Fix: Tesseract 3.04 is sensitive to order of configuration commands

“txt hocr” is not acceptable and does not produce expected output .txt
while “hocr text” works fine, so switch the order everywhere.

Should fix #169

ocrmypdf/exec/tesseract.py

@@ -43,7 +43,6 @@ HOCR_TEMPLATE = '''<?xml version="1.0" encoding="UTF-8"?>
  </body>
 </html>'''
 
-
 @lru_cache(maxsize=1)
 def version():
     args_tess = [
@@ -229,11 +228,13 @@ def generate_hocr(input_file, output_files, language: list, engine_mode,
 
     # Reminder: test suite tesseract spoofers will break after any changes
     # to the number of order parameters here
+    # Tesseract 3.04 requires the order here to be "hocr txt" and will fail
+    # on "txt hocr"
     args_tesseract.extend([
         input_file,
         prefix,
-        'txt',
-        'hocr'
+        'hocr',
+        'txt'
     ] + tessconfig)
     try:
         log.debug(args_tesseract)
@@ -316,8 +317,8 @@ def generate_pdf(*, input_image, skip_pdf, output_pdf, output_text,
     args_tesseract.extend([
         input_image,
         prefix,
-        'txt',
         'pdf',
+        'txt'
     ] + tessconfig)
 
     try:

tests/spoof/tesseract_big_image_error.py

@@ -20,11 +20,11 @@ def main():
     elif sys.argv[1] == '--list-langs':
         print('List of available languages (1):\neng', file=sys.stderr)
         sys.exit(0)
-    elif sys.argv[-1] == 'hocr':
+    elif sys.argv[-2] == 'hocr':
         print("Image too large: (33830, 14959)\n"
               "Error during processing.", file=sys.stderr)
         sys.exit(1)
-    elif sys.argv[-1] == 'pdf':
+    elif sys.argv[-2] == 'pdf':
         print("Image too large: (33830, 14959)\n"
               "Error during processing.", file=sys.stderr)
         sys.exit(1)

tests/spoof/tesseract_cache.py

@@ -35,10 +35,12 @@ def real_tesseract():
     return  # Not reachable
 
 def main():
-    operation = sys.argv[-1]
+    operation = sys.argv[-2]
     sidecar = False
-    if sys.argv[-2] == 'txt':
+    if sys.argv[-1] == 'txt':
         sidecar = True
+    elif sys.argv[-1] == 'stdout':
+        operation = 'stdout'
 
     # For anything unexpected operation, defer to real tesseract binary
     # Currently this includes all use of "--tesseract-config"

tests/spoof/tesseract_crash.py

@@ -28,10 +28,10 @@ def main():
     elif sys.argv[1] == '--list-langs':
         print('List of available languages (1):\neng', file=sys.stderr)
         sys.exit(0)
-    elif sys.argv[-1] == 'hocr':
+    elif sys.argv[-2] == 'hocr':
         print("KABOOM! Tesseract failed for some reason", file=sys.stderr)
         sys.exit(128 + signal.SIGSEGV)
-    elif sys.argv[-1] == 'pdf':
+    elif sys.argv[-2] == 'pdf':
         print("KABOOM! Tesseract failed for some reason", file=sys.stderr)
         sys.exit(128 + signal.SIGSEGV)
     elif sys.argv[-1] == 'stdout':

tests/spoof/tesseract_noop.py

@@ -52,7 +52,7 @@ def main():
     elif sys.argv[1] == '--list-langs':
         print('List of available languages (1):\neng', file=sys.stderr)
         sys.exit(0)
-    elif sys.argv[-1] == 'hocr':
+    elif sys.argv[-2] == 'hocr':
         inputf = sys.argv[-4]
         output = sys.argv[-3]
         with Image.open(inputf) as im, \
@@ -61,7 +61,7 @@ def main():
             f.write(HOCR_TEMPLATE.format(str(w), str(h)))
         with open(output + '.txt', 'w') as f:
             f.write('')
-    elif sys.argv[-1] == 'pdf':
+    elif sys.argv[-2] == 'pdf':
         inputf = sys.argv[-4]
         output = sys.argv[-3]
         pdf_bytes = img2pdf.convert([inputf], dpi=300)