added return end_obj method from process page so that we can iterate …

pdfminer/pdfinterp.py

@@ -837,8 +837,7 @@ def process_page(self, page):
             ctm = (1, 0, 0, 1, -x0, -y0)
         self.device.begin_page(page, ctm)
         self.render_contents(page.resources, page.contents, ctm=ctm)
-        self.device.end_page(page)
-        return
+        return self.device.end_page(page)
 
     # render_contents(resources, streams, ctm)
     #   Render the content streams.

pdfminer/pdfparser.py

@@ -52,6 +52,7 @@ def set_document(self, doc):
 
     KEYWORD_R = KWD(b'R')
     KEYWORD_NULL = KWD(b'null')
+    KEYWORD_OBJ = KWD(b'obj')
     KEYWORD_ENDOBJ = KWD(b'endobj')
     KEYWORD_STREAM = KWD(b'stream')
     KEYWORD_XREF = KWD(b'xref')
@@ -63,6 +64,14 @@ def do_keyword(self, pos, token):
         if token in (self.KEYWORD_XREF, self.KEYWORD_STARTXREF):
             self.add_results(*self.pop(1))
 
+        elif token is self.KEYWORD_OBJ:
+            # Handle implicit endobj in some bugged PDFs
+            if len(self.curstack) > 2:
+                stack = self.popall()
+                self.add_results(*stack[0:-2])
+                self.push(*stack[-2:])
+            self.push((pos, token))
+
         elif token is self.KEYWORD_ENDOBJ:
             self.add_results(*self.pop(4))