Source code for press.parsers.common

import re
from functools import partial

from lxml import etree
from litezip.main import COLLECTION_NSMAP


__all__ = (
    'make_elm_tree',
    'make_cnx_xpath',
    'parse_common_properties',
)


def _maybe(a_list):
    """Grab the first value if it exists."""
    try:
        return a_list[0]
    except IndexError:
        return None


def make_elm_tree(model):
    """Makes an element-like object (:mod:`lxml.etree`) from a litezip model
    (:class:`litezip.Collection` or :class:`litezip.Module`).

    """
    with model.file.open() as fb:
        elm_tree = etree.parse(fb)
    return elm_tree


def make_cnx_xpath(elm_tree):
    """Makes an xpath function that includes the CNX namespaces.

    :param elm_tree: the xml element to begin the xpath from
    :type elm_tree: an element-like object from :mod:`lxml.etree`

    """
    return partial(elm_tree.xpath, namespaces=COLLECTION_NSMAP)


def _squash_to_text(elm, remove_namespaces=False):
    if elm is None:
        return None
    value = [elm.text or '']
    for child in elm.getchildren():
        value.append(etree.tostring(child).decode('utf-8').strip())
    if remove_namespaces:
        value = [re.sub(' xmlns:?[^=]*="[^"]*"', '', v) for v in value]
    value = ''.join(value)
    return value


def parse_common_properties(elm_tree):
    """Given an element-like object (:mod:`lxml.etree`)
    lookup the common and return the properties.

    :param elm_tree: the root xml element
    :type elm_tree: an element-like object from :mod:`lxml.etree`
    :returns: common metadata properties
    :rtype: dict

    .. note::
       Press does not parse or update user information (aka "actor" in
       the xml). User modifications should be done using the "legacy"
       software.

    """
    xpath = make_cnx_xpath(elm_tree)
    role_xpath = lambda xp: tuple(xpath(xp)[0].split())  # noqa: E731

    props = {
        'id': _maybe(xpath('//md:content-id/text()')),
        'version': xpath('//md:version/text()')[0],
        'created': xpath('//md:created/text()')[0],
        'revised': xpath('//md:revised/text()')[0],
        'title': xpath('//md:title/text()')[0],
        'license_url': xpath('//md:license/@url')[0],
        'language': xpath('//md:language/text()')[0],
        'authors': role_xpath('//md:role[@type="author"]/text()'),
        'maintainers': role_xpath('//md:role[@type="maintainer"]/text()'),
        'licensors': role_xpath('//md:role[@type="licensor"]/text()'),
        'keywords': tuple(xpath('//md:keywordlist/md:keyword/text()')),
        'subjects': tuple(xpath('//md:subjectlist/md:subject/text()')),
        'abstract': _squash_to_text(_maybe(xpath('//md:abstract')),
                                    remove_namespaces=True),
    }
    return props