PDFDocument.py 8.95 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
##############################################################################
#
# Copyright (c) 2002-2006 Nexedi SARL and Contributors. All Rights Reserved.
#
# WARNING: This program as such is intended to be used by professional
# programmers who take the whole responsability of assessing all potential
# consequences resulting from its eventual inadequacies and bugs
# End users who are looking for a ready-to-use solution with commercial
# garantees and support are strongly adviced to contract a Free Software
# Service Company
#
# 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., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
#
##############################################################################

28 29
import tempfile, os, cStringIO

30
from AccessControl import ClassSecurityInfo
31 32
from Products.CMFCore.utils import getToolByName

33
from Products.ERP5Type import Permissions, PropertySheet, Constraint, interfaces
34 35
from Products.ERP5Type.Cache import CachingMethod
from Products.ERP5.Document.Image import Image
36
from Products.ERP5.Document.Document import ConversionCacheMixin, ConversionError
37

38
from zLOG import LOG, WARNING
39

40
class PDFDocument(Image, ConversionCacheMixin):
41
  """
42 43 44
  PDFDocument is a subclass of Image which is able to
  extract text content from a PDF file either as text
  or as HTML.
45 46
  """
  # CMF Type Definition
47
  meta_type = 'ERP5 PDF Document'
48 49 50 51 52 53 54 55 56 57
  portal_type = 'PDF'
  isPortalContent = 1
  isRADContent = 1

  # Declarative security
  security = ClassSecurityInfo()
  security.declareObjectProtected(Permissions.AccessContentsInformation)

  # Default Properties
  property_sheets = ( PropertySheet.Base
58
                    , PropertySheet.XMLObject
59 60 61 62 63 64
                    , PropertySheet.CategoryCore
                    , PropertySheet.DublinCore
                    , PropertySheet.Version
                    , PropertySheet.Reference
                    , PropertySheet.Document
                    , PropertySheet.Data
65 66 67
                    , PropertySheet.ExternalDocument
                    , PropertySheet.Url
                    , PropertySheet.Periodicity
68 69
                    )

70 71 72 73
  searchable_property_list = ('asText', 'title', 'description', 'id', 'reference',
                              'version', 'short_title',
                              'subject', 'source_reference', 'source_project_title',)

74
  security.declareProtected(Permissions.View, 'index_html')
75 76
  def index_html(self, REQUEST, RESPONSE, display=None, format='', quality=75, 
                                          resolution=None, frame=0):
77
    """
78 79 80
      Returns data in the appropriate format (graphical)
      it is always a zip because multi-page pdfs are converted into a zip
      file of many images
81
    """
82 83
    if format is None:
      RESPONSE.setHeader('Content-Type', 'application/pdf')
84
      return str(self.data)
85 86 87 88 89 90 91
    if format in ('html', 'txt', 'text'):
      mime, data = self.convert(format)
      RESPONSE.setHeader('Content-Length', len(data))
      RESPONSE.setHeader('Content-Type', '%s;charset=UTF-8' % mime)
      RESPONSE.setHeader('Accept-Ranges', 'bytes')
      return data
    return Image.index_html(self, REQUEST, RESPONSE, display=display,
92 93
                            format=format, quality=quality,
                            resolution=resolution, frame=frame)
94 95 96 97 98 99 100

  # Conversion API
  security.declareProtected(Permissions.ModifyPortalContent, 'convert')
  def convert(self, format, **kw):
    """
    Implementation of conversion for PDF files
    """
101
    if format == 'html':
102 103 104
      try:
        return self.getConversion(format=format)
      except KeyError:
105 106
        data = self._convertToHTML()
        self.setConversion(data, mime='text/html', format=format)
107
        return (mime, aq_base(data))
108
    elif format in ('txt', 'text'):
109 110 111
      try:
        return self.getConversion(format='txt')
      except KeyError:
112 113
        data = self._convertToText()
        self.setConversion(data, mime='text/plain', format='txt')
114
        return (mime, aq_base(data))
115 116 117 118 119
    else:
      return Image.convert(self, format, **kw)

  security.declareProtected(Permissions.ModifyPortalContent, 'populateContent')
  def populateContent(self):
120
    """
121 122 123
      Convert each page to an Image and populate the
      PDF directory with converted images. May be useful
      to provide online PDF reader
124
    """
125
    raise NotImplementedError
126 127

  security.declarePrivate('_convertToText')
128
  def _convertToText(self):
129
    """
130
      Convert the PDF text content to text with pdftotext
131
    """
132 133
    if not self.data:
      return ''
134
    tmp = tempfile.NamedTemporaryFile()
135
    tmp.write(str(self.data))
136 137 138 139 140 141
    tmp.seek(0)
    cmd = 'pdftotext -layout -enc UTF-8 -nopgbrk %s -' % tmp.name
    r = os.popen(cmd)
    h = r.read()
    tmp.close()
    r.close()
142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166
    
    if h != '':
      return h
    else:
      # Try to use OCR
      # As high dpi images are required, it may take some times to convert the
      # pdf. 
      # It may be required to use activities to fill the cache and at the end, 
      # to calculate the final result
      text = ''
      content_information = self.getContentInformation()
      page_count = int(content_information.get('Pages', 0))
      for page_number in range(page_count):
        src_mimetype, png_data = self.convert(
            'png', quality=100, resolution=300,
            frame=page_number, display='identical')
        if not src_mimetype.endswith('png'):
          continue
        content = '%s' % png_data
        mime_type = getToolByName(self, 'mimetypes_registry').\
                                    lookupExtension('name.%s' % 'txt')
        if content is not None:
          portal_transforms = getToolByName(self, 'portal_transforms')
          result = portal_transforms.convertToData(mime_type, content,
                                                   context=self,
167
                                                   filename=self.getTitleOrId(),
168 169
                                                   mimetype=src_mimetype)
          if result is None:
170 171
            raise ConversionError('PDFDocument conversion error. '
                                  'portal_transforms failed to convert to %s: %r' % (mime_type, self))
172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187
          text += result
      return text

  security.declareProtected('View', 'getSizeFromImageDisplay')
  def getSizeFromImageDisplay(self, image_display):
    """
    Return the size for this image display, or None if this image display name
    is not known. If the preference is not set, (0, 0) is returned.
    """
    # identical parameter can be considered as a hack, in order not to
    # resize the image to prevent text distorsion when using OCR.
    # A cleaner API is required.
    if image_display == 'identical':
      return (self.getWidth(), self.getHeight())
    else:
      return Image.getSizeFromImageDisplay(self, image_display)
188 189 190 191 192

  security.declarePrivate('_convertToHTML')
  def _convertToHTML(self):
    """
    Convert the PDF text content to HTML with pdftohtml
193 194 195

    NOTE: XXX check that command exists and was executed
    successfully
196
    """
197 198
    if not self.data:
      return ''
199
    tmp = tempfile.NamedTemporaryFile()
200
    tmp.write(str(self.data))
201 202 203 204 205 206 207
    tmp.seek(0)
    cmd = 'pdftohtml -enc UTF-8 -stdout -noframes -i %s' % tmp.name
    r = os.popen(cmd)
    h = r.read()
    tmp.close()
    r.close()
    h = h.replace('<BODY bgcolor="#A0A0A0"', '<BODY ') # Quick hack to remove bg color - XXX
208
    h = h.replace('href="%s.html' % tmp.name.split(os.sep)[-1], 'href="asEntireHTML') # Make links relative
209 210 211 212 213 214 215
    return h

  security.declareProtected(Permissions.AccessContentsInformation, 'getContentInformation')
  def getContentInformation(self):
    """
    Returns the information about the PDF document with
    pdfinfo.
216 217 218

    NOTE: XXX check that command exists and was executed
    successfully
219
    """
220 221 222 223
    try:
      return self._content_information.copy()
    except AttributeError:
      pass
224
    tmp = tempfile.NamedTemporaryFile()
225
    tmp.write(str(self.data))
226 227 228 229 230 231 232 233 234 235 236 237
    tmp.seek(0)
    cmd = 'pdfinfo -meta -box %s' % tmp.name
    r = os.popen(cmd)
    h = r.read()
    tmp.close()
    r.close()
    result = {}
    for line in h.splitlines():
      item_list = line.split(':')
      key = item_list[0].strip()
      value = ':'.join(item_list[1:]).strip()
      result[key] = value
238 239 240 241 242 243
    self._content_information = result
    return result.copy()

  def _setFile(self, data, precondition=None):
    try:
      del self._content_information
Yusei Tahara's avatar
Yusei Tahara committed
244
    except (AttributeError, KeyError):
245 246
      pass
    Image._setFile(self, data, precondition)