Migrate from xhtml2odt to pandoc for ODT generation (refs #4178)

1 files changed, 24 insertions, 35 deletions

diff --git a/ishtar_common/views_item.py b/ishtar_common/views_item.py
index b286ab3bf..513035903 100644
--- a/ishtar_common/views_item.py
+++ b/ishtar_common/views_item.py
@@ -1,13 +1,13 @@
 #!/usr/bin/env python
 # -*- coding: utf-8 -*-
 
+from copy import copy, deepcopy
 import csv
 import datetime
 import json
 import logging
-import optparse
 import re
-from copy import copy, deepcopy
+import subprocess
 from tempfile import NamedTemporaryFile
 
 from django.conf import settings
@@ -26,7 +26,6 @@ from tidylib import tidy_document as tidy
 from unidecode import unidecode
 from weasyprint import HTML, CSS
 from weasyprint.fonts import FontConfiguration
-from xhtml2odt import xhtml2odt
 
 from ishtar_common.utils import check_model_access_control, CSV_OPTIONS, \
     get_all_field_names
@@ -180,41 +179,31 @@ def show_item(model, name, extra_dct=None):
             tpl = loader.get_template('ishtar/sheet_%s.html' % name)
             context_instance['output'] = 'ODT'
             content = tpl.render(context_instance, request)
+            tidy_options = {'output-xhtml': 1, 'indent': 1,
+                            'tidy-mark': 0, 'doctype': 'auto',
+                            'add-xml-decl': 1, 'wrap': 1}
+            html, errors = tidy(content, options=tidy_options)
+            html = html.encode('utf-8').replace(" ", " ")
+            html = re.sub('<pre([^>]*)>\n', '<pre\\1>', html)
+
+            odt = NamedTemporaryFile()
+            html_source = NamedTemporaryFile()
+            with open(html_source.name, 'w') as html_file:
+                html_file.write(html)
+
+            pandoc_args = ["pandoc", "-f", "html", "-t", "odt",
+                           "-o", odt.name, html_source.name]
             try:
-                tidy_options = {'output-xhtml': 1, 'indent': 1,
-                                'tidy-mark': 0, 'doctype': 'auto',
-                                'add-xml-decl': 1, 'wrap': 1}
-                html, errors = tidy(content, options=tidy_options)
-                html = html.encode('utf-8').replace("&nbsp;", "&#160;")
-                html = re.sub('<pre([^>]*)>\n', '<pre\\1>', html)
-
-                odt = NamedTemporaryFile()
-                options = optparse.Values()
-                options.with_network = True
-                for k, v in (('input', ''),
-                             ('output', odt.name),
-                             ('template', settings.ODT_TEMPLATE),
-                             ('with_network', True),
-                             ('top_header_level', 1),
-                             ('img_width', '8cm'),
-                             ('img_height', '6cm'),
-                             ('verbose', False),
-                             ('replace_keyword', 'ODT-INSERT'),
-                             ('cut_start', 'ODT-CUT-START'),
-                             ('htmlid', None),
-                             ('url', "#")):
-                    setattr(options, k, v)
-                odtfile = xhtml2odt.ODTFile(options)
-                odtfile.open()
-                odtfile.import_xhtml(html)
-                odtfile = odtfile.save()
-            except xhtml2odt.ODTExportError:
-                return HttpResponse(content, content_type="application/xhtml")
+                subprocess.check_call(pandoc_args)
+            except subprocess.CalledProcessError:
+                return HttpResponse(content,
+                                    content_type="application/xhtml")
             response = HttpResponse(
                 content_type='application/vnd.oasis.opendocument.text')
-            response['Content-Disposition'] = 'attachment; filename=%s.odt' % \
-                                              filename
-            response.write(odtfile)
+            response['Content-Disposition'] = \
+                'attachment; filename={}.odt'.format(filename)
+            with open(odt.name, 'r') as odt_file:
+                response.write(odt_file.read())
             return response
         elif doc_type == 'pdf':
             tpl = loader.get_template('ishtar/sheet_%s_pdf.html' % name)