From ab0e5fa4256095e8d027aca65664b550d2b4ac0e Mon Sep 17 00:00:00 2001
From: "James R. Barlow" <jim@purplerock.ca>
Date: Fri, 19 Feb 2016 03:58:39 -0800
Subject: [PATCH] Improve error checking for tesseract -psm 0 (orientation)
 errors

---
 ocrmypdf/tesseract.py | 17 +++++++----------
 tests/test_main.py    |  9 +++++++++
 2 files changed, 16 insertions(+), 10 deletions(-)

diff --git a/ocrmypdf/tesseract.py b/ocrmypdf/tesseract.py
index 0e1a60cf..0a205621 100644
--- a/ocrmypdf/tesseract.py
+++ b/ocrmypdf/tesseract.py
@@ -90,19 +90,16 @@ def get_orientation(input_file, language: list, timeout: float, log):
         'stdout'
     ]
 
-    p = Popen(args_tesseract, close_fds=True, stdout=PIPE, stderr=STDOUT,
-              universal_newlines=True)
     try:
-        stdout, _ = p.communicate(timeout=timeout)
+        stdout = check_output(
+            args_tesseract, close_fds=True, stderr=STDOUT,
+            universal_newlines=True, timeout=timeout)
     except TimeoutExpired:
-        p.kill()
-        stdout, _ = p.communicate()
         return OrientationConfidence(angle=0, confidence=0.0)
+    except CalledProcessError as e:
+        tesseract_log_output(log, e.output, input_file)
+        raise e from e
     else:
-        if p.returncode != 0:
-            log.error(stdout)
-            return OrientationConfidence(angle=0, confidence=0.0)
-
         osd = {}
         for line in stdout.splitlines():
             line = line.strip()
@@ -128,7 +125,7 @@ def tesseract_log_output(log, stdout, input_file):
             log.warning(prefix + "lots of diacritics - possibly poor OCR")
         elif line.startswith('OSD: Weak margin'):
             log.warning(prefix + "unsure about page orientation")
-        elif 'error' in line.lower():
+        elif 'error' in line.lower() or 'exception' in line.lower():
             log.error(prefix + line.strip())
         else:
             log.info(prefix + line.strip())
diff --git a/tests/test_main.py b/tests/test_main.py
index 217908b8..c6381137 100644
--- a/tests/test_main.py
+++ b/tests/test_main.py
@@ -460,3 +460,12 @@ def test_tesseract_crash(renderer, spoof_tesseract_crash):
     assert sh.returncode == ExitCode.child_process_error
     assert not os.path.exists(_outfile('wontwork.pdf'))
     assert "ERROR" in err
+
+
+def test_tesseract_crash_autorotate(spoof_tesseract_crash):
+    sh, out, err = run_ocrmypdf_env(
+        'ccitt.pdf', 'wontwork.pdf',
+        '-r', env=spoof_tesseract_crash)
+    assert sh.returncode == ExitCode.child_process_error
+    assert not os.path.exists(_outfile('wontwork.pdf'))
+    assert "ERROR" in err