Merge #3969: M10: a symbolic TrueType font without an encoding reads its glyph names (#3962)

# Conflicts:
#	vendor/README.md

crates/fonts/src/pdf.rs

@@ -1423,6 +1423,115 @@ pub fn glyph_unicode(name: &str) -> Option<char> {
     chars.next().filter(|_| chars.len() == 0)
 }
 
+/// What each code of an embedded TrueType font that isn't flagged non-symbolic and names no base
+/// encoding shows, by the `post` table name of the glyph its `cmap` selects (the way the renderer
+/// selects it: the code, then at 0xF000, 0xF100 and 0xF200). Such a font's codes need not be
+/// Standard or WinAnsi: FrameMaker files distilled with Distiller 3 number them in the Macintosh
+/// glyph order, 29 below ASCII, so `Solid` read as `6ROLG` and typing set the wrong glyphs (#3962).
+/// Codes without a glyph, and fonts whose `post` table names no glyphs, are left out (the latter
+/// to [`symbolic_mac_roman_texts`]). A code whose glyph has no name with a Unicode value reads
+/// as WinAnsi, as the renderer reads it (vendored hayro-interpret patch (79)), so the editor and
+/// extraction agree: `€` for code 0x80, which Standard leaves without a character.
+fn symbolic_glyph_texts(doc: &Document, font: &Dict) -> Vec<(u32, String)> {
+    use skrifa::GlyphId;
+    use skrifa::raw::{FontData, FontRead as _, TableProvider as _, tables::cmap::PlatformId, tables::post::Post, types::Tag};
+    let Some(descriptor) = dict(doc, font.get(b"FontDescriptor")) else { return Vec::new() };
+    // Flag bit 6, Nonsymbolic (ISO 32000-2 Table 121).
+    if descriptor_flags(doc, &descriptor) & 32 != 0 {
+        return Vec::new();
+    }
+    let Some(object) = descriptor.get(b"FontFile2") else { return Vec::new() };
+    let object = doc.resolve(object);
+    let Object::Stream(stream) = object.as_ref() else { return Vec::new() };
+    let Ok(bytes) = doc.with_direct_parms(stream).decoded_within(MAX_SFNT) else { return Vec::new() };
+    let Ok(face) = skrifa::FontRef::from_index(&bytes, 0) else { return Vec::new() };
+    let Ok(cmap) = face.cmap() else { return Vec::new() };
+    // A `post` version written with its 16-bit halves swapped (`00 00 00 02` for 2.0, as ScanSoft
+    // PDF Create! writes it) is read as the version it means, as the renderer reads it (#2946,
+    // vendored hayro-interpret patch (49)); otherwise its names would be lost here alone.
+    let repaired = face.table_data(Tag::new(b"post")).and_then(|data| match data.as_bytes() {
+        [0, 0, 0, n @ 1..=3, rest @ ..] => Some([&[0, *n, 0, 0], rest].concat()),
+        _ => None,
+    });
+    let post = match &repaired {
+        Some(bytes) => Post::read(FontData::new(bytes)),
+        None => face.post(),
+    };
+    let Ok(post) = post else { return Vec::new() };
+    // A font that names no glyphs (no `post` table of version 1.0, or one of version 2.0 with an
+    // empty name index) is left to `symbolic_mac_roman_texts` and the encoding's reading.
+    let names_glyphs = match post.version() {
+        skrifa::raw::types::Version16Dot16::VERSION_1_0 => true,
+        skrifa::raw::types::Version16Dot16::VERSION_2_0 => post.glyph_name_index().is_some_and(|index| !index.is_empty()),
+        _ => false,
+    };
+    if !names_glyphs {
+        return Vec::new();
+    }
+    let custom = post_custom_names(&post);
+    (0..=255u32)
+        .filter_map(|code| {
+            // The records in order, as the renderer's `map_code` reads a symbolic font: a Windows
+            // symbol or Unicode cmap with the code and the code at 0xF000, 0xF100 and 0xF200, a
+            // Macintosh or Unicode-platform one with the code alone; the first glyph found wins.
+            let glyph = cmap.encoding_records().iter().find_map(|record| {
+                let offsets: &[u32] = match (record.platform_id(), record.encoding_id()) {
+                    (PlatformId::Windows, 0 | 1) => &[0, 0xF000, 0xF100, 0xF200],
+                    (PlatformId::Macintosh | PlatformId::Unicode, 0) => &[0],
+                    _ => return None,
+                };
+                let subtable = record.subtable(cmap.offset_data()).ok()?;
+                offsets.iter().find_map(|o| subtable.map_codepoint(code + o).filter(|g| *g != GlyphId::NOTDEF))
+            })?;
+            let glyph = u16::try_from(glyph.to_u32()).ok()?;
+            let winansi = || {
+                let u = crate::encodings::WIN_ANSI.get(usize::try_from(code).ok()?).copied().filter(|u| *u != 0)?;
+                char::from_u32(u).map(String::from)
+            };
+            let text = post_name(&post, &custom, glyph).filter(|n| uppercase_hex_forms(n)).and_then(glyph_text);
+            Some((code, text.or_else(winansi)?))
+        })
+        .collect()
+}
+
+/// Whether the `uni` and `u` components of a glyph name spell their hex digits in uppercase, as
+/// the glyph list specification's forms do (ISO 32000-2 §9.10.2): `uni1eae` names nothing, so
+/// its glyph's code reads as WinAnsi, as the renderer reads it (vendored hayro-interpret patch
+/// (79)). A component that is a list name (`a`, `ecircumflex`) is read as such whatever its case.
+fn uppercase_hex_forms(name: &str) -> bool {
+    let stem = name.split('.').next().unwrap_or_default();
+    stem.split('_').all(|part| {
+        // Outside the list, a component `glyph_text` reads is `u` or `uni` and hex digits, so a
+        // lowercase `a` to `f` in it is a lowercase digit.
+        crate::encodings::NAMES.binary_search_by(|(n, _)| (*n).cmp(part)).is_ok() || !part.bytes().any(|b| matches!(b, b'a'..=b'f'))
+    })
+}
+
+/// The names a version 2.0 `post` table spells out, in order, read once. `Post::glyph_name` walks
+/// this data from its start for each glyph it is asked about, which over a font program of up to
+/// 16 MiB, asked about 256 codes, is work an untrusted file can make large. Reading stops at the
+/// first name that can't be read, as each lookup would.
+fn post_custom_names<'a>(post: &skrifa::raw::tables::post::Post<'a>) -> Vec<&'a str> {
+    post.string_data().map(|names| names.iter().map_while(|name| name.ok().map(|n| n.as_str())).collect()).unwrap_or_default()
+}
+
+/// `glyph`'s name in `post`, as `Post::glyph_name` gives it, with the table's own names from
+/// [`post_custom_names`].
+fn post_name<'a>(post: &skrifa::raw::tables::post::Post<'a>, custom: &[&'a str], glyph: u16) -> Option<&'a str> {
+    use skrifa::raw::{tables::post::DEFAULT_GLYPH_NAMES, types::Version16Dot16};
+    match post.version() {
+        Version16Dot16::VERSION_1_0 => DEFAULT_GLYPH_NAMES.get(usize::from(glyph)).copied(),
+        Version16Dot16::VERSION_2_0 => {
+            let index = usize::from(post.glyph_name_index()?.get(usize::from(glyph))?.get());
+            match index.checked_sub(DEFAULT_GLYPH_NAMES.len()) {
+                None => DEFAULT_GLYPH_NAMES.get(index).copied(),
+                Some(i) => custom.get(i).copied(),
+            }
+        }
+        _ => None,
+    }
+}
+
 /// Code → Unicode for a font: its ToUnicode CMap (for a composite font, also the name
 /// `/Identity-H` or `/Identity-V`, read as code = Unicode), else (simple fonts) its encoding.
 fn unicode_map(doc: &Document, font: &Dict, composite: bool) -> Unicode {
@@ -1465,7 +1574,12 @@ fn unicode_map(doc: &Document, font: &Dict, composite: bool) -> Unicode {
                 out.encoding.insert(code as u32, c.to_string());
             }
         }
+        // A font named as Symbol or ZapfDingbats keeps reporting nothing: its names may be as
+        // un-Latin as its codes (WordPerfect's WPTypographicSymbols calls its apostrophe `equal`).
+        // A font whose `post` table names its glyphs reads by those names; one that names none
+        // and keeps the Macintosh order reads as Mac OS Roman, over the WinAnsi fallback.
         if base_name.is_none() && !symbolic_builtin {
+            out.encoding.extend(symbolic_glyph_texts(doc, font));
             out.encoding.extend(symbolic_mac_roman_texts(doc, font));
         }
         if let Some(Object::Dict(d)) = enc.as_deref()

crates/fonts/src/pdf/tests.rs

@@ -214,6 +214,194 @@ fn mac_subset_cmap_uses_pdf_unicode_without_changing_source_program() {
     assert!(append_editor_cmap(&mut broken, &std::collections::BTreeMap::from([(65, 1)])).is_none());
 }
 
+/// The names read once give what `Post::glyph_name` gives, for every glyph of a font with its
+/// own names (Inter's version 2.0 table), and none past the last glyph.
+#[test]
+fn post_names_read_once_match_the_table() {
+    use skrifa::raw::TableProvider as _;
+    let face = skrifa::FontRef::new(include_bytes!("../../../../assets/fonts/Inter-Regular.ttf")).unwrap();
+    let post = face.post().unwrap();
+    let custom = post_custom_names(&post);
+    assert!(custom.len() > 100, "Inter names glyphs of its own: {}", custom.len());
+    let glyphs = face.maxp().unwrap().num_glyphs();
+    for g in 0..glyphs {
+        assert_eq!(post_name(&post, &custom, g), post.glyph_name(skrifa::raw::types::GlyphId16::new(g)), "glyph {g}");
+    }
+    assert_eq!(post_name(&post, &custom, glyphs), None);
+}
+
+/// #3962: a symbolic TrueType subset with no `/Encoding` whose codes are glyph positions in
+/// the Macintosh order, 29 below ASCII, mapped by a Mac (1,0) cmap. Its text is what the
+/// `post` table names the glyphs its codes select, and typing encodes back to those codes.
+/// A (3,0) symbol cmap at 0xF000 + code is read too: code 0x20, `.notdef` in the Mac table,
+/// shows `k` there, code 0xC0 `é`, and code 0xC1 the variant `a.1`, whose suffix is dropped.
+/// Records are read in order, each with its offsets, as the renderer reads them: 0xC2 is the
+/// symbol cmap's `x`, not a later (3,1) cmap's `z`.
+#[test]
+fn a_symbolic_truetype_font_without_an_encoding_reads_its_glyph_names() {
+    use skrifa::{MetadataProvider as _, raw::TableProvider as _};
+    const TEXT: &str = "Solid state";
+    let original = include_bytes!("../../../../assets/fonts/Inter-Regular.ttf");
+    let face = skrifa::FontRef::new(original).unwrap();
+    let charmap = face.charmap();
+    let post = face.post().unwrap();
+    let variant = (1..face.maxp().unwrap().num_glyphs())
+        .find(|&g| post.glyph_name(skrifa::raw::types::GlyphId16::new(g)) == Some("a.1"))
+        .expect("Inter has a.1");
+    // A glyph whose name has no Unicode value: its code reads as WinAnsi, as the renderer reads it.
+    let unnamed = (1..face.maxp().unwrap().num_glyphs())
+        .find(|&g| post.glyph_name(skrifa::raw::types::GlyphId16::new(g)).is_some_and(|n| !n.starts_with('.') && glyph_text(n).is_none()))
+        .expect("Inter has a glyph whose name has no text");
+    // A format 6 Mac cmap over codes 3..=0x57: each letter at its code minus 29.
+    let (first, last) = (3u16, 0x57u16);
+    let mut glyphs = vec![0u16; usize::from(last - first + 1)];
+    for c in TEXT.chars() {
+        glyphs[usize::from(c as u16 - 29 - first)] = u16::try_from(charmap.map(c).unwrap().to_u32()).unwrap();
+    }
+    let mac_len = 10 + 2 * glyphs.len() as u16;
+    let gid = |c: char| u16::try_from(charmap.map(c).unwrap().to_u32()).unwrap();
+    let mut mac = vec![0, 6];
+    mac.extend_from_slice(&mac_len.to_be_bytes());
+    mac.extend_from_slice(&[0, 0]);
+    mac.extend_from_slice(&first.to_be_bytes());
+    mac.extend_from_slice(&(glyphs.len() as u16).to_be_bytes());
+    glyphs.iter().for_each(|g| mac.extend_from_slice(&g.to_be_bytes()));
+    // A format 4 subtable: one segment per code, then the closing 0xFFFF segment.
+    let format4 = |map: &[(u16, u16)]| {
+        let segments: Vec<(u16, u16)> = map.iter().copied().chain([(0xFFFF, 0)]).collect();
+        let n = segments.len() as u16;
+        let words: Vec<u16> = [4, 16 + 8 * n, 0, 2 * n, 2, 0, 2 * n - 2]
+            .into_iter()
+            .chain(segments.iter().map(|s| s.0))
+            .chain([0])
+            .chain(segments.iter().map(|s| s.0))
+            .chain(segments.iter().map(|&(code, glyph)| if code == 0xFFFF { 1 } else { glyph.wrapping_sub(code) }))
+            .chain(segments.iter().map(|_| 0))
+            .collect();
+        words.iter().flat_map(|w| w.to_be_bytes()).collect::<Vec<u8>>()
+    };
+    // A glyph whose name the table spells `uni1eae`, lowercase, below: the glyph list
+    // specification's hex digits are uppercase (ISO 32000-2 §9.10.2), so it names nothing.
+    let lowercase = (1..face.maxp().unwrap().num_glyphs())
+        .find(|&g| post.glyph_name(skrifa::raw::types::GlyphId16::new(g)) == Some("uni1EAE"))
+        .expect("Inter has uni1EAE");
+    // Symbol (3,0): 0xF020 → k, 0xF0C0 → é, 0xF0C1 → a.1, 0xF0C2 → x, 0xF0C4 → the unnamed glyph,
+    // 0xF0C5 → the one spelt in lowercase.
+    let symbol = format4(&[
+        (0xF020, gid('k')),
+        (0xF0C0, gid('é')),
+        (0xF0C1, variant),
+        (0xF0C2, gid('x')),
+        (0xF0C3, gid('w')),
+        (0xF0C4, unnamed),
+        (0xF0C5, lowercase),
+    ]);
+    // Unicode (3,1), after it: 0xC2 → z. The renderer reads the (3,0) record, offsets and
+    // all, before this one, so code 0xC2 is `x`.
+    let unicode = format4(&[(0xC2, gid('z'))]);
+    // A Unicode-platform (0,3) record, first, which the renderer doesn't read: 0xC3 → q there,
+    // so code 0xC3 is the symbol cmap's `w`.
+    let platform0 = format4(&[(0xC3, gid('q'))]);
+    // Four records, in platform order as the cmap table requires.
+    let at = |n: usize| (36 + n) as u32;
+    let mut cmap = vec![0, 0, 0, 4];
+    for (platform, encoding, offset) in [
+        (0u16, 3u16, at(0)),
+        (1, 0, at(platform0.len())),
+        (3, 0, at(platform0.len() + mac.len())),
+        (3, 1, at(platform0.len() + mac.len() + symbol.len())),
+    ] {
+        cmap.extend_from_slice(&platform.to_be_bytes());
+        cmap.extend_from_slice(&encoding.to_be_bytes());
+        cmap.extend_from_slice(&offset.to_be_bytes());
+    }
+    cmap.extend(platform0);
+    cmap.extend(mac);
+    cmap.extend(symbol);
+    cmap.extend(unicode);
+    let mut subset = original.to_vec();
+    let count = u16::from_be_bytes(subset[4..6].try_into().unwrap()) as usize;
+    let entry = (0..count).map(|i| 12 + i * 16).find(|i| &subset[*i..*i + 4] == b"cmap").unwrap();
+    let offset = subset.len() as u32;
+    subset[entry + 8..entry + 12].copy_from_slice(&offset.to_be_bytes());
+    subset[entry + 12..entry + 16].copy_from_slice(&(cmap.len() as u32).to_be_bytes());
+    subset.extend_from_slice(&cmap);
+    // The `post` table's own spelling of that name, a Pascal string, lowercased in place.
+    let spelt = b"\x07uni1EAE";
+    assert_eq!(subset.windows(spelt.len()).filter(|w| w == spelt).count(), 1);
+    let spot = subset.windows(spelt.len()).position(|w| w == spelt).unwrap();
+    subset[spot..spot + spelt.len()].copy_from_slice(b"\x07uni1eae");
+    let codes: Vec<u8> = TEXT.bytes().map(|b| b - 29).collect();
+    let metrics = |flags: i64, encoding: Option<Object>| {
+        let mut doc = Document::new_empty();
+        let program = doc.add(Object::Stream(Stream::flate(Dict::new(), &subset)));
+        let mut descriptor = Dict::new();
+        descriptor.set(b"Flags".to_vec(), Object::Int(flags));
+        descriptor.set(b"FontFile2".to_vec(), Object::Ref(program));
+        let mut entries = vec![("Subtype", Object::name("TrueType")), ("FontDescriptor", Object::Dict(descriptor))];
+        if let Some(encoding) = encoding {
+            entries.push(("Encoding", encoding));
+        }
+        let f = font(&mut doc, entries);
+        Metrics::from_dict(&doc, &f)
+    };
+    let symbolic = metrics(4, None);
+    assert_eq!(symbolic.decode(&codes), TEXT);
+    // The same font with its `post` version halves swapped (`00 00 00 02`) reads the same,
+    // as the renderer reads it (#2946).
+    let post = (0..count).map(|i| 12 + i * 16).find(|i| &subset[*i..*i + 4] == b"post").unwrap();
+    let at = u32::from_be_bytes(subset[post + 8..post + 12].try_into().unwrap()) as usize;
+    assert_eq!(subset[at..at + 4], [0, 2, 0, 0], "Inter's post is version 2.0");
+    let mut swapped = subset.clone();
+    swapped[at..at + 4].copy_from_slice(&[0, 0, 0, 2]);
+    let mut doc = Document::new_empty();
+    let program = doc.add(Object::Stream(Stream::flate(Dict::new(), &swapped)));
+    let mut descriptor = Dict::new();
+    descriptor.set(b"Flags".to_vec(), Object::Int(4));
+    descriptor.set(b"FontFile2".to_vec(), Object::Ref(program));
+    let f = font(&mut doc, vec![("Subtype", Object::name("TrueType")), ("FontDescriptor", Object::Dict(descriptor))]);
+    assert_eq!(Metrics::from_dict(&doc, &f).decode(&codes), TEXT, "a swapped post version");
+    assert_eq!(symbolic.decode(&[0x20, 0xC0, 0xC1, 0xC2, 0xC3]), "kéaxw");
+    assert_eq!(symbolic.decode(&[0xC4]), "Ä", "a glyph name with no text reads as WinAnsi, not Standard's `tilde`");
+    assert_eq!(symbolic.decode(&[0xC5]), "Å", "`uni1eae` names nothing, so the code reads as WinAnsi, not U+1EAE");
+    assert_eq!(symbolic.encode("Solid").as_deref(), Some(&codes[..5]), "typing sets the same glyphs");
+    // A named encoding, or the Nonsymbolic flag, says what the codes are: they read through it.
+    assert_eq!(metrics(4, Some(Object::name("WinAnsiEncoding"))).decode(&codes), "6ROLGVWDWH");
+    // `/Differences` still names its codes: 0x36 shows `A` there, not the glyph's `S`.
+    let mut differences = Dict::new();
+    differences.set(b"Differences".to_vec(), Object::Array(vec![Object::Int(0x36), Object::name("A")]));
+    assert_eq!(metrics(4, Some(Object::Dict(differences))).decode(&codes), "Aolid state");
+    assert_eq!(metrics(32, None).decode(&codes), "6ROLGVWDWH");
+    // A font named as a symbol font reports nothing, as before: its glyph names can't be
+    // trusted to be Latin either.
+    let mut doc = Document::new_empty();
+    let program = doc.add(Object::Stream(Stream::flate(Dict::new(), &subset)));
+    let mut descriptor = Dict::new();
+    descriptor.set(b"Flags".to_vec(), Object::Int(4));
+    descriptor.set(b"FontFile2".to_vec(), Object::Ref(program));
+    let f = font(
+        &mut doc,
+        vec![
+            ("Subtype", Object::name("TrueType")),
+            ("BaseFont", Object::name("KNMDCP+WPTypographicSymbols")),
+            ("FontDescriptor", Object::Dict(descriptor)),
+        ],
+    );
+    assert_eq!(Metrics::from_dict(&doc, &f).decode(&codes), "");
+}
+
+/// The glyph list specification's `uni` and `u` forms spell their hex digits in uppercase
+/// (ISO 32000-2 §9.10.2); list names keep their own case.
+#[test]
+fn glyph_names_with_lowercase_hex_digits_name_nothing() {
+    for name in ["uni1EAE", "u1F600", "a", "ecircumflex", "f_i", "a.ss01", "uni0041_uni0042"] {
+        assert!(uppercase_hex_forms(name), "{name}");
+    }
+    for name in ["uni1eae", "u1f600", "uni20ac.alt", "f_uni20ac", "uni0041_uni00e9"] {
+        assert!(!uppercase_hex_forms(name), "{name}");
+    }
+}
+
 fn metrics_with_flags(flags: Option<Object>) -> Metrics {
     let mut doc = Document::new_empty();
     let mut descriptor = Dict::new();

crates/render/src/raster/tests.rs

@@ -2091,6 +2091,125 @@ fn rounded_pixel_budget_invalid_requests_never_reach_allocator() {
     }
 }
 
+/// #3962: FrameMaker files distilled with Distiller 3 embed TrueType subsets that are symbolic,
+/// with no `/Encoding` and no `/ToUnicode`, whose codes are glyph positions in the Macintosh
+/// order, 29 below ASCII; the font's own `cmap` maps them to the right glyphs. Read as WinAnsi
+/// the text came out shifted (`Solid` as `6ROLG`), in the editor and in Word export alike. The
+/// `post` table names the glyph each code selects, which gives the text (vendored
+/// hayro-interpret patch 79). A font with an `/Encoding` still reads its codes through it.
+#[test]
+fn a_symbolic_truetype_font_without_an_encoding_reads_its_glyph_names() {
+    use skrifa::MetadataProvider;
+    use skrifa::raw::TableProvider;
+    const TEXT: &str = "Solid state";
+    // Four more codes: 0x3D selects `W.cv07`, a variant whose suffix is dropped (`W`); 0x3E a
+    // glyph named `uniXXXX`, that code point; 0x3F `EMquad`, a name with no text, so the code
+    // reads as WinAnsi (`?`); 0x40 the ligature `less_equal.liga`, two characters (`<=`).
+    const SUFFIXED: u16 = 0x3D;
+    const UNI: u16 = 0x3E;
+    const UNNAMED: u16 = 0x3F;
+    const LIGATURE: u16 = 0x40;
+    // 0x41 selects a glyph whose name is spelt `uni01cd`, lowercase: the glyph list
+    // specification's hex digits are uppercase (ISO 32000-2 §9.10.2), so the name stands for no
+    // text and the code reads as WinAnsi (`A`), not as U+01CD.
+    const LOWERCASE: u16 = 0x41;
+    let font = include_bytes!("../../../../assets/fonts/JetBrainsMono-Regular.ttf");
+    let face = skrifa::FontRef::new(font).expect("the test font parses");
+    let charmap = face.charmap();
+    let post = face.post().expect("the test font has a post table");
+    let glyphs = face.maxp().expect("the test font has a maxp table").num_glyphs();
+    let named = |name: &str| {
+        (1..glyphs).find(|&g| post.glyph_name(skrifa::raw::types::GlyphId16::new(g)) == Some(name)).expect("the test font has the glyph")
+    };
+    let (uni, uni_char) = (1..glyphs)
+        .find_map(|g| {
+            let name = post.glyph_name(skrifa::raw::types::GlyphId16::new(g))?;
+            let c = char::from_u32(u32::from_str_radix(name.strip_prefix("uni")?, 16).ok()?)?;
+            (!c.is_control()).then_some((g, c))
+        })
+        .expect("the test font names a glyph uniXXXX");
+    // Codes 29 below ASCII, mapped by a (3,0) symbol cmap at 0xF000 + code.
+    let mut map: Vec<(u16, u16)> =
+        TEXT.chars().map(|c| (0xF000 + c as u16 - 29, charmap.map(c).expect("the test font has the letter").to_u32() as u16)).collect();
+    map.push((0xF000 + SUFFIXED, named("W.cv07")));
+    map.push((0xF000 + UNI, uni));
+    map.push((0xF000 + UNNAMED, named("EMquad")));
+    map.push((0xF000 + LIGATURE, named("less_equal.liga")));
+    let lowercase = named("uni01CD");
+    assert_ne!(lowercase, uni, "the lowercased name is not the `uniXXXX` glyph read above");
+    map.push((0xF000 + LOWERCASE, lowercase));
+    map.sort_unstable();
+    map.dedup();
+    // A format 4 subtable: one segment per code, then the closing 0xFFFF segment.
+    let segments: Vec<(u16, u16)> = map.iter().copied().chain([(0xFFFF, 0)]).collect();
+    let n = segments.len() as u16;
+    let mut sub = Vec::new();
+    let words = |v: &mut Vec<u8>, w: &[u16]| w.iter().for_each(|w| v.extend_from_slice(&w.to_be_bytes()));
+    words(&mut sub, &[4, 16 + 8 * n, 0, 2 * n, 2, 0, 2 * n - 2]);
+    words(&mut sub, &segments.iter().map(|s| s.0).collect::<Vec<_>>());
+    words(&mut sub, &[0]);
+    words(&mut sub, &segments.iter().map(|s| s.0).collect::<Vec<_>>());
+    words(&mut sub, &segments.iter().map(|&(code, gid)| if code == 0xFFFF { 1 } else { gid.wrapping_sub(code) }).collect::<Vec<_>>());
+    words(&mut sub, &vec![0; segments.len()]);
+    let mut cmap = Vec::new();
+    words(&mut cmap, &[0, 1, 3, 0]);
+    cmap.extend_from_slice(&12u32.to_be_bytes());
+    cmap.extend(sub);
+    // The new cmap goes at the end; the table directory's `cmap` entry points at it.
+    let mut program = font.to_vec();
+    while !program.len().is_multiple_of(4) {
+        program.push(0);
+    }
+    let at = program.len();
+    program.extend(&cmap);
+    let tables = usize::from(u16::from_be_bytes([font[4], font[5]]));
+    let entry = (0..tables).map(|i| 12 + 16 * i).find(|&e| &font[e..e + 4] == b"cmap").expect("the test font has a cmap");
+    program[entry + 8..entry + 12].copy_from_slice(&(at as u32).to_be_bytes());
+    program[entry + 12..entry + 16].copy_from_slice(&(cmap.len() as u32).to_be_bytes());
+    // The `post` table's own spelling of the name, a Pascal string, lowercased in place.
+    let spelt = b"\x07uni01CD";
+    let spot = program.windows(spelt.len()).position(|w| w == spelt).expect("the post table spells the name");
+    assert_eq!(program.windows(spelt.len()).filter(|w| w == spelt).count(), 1);
+    program[spot..spot + spelt.len()].copy_from_slice(b"\x07uni01cd");
+
+    let text_of = |encoding: &str| -> String {
+        let codes: String = TEXT.chars().map(|c| format!("{:02X}", c as u8 - 29)).collect::<String>()
+            + &format!("{SUFFIXED:02X}{UNI:02X}{UNNAMED:02X}{LIGATURE:02X}{LOWERCASE:02X}");
+        let content = format!("BT /F1 14 Tf 10 40 Td <{codes}> Tj ET");
+        let objects: Vec<Vec<u8>> = vec![
+            b"<< /Type /Catalog /Pages 2 0 R >>".to_vec(),
+            b"<< /Type /Pages /Kids [3 0 R] /Count 1 >>".to_vec(),
+            b"<< /Type /Page /Parent 2 0 R /MediaBox [0 0 300 80] /Contents 4 0 R /Resources << /Font << /F1 5 0 R >> >> >>".to_vec(),
+            format!("<< /Length {} >>\nstream\n{content}\nendstream", content.len()).into_bytes(),
+            format!("<< /Type /Font /Subtype /TrueType /BaseFont /ABCDEF+JetBrainsMono-Regular /FirstChar 0 /LastChar 255 /Widths [{}] /FontDescriptor 6 0 R {encoding}>>", vec!["600"; 256].join(" ")).into_bytes(),
+            b"<< /Type /FontDescriptor /FontName /ABCDEF+JetBrainsMono-Regular /Flags 4 /FontBBox [-500 -300 1200 1000] /ItalicAngle 0 /Ascent 1020 /Descent -300 /CapHeight 730 /StemV 80 /FontFile2 7 0 R >>".to_vec(),
+            [format!("<< /Length {} >>\nstream\n", program.len()).into_bytes(), program.clone(), b"\nendstream".to_vec()].concat(),
+        ];
+        let mut pdf = b"%PDF-1.3\n".to_vec();
+        let mut offsets = Vec::new();
+        for (i, object) in objects.iter().enumerate() {
+            offsets.push(pdf.len());
+            pdf.extend(format!("{} 0 obj\n", i + 1).into_bytes());
+            pdf.extend(object);
+            pdf.extend(b"\nendobj\n");
+        }
+        let xref = pdf.len();
+        pdf.extend(format!("xref\n0 {}\n0000000000 65535 f \n", objects.len() + 1).into_bytes());
+        for offset in offsets {
+            pdf.extend(format!("{offset:010} 00000 n \n").into_bytes());
+        }
+        pdf.extend(format!("trailer << /Size {} /Root 1 0 R >>\nstartxref\n{xref}\n%%EOF\n", objects.len() + 1).into_bytes());
+        let mut r = PageRenderer::new(Arc::new(pdf), RenderConfig::default());
+        let page = r.filtered_text(0, crate::text::TextFilter::default()).expect("the page has text");
+        page.glyphs.iter().map(|g| g.text.as_str()).collect()
+    };
+    assert_eq!(text_of(""), format!("Solid stateW{uni_char}?<=A"));
+    // An encoding says what the codes are, so they read through it.
+    assert_eq!(text_of("/Encoding /WinAnsiEncoding "), "6ROLGVWDWH=>?@A");
+    // `/Differences` still names its code: 0x36 reads `A`, not the glyph's `S`.
+    assert_eq!(text_of("/Encoding << /Differences [54 /A] >> "), format!("Aolid stateW{uni_char}?<=A"));
+}
+
 /// Issue #102: an A4 page (595.28×841.89 pt) rendered 595×841 at 72 dpi, dropping the last
 /// partial row and column. Like pdftoppm it is now 596×842, the edge pixels partly covered.
 #[test]

vendor/README.md

@@ -39,7 +39,7 @@ 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 `.nLine truncated
+| hayro-interpret | 0.7.0 | Apache-2.0 OR MIT | (1) select `/AP /N` appearance states via `/AS` (checkboxes, radios, stateful icons) and honour the NoView flag; (2) Type3 Unicode fallback to `/Encoding` glyph names, then printable ASCII; (3) `Type3Glyph::advance_width`; (4) `InterpreterSettings::ocg_overrides` for viewer layer toggles; (5) soft masks whose group lacks `/CS` were dropped (Chrome gradient text); (6) paint nesting cap `MAX_PAINT_NESTING` for tiling patterns and Type 3 glyphs (self-referencing content overflowed the stack); (7) CID width ranges in `/W` and `/W2` clamped to `MAX_CID` (a `0 4294967295 w` range hung); (8) `InterpreterSettings::hide_comments` skips markup annotations but keeps widgets and links (View ▸ Hide all comments); (9) `image_size_ok`: images and masks over 2^28 pixels are not decoded (a fuzzed `/W 4294967295` stencil mask allocated 4 GiB); (10) `MAX_NESTED_PAINTS`: nested paints (forms, Type 3 glyphs and tiling patterns inside one another) share a per-page budget of 50,000 (a glyph showing eight glyphs of its own font, sixteen deep, hung); (11) depends on skrifa 0.47 instead of 0.42 (upstream read-fonts fixed a Type 1 tokenizer that recursed through runs of numbers: a fuzzed font hung or overflowed the stack), with the one API change (`Charset::glyph_id` returns `Option`); (12) annotation appearances whose `/BBox` or `/Matrix` collapses to a line or a point are skipped (mapping them onto `/Rect` divided by zero and handed the device infinite or NaN matrices); (13) a Highlight annotation without a usable normal appearance is drawn from its `/QuadPoints`, `/C` and `/CA` with Multiply blending, as pdfcraft-annot draws highlights (it was not drawn at all); (14) simple-font `/Widths` (`read_widths`): entries only for codes 0–255 (one entry per code below `/FirstChar` was allocated: `4294967295` asked for 34 GB), an entry that isn't a number leaves its code missing instead of ending the array, and a `/LastChar` below `/FirstChar` gives no widths instead of discarding the font for a standard one; (15) the appearance box is mapped onto `/Rect` corner to corner: the translation now accounts for the scale (an appearance whose box didn't start at the origin and differed in size from `/Rect` was drawn shifted); (16) a clip set by `W`/`W*` takes effect after whichever path-painting operator ends the path (`S`, `f`, `B`, … as well as `n`; ISO 32000-2 §8.5.4): it was applied only on `n`, so `re W* S` clipped nothing and the clip stayed pending for a later, unrelated `n`; (17) `InterpreterSettings::cancelled`: once set, interpretation stops before the next content operator (a render pool sets it for every worker when dropped, after an edit or on closing, and for one worker when the watchdog gives up on its render, so renders whose answers nobody can receive stop instead of running to the end); (18) `MAX_PAGE_CONTENT`: the content a page holds at once (its contents, and the forms, tiling patterns and Type 3 glyph procedures being painted) stays within 512 MiB (`decode_content`; once it runs out, `InterpreterWarning::ContentTruncated` is reported (once per page by pdfcraft-render) so a truncated page is not a clean render; each level gives its bytes back when it finishes): forms nest fifty deep and each level kept its content while the next ran, so a form painting itself around a stream inflating to 256 MiB would have held over 12 GB; and Type 0 sampled functions take at most `MAX_SAMPLE_ENTRIES` (2^21) table points and `MAX_SAMPLE_VALUES` (2^23) samples, read no further than `/Size` asks (256 MiB of 1-bit samples asked for over 8 GB); (19) mesh shadings (types 4–7) are sampled only inside the area being drawn (`ShadingPattern::encode_within`, which the vendored hayro renderer calls with that area plus a pixel), within 8 visits per pixel of it plus 4 per triangle (`TriangleSampler`), and patches are triangulated one at a time: points far off the page were sampled over their whole bounding box, up to 65536 × 65536 entries (a fuzzed 147-byte patch stream; a synthetic page passed 6 GB in six seconds), ten thousand overlapping triangles ran for minutes, and every patch became 722 triangles before any limit; meshes hold at most `MAX_MESH_PATCHES` (2^16) patches and `MAX_MESH_TRIANGLES` (2^20) triangles, a lattice needs two to 2^19 vertices per row (`/VerticesPerRow 0` pushed empty rows for ever; one huge row read the whole stream), and a page samples at most `MAX_PAGE_MESH_TRIANGLES` (2^25) mesh triangles in all (a mesh painted a thousand times was triangulated a thousand times); each cut logs a warning; (20) PDF function construction is limited to 64 levels and 10,000 total nodes with cycle detection; calculator parsing and evaluation share procedure depth/work budgets, integer division/bitshift/rotation/index are panic-free, and fixed-size stack errors propagate; (21) a Type 0 font drawn with a substitute that isn't CID-keyed (it isn't embedded, or its program doesn't load) draws `.nLine truncated
 | hayro | 0.7.1 | Apache-2.0 OR MIT | (1) `RenderSettings::{x_offset, y_offset}` to render one tile of a large page; (2) images over `MAX_IMAGE_PIXELS` (2^28) are skipped (a fuzzed `/W 4294967295` hung in resampling); (3) `tiling_cell_scale` keeps a tiling pattern's cell pixmap within 3000 px a side whatever its `/XStep`/`/YStep` (a fuzzed step far beyond `/BBox` allocated 2 GB); (4) stroke widths clamped to 4 × (canvas width + height) device pixels (a fuzzed `/LW 9223372036854775807` allocated 10 GB in stroke expansion); (5) `usable_dash_pattern`: a stroke whose dash pattern could cut it into more than `MAX_DASHES_PER_STROKE` (250,000) pieces, counting kurbo's restart of the pattern at every subpath and `ClosePath`, is drawn solid (a fuzzed `/D [[11] 0]` on a line from x = 92234775807 allocated over 5 GB, `[0.000001] 0 d` on a 20 pt line over 10 GB); (6) a dash pattern with a negative or non-finite entry, or a period that isn't positive, is drawn solid (`[-1 -1] 0 d` looped for ever in kurbo); (7) `render_size` and `render_into`: render into a buffer the caller allocates (PdfCraft renders into pixel-aligned memory that the GUI takes over as texture data); `render` is `render_into` on a new pixmap; (8) checked image source/target/RGBA, alpha-premultiplication scratch and intermediate sizes before resampling, finite transforms and exact source/mask buffers; sources wider than u16 may shrink before backend limits, with a 2^20 source-axis cap and conservative 128 MiB filter-planning estimate; an over-limit or failed resize keeps the original image resolution and geometry when the source fits the backend, otherwise skips it, with a logged warning; staging, resampling output/scratch, RGBA conversion and padding buffers use fallible reserves; (9) `may_paint_canvas`: a fill or stroke whose control-point box, grown by how far its paint reaches past the path (half the stroke width times the larger of the miter limit and √2) plus 0.5 user units and 2 pixels, misses the canvas is not drawn, so a tile no longer strokes and fills every path of its page (item 5 of #623); `set_skip_offscreen_paths` turns this off for the regression test; (10) `/Interpolate` governs the pre-resize too (#624): a minified image that does not request interpolation is pre-resized with nearest-neighbour (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<ColorReplacement>` argument (pass `None` for true colours); the upstream-only `[dev-dependencies.log]` entry is removed from `Cargo.toml` (log is a normal dependency since the logging patch), so `cargo test -p hayro` runs this crate's unit tests; (14) a shrunk 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 stroLine truncated
 | hayro-jbig2 | 0.3.0 | Apache-2.0 OR MIT | (1) at most 2^28 pixels per bitmap and page bitmap, and 2^25 cells per halftone grid and grayscale image (each side could be 65535: 4.3 billion pixels decoded one by one, and a 4-byte value per cell); (2) at most 2^26 symbol pixels per symbol dictionary (a fuzzed dictionary decoded for minutes); (3) a symbol dictionary's height classes are capped at one per new symbol and its export runs at `2 × symbols + 1`, each plus `EXTRA_EMPTY_STEPS` (1024): an empty height class or a zero-length export run doesn't advance, and past the end of its data the arithmetic decoder can decode them for ever (one changed byte hung the renderer); (4) a segment header's long-form referred-to segment count is capped at `u16::MAX` and a larger count is refused before its retention or referred-to bytes are read: the 29-bit count went straight to `Vec::with_capacity`, a reservation of about 2 GiB whose failure aborts the process (from storytold/craft-libs#176); (5) what a page's segments hold for later ones (intermediate regions, pattern and symbol dictionaries with their headers and retained contexts, and Huffman tables) is at most 64 MiB, two square bitmaps of 16,384 a side (`MAX_RETAINED_WORDS`), and a dictionary's new and exported symbols are counted against it as they are made: each was bounded on its own, but a few dozen bytes an intermediate region held 33 MB each, and 8 MiB symbols exported again and again killed the process, so a few kilobytes could exhaust memory; (6) a segment may refer to at most 2^20 symbols, checked before any are collected: references may repeat, and 65,535 of them to one dictionary, in a 78 KB stream, asked for 17 GB; (7) a pattern dictionary keeps at most 16 MiB of shifted copies, which only speed up drawing, and none that wouldn't fit what the page has left (`MAX_SHIFTED_WORDS`); (8) a custom Huffman table has at most 2^16 lines, and a stream with its globals at most 2^20 segments and 2^22 references, checked as they are read: both grew before anything counted them, and PDFs of a kilobyte or so reached 5 GB and more behind nested Flate; (9) a page past the bounds of (1) is refused when the stream is opened: a caller sizes its output from the header before the page bitmap is made, and an Indexed image in a 743-byte PDF asked for 137 GB, which aborts; (10) a stream may ask for at most 2^30 units of decoding work, four of the largest pages, about 6 s to 17 s (`MAX_WORK`): decoding a bitmap costs a pixel each and 4 a row, copying one a word each and 1 a row, and each integer an item reads costs 38, what one takes at worst, so one with no pixels pays too, before it is decoded: each bitmap was bounded, but nothing bounded how many there were, and a 795-byte PDF took 265 s; (11) `Image::decode_within` spends from what a run of images may still do, starting with its page's pixels, and `charge_reading` charges reading a stream and its globals before anything is parsed, so an operation over a document can bound what `MAX_WORK` bounds for one image, and images that fail or decode nothing add up too: a file of thousands of images made to reach that limit took as many times as long; (12) adding to a location, a reference offset, a symbol's corner or a Huffman range line near the end of an `i32`, and negating a decoded integer of 2^31, overflowed: a panic wherever overflow is checked. The refinement context, a text region's corners, the integer decoder and the Huffman table wrap, as a release build did, and `Bitmap::combine` saturates, which comes to the same, as a location that overflows is past the end of any page. What is drawn doesn't change; (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, TaLine truncated
 | lopdf | 0.45.0 | MIT | (1) PNG predictor rows longer than the data are refused before their buffers are allocated (a fuzzed `/Columns 4294967295` allocated two 4 GiB rows), and empty frames return without row buffers; (2) predictor row widths use saturating arithmetic or reject overflow; (3) sub-byte TIFF Predictor 2 scratch is bounded by the samples present in the input instead of the declared `/Colors`, aligned with [lopdf#589](https://github.com/J-F-Liu/lopdf/pull/589); (4) the object lookup keeps its cycle history through compressed objects, requires an ordinary in-file container, and caps the chain of indirect lookups at 16 (a chain of new objects, each a `/Length` compressed in a container whose own `/Length` names the next, or a stream whose `/Length` names the next stream, overflowed the stack: a dev build's 2 MiB thread within 800 hops); (5) object-stream members are parsed within their own extent (up to the next distinct offset), an entry whose offset repeats an earlier entry's is skipped (so a repeated object number can resolve to its earlier entry), and offsets are processed in index order; (6) `find_latest_trailer`: a rebuild takes a `/Type /XRef` stream dictionary as the trailer when the file has no `trailer` keyword (the latest in the file wins, as a later `trailer` would), so a concatenated or truncated file keeps its /Info and /ID instead of failing to load; a candidate whose /Root is the catalog beats one whose /Root is another dictionary, and an in-file /Root must match its generation (#2365) | (1) `inspect::tests::png_predictor_rows_longer_than_the_data_are_refused_before_allocating`, `inspect::tests::empty_png_predictor_frames_do_not_allocate_rows`; (2) `inspect::tests::tiff_subbyte_predictor_row_width_overflow_is_refused`; (3) `inspect::tests::tiff_subbyte_predictor_scratch_is_bounded_by_available_samples`, `inspect::tests::tiff_subbyte_predictors_keep_components_rows_and_padding`; (4) `inspect::tests::compressed_object_self_container_does_not_recurse`, `inspect::tests::compressed_object_mutual_containers_do_not_recurse`, `inspect::tests::object_stream_length_cycle_does_not_recurse`, `inspect::tests::compressed_object_container_chain_does_not_recurse`, `inspect::tests::compressed_length_chain_does_not_recurse`, `inspect::tests::plain_stream_length_chain_does_not_recurse`, `inspect::tests::compressed_objects_and_indirect_stream_length_still_load`, `inspect::tests::compressed_length_in_a_container_with_an_indirect_length_still_loads`; (5) `inspect::tests::object_stream_member_extents_bound_owned_strings`, `inspect::tests::object_stream_member_extents_preserve_index_order_and_whitespace`; (6) `inspect::tests::rebuilt_file_keeps_the_info_of_its_cross_reference_stream`, `inspect::tests::rebuilt_file_prefers_a_trailer_whose_root_is_the_catalog` |

vendor/hayro-interpret/src/font/true_type.rs

@@ -191,11 +191,23 @@ impl TrueTypeFont {
         }
 
         match &self.kind {
-            // PdfCraft patch (#3169): a name outside the glyph list reads by the glyph list
-            // specification's full mapping (`f_i`, `hyphen.case`).
-            Kind::Embedded(e) => e.code_to_unicode_name(code as u8).and_then(|name| {
-                glyph_name_to_unicode(name).map(BfString::Char).or_else(|| glyph_name_to_text(name))
-            }),
+            Kind::Embedded(e) => {
+                // PdfCraft patch (79): see `EmbeddedKind::symbolic_text`.
+                if let Some(text) = e.symbolic_text(code as u8) {
+                    let mut chars = text.chars();
+                    return Some(match (chars.next(), chars.next()) {
+                        (Some(c), None) => BfString::Char(c),
+                        _ => BfString::String(text.to_string()),
+                    });
+                }
+                // PdfCraft patch (#3169): a name outside the glyph list reads by the glyph list
+                // specification's full mapping (`f_i`, `hyphen.case`).
+                e.code_to_unicode_name(code as u8).and_then(|name| {
+                    glyph_name_to_unicode(name)
+                        .map(BfString::Char)
+                        .or_else(|| glyph_name_to_text(name))
+                })
+            }
             Kind::Standard(s) => s.char_code_to_unicode(code as u8).map(BfString::Char),
         }
 
@@ -221,6 +233,9 @@ struct EmbeddedKind {
     cached_mappings: RefCell<HashMap<u8, GlyphId>>,
     /// PostScript name from the PDF.
     postscript_name: Option<String>,
+    /// PdfCraft patch (79): the text each glyph's `post` table name stands for, for the names
+    /// that stand for any (`glyph_name_text`).
+    text_names: HashMap<GlyphId, String>,
     /// PdfCraft patch (101): whether the font may take the Mac OS Roman reading at all, worked out
     /// once when it is loaded (`EmbeddedKind::mac_roman_name`).
     mac_roman_eligible: bool,
@@ -240,6 +255,10 @@ impl EmbeddedKind {
             .and_then(|d| OpenTypeFontBlob::new(Arc::new(d.to_vec()), 0))?;
 
         let glyph_names = base_font.glyph_names();
+        let text_names = glyph_names
+            .iter()
+            .filter_map(|(name, glyph)| Some((*glyph, glyph_name_text(name)?)))
+            .collect();
 
         let cff_font_blob = base_font
             .font_ref()
@@ -283,6 +302,7 @@ impl EmbeddedKind {
             encoding,
             cached_mappings: RefCell::new(HashMap::new()),
             postscript_name,
+            text_names,
             mac_roman_eligible: !names_base_encoding && !symbolic_builtin,
         })
     }
@@ -325,6 +345,29 @@ impl EmbeddedKind {
             .or_else(|| win_ansi::get(code))
     }
 
+    /// PdfCraft patch (79): a symbolic font without an encoding selects its glyphs through its
+    /// own `cmap`, so its codes need not mean anything in WinAnsi. FrameMaker files distilled
+    /// with Distiller 3 number them in the Macintosh glyph order, 29 below ASCII, and read as
+    /// WinAnsi `Solid` came out as `6ROLG` (#3962). The name the `post` table gives the glyph the
+    /// code selects says what it is, as MuPDF reads it, and as pdfcraft-fonts reads it for the
+    /// editor. `None` when the font is non-symbolic, has an encoding or a `/Differences` entry
+    /// for the code, selects no glyph for it, or names it with no text: the code is then read as
+    /// before.
+    fn symbolic_text(&self, code: u8) -> Option<&str> {
+        if self.is_non_symbolic()
+            || !matches!(self.encoding, Encoding::BuiltIn)
+            || self.differences.contains_key(&code)
+            || self.cff_blob.is_some()
+        {
+            return None;
+        }
+        let glyph = self.map_code(code);
+        if glyph == GlyphId::NOTDEF {
+            return None;
+        }
+        self.text_names.get(&glyph).map(String::as_str)
+    }
+
     fn outline_glyph(&self, glyph: GlyphId) -> BezPath {
         self.base_font.outline_glyph(glyph)
     }
@@ -510,6 +553,38 @@ pub(crate) enum Width {
     Missing,
 }
 
+/// PdfCraft patch (79): the text a glyph name stands for, by the rules of the Adobe Glyph List
+/// specification (ISO 32000-2 §9.10.2): drop any suffix from the first `.` (`a.ss01` is `a`),
+/// then each `_`-separated component is a list name, `uni` and groups of exactly four
+/// uppercase hex digits, or `u` and four to six. A value in the surrogate range or past
+/// U+10FFFF is no character. `None` when any component is none of these (`uni20ac`, `uni65`,
+/// `u65`). Each component gives at least one character, so the text is never empty.
+fn glyph_name_text(name: &str) -> Option<String> {
+    // An empty stem (`.notdef`) is no list name and no number, so it gives `None` below.
+    let stem = name.split('.').next()?;
+    // `from_str_radix` alone would also take a sign (`uni+041`) and lowercase digits, which the
+    // specification's forms don't have.
+    let hex = |h: &str| !h.is_empty() && h.bytes().all(|b| matches!(b, b'0'..=b'9' | b'A'..=b'F'));
+    let mut out = String::new();
+    for part in stem.split('_') {
+        if let Some(text) = glyph_names::get(part) {
+            out.push_str(text);
+        } else if let Some(digits) = part.strip_prefix("uni").filter(|h| hex(h)) {
+            // A group short of four digits isn't there to take, so `uni041` gives `None`.
+            for i in (0..digits.len()).step_by(4) {
+                let unit = digits.get(i..i.saturating_add(4))?;
+                out.push(char::from_u32(u32::from_str_radix(unit, 16).ok()?)?);
+            }
+        } else {
+            let digits = part
+                .strip_prefix('u')
+                .filter(|h| (4..=6).contains(&h.len()) && hex(h))?;
+            out.push(char::from_u32(u32::from_str_radix(digits, 16).ok()?)?);
+        }
+    }
+    Some(out)
+}
+
 pub(crate) fn read_widths(dict: &Dict<'_>, descriptor: &Dict<'_>) -> Option<(Vec<Width>, f32)> {
     let mut widths = Vec::new();
 
@@ -601,3 +676,53 @@ pub(crate) fn read_encoding(dict: &Dict<'_>) -> (Encoding, HashMap<u8, String>)
         )
     }
 }
+
+#[cfg(test)]
+mod tests {
+    use super::glyph_name_text;
+
+    /// PdfCraft patch (79): glyph names read by the glyph list specification's rules.
+    #[test]
+    fn glyph_names_read_by_the_adobe_glyph_list_rules() {
+        for (name, expected) in [
+            ("a", "a"),
+            ("a.ss01", "a"),
+            ("less_equal.liga", "<="),
+            ("uni0041", "A"),
+            ("uni00410042", "AB"),
+            ("u0041", "A"),
+            ("u1F600", "\u{1F600}"),
+        ] {
+            assert_eq!(glyph_name_text(name).as_deref(), Some(expected), "{name}");
+        }
+        for name in [
+            "",
+            ".notdef",
+            ".alt",
+            "EMquad",
+            "uni041",
+            "uni00410",
+            "uni004G",
+            "uni+041",
+            "u+041",
+            "uniD800",
+            "u041",
+            "u0000041",
+            "u004G",
+            "a_EMquad",
+            // PdfCraft patch (79): the glyph list specification's hex digits are uppercase only.
+            "uni20ac",
+            "uni004100e9",
+            "u1f600",
+            "u20ac",
+            "a_uni20ac",
+            "uni65",
+            "u65",
+            "uni+065",
+            "uDFFF",
+            "u110000",
+        ] {
+            assert_eq!(glyph_name_text(name), None, "{name}");
+        }
+    }
+}