Allow pre conversion to happen on any ERP5 Web Page's content which include...

Allow pre conversion to happen on any ERP5 Web Page's content  which include relative images (through <img> tags).

bt5/erp5_dms_conversion_catalog/ExtensionTemplateItem/DocumentConversion.py

+import re
+
+def Base_extractImageUrlList(self, text_content=None):
+  """
+    Extract list of image URLS used in a Text document (i.e. Web Page)
+  """
+  if text_content is None:
+    text_content = self.getTextContent()
+  if text_content is not None:
+    return re.findall('src=[\"\'](.[^\"\']+)[\"\']', text_content, re.I) 
+  return []

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/Base_callPreConvert.xml

@@ -63,19 +63,21 @@ if portal_type not in allowed_portal_type_list:\n
   # no need to convert any non DMS types\n
   return\n
 \n
-format = portal.portal_preferences.getPreferredImageFormat()\n
-quality = portal.portal_preferences.getPreferredImageQuality()\n
-preferred_image_size = portal.portal_preferences.getPreferredImageSize()\n
+if kw=={}:\n
+  # use default set of system preferences\n
+  format = portal.portal_preferences.getPreferredImageFormat()\n
+  quality = portal.portal_preferences.getPreferredImageQuality()\n
+  # thumbnail is required always\n
+  display_list = ["thumbnail"]\n
+  kw = dict(format=format, quality=quality, display_list = display_list)\n
 \n
-# thumbnail is required always\n
-display_list = ["thumbnail"]\n
 method = context.getTypeBasedMethod(\'preConvert\')\n
-return method(**dict(format=format, quality=quality, display_list = display_list))\n
+return method(**kw)\n
 </string> </value>
         </item>
         <item>
             <key> <string>_params</string> </key>
-            <value> <string></string> </value>
+            <value> <string>**kw</string> </value>
         </item>
         <item>
             <key> <string>id</string> </key>

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/Base_extractImageUrlList.xml

+<?xml version="1.0"?>
+<ZopeData>
+  <record id="1" aka="AAAAAAAAAAE=">
+    <pickle>
+      <global name="ExternalMethod" module="Products.ExternalMethod.ExternalMethod"/>
+    </pickle>
+    <pickle>
+      <dictionary>
+        <item>
+            <key> <string>_function</string> </key>
+            <value> <string>Base_extractImageUrlList</string> </value>
+        </item>
+        <item>
+            <key> <string>_module</string> </key>
+            <value> <string>DocumentConversion</string> </value>
+        </item>
+        <item>
+            <key> <string>id</string> </key>
+            <value> <string>Base_extractImageUrlList</string> </value>
+        </item>
+        <item>
+            <key> <string>title</string> </key>
+            <value> <string></string> </value>
+        </item>
+      </dictionary>
+    </pickle>
+  </record>
+</ZopeData>

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/Image_preConvert.xml

@@ -53,6 +53,11 @@
             <value> <string>"""\n
   Do actual conversion of any Image type.\n
 """\n
+if quality is None:\n
+  # it\'s required so fall back to system preferences as\n
+  # directly accessed over URL will do the same\n
+  quality = context.getDefaultImageQuality(format)\n
+\n
 # UI uses \'large\' display\n
 display_list.append(\'large\')\n
 context.Base_preConvert(format, quality, display_list)\n
@@ -60,7 +65,7 @@ context.Base_preConvert(format, quality, display_list)\n
         </item>
         <item>
             <key> <string>_params</string> </key>
-            <value> <string>format, quality, display_list</string> </value>
+            <value> <string>format, quality=None, display_list=[]</string> </value>
         </item>
         <item>
             <key> <string>_proxy_roles</string> </key>

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/OOoDocument_preConvert.xml

@@ -53,7 +53,12 @@
             <value> <string>"""\n
   Do actual conversion of OOo types.\n
 """\n
+portal = context.getPortalObject()\n
+\n
 if context.hasBaseData():\n
+  if quality is None:\n
+    quality = portal.portal_preferences.getPreferredImageQuality()\n
+\n
   # empty documents do not need pre conversion\n
   context.activate(serialization_tag=\'pre_convert\').convert(**{\'format\': \'html\'})\n
   context.activate(serialization_tag=\'pre_convert\').Base_preConvert(format, quality, display_list)\n
@@ -61,7 +66,7 @@ if context.hasBaseData():\n
         </item>
         <item>
             <key> <string>_params</string> </key>
-            <value> <string>format, quality, display_list</string> </value>
+            <value> <string>format, quality=None, display_list=[]</string> </value>
         </item>
         <item>
             <key> <string>_proxy_roles</string> </key>

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/WebPage_preConvert.xml

@@ -50,7 +50,9 @@
         </item>
         <item>
             <key> <string>_body</string> </key>
-            <value> <string>"""\n
+            <value> <string encoding="cdata"><![CDATA[
+
+"""\n
   Do actual conversion of Web Page. As it involves cloudoo use serialization tag and activity.\n
 """\n
 portal = context.getPortalObject()\n
@@ -60,7 +62,12 @@ format_kw = {\'format\': format,\n
 for display in display_list:\n
   format_kw[\'display\'] = display\n
   context.activate(serialization_tag=\'pre_convert\').convert(**format_kw)\n
-</string> </value>
+\n
+# try to convert all relative referenced (i.e. by <img> tag) documents\n
+context.activate().WebPage_preConvertReferencedImageList(**format_kw)\n
+
+
+]]></string> </value>
         </item>
         <item>
             <key> <string>_params</string> </key>

bt5/erp5_dms_conversion_catalog/SkinTemplateItem/portal_skins/erp5_dms_conversion_catalog/WebPage_preConvertReferencedImageList.xml

+<?xml version="1.0"?>
+<ZopeData>
+  <record id="1" aka="AAAAAAAAAAE=">
+    <pickle>
+      <global name="PythonScript" module="Products.PythonScripts.PythonScript"/>
+    </pickle>
+    <pickle>
+      <dictionary>
+        <item>
+            <key> <string>Script_magic</string> </key>
+            <value> <int>3</int> </value>
+        </item>
+        <item>
+            <key> <string>_bind_names</string> </key>
+            <value>
+              <object>
+                <klass>
+                  <global name="NameAssignments" module="Shared.DC.Scripts.Bindings"/>
+                </klass>
+                <tuple/>
+                <state>
+                  <dictionary>
+                    <item>
+                        <key> <string>_asgns</string> </key>
+                        <value>
+                          <dictionary>
+                            <item>
+                                <key> <string>name_container</string> </key>
+                                <value> <string>container</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_context</string> </key>
+                                <value> <string>context</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_m_self</string> </key>
+                                <value> <string>script</string> </value>
+                            </item>
+                            <item>
+                                <key> <string>name_subpath</string> </key>
+                                <value> <string>traverse_subpath</string> </value>
+                            </item>
+                          </dictionary>
+                        </value>
+                    </item>
+                  </dictionary>
+                </state>
+              </object>
+            </value>
+        </item>
+        <item>
+            <key> <string>_body</string> </key>
+            <value> <string encoding="cdata"><![CDATA[
+
+"""\n
+  Get all images inside a document and try to pre convert relative ones\n
+"""\n
+portal = context.getPortalObject()\n
+\n
+MARKER = (None, \'\',)\n
+API_ARGUMENT_LIST = [\'format\', \'display\', \'quality\', \'resolution\']\n
+validation_state = (\'released\', \'released_alive\', \'published\', \'published_alive\',\n
+                    \'shared\', \'shared_alive\', \'public\', \'validated\')\n
+\n
+def convertUrlArgumentsToDict(convert_string):\n
+  convert_kw = {}\n
+  # some editors when creating wbe page content do escape \'&\' properly\n
+  convert_string = convert_string.replace(\'&amp;\', \'&\')\n
+\n
+  # convert from URL string to python arguments dict\n
+  for pair in convert_string.split(\'&\'):\n
+    arg_list = pair.split(\'=\')\n
+    if len(arg_list)==2:\n
+      convert_kw[arg_list[0]] = arg_list[1]\n
+  return convert_kw\n
+\n
+image_url_list = context.Base_extractImageUrlList()\n
+for image_url in image_url_list:\n
+  if not image_url.startswith(\'http://\') and not image_url.startswith(\'https://\'):\n
+    # try to use only relative URLs\n
+    part_list = image_url.split(\'?\')\n
+    if len(part_list)==2:\n
+      # don\'t deal with bad URLs (having more than one \'?\') inside\n
+      reference = part_list[0]\n
+      convert_string = part_list[1]\n
+\n
+      # check we have locally such a reference so we can convert it\n
+      kw = {\'portal_type\': portal.getPortalDocumentTypeList() + portal.getPortalEmbeddedDocumentTypeList(),\n
+            \'reference\': reference,\n
+            \'validation_state\': validation_state}\n
+\n
+      document = portal.portal_catalog.getResultValue(**kw)\n
+      if document is not None:\n
+        # try to pre convert it based on extracted URL\'s arguments\n
+        convert_kw = convertUrlArgumentsToDict(convert_string)\n
+\n
+        # XXX: we do check if "data" methods exists on pretending to be Document portal types\n
+        # we need a way to do this by introspection\n
+        if ((getattr(document, "getData", None) is not None and document.getData() not in MARKER) or \\\n
+           (getattr(document, "getBaseData", None) is not None and document.getBaseData() not in MARKER)):\n
+          if \'display\' in convert_kw.keys():\n
+            # conversion script aggregate all possible display options into a list\n
+            convert_kw[\'display_list\'] = [convert_kw.pop(\'display\')]\n
+\n
+          # only certain arguments make sense due to API so leave only them\n
+          for key in convert_kw.keys():\n
+            if key not in API_ARGUMENT_LIST:\n
+              convert_kw.pop(key)\n
+\n
+          # do real conversion\n
+          format = convert_kw.get(\'format\')\n
+          if format not in MARKER:\n
+            # format is mandatory if it\'s missing then anyway URL request will fail so\n
+            # don\'t bother create an activity\n
+            document.activate(priority=4, tag="conversion").Base_callPreConvert(**convert_kw)\n
+
+
+]]></string> </value>
+        </item>
+        <item>
+            <key> <string>_params</string> </key>
+            <value> <string>**kw</string> </value>
+        </item>
+        <item>
+            <key> <string>id</string> </key>
+            <value> <string>WebPage_preConvertReferencedImageList</string> </value>
+        </item>
+      </dictionary>
+    </pickle>
+  </record>
+</ZopeData>

bt5/erp5_dms_conversion_catalog/bt/revision

-18
\ No newline at end of file
+19
\ No newline at end of file

bt5/erp5_dms_conversion_catalog/bt/template_extension_id_list

+DocumentConversion
\ No newline at end of file