def resolve_uid(self, x):
"""Parse HTML and update with URLs pointing to Plone objects.
ex. url: "http://worpress.com/wp-content/uploads/2010/04/image.jpg"
becomes: "resolveuid/c82a53270c904cfbbfd1a0d4cef90676"
:param x: [required] Parsed Regex
:type x: type Regex Match object
:returns: the tag with an internal url
:rtype: str
"""
start = x.group(1) # Start of tag ex.: '<img src="'
url = x.group(2) # URL
end = x.group(3) # End of tag ex.: '" />'
url = fix_id(url)
o = urlparse(url)
internal_url = o.netloc == self.domain
is_site_root = o.path == '' or o.path == '/'
# links to external URL or to site root are ignored
if not internal_url or is_site_root:
return x.group(0) # return unchanged
path = str(o.path).strip(' ').lstrip('/')
obj = traverse(self.context, path, None)
if obj is None: # object not found
logger.warn('Could not resolve UUID: {0}'.format(url))
return x.group(0) # return unchanged
# Create internal URL
uuid = obj.UID()
return '{0}resolveuid/{1}{2}'.format(start, uuid, end)
def _skip(row, skip):
"""Test if we will need to skip row processing by dealing with the
following cases:
- parsing errors
- items with revision type
- items with draft status
- explicit request
:param row: [required] row to be analized
:type row: dictionary
:param skip: [required] list of item ID to be explicitly skiped
:type skip: list
:returns: True if we will skip the row
:rtype: bool
"""
if row['ID'] in skip:
logger.info('Skipping row ID: ' + row['ID'])
return True
elif len(row) != 23 and 'publish' in row.values():
logger.warn('Parsing error on row ID: ' + row['ID'])
return True
elif row['post_type'] not in KNOWN_POST_TYPES:
logger.warn('Parsing error on row ID: ' + row['ID'])
return True
elif row['post_type'] == 'revision':
logger.debug('Revision type on row ID: ' + row['ID'])
return True
elif row['post_status'] == 'draft':
logger.debug('Draft status on row ID: ' + row['ID'])
return True
return False
def __iter__(self):
fetch_errors = [] # record all errors
for item in self.previous:
if '_guid' not in item:
yield item
continue
url = item['_guid']
path = item['_path'] # TODO: read path key from options
if not path: # not enough information
yield item
continue
obj = self.context.unrestrictedTraverse(
path.encode().lstrip('/'), None)
# if object exists we will try to avoid downloading it again
if obj is not None:
if obj.portal_type not in ('File', 'Image'): # not an attachment
yield item
continue
# request only the header to check it
try:
r = requests.head(url)
except ConnectionError: # skip on connection error
fetch_errors.append(url)
yield item
continue
# content-length header could be missing if remote web
# server is misconfigured for some mime types
size = int(r.headers.get('content-length', 0))
if size == obj.size(): # already downloaded it
yield item
continue
try:
r = requests.get(url)
except RequestException: # skip on timeouts and other errors
fetch_errors.append(url)
yield item
continue
if r.status_code != 200: # log error and skip item
fetch_errors.append(url)
msg = u'Error {0} when fetching {1}'.format(r.status_code, url)
logger.warn(msg)
yield item
continue
item['_data'] = r.content
yield item
def add_related_content(self, obj, item):
"""Look into WordPress list of related content and create Plone
related content list.
:param obj: [required] object to add related content
:type obj: type constructor parameter
:param item: [required] transmogrify item
:type item: dict
"""
# Get the string with URLs from related content
pinged = item.get('_pinged', '')
if pinged == '':
return # No related content
# The URL is formated with multiple URLs together without
# separator. To break it into a list, I need to split on
# http and reconstruct the url
# TODO: handle HTTPS scheme
related_urls = set('http{0}'.format(url.rstrip('/'))
for url in pinged.split('http')[1:])
# Create a list of related items to update object's field
related_items = []
for url in related_urls:
# Parse URL and check domain
url = fix_id(url)
o = urlparse(url)
if o.netloc != self.domain:
continue
path = str(o.path).strip(' ').lstrip('/')
related_obj = traverse(self.context, path, None)
if related_obj is None: # object not found
logger.warn('Broken link: {0}'.format(url))
continue
# Get related item ID
intids = getUtility(IIntIds)
to_id = intids.getId(related_obj)
related_items.append(RelationValue(to_id))
# No related content
if len(related_items) == 0:
return
obj.relatedItems = related_items
def __iter__(self):
for item in self.previous:
yield item
filename = os.path.join(self.source, 'wp_posts.csv')
assert os.path.isfile(filename), 'Missing file: ' + filename
with open(filename) as csvfile:
csv.field_size_limit(self.field_size_limit)
reader = csv.DictReader(csvfile, **csv_options)
for row in reader:
if _skip(row, self.skip): # should we process this row?
continue
item = dict()
post_type = row['post_type']
if post_type == 'post':
# posts are imported as portal_type
item['portal_type'] = self.portal_type
elif post_type == 'page':
# pages are imported as Page
item['portal_type'] = 'Page'
elif post_type == 'attachment':
# attachments are imported as Image or File
is_image = row['post_mime_type'].startswith('image')
item['portal_type'] = 'Image' if is_image else 'File'
item['_mimetype'] = row['post_mime_type']
item['_guid'] = row['guid'] # store for later
if post_type != 'attachment':
# for posts and pages the id is the post name
item_id = row['post_name']
# Zope ids need to be ASCII
item_id = fix_id(item_id)
item['title'] = strip_tags(row['post_title'])
else:
# for attachments we need to parse the guid
# and use the file name as title
url = urlparse(row['guid'])
item_id = item['title'] = url.path.split('/')[-1]
item_id = fix_id(item_id)
# on Zope ids can't start with "_"
if bad_id(item_id) is not None:
logger.warn('Invalid object id on row ID: ' + row['ID'])
continue
# WordPress stores only publication and modification times
# we use publication date as creation date
item['creation_date'] = item['effective_date'] = row['post_date']
item['modification_date'] = row['post_modified']
try:
item['_path'] = self.get_path(row['ID'], item_id, post_type, item)
except KeyError:
# files defining taxonomies are probably outdated
logger.warn('No taxonomies found for row ID: ' + row['ID'])
continue
item['description'] = row['post_excerpt']
# quotes are escaped; we need to fix that
item['text'] = row['post_content'].replace('\\"', '"')
# TODO: validate HTML to avoid post-processing surprises
# use display_name instead of author_id, if match found
author_id = row['post_author']
item['creators'] = self.display_names.get(author_id, author_id)
if row['post_status'] == 'publish':
item['_transitions'] = 'publish'
item['_pinged'] = row['pinged'] # store for later
yield item
