Angelo Gladding

dlv5vbq7lzlthol5 4b942a3185b37d00

South Pasadena, California, United States currently feels like 56°F

Home CodemfFiles

Download raw file

a Microformats parser and utilities


import mf2py
import mf2util as util
import uri

__all__ = ["parse", "util"]

def parse(doc=None, url=None, html_parser=None):
    return mf2py.parse(doc, str(url), html_parser)

def representative_hcard(parsed, source_url):
    """Find the representative h-card for a URL

    :param dict parsed: an mf2 parsed dict
    :param str source_url: the source of the parsed document.
    :return: the representative h-card if one is found
    source_url = uri.parse(source_url).minimized

    def get_normal(hcard, prop):
        for _url in hcard["properties"].get(prop, []):
                yield uri.parse(_url).minimized
            except ValueError:

    hcards = [h for h in util.find_all_entries(parsed, ['h-card'],
              if (h["properties"].get("name", [""])[0] or
                  h["properties"].get("nickname", [""])[0])]

    # uid and url both match source_url
    for hcard in hcards:
        if (source_url in get_normal(hcard, "uid")
                and source_url in get_normal(hcard, "url")):
            return hcard

    # url that is also a rel=me
    for hcard in hcards:
        rel_mes = set()
        for rel_me in parsed.get("rels", {}).get("me", []):
                rel_me = uri.parse(rel_me)
            except ValueError:
            if isinstance(rel_me, (uri.HTTPURI, uri.HTTPSURI)):
        if any(url in rel_mes for url in get_normal(hcard, "url")):
            return hcard

    # single hcard with matching url
    found = []
    count = 0
    for hcard in hcards:
        # if source_url in hcard['properties'].get('url', []):
        for card_url in get_normal(hcard, "url"):
            if card_url.rstrip("/") == source_url:
                count += 1
    if count:
        return found[0]

util.representative_hcard = representative_hcard

def interpret_entry(
        parsed, source_url, base_href=None, hentry=None,
        use_rel_syndication=True, want_json=False, fetch_mf2_func=None):
    given a document containing an h-entry, return a dictionary

        {'type': 'entry',
         'url': permalink of the document (may be different than source_url),
         'published': datetime or date,
         'updated': datetime or date,
         'name': title of the entry,
         'content': body of entry (contains HTML),
         'author': {
          'name': author name,
          'url': author url,
          'photo': author photo
         'syndication': [
           'syndication url',
         'in-reply-to': [...],
         'like-of': [...],
         'repost-of': [...]}

    :param dict parsed: the result of parsing a document containing mf2 markup
    :param str source_url: the URL of the parsed document, used by the
      authorship algorithm
    :param str base_href: (optional) the href value of the base tag
    :param dict hentry: (optional) the item in the above document
      representing the h-entry. if provided, we can avoid a redundant
      call to find_first_entry
    :param boolean use_rel_syndication: (optional, default True) Whether
      to include rel=syndication in the list of syndication sources. Sometimes
      useful to set this to False when parsing h-feeds that erroneously include
      rel=syndication on each entry.
    :param boolean want_json: (optional, default False) if true, the result
      will be pure json with datetimes as strings instead of python objects
    :param callable fetch_mf2_func: (optional) function to fetch mf2 parsed
      output for a given URL.
    :return: a dict with some or all of the described properties


    # find the h-entry if it wasn't provided
    if not hentry:
        hentry = util.find_first_entry(parsed, ['h-entry'])
        if not hentry:
            return {}

    result = _interpret_common_properties(
        parsed, source_url, base_href, hentry, use_rel_syndication, want_json,
    if 'h-cite' in hentry.get('type', []):
        result['type'] = 'cite'
        result['type'] = 'entry'

    # NOTE patch start
    if "category" in hentry["properties"]:
        result["category"] = hentry["properties"]["category"]
    if "pubkey" in hentry["properties"]:
        result["pubkey"] = hentry["properties"]["pubkey"]
    if "vote" in hentry["properties"]:
        result["vote"] = hentry["properties"]["vote"]
    # NOTE patch end

    title = util.get_plain_text(hentry['properties'].get('name'))
    if title and util.is_name_a_title(title, result.get('content-plain')):
        result['name'] = title

    for prop in ('in-reply-to', 'like-of', 'repost-of', 'bookmark-of',
                 'vote-on', 'comment', 'like', 'repost'):  # NOTE added vote-on
        for url_val in hentry['properties'].get(prop, []):
            if isinstance(url_val, dict):
                result.setdefault(prop, []).append(
                    util.interpret(parsed, source_url, base_href, url_val,
                result.setdefault(prop, []).append({
                    'url': url_val,

    return result

util.interpret_entry = interpret_entry

def _interpret_common_properties(
        parsed, source_url, base_href, hentry, use_rel_syndication,
        want_json, fetch_mf2_func):
    result = {}
    props = hentry['properties']

    for prop in ('url', 'uid', 'photo', 'featured' 'logo'):
        value = util.get_plain_text(props.get(prop))
        if value:
            result[prop] = value

    for prop in ('start', 'end', 'published', 'updated', 'deleted'):
        date_str = util.get_plain_text(props.get(prop))
        if date_str:
            if want_json:
                result[prop] = date_str
                result[prop + '-str'] = date_str
                    date = util.parse_datetime(date_str)
                    if date:
                        result[prop] = date
                except ValueError:
                    util.logging.warn('Failed to parse datetime %s', date_str)

    author = util.find_author(parsed, source_url, hentry, fetch_mf2_func)
    if author:
        result['author'] = author

    content_prop = props.get('content')
    content_value = None
    if content_prop:
        if isinstance(content_prop[0], dict):
            content_html = content_prop[0].get('html', '').strip()
            content_value = content_prop[0].get('value', '').strip()
            content_value = content_html = content_prop[0]
        result['content'] = util.convert_relative_paths_to_absolute(
            source_url, base_href, content_html)
        result['content-plain'] = content_value

    summary_prop = props.get('summary')
    if summary_prop:
        if isinstance(summary_prop[0], dict):
            result['summary'] = summary_prop[0]['value']
            result['summary'] = summary_prop[0]

    # Collect location objects, then follow this algorithm to consolidate
    # their properties:
    # //
    location_stack = [props]

    for prop in 'location', 'adr':
        vals = props.get(prop)
        if vals:
            if isinstance(vals[0], util.string_type):
                location_stack.append({'name': vals})
                location_stack.append(vals[0].get('properties', {}))

    geo = props.get('geo')
    if geo:
        if isinstance(geo[0], dict):
            location_stack.append(geo[0].get('properties', {}))
            if geo[0].startswith('geo:'):
                # a geo: URL. try to parse it.
                # //
                parts = geo[0][len('geo:'):].split(';')[0].split(',')
                if len(parts) >= 2:
                        'latitude': [parts[0]],
                        'longitude': [parts[1]],
                        'altitude': [parts[2]] if len(parts) >= 3 else [],

    for prop in util.LOCATION_PROPERTIES:
        for obj in location_stack:
            if obj and obj.get(prop) and not (obj == props and prop == 'name'):
                result.setdefault('location', {})[prop] = obj[prop][0]

    if use_rel_syndication:
        result['syndication'] = list(set(
            parsed.get('rels', {}).get('syndication', []) +
            hentry['properties'].get('syndication', [])))
        result['syndication'] = hentry['properties'].get('syndication', [])

    # TODO patch start
    checkin_prop = props.get('checkin')
    if checkin_prop:
        if isinstance(checkin_prop[0], dict):
            props = checkin_prop[0]['properties']
            result['checkin'] = {"name": props["name"][0]}
                result.update({"latitude": props["latitude"][0],
                               "longitude": props["longitude"][0]})
            except KeyError:
            result['checkin'] = checkin_prop[0]
    # TODO patch end

    return result

util._interpret_common_properties = _interpret_common_properties