From b48d7f27b6a8f71331d25029d61b46b815762b3d Mon Sep 17 00:00:00 2001
From: "plvines (corpora)" <paul.vines@baesystems.com>
Date: Sat, 15 Feb 2020 22:29:57 +0000
Subject: [PATCH] Added relaxation/violation for ANY DATA (not just newlines)
after final %%EOF
---
pdf.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/pdf.c b/pdf.c
index 96f6879..01d13bc 100644
--- a/pdf.c
+++ b/pdf.c
@@ -700,9 +700,11 @@ init_parser(struct Env *aux)
// XXX should lws be allowed before EOF marker?
// NB: lws before xref offset is allowed, cf. p.48 (example 4)
LIT("%%EOF"),
- CHX(VIOL(SEQ(nl, h_many1(nl)),
+ CHX(VIOL(SEQ(nl, h_many1(nl), end),
"(offset FROM END) Multiple newlines after final %%EOF (severity=4)"),
- h_many(nl))));
+ SEQ(h_many(nl), end),
+ VIOL(SEQ(h_many1(h_ch_range(0, 255)), end),
+ "(offset FROM END) Data after final %%EOF (severity=7)"))));
H_RULE(xr_td, SEQ(xrefs, KW("trailer"), ws, dict));
@@ -711,12 +713,13 @@ init_parser(struct Env *aux)
H_RULE(tail, SEQ(body, h_optional(xr_td), startxr));
// XXX the real world likes to omit 'startxr' from all but the
// last trailer. we should accept-and-warn in that case.
- H_RULE(final_eof_junk, CHX(VIOL(h_many1(nl), "Multiple newlines after final %%EOF (severity=4)"),
+ H_RULE(final_eof_junk, CHX(VIOL(SEQ(h_many1(nl), end), "Multiple newlines after final %%EOF (severity=4)"),
+ VIOL(h_many1(h_ch_range(0, 255)), "Data after final %%EOF (severity=7)"),
end));
H_RULE(pdf, SEQ(header, OPT(hdr_junk), h_many1(tail), final_eof_junk));
/* debug parser to consume as much as possible */
- H_RULE(pdfdbg, SEQ(header, h_many(tail), body, OPT(xr_td), OPT(startxr)));
+ H_RULE(pdfdbg, SEQ(header, OPT(hdr_junk), h_many(tail), body, OPT(xr_td), OPT(startxr), final_eof_junk));
/* global parser variables */
--
GitLab