diff --git a/crates/fonts/src/embedding.rs b/crates/fonts/src/embedding.rs index a5bb011e00..23f34dbb36 100644 --- a/crates/fonts/src/embedding.rs +++ b/crates/fonts/src/embedding.rs @@ -1,9 +1,10 @@ //! Unicode embedding of an explicitly supplied, embeddable TrueType face. Font files are //! document inputs or the optional craft-fonts build input, never system-interface fallbacks. +use std::collections::HashMap; use std::fmt::Write as _; -use pdfcraft_cos::{Dict, Document, Object, PdfString, Stream}; +use pdfcraft_cos::{Dict, Document, ObjRef, Object, PdfString, Stream}; use skrifa::{ FontRef, MetadataProvider, instance::{LocationRef, Size}, @@ -173,7 +174,8 @@ pub fn embed_mapping( return Err("the font family name is invalid".into()); } let glyph_map = if cff { cid_encoding.as_bytes() } else { gids.as_slice() }; - if let Some(resource) = same_resource(doc, fonts, &name, bytes, glyph_map, cmap.as_bytes()) { + let mut programs = ProgramMatch::new(bytes); + if let Some(resource) = same_resource(doc, fonts, &name, &mut programs, glyph_map, cmap.as_bytes()) { return Ok(resource); } let mut descriptor = Dict::new(); @@ -193,12 +195,7 @@ pub fn embed_mapping( ); // Reuse the program another resource of this page already has, else one registered on another // page; source subset programs stay untouched. - let existing = written(doc, fonts).find(|w| w.has_program(doc, bytes)).map(|w| w.program).or_else(|| { - doc.scan_objects().take(8192).find_map(|(r, o)| match o.as_ref() { - Object::Stream(s) if is_program(doc, s, bytes) => Some(Object::Ref(r)), - _ => None, - }) - }); + let existing = existing_program(doc, fonts, &mut programs); let program = existing.unwrap_or_else(|| { let mut d = Dict::new(); d.set(b"Length1".to_vec(), Object::Int(bytes.len() as i64)); @@ -246,12 +243,64 @@ pub fn embed_mapping( Ok(resource) } -/// The decoded data of the stream `object` refers to. -fn stream_data(doc: &Document, object: Option<&Object>) -> Option> { - match doc.resolve(object?).as_ref() { - Object::Stream(s) => s.decoded_within(17 << 20).ok(), - _ => None, +/// Whether the stream `object` refers to decodes to `expected`. +fn stream_is(doc: &Document, object: Option<&Object>, expected: &[u8]) -> bool { + object.is_some_and(|o| match doc.resolve(o).as_ref() { + Object::Stream(s) => s.decoded_within(expected.len()).is_ok_and(|d| d == expected), + _ => false, + }) +} + +/// Full programs one embedding decodes at most while looking for its own. A real match passes +/// the cheap checks and decodes once; many candidates that pass them come from a hostile file. +const MAX_PROGRAM_DECODES: usize = 16; + +/// Recognises the full program written for `bytes`: each candidate stream decodes at most once +/// and within `bytes.len()`, and at most [`MAX_PROGRAM_DECODES`] decode in all. +struct ProgramMatch<'a> { + bytes: &'a [u8], + known: HashMap, + decodes: usize, +} + +impl<'a> ProgramMatch<'a> { + fn new(bytes: &'a [u8]) -> Self { + Self { bytes, known: HashMap::new(), decodes: 0 } } + + /// Whether `r`, holding `object`, is the program. + fn is(&mut self, doc: &Document, r: ObjRef, object: &Object) -> bool { + if let Some(&known) = self.known.get(&r) { + return known; + } + let found = match object { + Object::Stream(s) + if s.dict.get(b"PCFullFont").is_some() + && s.dict.get(b"Length1").and_then(|l| doc.resolve(l).as_int()) == i64::try_from(self.bytes.len()).ok() + && self.decodes < MAX_PROGRAM_DECODES => + { + self.decodes = self.decodes.saturating_add(1); + s.decoded_within(self.bytes.len()).is_ok_and(|data| data == self.bytes) + } + _ => false, + }; + self.known.insert(r, found); + found + } + + /// Whether the resource's program is this one. + fn written(&mut self, doc: &Document, w: &Written) -> bool { + w.program.as_ref().is_some_and(|r| self.is(doc, r, &*doc.get(r))) + } +} + +/// The program for `programs.bytes` that this page's `PCFull` resources, or else the first +/// objects of the document, already have. +fn existing_program(doc: &Document, fonts: &Dict, programs: &mut ProgramMatch) -> Option { + written(doc, fonts) + .find(|w| programs.written(doc, w)) + .map(|w| w.program) + .or_else(|| doc.scan_objects().take(8192).find_map(|(r, o)| programs.is(doc, r, &*o).then_some(Object::Ref(r)))) } /// A `PCFull` resource that [`embed_mapping`] wrote. @@ -263,21 +312,6 @@ struct Written { program: Object, } -impl Written { - /// Whether its program is `bytes`. - fn has_program(&self, doc: &Document, bytes: &[u8]) -> bool { - matches!(doc.resolve(&self.program).as_ref(), Object::Stream(s) if is_program(doc, s, bytes)) - } -} - -/// Whether `s` is a full program written for `bytes`, compressed or not. The length it was -/// written with is compared before decoding. -fn is_program(doc: &Document, s: &Stream, bytes: &[u8]) -> bool { - s.dict.get(b"PCFullFont").is_some() - && s.dict.get(b"Length1").and_then(|l| doc.resolve(l).as_int()) == i64::try_from(bytes.len()).ok() - && s.decoded_within(17 << 20).is_ok_and(|data| data == bytes) -} - /// The `PCFull` resources in `fonts`, in order. fn written<'a>(doc: &'a Document, fonts: &'a Dict) -> impl Iterator + 'a { fonts.iter().filter(|(key, _)| key.starts_with(b"PCFull")).take(256).filter_map(move |(key, value)| { @@ -291,14 +325,14 @@ fn written<'a>(doc: &'a Document, fonts: &'a Dict) -> impl Iterator Option { +fn same_resource(doc: &Document, fonts: &Dict, name: &str, programs: &mut ProgramMatch, glyph_map: &[u8], unicode: &[u8]) -> Option { written(doc, fonts) .find(|w| { let glyphs = w.cid.get(b"CIDToGIDMap").or_else(|| w.top.get(b"Encoding")); w.top.name(b"BaseFont") == Some(name.as_bytes()) - && stream_data(doc, w.top.get(b"ToUnicode")).is_some_and(|d| d == unicode) - && stream_data(doc, glyphs).is_some_and(|d| d == glyph_map) - && w.has_program(doc, bytes) + && stream_is(doc, w.top.get(b"ToUnicode"), unicode) + && stream_is(doc, glyphs, glyph_map) + && programs.written(doc, w) }) .map(|w| w.resource) } @@ -385,6 +419,48 @@ mod tests { assert_eq!(programs(&doc), 2); } + #[test] + fn the_program_search_decodes_each_candidate_once_and_few_in_all() { + let bytes = b"the program".as_slice(); + let decoy = |doc: &mut Document| { + let mut d = Dict::new(); + d.set(b"PCFullFont".to_vec(), Object::Bool(true)); + d.set(b"Length1".to_vec(), Object::Int(bytes.len() as i64)); + doc.add(Object::Stream(Stream::flate(d, &[0; 4096]))) + }; + let resource = |doc: &mut Document, program: ObjRef| { + let mut descriptor = Dict::new(); + descriptor.set(b"FontFile2".to_vec(), Object::Ref(program)); + let mut cid = Dict::new(); + cid.set(b"FontDescriptor".to_vec(), Object::Dict(descriptor)); + let mut top = Dict::new(); + top.set(b"DescendantFonts".to_vec(), Object::Array(vec![Object::Dict(cid)])); + doc.add(Object::Dict(top)) + }; + // 200 resources sharing one decoy program decode it once. + let mut doc = Document::new_empty(); + let shared = decoy(&mut doc); + let mut fonts = Dict::new(); + for i in 0..200 { + let r = resource(&mut doc, shared); + fonts.set(format!("PCFull{i}").into_bytes(), Object::Ref(r)); + } + let mut programs = ProgramMatch::new(bytes); + assert_eq!(existing_program(&doc, &fonts, &mut programs), None); + assert_eq!(programs.decodes, 1); + // 100 distinct decoys decode a bounded number of times. + let mut doc = Document::new_empty(); + let mut fonts = Dict::new(); + for i in 0..100 { + let program = decoy(&mut doc); + let r = resource(&mut doc, program); + fonts.set(format!("PCFull{i}").into_bytes(), Object::Ref(r)); + } + let mut programs = ProgramMatch::new(bytes); + assert_eq!(existing_program(&doc, &fonts, &mut programs), None); + assert_eq!(programs.decodes, MAX_PROGRAM_DECODES); + } + #[test] fn the_full_program_is_written_compressed_and_another_page_reuses_it() { let bytes = include_bytes!("../../../assets/fonts/Inter-Regular.ttf");