Add workaround for tesseract 4.0

It needs to have C locale for it's execution due to code relying on
locale dependant functions for parsing trained data (eg. sscanf).

This is really a workaround, but given that tesseract 4.0 is shipped
with upcoming Debian stable, we will have to live with this for quite
some time.

Fixes #2581

Signed-off-by: Michal Čihař <[email protected]>

docs/changes.rst

@@ -20,6 +20,7 @@ Released on ? 2019.
 * Added check for Kashida letters.
 * Added option to squash commits based on authors.
 * Improved support for xlsx file format.
+* Compatibility with tesseract 4.0.
 
 weblate 3.4
 -----------

weblate/screenshots/views.py

@@ -30,8 +30,11 @@
 
 from PIL import Image
 
+from weblate.utils.locale import c_locale
+
 try:
-    from tesserocr import PyTessBaseAPI, RIL
+    with c_locale():
+        from tesserocr import PyTessBaseAPI, RIL
     HAS_OCR = True
 except ImportError:
     HAS_OCR = False
@@ -273,7 +276,7 @@ def ocr_search(request, pk):
     results = set()
 
     # Extract and match strings
-    with PyTessBaseAPI() as api:
+    with c_locale(), PyTessBaseAPI() as api:
         for image in (original_image, scaled_image):
             for match in ocr_extract(api, image, strings):
                 results.add(sources[match])

weblate/utils/locale.py

@@ -0,0 +1,36 @@
+# -*- coding: utf-8 -*-
+#
+# Copyright © 2012 - 2019 Michal Čihař <[email protected]>
+#
+# This file is part of Weblate <https://weblate.org/>
+#
+# This program is free software: you can redistribute it and/or modify
+# it under the terms of the GNU General Public License as published by
+# the Free Software Foundation, either version 3 of the License, or
+# (at your option) any later version.
+#
+# This program is distributed in the hope that it will be useful,
+# but WITHOUT ANY WARRANTY; without even the implied warranty of
+# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+# GNU General Public License for more details.
+#
+# You should have received a copy of the GNU General Public License
+# along with this program.  If not, see <https://www.gnu.org/licenses/>.
+#
+
+from __future__ import absolute_import
+
+from locale import setlocale, getlocale, LC_ALL
+from contextlib import contextmanager
+
+
+@contextmanager
+def c_locale():
+    """Context to execute something in C locale."""
+    try:
+        currlocale = getlocale()
+    except ValueError:
+        currlocale = ('C', 'UTF-8')
+    setlocale(LC_ALL, "C")
+    yield
+    setlocale(LC_ALL, currlocale)