vec::new(); + out.try_reserve_exact(dst as usize).ok()?; + let last = u64::from(src - 1) << 16; + for t in 0..dst { + if !smooth { + let s = nearest_source_index(src, dst, t); + out.push((s, s, 0)); + continue; + } + // In u128, as `(2t + 1)·src·2^16` can pass 2^64 for a long side; the quotient is at + // most `src·2^16`, so it fits back in a u64. + let centre = ((2 * u128::from(t) + 1) * u128::from(src) << 16) / (2 * u128::from(dst)); + let at = u64::try_from(centre).ok()?.saturating_sub(u64::from(GRID_ONE / 2)).min(last); + let s = (at >> 16) as usize; + out.push((s, (s + 1).min(src as usize - 1), (at & 0xFFFF) as u32)); + } + Some(out) +} + /// PdfCraft patch (#624): the source index that PicScale's nearest-neighbour plan copies for /// destination `index` of `dst` samples taken from `src`. It is the same fixed-point centre /// sampling as `ResampleNearestPlan`, so these are the samples a nearest resize would copy. @@ -236,6 +297,255 @@ fn nearest_premultiplied_rgba( Some(out) } +/// PdfCraft patch: the shorter side, in samples, from which an image that did not ask for +/// `/Interpolate` is still smoothed when enlarged. A page scanned at 72 dpi is 612 samples wide; +/// icons, QR codes and pixel art are usually far smaller and keep their hard pixels. +const SMOOTHED_IMAGE_SIDE: u32 = 512; + +/// PdfCraft patch: the scale above which such an image counts as enlarged. A scan drawn at its own +/// resolution can come out a few ulps above 1 (the scale is an `f32`: a 192 dpi scan on a 2× +/// display at 100%), and sampled bilinearly at a half-pixel offset it blurred by half a pixel. +/// Below 1% nearest-neighbour repeats at most one sample in a hundred. +const SMOOTHED_IMAGE_SCALE: f32 = 1.01; + +/// PdfCraft patch: fixed-point unit of [`area_shrink`]'s weights. +const AREA_ONE: u64 = 1 << 16; + +/// PdfCraft patch: the source samples target sample `t` covers when `src` samples shrink to `dst` +/// (`dst <= src`), with their weights in [`AREA_ONE`] units. Target `t` spans `[t·src, (t+1)·src)` +/// and source `s` spans `[s·dst, (s+1)·dst)`, both in units of 1/(src·dst) of the axis, so every +/// overlap is exact. The weights are differences of one rounded running sum, so they add up to +/// exactly [`AREA_ONE`] and a flat area keeps its exact value. +fn area_taps(src: u32, dst: u32, t: u32) -> impl Iterator { + let (src, dst, t) = (u64::from(src), u64::from(dst), u64::from(t)); + // `t < dst <= src <= 2^28` (MAX_IMAGE_PIXELS), so every product below fits in a u64. + let (lo, hi) = (t * src, (t + 1) * src); + let fixed = move |x: u64| (x * AREA_ONE + src / 2) / src; + (lo / dst..hi.div_ceil(dst)).map(move |s| { + let from = (s * dst).max(lo) - lo; + let to = ((s + 1) * dst).min(hi) - lo; + (s as usize, fixed(to) - fixed(from)) + }) +} + +/// PdfCraft patch: [`area_taps`] of every target sample of one axis, worked out once: for each +/// target, the first source sample it covers and the weights of the samples from there on (the +/// samples a target covers are consecutive). There are at most `src + dst` taps in all. +struct AreaAxis { + first: Vec, + /// Where each target's weights start in `weights`, and where the last one's end. + offsets: Vec, + weights: Vec, +} + +impl AreaAxis { + fn new(src: u32, dst: u32) -> Option { + let n = dst as usize; + let (mut first, mut offsets, mut weights) = (Vec::new(), Vec::new(), Vec::new()); + first.try_reserve_exact(n).ok()?; + offsets.try_reserve_exact(n.checked_add(1)?).ok()?; + weights.try_reserve_exact((src as usize).checked_add(n.checked_mul(2)?)?).ok()?; + offsets.push(0); + for t in 0..dst { + let mut start = None; + for (s, w) in area_taps(src, dst, t) { + start.get_or_insert(s); + // A weight is at most AREA_ONE, so it fits in a u32. + weights.push(w as u32); + } + first.push(start?); + offsets.push(weights.len()); + } + Some(Self { first, offsets, weights }) + } + + /// Target `t`'s first source sample and its weights. + fn get(&self, t: usize) -> Option<(usize, &[u32])> { + let (lo, hi) = (*self.offsets.get(t)?, *self.offsets.get(t.checked_add(1)?)?); + Some((*self.first.get(t)?, self.weights.get(lo..hi)?)) + } +} + +/// PdfCraft patch: shrinks `N`-channel samples from `source` to `target` (no side larger than its +/// source side) by area averaging: each target sample is the mean of the source it covers, +/// weighted by how much of it it covers. Nothing is invented between samples, which is what +/// `/Interpolate` governs (ISO 32000-2 §8.9.5.3, for images drawn above their resolution), and +/// nothing is dropped: nearest-neighbour keeps one sample in N per axis, so a scan shown at screen +/// size lost the strokes of its text. Premultiplied input stays premultiplied. `None` if nothing +/// shrinks, the sizes don't fit or a buffer cannot be allocated. +fn area_shrink(data: &[u8], source: (u32, u32), target: (u32, u32)) -> Option> { + if data.len() != source_byte_len(source.0, source.1, N, MAX_IMAGE_PIXELS)? { + return None; + } + let row_len = (source.0 as usize).checked_mul(N)?; + area_shrink_rows::(source, target, |s, row| { + let start = s.checked_mul(row_len)?; + row.copy_from_slice(data.get(start..start.checked_add(row_len)?)?); + Some(()) + }) +} + +/// PdfCraft patch: [`area_shrink`] of a colour image under its same-size soft mask (a stencil's +/// ink under the stencil, or an image under a mask sampled onto its grid), premultiplied a source +/// row at a time as it is averaged. The result equals premultiplying the expanded RGBA and +/// shrinking that, without building it: a 10718 × 13701 stencil shown at page size peaked at +/// 1.16 GB with the full-resolution RGBA, which is 587 MB. +fn area_shrink_premultiplied( + color: &ImageData, + alpha: &[u8], + source: (u32, u32), + target: (u32, u32), +) -> Option> { + let (samples, channels) = match color { + ImageData::Luma(luma) => (luma.data.as_slice(), 1), + ImageData::Rgb(rgb) => (rgb.data.as_slice(), 3), + }; + if samples.len() != source_byte_len(source.0, source.1, channels, MAX_IMAGE_PIXELS)? + || alpha.len() != source_byte_len(source.0, source.1, 1, MAX_IMAGE_PIXELS)? + { + return None; + } + let width = source.0 as usize; + area_shrink_rows::<4>(source, target, |s, row| { + let start = s.checked_mul(width)?; + let end = start.checked_add(width)?; + let colors = samples.get(start.checked_mul(channels)?..end.checked_mul(channels)?)?; + let alphas = alpha.get(start..end)?; + for ((out, px), a) in row.chunks_exact_mut(4).zip(colors.chunks_exact(channels)).zip(alphas) { + let [r, g, b] = match px { + [v] => [*v; 3], + [r, g, b] => [*r, *g, *b], + _ => return None, + }; + out.copy_from_slice(&AlphaColor::from_rgba8(r, g, b, *a).premultiply().to_rgba8().to_u8_array()); + } + Some(()) + }) +} + +/// PdfCraft patch: the area shrink behind [`area_shrink`], reading source row `s` through +/// `read_row(s, row)` into a buffer of one row. Each target row averages the source rows it +/// covers, and that row is shrunk across at once with taps worked out once per image, so work +/// memory is one source row of samples, one of `u32` accumulators, one shrunk row and the +/// horizontal taps (at most `sw + 2 × tw`): never the image. A side over +/// `MAX_RESAMPLING_SOURCE_SIDE` is refused, so none of it exceeds a few times one source row. +fn area_shrink_rows( + source: (u32, u32), + target: (u32, u32), + mut read_row: impl FnMut(usize, &mut [u8]) -> Option<()>, +) -> Option> { + let ((sw, sh), (tw, th)) = (source, target); + // A side over MAX_RESAMPLING_SOURCE_SIDE would make the accumulator row and the tap table + // several times larger than the samples (a 2^27 × 2 image of zeros is a tiny file). + if tw == 0 + || th == 0 + || tw > sw + || th > sh + || (tw, th) == (sw, sh) + || sw > MAX_RESAMPLING_SOURCE_SIDE + || sh > MAX_RESAMPLING_SOURCE_SIDE + { + return None; + } + let row_len = (sw as usize).checked_mul(N)?; + let zeroed = |len: usize| { + let mut v = image_buffer(len)?; + v.resize(len, 0); + Some(v) + }; + let (mut source_row, mut shrunk_row) = (zeroed(row_len)?, image_buffer(row_len)?); + let mut acc: Vec = Vec::new(); + acc.try_reserve_exact(row_len).ok()?; + acc.resize(row_len, 0); + // The taps depend only on the widths, so they are worked out once, not again for each row. + let taps = if tw == sw { None } else { Some(AreaAxis::new(sw, tw)?) }; + let out_len = match taps { + None => source_byte_len(tw, th, N, MAX_IMAGE_PIXELS)?, + Some(_) => image_byte_len(tw, th, N, MAX_IMAGE_PIXELS)?, + }; + let mut out = image_buffer(out_len)?; + // Weights sum to AREA_ONE (2^16) and samples are at most 255, so every sum fits in a u32. + let round = |acc: u32| ((acc + (AREA_ONE as u32) / 2) >> 16).min(255) as u8; + // The last source row read: neighbouring target rows share the row on their boundary. + let mut loaded = None; + for t in 0..th { + // Vertical: the whole source rows target row `t` covers, averaged. + let row: &[u8] = if th == sh { + read_row(t as usize, &mut source_row)?; + &source_row + } else { + acc.fill(0); + for (s, w) in area_taps(sh, th, t) { + if loaded != Some(s) { + read_row(s, &mut source_row)?; + loaded = Some(s); + } + // A weight is at most AREA_ONE, so it fits in a u32. + let w = w as u32; + for (a, v) in acc.iter_mut().zip(&source_row) { + *a += w * u32::from(*v); + } + } + shrunk_row.clear(); + shrunk_row.extend(acc.iter().map(|a| round(*a))); + &shrunk_row + }; + // Horizontal: that row shrunk to `tw` samples. + let Some(taps) = &taps else { + out.extend_from_slice(row); + continue; + }; + for x in 0..tw as usize { + let (first, weights) = taps.get(x)?; + let samples = row.get(first * N..(first + weights.len()) * N)?; + let mut acc = [0_u32; N]; + for (w, px) in weights.iter().zip(samples.chunks_exact(N)) { + for (a, v) in acc.iter_mut().zip(px) { + *a += w * u32::from(*v); + } + } + out.extend(acc.map(round)); + } + } + Some(out) +} + +/// PdfCraft patch: [`area_shrink`] for an image's pixel format. If it cannot run, the image keeps +/// its source resolution, as when [`Renderer::resize_image_data`] fails. +fn area_resize( + data: Vec, + source: (u32, u32), + target: (u32, u32), + format: ImagePixelFormat, +) -> (Vec, u32, u32) { + let shrunk = match format { + ImagePixelFormat::Luma => area_shrink::<1>(&data, source, target), + ImagePixelFormat::Rgb => area_shrink::<3>(&data, source, target), + ImagePixelFormat::Rgba => area_shrink::<4>(&data, source, target), + }; + match shrunk { + Some(out) => (out, target.0, target.1), + None => { + log::warn!( + "Image area shrink failed or exceeded its buffer limit; falling back to source data" + ); + (data, source.0, source.1) + } + } +} + +/// PdfCraft patch: premultiplies straight RGBA samples in place. +fn premultiply_rgba(rgba: &mut [u8]) { + let (chunks, _) = rgba.as_chunks_mut::<4>(); + for chunk in chunks { + let [r, g, b, a] = *chunk; + *chunk = AlphaColor::from_rgba8(r, g, b, a) + .premultiply() + .to_rgba8() + .to_u8_array(); + } +} + /// PdfCraft patch: most pieces one dashed stroke may be cut into. Stroke expansion keeps every /// dash, so a pattern that is tiny next to its path asks for billions: a fuzzed `/D [[11] 0]` on /// a line from x = 92234775807 allocated over 5 GB, and `[0.000001] 0 d` on a 20 pt line over @@ -423,9 +733,12 @@ fn snap_minified_image( planned: (u32, u32), ) -> Option<(u32, u32, Affine)> { let [a, b, c, d, e, f] = transform.as_coeffs(); + // Halves round up, not away from zero: a tile's origin moves edges by whole pixels, and only + // then do they snap where the whole page snaps them (-2.5 must go to -2 as 2.5 goes to 3). + let snap = |v: f64| (v + 0.5).floor(); // The device span of a source edge from `origin` across `side` samples of `scale`, rounded. let edge = |origin: f64, scale: f64, side: u32| { - let (p, q) = (origin.round(), (origin + scale * f64::from(side)).round()); + let (p, q) = (snap(origin), snap(origin + scale * f64::from(side))); let (lo, hi) = (p.min(q), p.max(q)); (hi - lo >= 1.0 && lo.abs() < 1e6 && hi.abs() < 1e6).then_some((lo, hi)) }; @@ -935,17 +1248,51 @@ impl Renderer { return; } let interpolate = image_data.interpolate(); - if let Some(a) = &alpha_data { + let requested_resize = x_scale < 1.0 || y_scale < 1.0; + // PdfCraft patch: an image that did not ask for interpolation and is drawn below its + // resolution is shrunk by area averaging (`area_shrink`), not nearest-neighbour, which + // broke up the text of scans shown at screen size. Type 3 glyph images keep the nearest + // shrink of #624: they are drawn with a frame of their own and never snapped. + let area = !interpolate + && requested_resize + && !self.in_type3_glyph + // Wider sources keep #624's bounded nearest-neighbour shrink (`area_shrink` refuses + // them): skipping them, as a refused shrink at source size would, loses the image. + && source_width.max(source_height) <= MAX_RESAMPLING_SOURCE_SIDE; + let mut alpha_data = alpha_data; + if let Some(a) = alpha_data.take() { if source_byte_len(a.width, a.height, 1, MAX_IMAGE_PIXELS) != Some(a.data.len()) { log::warn!("Image has invalid or over-limit alpha dimensions/data and was skipped"); return; } if a.width != source_width || a.height != source_height || a.interpolate != interpolate { - if let Some(alpha) = alpha_data { - self.draw_image_with_alpha_mask(image_data, alpha); + // PdfCraft patch: a soft mask no larger than its area-shrunk image is sampled onto + // the image's grid, so colour and mask are averaged together, premultiplied. Drawn + // as a mask layer over the image instead, the averaged colour (transparent samples + // included) was multiplied by the averaged mask, and the colour of transparent + // samples showed at the mask's edges. A larger mask carries detail the image does + // not (#1649) and keeps the layer. + let on_grid = (area && a.width <= source_width && a.height <= source_height) + .then(|| mask_on_image_grid(&a, source_width, source_height)) + .flatten(); + match on_grid { + Some(data) => { + alpha_data = Some(LumaData { + data, + width: source_width, + height: source_height, + interpolate, + ..a + }); + } + None => { + self.draw_image_with_alpha_mask(image_data, a); + return; + } } - return; + } else { + alpha_data = Some(a); } } let mut quality = if interpolate { @@ -955,7 +1302,6 @@ impl Renderer { }; let has_alpha = alpha_data.is_some(); let mut may_have_transparency = has_alpha || self.force_images_may_have_transparency(); - let requested_resize = x_scale < 1.0 || y_scale < 1.0; let (new_width, new_height) = match image_resampling_size( source_width, source_height, @@ -985,12 +1331,28 @@ impl Renderer { if requested_resize && self.in_type3_glyph && interpolate { quality = ImageQuality::High; } - // PdfCraft patch: an interpolated image that is shrunk, with its axes along the device's - // (upright, mirrored or a quarter turn), lands on whole device pixels. It is shrunk once to - // exactly its snapped device rectangle and drawn 1:1 from the snapped corner. Otherwise the - // shrunk raster (rounded up to a whole size) is sampled a second time at a fractional scale - // and offset, which blurs scans and photos. - let snapped = if interpolate + // PdfCraft patch: an image the size of a page scan is smoothed when drawn above its + // resolution, even without `/Interpolate`: bilinearly, as images that ask for it are. + // Scans rarely ask, and drawn nearest-neighbour each sample of a zoomed scan became a + // block, so the edges of its text were staircases. Bicubic looked the same on a zoomed + // 300 dpi scan and drew its tiles five times slower. Smaller images (icons, QR codes, + // pixel art) keep their hard pixels. A scan shrunk on one side and stretched on the other + // is smoothed too: the area shrink leaves the shrunk side at 1:1, where bilinear adds + // nothing, and the stretched side would otherwise be drawn nearest-neighbour. + if !interpolate + && !self.in_type3_glyph + && (x_scale > SMOOTHED_IMAGE_SCALE || y_scale > SMOOTHED_IMAGE_SCALE) + && source_width.min(source_height) >= SMOOTHED_IMAGE_SIDE + { + quality = ImageQuality::Medium; + } + // PdfCraft patch: an image that is shrunk (filtered or area-averaged), with its axes along + // the device's (upright, mirrored or a quarter turn), lands on whole device pixels. It is + // shrunk once to exactly its snapped device rectangle and drawn 1:1 from the snapped + // corner. Otherwise the shrunk raster (rounded up to a whole size) is sampled a second time + // at a fractional scale and offset, which blurs scans and photos (or, by the nearest + // sampler of a non-interpolated image, drops samples again). + let snapped = if (interpolate || area) && requested_resize && !self.in_type3_glyph && (new_width, new_height) != (source_width, source_height) @@ -1000,13 +1362,31 @@ impl Renderer { None }; let (new_width, new_height) = snapped.map_or((new_width, new_height), |s| (s.0, s.1)); + // PdfCraft patch: an area shrink never enlarges. A side drawn at or above its resolution + // (an image shrunk across and stretched down, as a 204 × 98 dpi fax is, or a snapped side + // rounded past its source) keeps its samples and is enlarged by the sampler; the snap + // then no longer matches, so the full transform places the image. + let (new_width, new_height) = if area { + (new_width.min(source_width), new_height.min(source_height)) + } else { + (new_width, new_height) + }; let needs_resize = (new_width, new_height) != (source_width, source_height); - // PdfCraft patch (#624): a minified image with a soft mask that is not interpolated is - // nearest-sampled straight into premultiplied target RGBA. That skips the full-resolution - // RGBA copy below and the premultiply pass after it. + // PdfCraft patch (#624): a minified image with a soft mask that is neither interpolated nor + // area-averaged (a Type 3 glyph image, or a source wider than MAX_RESAMPLING_SOURCE_SIDE) + // is nearest-sampled straight into premultiplied target RGBA. That skips the + // full-resolution RGBA copy below and the premultiply pass after it. PdfCraft patch: an + // area-averaged one (a stencil's ink, or an image under its mask) skips them too, and is + // premultiplied a source row at a time as it is averaged (`area_shrink_premultiplied`). let direct = match &alpha_data { - Some(alpha) if needs_resize && !interpolate => nearest_premultiplied_rgba( + Some(alpha) if needs_resize && !interpolate && !area => nearest_premultiplied_rgba( + &image_data, + &alpha.data, + (source_width, source_height), + (new_width, new_height), + ), + Some(alpha) if needs_resize && area => area_shrink_premultiplied( &image_data, &alpha.data, (source_width, source_height), @@ -1014,7 +1394,7 @@ impl Renderer { ), _ => None, }; - let premultiplied = direct.is_some(); + let mut premultiplied = direct.is_some(); // Preserve the single-channel/RGB fast paths. With alpha, expand directly to RGBA; // every source and alpha length was validated before zipping or allocating. let (data, format) = match (direct, image_data, alpha_data) { @@ -1057,6 +1437,15 @@ impl Renderer { }; let (data, mut img_width, mut img_height) = if premultiplied { (data, new_width, new_height) + } else if needs_resize && area { + let mut data = data; + if matches!(format, ImagePixelFormat::Rgba) { + // Averaging straight colours would bleed the colour of transparent samples into + // their neighbours; premultiplied colours count each by its coverage. + premultiply_rgba(&mut data); + premultiplied = true; + } + area_resize(data, (source_width, source_height), (new_width, new_height), format) } else if needs_resize { self.resize_image_data( data, @@ -1103,14 +1492,7 @@ impl Renderer { rgba }; if has_alpha && !premultiplied { - let (chunks, _) = rgba_data.as_chunks_mut::<4>(); - for chunk in chunks { - let [r, g, b, a] = *chunk; - *chunk = AlphaColor::from_rgba8(r, g, b, a) - .premultiply() - .to_rgba8() - .to_u8_array(); - } + premultiply_rgba(&mut rgba_data); } // Type 3 glyphs get a transparent two-pixel frame for interpolation. Check the padded // RGBA allocation and u16 dimensions too, before narrowing or growing either side. @@ -2635,3 +3017,201 @@ mod snap_tests { assert!(snap_minified_image(&turned_noisy, source, (20_000, 200)).is_some()); } } + +/// PdfCraft patch: the area-averaging shrink of non-interpolated images. +#[cfg(test)] +mod area_shrink_tests { + use super::{ + AREA_ONE, AreaAxis, ImageData, LumaData, MAX_IMAGE_PIXELS, RgbData, area_shrink, + area_shrink_premultiplied, area_taps, mask_on_image_grid, premultiply_rgba, + }; + + /// Deterministic samples that vary from one to the next. + fn noise(len: usize, seed: u32) -> Vec { + let mut x = seed.wrapping_mul(2_654_435_761) | 1; + (0..len) + .map(|_| { + x ^= x << 13; + x ^= x >> 17; + x ^= x << 5; + (x >> 24) as u8 + }) + .collect() + } + + /// The shrink as two whole passes, every row averaged down and then every row across, each + /// rounded to a byte: what `area_shrink` did before it worked a row at a time. + fn two_pass_shrink(data: &[u8], (sw, sh): (u32, u32), (tw, th): (u32, u32)) -> Vec { + let round = |acc: u64| ((acc + AREA_ONE / 2) >> 16).min(255) as u8; + let row_len = sw as usize * N; + let tall: Vec = (0..th) + .flat_map(|t| { + (0..row_len).map(move |i| { + round(area_taps(sh, th, t).map(|(s, w)| w * u64::from(data[s * row_len + i])).sum()) + }) + }) + .collect(); + tall.chunks_exact(row_len) + .flat_map(|row| { + (0..tw).flat_map(move |t| { + (0..N).map(move |c| round(area_taps(sw, tw, t).map(|(s, w)| w * u64::from(row[s * N + c])).sum())) + }) + }) + .collect() + } + + /// Working a row at a time gives the same bytes as two whole passes, in both directions, one + /// direction alone, and with source rows shared by neighbouring targets. + #[test] + fn a_row_at_a_time_shrink_equals_two_whole_passes() { + let sizes = [((7, 5), (3, 2)), ((10, 3), (6, 3)), ((4, 9), (4, 4)), ((13, 11), (1, 1)), ((31, 17), (12, 7))]; + for (i, (source, target)) in sizes.into_iter().enumerate() { + let gray = noise((source.0 * source.1) as usize, i as u32); + assert_eq!(area_shrink::<1>(&gray, source, target), Some(two_pass_shrink::<1>(&gray, source, target)), "{source:?} → {target:?}"); + let rgba = noise((source.0 * source.1 * 4) as usize, i as u32 + 99); + assert_eq!(area_shrink::<4>(&rgba, source, target), Some(two_pass_shrink::<4>(&rgba, source, target)), "{source:?} → {target:?}"); + } + } + + /// Premultiplying a row at a time as it is averaged gives the same bytes as premultiplying the + /// whole expanded RGBA first and shrinking that, for gray and RGB colour. + #[test] + fn premultiplying_row_by_row_equals_premultiplying_the_whole_image() { + let sizes = [((7, 5), (3, 2)), ((10, 3), (6, 3)), ((4, 9), (4, 4)), ((31, 17), (12, 7))]; + for (i, (source, target)) in sizes.into_iter().enumerate() { + let (w, h) = source; + let pixels = (w * h) as usize; + let alpha = noise(pixels, i as u32 + 7); + let gray = noise(pixels, i as u32 + 11); + let rgb = noise(pixels * 3, i as u32 + 13); + let colors = [ + ImageData::Luma(LumaData { data: gray.clone(), width: w, height: h, interpolate: false, scale_factors: (1.0, 1.0) }), + ImageData::Rgb(RgbData { data: rgb.clone(), width: w, height: h, interpolate: false, scale_factors: (1.0, 1.0) }), + ]; + for color in &colors { + let mut whole: Vec = match color { + ImageData::Luma(_) => gray.iter().zip(&alpha).flat_map(|(g, a)| [*g, *g, *g, *a]).collect(), + ImageData::Rgb(_) => rgb.chunks_exact(3).zip(&alpha).flat_map(|(c, a)| [c[0], c[1], c[2], *a]).collect(), + }; + premultiply_rgba(&mut whole); + assert_eq!( + area_shrink_premultiplied(color, &alpha, source, target), + area_shrink::<4>(&whole, source, target), + "{source:?} → {target:?}" + ); + } + let short = alpha.get(1..).unwrap_or_default(); + assert_eq!(area_shrink_premultiplied(&colors[0], short, source, target), None, "alpha shorter than the image"); + } + } + + fn mask(data: &[u8], width: u32, height: u32) -> LumaData { + LumaData { data: data.to_vec(), width, height, interpolate: false, scale_factors: (1.0, 1.0) } + } + + /// Each image sample takes the mask sample under its centre, at whole and fractional ratios. + #[test] + fn a_mask_is_sampled_onto_the_image_grid() { + assert_eq!(mask_on_image_grid(&mask(&[10, 20], 2, 1), 4, 2), Some(vec![10, 10, 20, 20, 10, 10, 20, 20])); + // Centres of 7 samples over 3: 0.5/7·3 = 0.21 → 0, 1.5/7·3 = 0.64 → 0, 2.5/7·3 = 1.07 → 1, … + assert_eq!(mask_on_image_grid(&mask(&[1, 2, 3], 3, 1), 7, 1), Some(vec![1, 1, 2, 2, 2, 3, 3])); + } + + /// A mask that asks for `/Interpolate` is blended between the samples around each centre, + /// and held at its ends: the centres of 4 samples over 2 lie at -¼, ¼, ¾ and 1¼. + #[test] + fn an_interpolated_mask_is_blended_onto_the_image_grid() { + let smooth = |data: &[u8], width, height| LumaData { interpolate: true, ..mask(data, width, height) }; + assert_eq!(mask_on_image_grid(&smooth(&[0, 255], 2, 1), 4, 2), Some(vec![0, 64, 191, 255, 0, 64, 191, 255])); + // Centres of 7 samples over 3 lie 3/7 apart, from -2/7 to 2 2/7: a ramp, flat at the ends. + assert_eq!(mask_on_image_grid(&smooth(&[0, 70, 140], 3, 1), 7, 1), Some(vec![0, 10, 40, 70, 100, 130, 140])); + assert_eq!(mask_on_image_grid(&smooth(&[0, 255], 1, 2), 1, 4), Some(vec![0, 64, 191, 255])); + assert_eq!(mask_on_image_grid(&smooth(&[7, 9], 2, 1), 2, 1), Some(vec![7, 9]), "same size is unchanged"); + } + + #[test] + fn malformed_masks_are_not_sampled() { + assert_eq!(mask_on_image_grid(&mask(&[], 0, 1), 4, 1), None, "an empty mask"); + assert_eq!(mask_on_image_grid(&mask(&[1, 2], 2, 1), 0, 1), None, "an empty image"); + assert_eq!(mask_on_image_grid(&mask(&[1], 2, 1), 4, 1), None, "mask data shorter than its size"); + assert_eq!(mask_on_image_grid(&mask(&[1, 2, 3, 4], 1, 4), 1, 2), None, "a mask larger than the image"); + } + + /// Every target sample's weights add up to exactly one, and together the targets take every + /// source sample exactly once (each `dst / src` of it, in fixed point within rounding). + #[test] + fn taps_are_exact_partitions_of_unity() { + for src in 1..=40_u32 { + for dst in 1..=src { + let mut taken = vec![0_u64; src as usize]; + for t in 0..dst { + let taps: Vec<_> = area_taps(src, dst, t).collect(); + assert_eq!(taps.iter().map(|(_, w)| w).sum::(), AREA_ONE, "{src} → {dst}, target {t}"); + assert!(taps.iter().all(|(_, w)| *w > 0), "{src} → {dst}, target {t}: an empty tap"); + for (s, w) in taps { + taken[s] += w; + } + } + // Source s is dst/src of one target in all; scaled by src/dst, that is one. + for (s, w) in taken.iter().enumerate() { + let share = (*w * u64::from(src)) as f64 / (AREA_ONE * u64::from(dst)) as f64; + assert!((share - 1.0).abs() < 1e-3, "{src} → {dst}: source {s} weighs {share}"); + } + } + } + } + + /// The table the horizontal pass reads holds exactly the taps of every target, in order. + #[test] + fn the_tap_table_holds_every_targets_taps() { + for src in 1..=40_u32 { + for dst in 1..=src { + let axis = AreaAxis::new(src, dst).expect("table"); + for t in 0..dst { + let (first, weights) = axis.get(t as usize).expect("target"); + let expected: Vec<(usize, u64)> = area_taps(src, dst, t).collect(); + let got: Vec<(usize, u64)> = weights.iter().enumerate().map(|(k, w)| (first + k, u64::from(*w))).collect(); + assert_eq!(got, expected, "{src} → {dst}, target {t}"); + } + assert_eq!(axis.get(dst as usize), None, "{src} → {dst}: no target past the last"); + } + } + } + + /// The largest side an image may have still computes in u64 without overflowing. + #[test] + fn taps_at_the_pixel_limit_do_not_overflow() { + let side = MAX_IMAGE_PIXELS as u32; + for (dst, t) in [(1, 0), (3, 2), (side - 1, side - 2), (side, side - 1)] { + assert_eq!(area_taps(side, dst, t).map(|(_, w)| w).sum::(), AREA_ONE, "{side} → {dst}, target {t}"); + } + } + + #[test] + fn flat_images_keep_their_exact_value() { + for (source, target) in [((7, 5), (3, 2)), ((10, 3), (6, 3)), ((4, 9), (4, 4)), ((13, 11), (1, 1))] { + let data = vec![77_u8; (source.0 * source.1 * 3) as usize]; + let out = area_shrink::<3>(&data, source, target).expect("shrinks"); + assert_eq!(out.len(), (target.0 * target.1 * 3) as usize); + assert!(out.iter().all(|v| *v == 77), "{source:?} → {target:?}: {out:?}"); + } + } + + /// A side over the resampling limit is refused before any work buffer is allocated. + #[test] + fn sides_over_the_resampling_limit_are_refused() { + let wide = (super::MAX_RESAMPLING_SOURCE_SIDE + 1) as usize; + assert_eq!(area_shrink::<1>(&vec![0_u8; wide], (wide as u32, 1), (10, 1)), None, "too wide"); + assert_eq!(area_shrink::<1>(&vec![0_u8; wide], (1, wide as u32), (1, 10)), None, "too tall"); + } + + #[test] + fn malformed_shrinks_are_refused() { + let data = vec![0_u8; 4 * 4]; + assert_eq!(area_shrink::<1>(&data, (4, 4), (4, 4)), None, "nothing shrinks"); + assert_eq!(area_shrink::<1>(&data, (4, 4), (5, 2)), None, "a side grows"); + assert_eq!(area_shrink::<1>(&data, (4, 4), (0, 2)), None, "an empty side"); + assert_eq!(area_shrink::<1>(&data[1..], (4, 4), (2, 2)), None, "data shorter than the source"); + assert_eq!(area_shrink::<3>(&data, (4, 4), (2, 2)), None, "data of another channel count"); + } +}