From ddf109c3328e811604ec75a9834861aefc0d6b40 Mon Sep 17 00:00:00 2001
From: Pompolic <pompolic@special-circumstanc.es>
Date: Mon, 21 Feb 2022 18:51:40 +0100
Subject: [PATCH] Workaround: avoid crashing if /Contents in a page object is
an indirect array
---
pdf.c | 29 +++++++++++++++++++++--------
1 file changed, 21 insertions(+), 8 deletions(-)
diff --git a/pdf.c b/pdf.c
index 817fdcd..b78944e 100644
--- a/pdf.c
+++ b/pdf.c
@@ -4747,20 +4747,33 @@ void parse_pagenode (
parse_rsrcdict(arena, contents, myNode, aux);
//pp_ptnode(stdout, myNode);
}
+ /* Indirect reference to an array, which may in turn have indirect references */
+ // XXX: possible refactor: the branch with contents_t being a TT_SEQUENCE is meant to do the same thing, could be factored out into a function of its own
+ else if (contents->token_type == TT_SEQUENCE)
+ {
+ myNode->pn.textStream = NULL;
+ goto end; //XXX: This avoids crashing
+ }
else {
// DEBUG
//fprintf(stdout, "\n\nparse_pagenode: What is token 0 anyway?\n");
- HParsedToken *tok0 = H_INDEX_TOKEN(contents, 0);
- h_pprintln(stdout, tok0);
+ if(contents->seq->used >= 1)
+ {
+ HParsedToken *tok0 = H_INDEX_TOKEN(contents, 0);
+ h_pprintln(stdout, tok0);
+ }
- HParsedToken *res_strm = H_INDEX_TOKEN(contents, 1);
- if (res_strm->token_type == TT_SEQUENCE) { // this seems like a big assumption
- myNode->pn.textStream = res_strm;
+ if(contents->seq->used >= 2)
+ {
+ HParsedToken *res_strm = H_INDEX_TOKEN(contents, 1);
+ if (res_strm->token_type == TT_SEQUENCE) { // this seems like a big assumption
+ myNode->pn.textStream = res_strm;
- //fprintf(stdout, "parse_pagenode: Page node contents = %p\n", (void *)contents);
+ //fprintf(stdout, "parse_pagenode: Page node contents = %p\n", (void *)contents);
+ }
+ else
+ myNode->pn.textStream = NULL;
}
- else
- myNode->pn.textStream = NULL;
}
}
else if (contents_t->token_type == TT_SEQUENCE) {
--
GitLab