def __iter__(self):
for item in self.previous:
id = item.get("id", u"")
pathkey = self.pathkey(*item.keys())[0]
if not pathkey:
yield item; continue
path = item[pathkey]
_id = path.rstrip("/").rsplit("/")[-1]
if not id:
id = _id
# Removes accented characters from composed characters
# where possible - and supress non "alphanum_-" chars:
# (example: u"maçã" -> u"maca"
normal_path = path.rstrip("/").rsplit("/",1)[0] + \
"/" + normalize_url(_id)
normal_id = normalize_url(id)
plone = get_plone_root(self.transmogrifier.context)
if plone:
s_id = str(normal_id)
if (hasattr(plone, s_id) and
s_id not in plone.contentIds()):
# This condition would raise a "BadRequest"
# exception
normal_id += u"_1"
normal_path += u"_1"
if normal_path == path and normal_id == id:
yield item
continue
if "id" in item:
item["id"] = normal_id
item[pathkey] = normal_path
if not hasattr(self.transmogrifier, "sc_transmogrifier_changed_paths"):
self.transmogrifier.sc_transmogrifier_changed_paths = {}
self.transmogrifier.sc_transmogrifier_changed_paths[path] = normal_path
if not "_orig_path" in item:
item["_orig_path"] = normal_path
yield item
def transmogrify(self, item):
path = self.get_path(item)
text = item.get("text", u"")
if not text:
raise NothingToDoHere
# One never knows how those json read
# strings end up:
if not isinstance(text, unicode):
text = text.decode("utf-8")
image_matches = re.finditer(ur"""<img.*?src\s*?=\s*?"(.*?)".*?>""", text,
re.DOTALL | re.MULTILINE
)
images = []
for match in reversed(list(image_matches)):
# *** retrieve image URL and name data:
start_tag, end_tag, img_title, rel_url, url = \
get_image_refs(match, path, self.source_prefix)
if url == rel_url and not self.load_external_images:
continue
# *** retrieve remote image itself
real_url, image, view_parts = get_remote_image(
url, item, img_title,
self.use_jsonmigrator)
if real_url is None:
continue
image_filename = image["_filename"]
# *** Fix image references in item's text html content:
if real_url.startswith(self.source_prefix):
img_path = real_url[len(self.source_prefix):]
else: # the still-not-used external images case
# FIXME: won't work for non folderish content types:
img_path = path + "/" + url.rsplit("/")[-1]
img_ref = img_path
if self.embed_images:
image["_orig_path"] = img_path
img_path = path + "/" + image_filename
img_ref = image_filename
# Restore the reference to the
# special view used in plone, after the
# image file name:
if view_parts:
img_ref += "/%s/" % ("/".join(view_parts))
image["_path"] = img_path
if self.replace_references and img_path != rel_url:
img_ref_unicode = img_ref.decode("utf-8")
# FIXME: For now, sc.transmgorgrifier
# fix_path blueprint is not keeping
# references in texts in sync
# so, we assume it is being used on the pipeline
# and normalize the url here
# (should not affect well behaved urls anyway)
img_ref_unicode = normalize_url(img_ref_unicode)
text = text[:start_tag] + img_ref_unicode + text[end_tag:]
# *** Set image to be put in the pipeline
if self.use_wormhole:
self.storage["wormhole"].push(image)
else:
images.append(image)
item["text"] = text
return item, images
