From 3ce8fe79270e9a58ec17eb9e837f2cfdbc048322 Mon Sep 17 00:00:00 2001
From: Jonas Jenwald <jonas.jenwald@gmail.com>
Date: Mon, 4 Mar 2019 23:17:21 +0100
Subject: [PATCH] Handle corrupt ASCII85Decode inline images with whitespace
 "inside" of the EOD marker (issue 10614)

There's a number of things wrong with the PDF document, since its inline images are first all *a lot* larger than the 4 KB limit (as mandated by the specification, see https://www.adobe.com/content/dam/acom/en/devnet/acrobat/pdfs/PDF32000_2008.pdf#G7.1852045).

Furthermore the actual ASCII85Decode data is interspersed with *a lot* of needless whitespace, in particular also "inside" of the EOD (end-of-data) marker which thus completely breaks the detection.
Note that according to the specification, see https://www.adobe.com/content/dam/acom/en/devnet/acrobat/pdfs/PDF32000_2008.pdf#G6.1940130, this patch should be safe since it explicitly mentions that *all* whitespace should be ignored.
---
 src/core/parser.js            | 18 +++++++++++++-----
 test/pdfs/issue10614.pdf.link |  1 +
 test/test_manifest.json       |  8 ++++++++
 3 files changed, 22 insertions(+), 5 deletions(-)
 create mode 100644 test/pdfs/issue10614.pdf.link

diff --git a/src/core/parser.js b/src/core/parser.js
index c01001fca..8eb8a578c 100644
--- a/src/core/parser.js
+++ b/src/core/parser.js
@@ -336,14 +336,22 @@ var Parser = (function ParserClosure() {
      * Find the EOD (end-of-data) marker '~>' (i.e. TILDE + GT) of the stream.
      * @returns {number} The inline stream length.
      */
-    findASCII85DecodeInlineStreamEnd:
-        function Parser_findASCII85DecodeInlineStreamEnd(stream) {
+    findASCII85DecodeInlineStreamEnd(stream) {
       var TILDE = 0x7E, GT = 0x3E;
       var startPos = stream.pos, ch, length;
       while ((ch = stream.getByte()) !== -1) {
-        if (ch === TILDE && stream.peekByte() === GT) {
-          stream.skip();
-          break;
+        if (ch === TILDE) {
+          ch = stream.peekByte();
+          // Handle corrupt PDF documents which contains whitespace "inside" of
+          // the EOD marker (fixes issue10614.pdf).
+          while (isSpace(ch)) {
+            stream.skip();
+            ch = stream.peekByte();
+          }
+          if (ch === GT) {
+            stream.skip();
+            break;
+          }
         }
       }
       length = stream.pos - startPos;
diff --git a/test/pdfs/issue10614.pdf.link b/test/pdfs/issue10614.pdf.link
new file mode 100644
index 000000000..203d534fa
--- /dev/null
+++ b/test/pdfs/issue10614.pdf.link
@@ -0,0 +1 @@
+https://github.com/mozilla/pdf.js/files/2927954/FOX.ALEXANDER.F.VS.FOX.BARBARA.E.2015-004684-FC-04.Doc-46-Memorandum-of-Law.Fla.11th.Cir.Ct.May.31.2016.pdf
diff --git a/test/test_manifest.json b/test/test_manifest.json
index 5baf33905..b5d8649f0 100644
--- a/test/test_manifest.json
+++ b/test/test_manifest.json
@@ -1465,6 +1465,14 @@
        "link": false,
        "type": "text"
     },
+    {  "id": "issue10614",
+       "file": "pdfs/issue10614.pdf",
+       "md5": "c41da60ce9af100cb78e1c2a6ba18232",
+       "rounds": 1,
+       "link": true,
+       "lastPage": 1,
+       "type": "eq"
+    },
     {  "id": "issue6071",
        "file": "pdfs/issue6071.pdf",
        "md5": "2e08526d8e7c9ba4269fc12ef488d3eb",