added ocr module for abbyy

src/filters/rclocr.py

@@ -17,7 +17,12 @@
 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 ########################################################
 
-# Running OCR programs for Recoll
+# Running OCR programs for Recoll. This is excecuted from,
+# e.g. rclpdf.py if pdftotext returns no data.
+#
+# The script tries to retrieve the data from the ocr cache, else it
+# runs the configured OCR program and updates the cache. In both cases it writes
+# the resulting text to stdout.
 
 import os
 import sys
@@ -47,7 +52,8 @@ if incache:
     
 #### Data not in cache
 
-# Retrieve known ocr program names and try to load the corresponding module
+# Retrieve configured OCR program names and try to load the
+# corresponding module
 ocrprogs = config.getConfParam("ocrprogs")
 if not ocrprogs:
     _deb("No ocrprogs variable in recoll configuration")
@@ -59,7 +65,7 @@ for ocrprog in proglist:
     try:
         modulename = "rclocr" + ocrprog
         ocr = importlib.import_module(modulename)
-        if ocr.ocrpossible(path):
+        if ocr.ocrpossible(config, path):
             ok = True
             break
     except Exception as err:
@@ -72,6 +78,8 @@ if not ok:
 
 #_deb("Using ocr module %s" % modulename)
 
+# The OCR module will retrieve its specific parameters from the
+# configuration
 status, data = ocr.runocr(config, path)
 
 if not status:

src/filters/rclocrabbyy.py

@@ -0,0 +1,147 @@
+#!/usr/bin/env python3
+#################################
+# Copyright (C) 2020 J.F.Dockes
+#   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 2 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, write to the
+#   Free Software Foundation, Inc.,
+# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
+########################################################
+
+# Running abbyyocr for Recoll OCR (see rclocr.py)
+
+import os
+import sys
+import atexit
+import tempfile
+import subprocess
+import glob
+
+import rclexecm
+
+_mswindows = (sys.platform == "win32")
+if _mswindows:
+    ocrlangfile = "rclocrlang.txt"
+else:
+    ocrlangfile = ".rclocrlang"
+
+_okexts = ('.pdf', '.tif', '.tiff', '.jpg', '.png', '.jpeg')
+
+abbyyocrcmd = ""
+abbyocrdir = ""
+
+def _deb(s):
+    if not _mswindows:
+        print("rclocrabbyy: %s" % s, file=sys.stderr)
+
+# Return true if abbyy appears to be available
+def ocrpossible(config, path):
+    global abbyyocrcmd
+    if not abbyyocrcmd:
+        config.setKeyDir(os.path.dirname(path))
+        abbyyocrcmd = config.getConfParam("abbyyocrcmd")
+        if not abbyyocrcmd:
+            abbyyocrcmd = rclexecm.which("abbyyocr11")
+        if not abbyyocrcmd:
+            return False
+        global abbyyocrdir
+        abbyyocrdir = os.path.dirname(abbyyocrcmd)
+    
+    # Check input format
+    base,ext = os.path.splitext(path)
+    ext = ext.lower()
+    if ext in _okexts:
+        return True
+    return False
+
+
+# Try to guess tesseract language. This should depend on the input
+# file, but we have no general way to determine it. So use the
+# environment and hope for the best.
+def _guessocrlang(config, path):
+    ocrlang = ""
+
+    dirname = os.path.dirname(path)
+
+    # First look for a language def file in the file's directory
+    langfile = os.path.join(dirname, ocrlangfile)
+    if os.path.isfile(langfile):
+        ocrlang = open(langfile, "r").read().strip()
+    if ocrlang:
+        _deb("OCR lang from file: %s" % ocrlang)
+        return ocrlang
+
+    # Then look for a config file  option.
+    config.setKeyDir(dirname)
+    ocrlang = config.getConfParam("abbyylang")
+    if ocrlang:
+        _deb("OCR lang from config: %s" % ocrlang)
+        return ocrlang
+
+    # Half-assed trial to guess from LANG then default to english
+    try:
+        localelang = os.environ.get("LANG", "").split("_")[0]
+        if localelang == "en":
+            ocrlang = "English"
+        elif localelang == "de":
+            ocrlang = "German"
+        elif localelang == "fr":
+            ocrlang = "French"
+    except:
+        pass
+
+    if not ocrlang:
+        ocrlang = "English"
+    _deb("OCR lang (guessed): %s" % ocrlang)
+    return ocrlang
+
+
+# run ocr on the input path and output the result data.
+def runocr(config, path):
+    ocrlang = _guessocrlang(config, path)
+    my_env = os.environ.copy()
+    eldpn = "LD_LIBRARY_PATH"
+    if eldpn in my_env:
+        oldpath = ":" + my_env[eldpn]
+    else:
+        oldpath = ""
+    my_env[eldpn] = abbyyocrdir + oldpath
+
+    try:
+        out = subprocess.check_output(
+            [abbyyocrcmd, "-lpp", "BookArchiving_Accuracy",
+             "-rl", ocrlang,
+             "-tet", "UTF8",
+             "-f", "TextUnicodeDefaults",
+             "-if", path,
+             "-c"],
+            env = my_env,
+            stderr=subprocess.DEVNULL)
+    except Exception as e:
+        _deb("%s failed: %s" % (abbyyocrcmd,e))
+        return False, ""
+    return True, out
+            
+
+if __name__ == '__main__':
+    import rclconfig
+    config = rclconfig.RclConfig()
+    path =  sys.argv[1]
+    if ocrpossible(config, path):
+        ok, data = runocr(config, sys.argv[1])
+    else:
+        _deb("ocrpossible returned false")
+        sys.exit(1)
+    if ok:
+        sys.stdout.buffer.write(data)
+    else:
+        _deb("OCR program failed")

src/filters/rclocrtesseract.py

@@ -17,6 +17,8 @@
 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 ########################################################
 
+# Running tesseract for Recoll OCR (see rclocr.py)
+
 import os
 import sys
 import atexit
@@ -28,7 +30,7 @@ import rclexecm
 
 _mswindows = (sys.platform == "win32")
 if _mswindows:
-    ocrlangfile = ".rclocrlang"
+    ocrlangfile = "rclocrlang.txt"
 else:
     ocrlangfile = ".rclocrlang"
 
@@ -65,7 +67,7 @@ atexit.register(finalcleanup)
 
 # Return true if tesseract and the appropriate conversion program for
 # the file type (e.g. pdftoppt for pdf) appear to be available
-def ocrpossible(path):
+def ocrpossible(config, path):
     # Check for tesseract
     global tesseract
     tesseract = rclexecm.which("tesseract")
@@ -217,11 +219,12 @@ if __name__ == '__main__':
     import rclconfig
     config = rclconfig.RclConfig()
     path =  sys.argv[1]
-    if ocrpossible(path):
+    if ocrpossible(config, path):
-        data = runocr(config, sys.argv[1])
+        ok, data = runocr(config, sys.argv[1])
     else:
         _deb("ocrpossible returned false")
         sys.exit(1)
-    sys.stdout.buffer.write(data)
+    if ok:
-    
+        sys.stdout.buffer.write(data)
-
+    else:
+        _deb("OCR program failed")