matthias/mayan-edms
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Improved tesseract execution handling

Browse Source
This commit is contained in:
Roberto Rosario
2011-02-17 23:31:54 -04:00
parent 5e5fee7330
commit c18cb099c6
1 changed files with 21 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
apps/ocr/api.py
View File

@@ -1,5 +1,6 @@
#Some code from http://wiki.github.com/hoffstaetter/python-tesseract #Some code from http://wiki.github.com/hoffstaetter/python-tesseract
import codecs
import os import os
import subprocess import subprocess
import tempfile import tempfile
@@ -14,6 +15,11 @@ from documents.models import Document
from converter.api import convert_document_for_ocr from converter.api import convert_document_for_ocr
from ocr.conf.settings import TESSERACT_PATH from ocr.conf.settings import TESSERACT_PATH
from ocr.conf.settings import TESSERACT_LANGUAGE
class TesseractError(Exception):
pass
def cleanup(filename): def cleanup(filename):
@@ -23,51 +29,37 @@ def cleanup(filename):
except OSError: except OSError:
pass pass
class TesseractError(Exception):
pass
# def __init__(self, status, message):
# self.status = status
# self.message = message
def get_errors(error_string):
'''
returns all lines in the error_string that start with the string "error"
'''
lines = error_string.splitlines()
return lines[1]
#error_lines = (line for line in lines if line.find('error') >= 0)
#return '\n'.join(error_lines)
def run_tesseract(input_filename, output_filename_base, lang=None): def run_tesseract(input_filename, output_filename_base, lang=None):
command = [TESSERACT_PATH, input_filename, output_filename_base] command = [TESSERACT_PATH, input_filename, output_filename_base]
if lang is not None: if lang is not None:
command += ['-l', lang] command += ['-l', lang]
proc = subprocess.Popen(command, stderr=subprocess.PIPE) proc = subprocess.Popen(command, stderr=subprocess.PIPE, stdout=subprocess.PIPE)
return (proc.wait(), proc.stderr.read()) return_code = proc.wait()
if return_code != 0:
error_text = proc.stderr.read()
raise TesseractError(error_text)
def do_document_ocr(document): def do_document_ocr(document):
for page_index, document_page in enumerate(document.documentpage_set.all()): for page_index, document_page in enumerate(document.documentpage_set.all()):
imagefile = convert_document_for_ocr(document, page=page_index) imagefile = convert_document_for_ocr(document, page=page_index)
desc, filepath = tempfile.mkstemp() desc, filepath = tempfile.mkstemp()
try: try:
status, error_string = run_tesseract(imagefile, filepath) run_tesseract(imagefile, filepath, TESSERACT_LANGUAGE)
if status:
errors = get_errors(error_string)
raise TesseractError(errors)
finally:
ocr_output = os.extsep.join([filepath, 'txt']) ocr_output = os.extsep.join([filepath, 'txt'])
f = codecs.open(ocr_output, 'r', 'utf-8')
f = file(ocr_output)
try:
document_page = document.documentpage_set.get(page_number=page_index+1) document_page = document.documentpage_set.get(page_number=page_index+1)
document_page.content = f.read().strip() document_page.content = f.read().strip()
document_page.page_label = _(u'Text from OCR') document_page.page_label = _(u'Text from OCR')
document_page.save() document_page.save()
finally:
f.close() f.close()
cleanup(filepath)
cleanup(ocr_output) cleanup(ocr_output)
except TesseractError, e:
cleanup(filepath)
cleanup(imagefile)
raise TesseractError(e)
finally:
cleanup(filepath)
cleanup(imagefile) cleanup(imagefile)