aleleba/pdfcraft
1
0
Fork 0
mirror of https://github.com/storytold/pdfcraft.git synced 2026-10-28 12:08:46 -06:00
Code Issues Packages Projects Releases Wiki Activity

fix(render): read an annotation attachment's Contents only when it names it

Browse Source 
An annotation attachment's /Contents was read and charged to the text
budget before the file specification's /UF or /F, which then named the
attachment, so the discarded text could leave a later web address
unlisted. /Contents is now read only when neither name is there.
This commit is contained in:
aiongg
2026-10-22 22:02:10 +00:00
parent aaa4b99eb9
commit a0a66b982b
1 changed files with 41 additions and 5 deletions
Download Patch File Download Diff File Expand all files Collapse all files
crates/render/src/inspect.rs
+41 -5
View File
@@ -746,7 +746,8 @@ impl<'a> Inspector<'a> {
if let Some(ef) = names.get(b"EmbeddedFiles").ok().and_then(|o| self.dict(o)) {
let mut seen = HashSet::new();
for (name, spec) in self.name_tree(ef, &mut seen, 0) {
info.attachments.push(self.attachment(name.clone(), spec, AttachmentSource::Document { key: name }, &mut text_budget));
let fallback = |_: &mut usize| Some(name.clone());
info.attachments.push(self.attachment(spec, AttachmentSource::Document { key: name.clone() }, &mut text_budget, fallback));
}
}
info.has_javascript |= names.get(b"JavaScript").is_ok();
@@ -1212,8 +1213,8 @@ impl<'a> Inspector<'a> {
if subtype == "FileAttachment"
&& let Ok(fs) = d.get(b"FS")
{
let name = self.budgeted_text(d, b"Contents", text_budget).unwrap_or_else(|| "attachment".into());
info.attachments.push(self.attachment(name, fs, AttachmentSource::Annotation { page, index }, text_budget));
let fallback = |budget: &mut usize| self.budgeted_text(d, b"Contents", budget);
info.attachments.push(self.attachment(fs, AttachmentSource::Annotation { page, index }, text_budget, fallback));
}
if subtype == "Link" {
if let Some(link) = self.link(page, index, d, text_budget) {
@@ -1652,7 +1653,14 @@ impl<'a> Inspector<'a> {
out.sort_by(|a, b| a.name.cmp(&b.name));
}
fn attachment(&self, name: String, spec: &Object, source: AttachmentSource, text_budget: &mut usize) -> Attachment {
/// The attachment's name is the file specification's, else `fallback`'s, else a generic one.
fn attachment(
&self,
spec: &Object,
source: AttachmentSource,
text_budget: &mut usize,
fallback: impl FnOnce(&mut usize) -> Option<String>,
) -> Attachment {
let d = self.dict(spec);
let size =
d.and_then(|d| d.get(b"EF").ok()).and_then(|o| self.dict(o)).and_then(|ef| ef.get(b"F").ok().or_else(|| ef.get(b"UF").ok())).and_then(
@@ -1669,7 +1677,10 @@ impl<'a> Inspector<'a> {
_ => None,
},
);
let display = d.and_then(|d| self.budgeted_text(d, b"UF", text_budget).or_else(|| self.budgeted_text(d, b"F", text_budget))).unwrap_or(name);
let display = d
.and_then(|d| self.budgeted_text(d, b"UF", text_budget).or_else(|| self.budgeted_text(d, b"F", text_budget)))
.or_else(|| fallback(text_budget))
.unwrap_or_else(|| "attachment".into());
Attachment { name: display, description: d.and_then(|d| self.budgeted_text(d, b"Desc", text_budget)), size, source }
}
}
@@ -2761,6 +2772,31 @@ trailer << /Root 1 0 R >>
assert_eq!(budget, 7);
}
/// An annotation attachment's `/Contents` spends the text budget only when it names the
/// attachment, so long descriptions of named files leave a later web address listed.
#[test]
fn a_named_attachment_leaves_its_contents_unspent() {
let files = "<< /Subtype /FileAttachment /Rect [0 0 1 1] /Contents 6 0 R /FS << /F (a.txt) >> >> ".repeat(20);
// Longer than what 20 MiB of charged `/Contents` would leave of the 16 MiB budget.
let uri = format!("https://example.com/{}", "a".repeat(2 << 20));
let pdf = format!(
"%PDF-1.7\n1 0 obj << /Type /Catalog /Pages 2 0 R >> endobj\n\
2 0 obj << /Type /Pages /Kids [3 0 R] /Count 1 >> endobj\n\
3 0 obj << /Type /Page /Parent 2 0 R /MediaBox [0 0 200 200] /Annots [{files} \
<< /Subtype /Link /Rect [0 0 1 1] /A << /S /URI /URI ({uri}) >> >>] >> endobj\n\
6 0 obj ({}) endobj\n\
trailer << /Root 1 0 R >>\n%%EOF",
"x".repeat(1 << 20),
);
let doc = Document::load_mem(pdf.as_bytes()).expect("loads");
let page = PageInfo { width: 200.0, height: 200.0, crop: [0.0, 0.0, 200.0, 200.0], rotation: 0, label: String::new() };
let mut info = DocInfo { pages: vec![page], ..Default::default() };
Inspector::new(&doc).fill(&mut info);
assert!(info.attachments.iter().all(|a| a.name == "a.txt"));
assert_eq!(info.links.len(), 1);
assert!(matches!(&info.links[0].target, LinkTarget::Uri(u) if *u == uri));
}
/// Links and attachments sharing one long string each keep a copy only within
/// `MAX_LINK_ATTACHMENT_TEXT_BYTES`, and every one is still listed.
#[test]