def FirstInListLikeObject(self, value, otherwise = None): if bm_extract.is_list(value): if value: return value[0] return None if bm_extract.is_list_like(value): any = False for sub in value: return value return None return otherwise
def CustomizeAtomItem(self, itemd):
try:
author = itemd.pop("author")
if author:
itemd["author"] = bm_extract.as_string(author, "@@title")
if bm_extract.is_list(author) or bm_extract.is_list_like(author):
itemd["hcard:author"] = map(lambda a: hcard.decompose(a, "hcard"), author)
elif bm_extract.is_dict(author):
itemd["hcard:author"] = hcard.decompose(author, "hcard")
except KeyError:
pass
self.ExtractCategories(itemd)
return bm_api.APIBase.CustomizeAtomItem(self, itemd)
def TranscribeNode(self, e_node, o):
"""Convert a dictionary into XML"""
if bm_extract.is_list(o):
for sub in o:
self.TranscribeNode(e_node, sub)
elif bm_extract.is_dict(o):
#
# Get the attributes
#
ad = {}
for key, sub in o.iteritems():
if key.startswith("@@") or key.find(":@@") > -1:
continue
if key == '@':
e_node.text = bm_extract.coerce_string(sub, separator = ",")
continue
if key.startswith("@") or key.find(":@") > -1:
if bm_extract.is_atomic(sub):
e_node.set(key.replace('@', ''), bm_extract.coerce_string(sub))
## ad[key.replace('@', '')] = bm_extract.coerce_string(sub)
elif bm_extract.is_list(sub) or bm_extract.is_list_like(sub):
e_node.set(key.replace('@', ''), bm_extract.coerce_string(sub, separator = ","))
## ad[key.replace('@', '')] = bm_extract.coerce_string(sub, separator = ",")
#
# Note here that:
# - @@ means an attribute it hidden
# - @ are attributes, and are found in the previous step
# - lists are processed specially, as they result in repeated children
#
for key, sub in o.iteritems():
if key.startswith("@@") or key.find(":@@") > -1:
continue
if key.startswith("@") or key.find(":@") > -1:
continue
#
#
#
if bm_extract.is_list_like(sub):
sub = list(sub)
if bm_extract.is_list(sub):
any = False
for subsub in sub:
any = True
e_child = ElementTree.SubElement(e_node, key)
self.TranscribeNode(e_child, subsub)
if any:
continue
sub = None
#
#
#
e_child = ElementTree.SubElement(e_node, key)
self.TranscribeNode(e_child, sub)
elif bm_extract.is_list_like(o):
for sub in list(o):
self.TranscribeNode(e_node, sub)
elif bm_extract.is_none(o):
pass
else:
if e_node.text:
e_node.text += "\n"
e_node.text += bm_extract.coerce_string(o)
else:
e_node.text = bm_extract.coerce_string(o)
def ScrubEntry(self, itemd):
if bm_extract.is_dict(itemd):
nd = {}
seen_html = False
seen_rss = False
seen_url = False
for key, value in itemd.iteritems():
if self.AtomLike():
if key == "link":
key = "htmlUrl"
elif key == "feeds":
key = "rssUrl"
elif key == "content":
key = "description"
elif key == "title":
key = "text"
elif key == "category":
key = "tags"
value = ", ".join(map(lambda d: d["term"], value))
elif key == "links":
for ld in bm_extract.coerce_list(value):
if bm_extract.as_string(ld, "rel") == "alternate":
key = "rssUrl"
value = bm_extract.as_string(ld, "href")
#
# datetimes (?)
#
try:
created = itemd.pop("created")
itemd["created"] = bm_extract.coerce_datetime(created, otherwise = created, rfc822 = True)
except KeyError:
pass
if key == "rssUrl":
value = self.FirstInListLikeObject(value, value)
if value == None:
continue
seen_rss = True
elif key == "htmlUrl":
value = self.FirstInListLikeObject(value, value)
if value == None:
continue
seen_html = True
elif key == "url":
seen_url = True
if key in [ "items", "outline" ]:
nd["outline"] = self.ScrubEntry(value)
elif value == None:
pass
elif bm_extract.is_atomic(value):
nd['@%s' % key] = value
if seen_rss:
nd.setdefault("@type", "rss")
elif seen_html:
nd.setdefault("@type", "link")
elif seen_url:
nd.setdefault("@type", "link")
nd.setdefault("@text", "")
return nd
elif bm_extract.is_atomic(itemd):
return {
"@title" : bm_extract.coerce_string(itemd)
}
elif bm_extract.is_list(itemd) or bm_extract.is_list_like(itemd):
return map(self.ScrubEntry, itemd)
return itemd