Page MenuHomec4science
Files F74601076

hocrlib.py
No OneTemporary
Actions

File Metadata

Created
Sun, Jul 28, 16:20

hocrlib.py
View Options

## This file is part of Invenio.
## Copyright (C) 2010, 2011 CERN.
##
## Invenio 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.
##
## Invenio 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 Invenio; if not, write to the Free Software Foundation, Inc.,
## 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA.
"""hOCR parser and tools"""
from htmlentitydefs import entitydefs
import HTMLParser
import re
import os.path
from logging import info
from reportlab.pdfgen.canvas import Canvas
from reportlab.lib.units import inch
from reportlab.lib.colors import green, red
CFG_PPM_RESOLUTION = 300
_RE_PARSE_HOCR_BBOX = re.compile(r'\bbbox\s+(\d+)\s+(\d+)\s+(\d+)\s+(\d+)')
_RE_CLEAN_SPACES = re.compile(r'\s+')
def extract_hocr(hocr_text):
"""
Parse hocr_text and return a structure suitable to be used by create_pdf.
"""
class HOCRReader(HTMLParser.HTMLParser):
def __init__(self):
HTMLParser.HTMLParser.__init__(self)
self.lines = []
self.bbox = None
self.text = ""
self.image = ''
self.page_bbox = None
self.pages = []
self.started = False
def store_current_page(self):
if self.image:
self.store_current_line()
self.sort_current_lines()
self.pages.append((self.page_bbox, self.image, self.lines))
self.page_bbox = None
self.image = ''
self.lines = []
def sort_current_lines(self):
def line_cmp(a, b):
y0_a = a[0][1]
y0_b = b[0][1]
return cmp(y0_b, y0_a)
self.lines.sort(line_cmp)
def store_current_line(self):
if self.bbox:
self.lines.append((self.bbox, _RE_CLEAN_SPACES.sub(' ', self.text).strip()))
self.bbox = None
self.text = ""
def extract_hocr_properties(self, title):
properties = title.split(';')
ret = {}
for prop in properties:
prop = prop.strip()
key, value = prop.split(' ', 1)
key = key.strip().lower()
value = value.strip()
ret[key] = value
return ret
def handle_starttag(self, tag, attrs):
attrs = dict(attrs)
if tag == 'title':
self.started = False # prevents <title> of the OCR output to be put into text
elif attrs.get('class') == 'ocr_line':
self.started = True
self.store_current_line()
properties = self.extract_hocr_properties(attrs.get('title', ''))
try:
self.bbox = tuple(map(lambda x: int(x), properties['bbox'].split(' ', 4)))
except:
## If no bbox is retrievable, let's skip this line
pass
elif attrs.get('class') == 'ocr_page':
self.store_current_page()
properties = self.extract_hocr_properties(attrs.get('title', ''))
try:
self.page_bbox = tuple(map(lambda x: int(x), properties['bbox'].split(' ', 4)))
except:
## If no bbox is retrievable, let's skip this line
pass
try:
self.image = os.path.abspath(properties['image'])
except:
pass
def handle_entityref(self, name):
if self.started and name in entitydefs:
self.text += entitydefs[name].decode('latin1').encode('utf8')
def handle_data(self, data):
if self.started and data.strip():
self.text += data
def handle_charref(self, data):
if self.started:
try:
self.text += unichr(int(data)).encode('utf8')
except:
pass
def close(self):
HTMLParser.HTMLParser.close(self)
self.store_current_page()
hocr_reader = HOCRReader()
hocr_reader.feed(hocr_text)
hocr_reader.close()
return hocr_reader.pages
def start_pdf(filename, author=None, keywords=None, subject=None, title=None):
""" Starts a new pdf document
@param filename the name of the PDF generated in output.
@param author the author name.
@param subject the subject of the document.
@param title the title of the document.
"""
canvas = Canvas(filename)
if author:
canvas.setAuthor(author)
if keywords:
canvas.setKeywords(keywords)
if title:
canvas.setTitle(title)
if subject:
canvas.setSubject(subject)
canvas.setPageCompression(1)
return canvas
def add_pdf_page(canvas, hocr, font='Courier', draft=False):
""" Add one page of hOCR output into a searchable PDF.
@param canvas the pdf being produced
@param hocr the hocr structure as coming from extract_hocr.
@param working dir with the image and ocr output.
@param image the image that has been ocred
@param font the default font (e.g. Courier, Times-Roman).
@param draft whether to enable debug information in the output.
"""
ratio = float(CFG_PPM_RESOLUTION) / 72. # pix to pts
bbox, _, lines = hocr
img_width, img_height = bbox[2:]
page_size = (img_width / ratio, img_height / ratio)
canvas.setPageSize(page_size)
canvas.setFont(font, 12)
for bbox, line in lines:
if draft:
canvas.setFillColor(red)
x0, y0, x1, y1 = bbox
width = (x1 - x0) / ratio
height = ((y1 - y0) / ratio)
x0 = x0 / ratio
y0 = page_size[1] - (y0 / ratio) - height / 1.3
canvas.setFontSize(height)
text_width = canvas.stringWidth(line)
if text_width:
## If text_width != 0
text_object = canvas.beginText(x0, y0)
text_object.setHorizScale(1.0 * width / text_width * 100)
text_object.textOut(line)
canvas.drawText(text_object)
else:
info('%s, %s has width 0' % (bbox, line))
if draft:
canvas.setStrokeColor(green)
canvas.rect(x0, y0, width, height)
if draft:
canvas.circle(0, 0, 10, fill=1)
canvas.circle(0, page_size[1], 10, fill=1)
canvas.circle(page_size[0], 0, 10, fill=1)
canvas.circle(page_size[0], page_size[1], 10, fill=1)
canvas.setFillColor(green)
canvas.setStrokeColor(green)
canvas.circle(0, page_size[1] - img_height / ratio, 5, fill=1)
canvas.circle(img_width / ratio, img_height /ratio, 5, fill=1)
canvas.showPage()
canvas.save()
def close_pdf(canvas):
""" Finishes the pdf file.
@param canvas the pdf being produced
"""
canvas.save()
def create_pdf(hocr, output_pdf, font='Courier', draft=False):
""" transform hOCR information into a text-only PDF.
@param hocr the hocr structure as coming from extract_hocr.
@param filename the name of the PDF generated in output.
@param font the default font (e.g. Courier, Times-Roman).
@param draft whether to enable debug information in the output.
"""
canvas = start_pdf(output_pdf)
for page in hocr:
add_pdf_page(canvas, page, font, draft)
close_pdf(canvas)

Event Timeline

c4science · Help