From 6d0461435f9908025bb87749a89aa2988625966a Mon Sep 17 00:00:00 2001
From: "James R. Barlow" <jim@purplerock.ca>
Date: Thu, 24 May 2018 22:52:33 -0700
Subject: [PATCH] Use OperandGrouper whitelist

---
 src/ocrmypdf/pdfinfo.py | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/src/ocrmypdf/pdfinfo.py b/src/ocrmypdf/pdfinfo.py
index 4bf7b036..21b24294 100644
--- a/src/ocrmypdf/pdfinfo.py
+++ b/src/ocrmypdf/pdfinfo.py
@@ -150,8 +150,10 @@ def _interpret_contents(contentstream, initial_shorthand=UNIT_SQUARE):
     found_text = False
     text_operators = tuple(
         pikepdf.Operator(op) for op in ('Tj', 'TJ', '"', "'"))
+    operator_whitelist = """q Q Do cm TJ Tj " ' BI ID EI"""
 
-    for n, op in enumerate(pikepdf.parse_content_stream(contentstream)):
+    for n, op in enumerate(
+            pikepdf.parse_content_stream(contentstream, operator_whitelist)):
         operands, command = op
         if command == pikepdf.Operator('q'):
             stack.append(ctm)