Merge branch 'feature/pagesegmode' into develop

ocrmypdf/main.py

@@ -170,13 +170,16 @@ def check_pil_encoder(codec_name, friendly_name):
     "Advanced",
     "Advanced options for power users")
 advanced.add_argument(
-    '--tesseract-config', default=[], type=list, action='append',
+    '--tesseract-config', action='append', metavar='CFG', default=[],
     help="additional Tesseract configuration files")
+advanced.add_argument(
+    '--tesseract-pagesegmode', action='store', type=int, metavar='PSM',
+    help="set Tesseract page segmentation mode (see tesseract --help)")
 advanced.add_argument(
     '--pdf-renderer', choices=['auto', 'tesseract', 'hocr'], default='auto',
     help='choose OCR PDF renderer')
 advanced.add_argument(
-    '--tesseract-timeout', default=180.0, type=float,
+    '--tesseract-timeout', default=180.0, type=float, metavar='SECONDS',
     help='give up on OCR after the timeout, but copy the preprocessed page '
          'into the final output')
 
@@ -529,6 +532,7 @@ def ocr_tesseract_hocr(
         timeout=options.tesseract_timeout,
         pageinfo_getter=partial(get_pageinfo, input_file, pdfinfo,
                                 pdfinfo_lock),
+        pagesegmode=options.tesseract_pagesegmode,
         log=log
         )
 
@@ -635,6 +639,7 @@ def tesseract_ocr_and_render_pdf(
         language=options.language,
         tessconfig=options.tesseract_config,
         timeout=options.tesseract_timeout,
+        pagesegmode=options.tesseract_pagesegmode,
         log=log)
 
 

ocrmypdf/tesseract.py

@@ -77,17 +77,24 @@ def languages():
 
 
 def generate_hocr(input_file, output_hocr, language: list, tessconfig: list,
-                  timeout: float, pageinfo_getter, log):
+                  timeout: float, pageinfo_getter, pagesegmode: int, log):
 
     badxml = os.path.splitext(output_hocr)[0] + '.badxml'
 
     args_tesseract = [
         get_program('tesseract'),
-        '-l', '+'.join(language),
+        '-l', '+'.join(language)
+    ]
+
+    if pagesegmode is not None:
+        args_tesseract.extend(['-psm', str(pagesegmode)])
+
+    args_tesseract.extend([
         input_file,
         badxml,
         'hocr'
-    ] + tessconfig
+    ] + tessconfig)
+    print(args_tesseract)
     p = Popen(args_tesseract, close_fds=True, stdout=PIPE, stderr=PIPE,
               universal_newlines=True)
     try:
@@ -135,7 +142,7 @@ def generate_hocr(input_file, output_hocr, language: list, tessconfig: list,
 
 
 def generate_pdf(input_image, skip_pdf, output_pdf, language: list,
-                 tessconfig: list, timeout: float, log):
+                 tessconfig: list, timeout: float, pagesegmode: int, log):
     '''Use Tesseract to render a PDF.
 
     input_image -- image to analyze
@@ -148,11 +155,17 @@ def generate_pdf(input_image, skip_pdf, output_pdf, language: list,
 
     args_tesseract = [
         get_program('tesseract'),
-        '-l', '+'.join(language),
+        '-l', '+'.join(language)
+    ]
+
+    if pagesegmode is not None:
+        args_tesseract.extend(['-psm', str(pagesegmode)])
+
+    args_tesseract.extend([
         input_image,
         os.path.splitext(output_pdf)[0],  # Tesseract appends suffix
         'pdf'
-    ] + tessconfig
+    ] + tessconfig)
     p = Popen(args_tesseract, close_fds=True, stdout=PIPE, stderr=PIPE,
               universal_newlines=True)
 

tests/spoof/tesseract_cache.py

@@ -37,6 +37,11 @@ def main():
         m.update(lang.encode())
     except ValueError:
         pass
+    try:
+        psm = sys.argv[sys.argv.index('-psm') + 1]
+        m.update(psm.encode())
+    except ValueError:
+        pass
 
     input_file = sys.argv[-3]
     output_file = sys.argv[-2]

tests/test_main.py

@@ -56,8 +56,8 @@ def check_ocrmypdf(input_basename, output_basename, *args, env=None):
     input_file = _make_input(input_basename)
     output_file = _make_output(output_basename)
 
-    sh, _, err = run_ocrmypdf_sh(input_file, output_file, *args, env=env)
-    assert sh.returncode == 0, err
+    sh, out, err = run_ocrmypdf_sh(input_file, output_file, *args, env=env)
+    assert sh.returncode == 0, dict(stdout=out, stderr=err)
     assert os.path.exists(output_file), "Output file not created"
     assert os.stat(output_file).st_size > 100, "PDF too small or empty"
     return output_file
@@ -348,3 +348,18 @@ def test_encrypted():
     p, out, err = run_ocrmypdf_env('skew-encrypted.pdf', 'wont_be_created.pdf')
     assert p.returncode == ExitCode.input_file
     assert out.find('password')
+
+
+@pytest.mark.parametrize('renderer', [
+    'hocr',
+    'tesseract',
+    ])
+def test_pagesegmode(renderer, spoof_tesseract_cache):
+    check_ocrmypdf(
+        'skew.pdf', 'test_psm_%s.pdf' % renderer,
+        '--tesseract-pagesegmode', '7',
+        '-v', '1',
+        '--pdf-renderer', renderer, env=spoof_tesseract_cache)
+
+
+