OOoUtils.py 18.4 KB
Newer Older
Kevin Deldycke's avatar
Kevin Deldycke committed
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 28 29 30 31 32 33 34
##############################################################################
#
# Copyright (c) 2003-2005 Nexedi SARL and Contributors. All Rights Reserved.
#                         Kevin DELDYCKE    <kevin@nexedi.com>
#                         Guillaume MICHON  <guillaume@nexedi.com>
#
# 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.
#
##############################################################################

from Products.PythonScripts.Utility import allow_class
from ZPublisher.HTTPRequest import FileUpload
from xml.dom.ext.reader import PyExpat
from xml.dom import Node
from AccessControl import ClassSecurityInfo
35 36
from Globals import InitializeClass, get_request
from zipfile import ZipFile, ZIP_DEFLATED
37 38 39 40
try:
  from cStringIO import StringIO
except ImportError:
  from StringIO import StringIO
Kevin Deldycke's avatar
Kevin Deldycke committed
41 42
from zLOG import LOG
import imghdr
43
import random
Kevin Deldycke's avatar
Kevin Deldycke committed
44 45 46 47 48



class CorruptedOOoFile(Exception): pass

49 50 51 52 53 54 55 56 57 58 59 60
OOo_mimeType_dict = {
  'sxw' : 'application/vnd.sun.xml.writer',
  'stw' : 'application/vnd.sun.xml.writer.template',
  'sxg' : 'application/vnd.sun.xml.writer.global',
  'sxc' : 'application/vnd.sun.xml.calc',
  'stc' : 'application/vnd.sun.xml.calc.template',
  'sxi' : 'application/vnd.sun.xml.impress',
  'sti' : 'application/vnd.sun.xml.impress.template',
  'sxd' : 'application/vnd.sun.xml.draw',
  'std' : 'application/vnd.sun.xml.draw.template',
  'sxm' : 'application/vnd.sun.xml.math',
}
Kevin Deldycke's avatar
Kevin Deldycke committed
61

62 63 64 65 66 67 68 69 70
class OOoBuilder:
  """
  Tool that allows to reinject new files in a ZODB OOo document.
  """
  # Declarative security
  security = ClassSecurityInfo()

  security.declarePrivate('__init__')
  def __init__(self, document):
71
    if hasattr(document, 'data') :
72 73
      self._document = StringIO()
      self._document.write(document.data)
74 75 76
    elif hasattr(document, 'read') :
      self._document = document
    else :
77 78
      self._document = StringIO()
      self._document.write(document)
79
    self._image_count = 0    
80
    self._manifest_additions_list = []
81 82 83 84 85 86 87 88 89 90 91

  security.declarePublic('replace')
  def replace(self, filename, stream):
    """
    Replaces the content of filename by stream in the archive.
    Creates a new file if filename was not already there.
    """
    try:
      zf = ZipFile(self._document, mode='a', compression=ZIP_DEFLATED)
    except RuntimeError:
      zf = ZipFile(self._document, mode='a')
92 93 94 95 96 97 98
    try:
	    # remove the file first if it exists
	    fi = zf.getinfo(filename)
	    zf.filelist.remove( fi )
    except KeyError:
	    # This is a new file
	    pass
99 100
    zf.writestr(filename, stream)
    zf.close()
101 102 103 104 105 106 107 108 109 110 111 112
  
  security.declarePublic('extract')
  def extract(self, filename):
    """
    Extracts a file from the archive
    """
    try:
      zf = ZipFile(self._document, mode='r', compression=ZIP_DEFLATED)
    except RuntimeError:
      zf = ZipFile(self._document, mode='r')
    return zf.read(filename)
  
113 114 115 116 117 118 119 120 121 122
  security.declarePublic('getNameList')
  def getNameList(self):
    try:
      zf = ZipFile(self._document, mode='r', compression=ZIP_DEFLATED)
    except RuntimeError:
      zf = ZipFile(self._document, mode='r')
    li = zf.namelist()
    zf.close()
    return li

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148
  security.declarePublic('getMimeType')
  def getMimeType(self):
    return self.extract('mimetype')

  security.declarePublic('prepareContentXml')
  def prepareContentXml(self) :
    """
      extracts content.xml text and prepare it :
        - add tal namespace
        - indent the xml
    """
    import pprint
    content_xml = self.extract('content.xml')
    reader = PyExpat.Reader()
    document = reader.fromString(content_xml)
    document_element = document.documentElement
    from xml.dom.ext import PrettyPrint
    output = StringIO()
    PrettyPrint(document_element, output)
    return output.getvalue().replace(
      "office:version='1.0'",
      """ xmlns:tal='http://xml.zope.org/namespaces/tal'
          xmlns:i18n='http://xml.zope.org/namespaces/i18n'
          xmlns:metal='http://xml.zope.org/namespaces/metal'
          tal:attributes='dummy python:request.RESPONSE.setHeader("Content-Type", "text/html;; charset=utf-8")'
         office:version='1.0'""")
149

150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183
    """
    """

  security.declarePublic('addFileEntry')
  def addFileEntry(self, full_path, media_type, content=None):
      """ Add a file entry to the manifest and possibly is content """
      self.addManifest(full_path, media_type)
      if content:
          self.replace(full_path, content)

  security.declarePublic('addManifest')
  def addManifest(self, full_path, media_type):
    """ Add a path to the manifest """
    li = '<manifest:file-entry manifest:media-type="%s" manifest:full-path="%s"/>'%(media_type, full_path)
    self._manifest_additions_list.append(li)

  security.declarePublic('updateManifest')
  def updateManifest(self):
    """ Add a path to the manifest """
    MANIFEST_FILENAME = 'META-INF/manifest.xml'
    meta_infos = self.extract(MANIFEST_FILENAME)

    # prevent some duplicates
    for meta_line in meta_infos.split('\n'):
        for new_meta_line in self._manifest_additions_list:
            if meta_line.strip() == new_meta_line:
                self._manifest_additions_list.remove(new_meta_line)

    # add the new lines
    self._manifest_additions_list.append('</manifest:manifest>')
    meta_infos = meta_infos.replace( self._manifest_additions_list[-1], '\n'.join(self._manifest_additions_list) )
    self.replace(MANIFEST_FILENAME, meta_infos)
    self._manifest_additions_list = []

184
  security.declarePublic('addImage')
185 186 187 188 189 190 191 192 193 194
  def addImage(self, image, format='png'):
    """
    Add an image to the current document and return its id
    """
    count = self._image_count
    self._image_count += 1
    name = "Picture/%s.%s" % (count, format)
    self.replace(name, image)
    return "#%s" % name

195 196 197 198 199 200 201 202 203 204 205 206 207 208
  security.declarePublic('render')
  def render(self, name='', extension='sxw'):
    """
    returns the OOo document
    """
    request = get_request()
    request.response.setHeader('Content-type', OOo_mimeType_dict.get(extension, 'application/vnd.sun.xml.writer'))
    if name:
      request.response.setHeader('Content-Disposition', 'attachment; filename=%s.%s' % (name, extension))
    self._document.seek(0)
    return self._document.read()
    
InitializeClass(OOoBuilder)
allow_class(OOoBuilder)
Kevin Deldycke's avatar
Kevin Deldycke committed
209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228

class OOoParser:
  """
    General purpose tools to parse and handle OpenOffice v1.x documents.
  """


  # Declarative security
  security = ClassSecurityInfo()


  security.declarePrivate('__init__')
  def __init__(self):
    # Create the PyExpat reader
    self.reader = PyExpat.Reader()
    self.oo_content_dom = None
    self.oo_styles_dom  = None
    self.oo_files = {}
    self.pictures = {}
    self.ns = {}
Kevin Deldycke's avatar
Kevin Deldycke committed
229
    self.filename = None
Kevin Deldycke's avatar
Kevin Deldycke committed
230 231


232
  security.declareProtected('Import/Export objects', 'openFile')
233
  def openFile(self, file_descriptor):
Kevin Deldycke's avatar
Kevin Deldycke committed
234 235 236 237 238
    """
      Load all files in the zipped OpenOffice document
    """
    # Try to unzip the Open Office doc
    try:
239
      oo_unzipped = ZipFile(file_descriptor, mode="r")
Kevin Deldycke's avatar
Kevin Deldycke committed
240 241 242 243 244 245
    except:
      raise CorruptedOOoFile
    # Test the integrity of the file
    if oo_unzipped.testzip() != None:
      raise CorruptedOOoFile

Kevin Deldycke's avatar
Kevin Deldycke committed
246 247 248
    # Get the filename
    self.filename = file_descriptor.filename

Kevin Deldycke's avatar
Kevin Deldycke committed
249 250 251
    # List and load the content of the zip file
    for name in oo_unzipped.namelist():
      self.oo_files[name] = oo_unzipped.read(name)
252
    oo_unzipped.close()
Kevin Deldycke's avatar
Kevin Deldycke committed
253 254 255 256 257 258 259 260 261 262 263 264 265 266

    # Get the main content and style definitions
    self.oo_content_dom = self.reader.fromString(self.oo_files["content.xml"])
    self.oo_styles_dom  = self.reader.fromString(self.oo_files["styles.xml"])

    # Create a namespace table
    doc_ns = self.oo_styles_dom.getElementsByTagName("office:document-styles")
    for i in range(doc_ns[0].attributes.length):
        if doc_ns[0].attributes.item(i).nodeType == Node.ATTRIBUTE_NODE:
            name = doc_ns[0].attributes.item(i).name
            if name[:5] == "xmlns":
                self.ns[name[6:]] = doc_ns[0].attributes.item(i).value


Kevin Deldycke's avatar
Kevin Deldycke committed
267 268 269 270 271 272 273 274
  security.declarePublic('getFilename')
  def getFilename(self):
    """
      Return the name of the OpenOffice file
    """
    return self.filename


275 276
  security.declarePublic('getPicturesMapping')
  def getPicturesMapping(self):
Kevin Deldycke's avatar
Kevin Deldycke committed
277 278 279 280 281 282 283 284 285 286 287 288
    """
      Return a dictionnary of all pictures in the document
    """
    if len(self.pictures) <= 0:
      for file_name in self.oo_files:
        raw_data = self.oo_files[file_name]
        pict_type = imghdr.what(None, raw_data)
        if pict_type != None:
          self.pictures[file_name] = raw_data
    return self.pictures


289 290
  security.declarePublic('getContentDom')
  def getContentDom(self):
Kevin Deldycke's avatar
Kevin Deldycke committed
291 292 293 294 295 296
    """
      Return the DOM tree of the main OpenOffice content
    """
    return self.oo_content_dom


297 298
  security.declarePublic('getSpreadsheetsDom')
  def getSpreadsheetsDom(self, include_embedded=False):
299 300 301 302
    """
      Return a list of DOM tree spreadsheets (optionnaly included embedded ones)
    """
    spreadsheets = []
303
    spreadsheets = self.getPlainSpreadsheetsDom()
304
    if include_embedded == True:
305
      spreadsheets += self.getEmbeddedSpreadsheetsDom()
306 307 308
    return spreadsheets


309
  security.declarePublic('getSpreadsheetsMapping')
310
  def getSpreadsheetsMapping(self, include_embedded=False, no_empty_lines=False, normalize=True):
311 312 313
    """
      Return a list of table-like spreadsheets (optionnaly included embedded ones)
    """
314
    tables = {}
315
    tables = self.getPlainSpreadsheetsMapping(no_empty_lines, normalize)
316
    if include_embedded == True:
317
      embedded_tables = self.getEmbeddedSpreadsheetsMapping(no_empty_lines, normalize)
318 319
      tables = self._getTableListUnion(tables, embedded_tables)
    return tables
320 321


322 323
  security.declarePublic('getPlainSpreadsheetsDom')
  def getPlainSpreadsheetsDom(self):
324 325 326 327 328 329 330 331 332 333
    """
      Retrieve every spreadsheets from the document and get they DOM tree
    """
    spreadsheets = []
    # List all spreadsheets
    for table in self.oo_content_dom.getElementsByTagName("table:table"):
      spreadsheets.append(table)
    return spreadsheets


334
  security.declarePublic('getPlainSpreadsheetsMapping')
335
  def getPlainSpreadsheetsMapping(self, no_empty_lines=False, normalize=True):
336 337 338
    """
      Return a list of plain spreadsheets from the document and transform them as table
    """
339
    tables = {}
340
    for spreadsheet in self.getPlainSpreadsheetsDom():
341
      new_table = self.getSpreadsheetMapping(spreadsheet, no_empty_lines, normalize)
342
      if new_table != None:
343
        tables = self._getTableListUnion(tables, new_table)
344 345 346
    return tables


347 348
  security.declarePublic('getEmbeddedSpreadsheetsDom')
  def getEmbeddedSpreadsheetsDom(self):
Kevin Deldycke's avatar
Kevin Deldycke committed
349 350 351 352 353 354 355
    """
      Return a list of existing embedded spreadsheets in the file as DOM tree
    """
    spreadsheets = []
    # List all embedded spreadsheets
    emb_objects = self.oo_content_dom.getElementsByTagName("draw:object")
    for embedded in emb_objects:
356 357 358 359 360 361 362 363 364 365 366 367
        document = embedded.getAttributeNS(self.ns["xlink"], "href")
        if document:
            try:
                object_content = self.reader.fromString(self.oo_files[document[3:] + '/content.xml'])
                tables = object_content.getElementsByTagName("table:table")
                if tables:
                    for table in tables:
                        spreadsheets.append(table)
                else: # XXX: insert the link to OLE document ?
                    pass
            except:
                pass
Kevin Deldycke's avatar
Kevin Deldycke committed
368 369 370
    return spreadsheets


371
  security.declarePublic('getEmbeddedSpreadsheetsMapping')
372
  def getEmbeddedSpreadsheetsMapping(self, no_empty_lines=False, normalize=True):
Kevin Deldycke's avatar
Kevin Deldycke committed
373
    """
374
      Return a list of embedded spreadsheets in the document as table
Kevin Deldycke's avatar
Kevin Deldycke committed
375
    """
376
    tables = {}
377
    for spreadsheet in self.getEmbeddedSpreadsheetsDom():
378
      new_table = self.getSpreadsheetMapping(spreadsheet, no_empty_lines, normalize)
Kevin Deldycke's avatar
Kevin Deldycke committed
379
      if new_table != None:
380
        tables = self._getTableListUnion(tables, new_table)
Kevin Deldycke's avatar
Kevin Deldycke committed
381 382 383
    return tables


384
  security.declarePublic('getSpreadsheetMapping')
385
  def getSpreadsheetMapping(self, spreadsheet=None, no_empty_lines=False, normalize=True):
Kevin Deldycke's avatar
Kevin Deldycke committed
386 387
    """
      This method convert an OpenOffice spreadsheet to a simple table.
388
      This code is based on the oo2pt tool (http://cvs.sourceforge.net/viewcvs.py/collective/CMFReportTool/oo2pt).
Kevin Deldycke's avatar
Kevin Deldycke committed
389
    """
390
    if spreadsheet == None or spreadsheet.nodeName != 'table:table':
Kevin Deldycke's avatar
Kevin Deldycke committed
391 392
      return None

393
    table = []
Kevin Deldycke's avatar
Kevin Deldycke committed
394

395 396 397
    # Get the table name
    table_name = spreadsheet.getAttributeNS(self.ns["table"], "name")

398 399
    # Scan table and store usable informations
    for line in spreadsheet.getElementsByTagName("table:table-row"):
400 401 402 403 404 405

      # TODO : to the same as cell about abusive repeated lines

      line_group_found = line.getAttributeNS(self.ns["table"], "number-rows-repeated")
      if not line_group_found:
        lines_to_repeat = 1
406
      else:
407
        lines_to_repeat = int(line_group_found)
408

409
      for i in range(lines_to_repeat):
410 411
        table_line = []

412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433
        # Get all cells
        cells = line.getElementsByTagName("table:table-cell")
        cell_index_range = range(len(cells))

        for cell_index in cell_index_range:
          cell = cells[cell_index]

          # If the cell as no child, cells have no content
          # And if the cell is the last of the row, we don't need to add it to the line
          # So we can go to the next line (= exit this cells loop)
          #
          # I must do this test because sometimes the following cell group
          #   can be found in OOo documents : <table:table-cell table:number-columns-repeated='246'/>
          # This is bad because it create too much irrevelent content that slow down the process
          # So it's a good idea to break the loop in this case
          if cell.childNodes.length == 0 and cell_index == cell_index_range[-1]:
            break

          # Handle cells group
          cell_group_found = cell.getAttributeNS(self.ns["table"], "number-columns-repeated")
          if not cell_group_found:
            cells_to_repeat = 1
434
          else:
435
            cells_to_repeat = int(cell_group_found)
436

437 438 439
          # Ungroup repeated cells
          for j in range(cells_to_repeat):
            # Get the cell content
440 441 442 443 444 445 446 447 448 449
            cell_text = None
            text_tags = cell.getElementsByTagName("text:p")
            for text in text_tags:
              for k in range(text.childNodes.length):
                child = text.childNodes[k]
                if child.nodeType == Node.TEXT_NODE:
                  if cell_text == None:
                    cell_text = ''
                  cell_text += child.nodeValue

450
            # Add the cell to the line
451
            table_line.append(cell_text)
452

Kevin Deldycke's avatar
Kevin Deldycke committed
453 454 455 456 457 458 459 460 461
        # Delete empty lines if needed
        if no_empty_lines:
          empty_cell = 0
          for table_cell in table_line:
            if table_cell == None:
              empty_cell += 1
          if empty_cell == len(table_line):
            table_line = None

462
        # Add the line to the table
Kevin Deldycke's avatar
Kevin Deldycke committed
463 464
        if table_line != None:
          table.append(table_line)
465

466
    # Reduce the table to the minimum
467 468 469 470 471 472 473 474 475 476
    new_table = self._getReducedTable(table)

    # Get a homogenized table
    if normalize:
      table_size = self._getTableSizeDict(new_table)
      new_table = self._getNormalizedBoundsTable( table  = new_table
                                                , width  = table_size['width']
                                                , height = table_size['height']
                                                )
    return {table_name: new_table}
Kevin Deldycke's avatar
Kevin Deldycke committed
477 478


479 480
  security.declarePrivate('_getReducedTable')
  def _getReducedTable(self, table):
Kevin Deldycke's avatar
Kevin Deldycke committed
481
    """
482
      Reduce the table to its minimum size
Kevin Deldycke's avatar
Kevin Deldycke committed
483 484 485 486 487
    """
    empty_lines = 0
    no_more_empty_lines = 0

    # Eliminate all empty cells at the ends of lines and columns
488
    # Browse the table starting from the bottom for easy empty lines count
489
    for line in range(len(table)-1, -1, -1):
Kevin Deldycke's avatar
Kevin Deldycke committed
490
      empty_cells = 0
491
      line_content = table[line]
Kevin Deldycke's avatar
Kevin Deldycke committed
492
      for cell in range(len(line_content)-1, -1, -1):
493
        if line_content[cell] in ('', None):
Kevin Deldycke's avatar
Kevin Deldycke committed
494 495 496
          empty_cells += 1
        else:
          break
497

Kevin Deldycke's avatar
Kevin Deldycke committed
498 499 500 501
      if (not no_more_empty_lines) and (empty_cells == len(line_content)):
        empty_lines += 1
      else:
        line_size = len(line_content) - empty_cells
502
        table[line] = line_content[:line_size]
Kevin Deldycke's avatar
Kevin Deldycke committed
503 504
        no_more_empty_lines = 1

505 506 507
    table_height = len(table) - empty_lines

    return table[:table_height]
Kevin Deldycke's avatar
Kevin Deldycke committed
508

509 510 511 512 513 514

  security.declarePrivate('_getTableSizeDict')
  def _getTableSizeDict(self, table):
    """
      Get table dimension as dictionnary contain both height and width
    """
Kevin Deldycke's avatar
Kevin Deldycke committed
515
    max_cols = 0
516 517 518 519
    for line_index in range(len(table)):
      line = table[line_index]
      if len(line) > max_cols:
        max_cols = len(line)
Kevin Deldycke's avatar
Kevin Deldycke committed
520

521 522 523
    return { 'width' : max_cols
           , 'height': len(table)
           }
Kevin Deldycke's avatar
Kevin Deldycke committed
524 525


526 527
  security.declarePrivate('_getNormalizedBoundsTable')
  def _getNormalizedBoundsTable(self, table, width=0, height=0):
Kevin Deldycke's avatar
Kevin Deldycke committed
528
    """
529
      Add necessary cells and lines to obtain given bounds
Kevin Deldycke's avatar
Kevin Deldycke committed
530
    """
531 532
    while height > len(table):
      table.append([])
Kevin Deldycke's avatar
Kevin Deldycke committed
533
    for line in range(height):
534 535 536 537
      while width > len(table[line]):
        table[line].append(None)
    return table

Kevin Deldycke's avatar
Kevin Deldycke committed
538

539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555
  security.declarePrivate('_getTableListUnion')
  def _getTableListUnion(self, list1, list2):
    """
      Coerce two dict containing tables structures.
      We need to use this method because a OpenOffice document can hold
        several embedded spreadsheets with the same id. This explain the
        use of random suffix in such extreme case.
    """
    for list2_key in list2.keys():
      # Generate a new table ID if needed
      new_key = list2_key
      while new_key in list1.keys():
        new_key = list2_key + '_' + str(random.randint(1000,9999))
      list1[new_key] = list2[list2_key]
    return list1


556

Kevin Deldycke's avatar
Kevin Deldycke committed
557 558
InitializeClass(OOoParser)
allow_class(OOoParser)