Merge pull request #136 from maxpmaxp/issue-135-linearized-pdf-traile…

…r-parsing-issue

ISSUE-135: linearized PDF trailer parsing issue

CHANGELOG.txt

@@ -1,3 +1,7 @@
+pdfreader 0.1.16dev
+---------------------------
+ - issue #135 - trailer parsing issue for linearized PDFs
+
 pdfreader 0.1.15
 ---------------------------
 - don't raise error on unsupported stream filters, return raw unfiltered data instead

pdfreader/parsers/document.py

@@ -1,3 +1,4 @@
+import logging
 import re
 
 from ..constants import WHITESPACE_CODES, EOL, SP
@@ -558,14 +559,19 @@ def locate_backwards_from_trailer(self, num, gen):
         chunk_len = len(expected_obj)
         obj_header = self.read(chunk_len)
 
-        while obj_header != expected_obj:
-            self.skip_backwards_until(b"endobj")
-            self.skip_backwards_until(b"endobj")
-            self.read(6)
-            self.maybe_spaces_or_comments()
-            obj_header = self.read(chunk_len)
-            if len(obj_header) < chunk_len:
-                return
+        try:
+            while obj_header != expected_obj:
+                self.skip_backwards_until(b"endobj")
+                self.skip_backwards_until(b"endobj")
+                self.read(6)
+                self.maybe_spaces_or_comments()
+                obj_header = self.read(chunk_len)
+                if len(obj_header) < chunk_len:
+                    return
+        except TypeError:
+            # Issue 135: https://github.com/maxpmaxp/pdfreader/issues/135
+            logging.debug("Failed to locate backwards from trailer")
+            return
 
         self.read_backward(chunk_len)
         obj = self.indirect_object()

setup.py

@@ -9,7 +9,7 @@
 if sys.version_info[:2] < (3, 6):
     warnings.warn("Python version >= 3.6 required.")
 
-version = '0.1.15'
+version = '0.1.16dev'
 
 
 import os.path