def parse(location, handler):
"""
Given the location of an XML file and a handler function accepting a
etree document, parse the file at location and invoke the handler on
the etree doc. If parsing fails while calling handler, another
approach to parsing is used.
This is a workaround some lxml bug/weirdness wrt unicode in the 2.3
version in use.
The `handler` function must have no side effects and can be called
again on failures without risk.
Try first to call lxml from a location then try from a string to
deal with weird encodings
"""
try:
parser = etree.XMLParser(recover=True, remove_blank_text=True, resolve_entities=False)
xdoc = etree.parse(location, parser)
return handler(xdoc)
except:
parser = etree.XMLParser(recover=True, remove_blank_text=True, resolve_entities=False)
text = analysis.unicode_text(location)
xdoc= etree.fromstring(_as_unicode_bytes(text), parser)
return handler(xdoc)
def __init__(self, location):
# NOTE: most of this is copied over from Pom.__init__
try:
with codecs.open(location, 'rb', encoding='UTF-8') as fh:
xml = fh.read()
except UnicodeDecodeError as _a:
xml = analysis.unicode_text(location)
xml = xml[xml.find('<project'):]
xml = STRIP_NAMESPACE_RE.sub('<project>', xml, 1)
self._xml = etree.fromstring(xml, parser=POM_PARSER)
# FXIME: we do not use a client for now. there are pending issues at pymaven to address this
self._client = None
self.model_version = self._get_attribute('modelVersion')
self.group_id = self._get_attribute('groupId')
self.artifact_id = self._get_attribute('artifactId')
self.version = self._get_attribute('version')
self.classifier = self._get_attribute('classifier')
self.packaging = self._get_attribute('packaging') or 'jar'
self.name = self._get_attribute('name')
self.description = self._get_attribute('description')
self.inception_year = self._get_attribute('inceptionYear')
self.url = self._get_attribute('url')
self.organization_name = self._get_attribute('organization/name')
self.organization_url = self._get_attribute('organization/url')
self.licenses = list(self._find_licenses())
self.developers = list(self._find_parties('developers/developer'))
self.contributors = list(
self._find_parties('contributors/contributor'))
self.mailing_lists = list(self._find_mailing_lists())
self.scm = self._find_scm()
self.issue_management = self._find_issue_management()
self.ci_management = self._find_ci_management()
self.distribution_management = self._find_distribution_management()
self.repositories = list(
self._find_repositories('repositories/repository'))
self.plugin_repositories = list(
self._find_repositories('pluginRepositories/pluginRepository'))
self.modules = self._get_attributes_list('modules/module')
# FIXME: this attribute should be collected with the parent but
# is not retrieved yet by pymaven it points to the relative path
# where to find the full parent POM
self.parent_relative_path = self._get_attribute(
'relativePath') # or '../pom.xml'
# FIXME: Other types that are not collected for now (or
# indirectly through dependencies management) include: build,
# reporting, profiles, etc
# dynamic attributes
self._parent = None
self._dep_mgmt = None
self._dependencies = None
self._properties = None
def _parse_as_string(location):
"""
Return an etree doc from the XML file at `location` trying hard to get
unicode.
"""
parser = etree.XMLParser(recover=True,
remove_blank_text=True,
resolve_entities=False)
text = analysis.unicode_text(location)
return etree.fromstring(_as_unicode_bytes(text), parser)
def __init__(self, location=None, text=None):
"""
Build a POM from a location or unicode text.
"""
assert (location or text) and (not (location and text))
if location:
try:
with io.open(location, encoding='utf-8') as fh:
xml_text = fh.read()
except UnicodeDecodeError as _a:
xml_text = analysis.unicode_text(location)
else:
xml_text = text
xml_text = strip_namespace(xml_text)
xml_text = xml_text.encode('utf-8')
if TRACE:
logger.debug('MavenPom.__init__: xml_text: {}'.format(xml_text))
self._pom_data = etree.fromstring(xml_text, parser=pom.POM_PARSER)
# collect and then remove XML comments from the XML elements tree
self.comments = self._get_comments()
etree.strip_tags(self._pom_data, etree.Comment)
# FIXME: we do not use a client for now.
# There are pending issues at pymaven to address this
self._client = None
self.model_version = self._get_attribute('modelVersion')
if not self.model_version:
# for older POM version 3
self.model_version = self._get_attribute('pomVersion')
self.group_id = self._get_attribute('groupId')
self.artifact_id = self._get_attribute('artifactId')
if TRACE:
logger.debug('MavenPom.__init__: self.artifact_id: {}'.format(self.artifact_id))
self.version = self._get_attribute('version')
self.classifier = self._get_attribute('classifier')
self.packaging = self._get_attribute('packaging') or 'jar'
self.name = self._get_attribute('name')
self.description = self._get_attribute('description')
self.inception_year = self._get_attribute('inceptionYear')
self.url = self._get_attribute('url')
self.organization_name = self._get_attribute('organization/name')
self.organization_url = self._get_attribute('organization/url')
self.licenses = list(self._find_licenses())
self.developers = list(self._find_parties('developers/developer'))
self.contributors = list(self._find_parties('contributors/contributor'))
self.mailing_lists = list(self._find_mailing_lists())
self.scm = self._find_scm()
self.issue_management = self._find_issue_management()
self.ci_management = self._find_ci_management()
self.distribution_management = self._find_distribution_management()
self.repositories = list(self._find_repositories('repositories/repository'))
self.plugin_repositories = list(self._find_repositories('pluginRepositories/pluginRepository'))
self.modules = self._get_attributes_list('modules/module')
# FIXME: this attribute should be collected with the parent but
# is not retrieved yet by pymaven it points to the relative path
# where to find the full parent POM
self.parent_relative_path = self._get_attribute('relativePath') # or '../pom.xml_text'
# FIXME: Other types that are not collected for now (or
# indirectly through dependencies management) include: build,
# reporting, profiles, etc
# dynamic attributes
self._parent = None
self._dep_mgmt = None
self._dependencies = None
self._properties = None
def __init__(self, location=None, text=None):
"""
Build a POM from a location or unicode text.
"""
assert (location or text) and (not (location and text))
# NOTE: most of this is derived from pymaven.Pom.__init__
if location:
try:
with io.open(location, encoding='utf-8') as fh:
xml = fh.read()
except UnicodeDecodeError as _a:
xml = analysis.unicode_text(location)
else:
xml = text
xml = xml[xml.find('<project'):]
xml = STRIP_NAMESPACE_RE.sub('<project>', xml, 1)
parser = etree.XMLParser(
recover=True,
# we keep comments in case there is a license in the comments
remove_comments=False,
remove_pis=True,
remove_blank_text=True, resolve_entities=False
)
self._xml = etree.fromstring(xml, parser=parser)
# collect and then remove XML comments from the XML elements tree
self.comments = self._get_comments()
etree.strip_tags(self._xml, etree.Comment)
# FIXME: we do not use a client for now. There are pending issues at pymaven to address this
self._client = None
self.model_version = self._get_attribute('modelVersion')
if not self.model_version:
# for version 3
self.model_version = self._get_attribute('pomVersion')
self.group_id = self._get_attribute('groupId')
self.artifact_id = self._get_attribute('artifactId')
if TRACE: logger.debug('MavenPom.__init__: self.artifact_id: {}'.format(self.artifact_id))
self.version = self._get_attribute('version')
self.classifier = self._get_attribute('classifier')
self.packaging = self._get_attribute('packaging') or 'jar'
self.name = self._get_attribute('name')
self.description = self._get_attribute('description')
self.inception_year = self._get_attribute('inceptionYear')
self.url = self._get_attribute('url')
self.organization_name = self._get_attribute('organization/name')
self.organization_url = self._get_attribute('organization/url')
self.licenses = list(self._find_licenses())
self.developers = list(self._find_parties('developers/developer'))
self.contributors = list(self._find_parties('contributors/contributor'))
self.mailing_lists = list(self._find_mailing_lists())
self.scm = self._find_scm()
self.issue_management = self._find_issue_management()
self.ci_management = self._find_ci_management()
self.distribution_management = self._find_distribution_management()
self.repositories = list(self._find_repositories('repositories/repository'))
self.plugin_repositories = list(self._find_repositories('pluginRepositories/pluginRepository'))
self.modules = self._get_attributes_list('modules/module')
# FIXME: this attribute should be collected with the parent but
# is not retrieved yet by pymaven it points to the relative path
# where to find the full parent POM
self.parent_relative_path = self._get_attribute('relativePath') # or '../pom.xml'
# FIXME: Other types that are not collected for now (or
# indirectly through dependencies management) include: build,
# reporting, profiles, etc
# dynamic attributes
self._parent = None
self._dep_mgmt = None
self._dependencies = None
self._properties = None
def __init__(self, location=None, text=None):
"""
Build a POM from a location or unicode text.
"""
assert (location or text) and (not (location and text))
# NOTE: most of this is copied over from Pom.__init__
if location:
try:
with codecs.open(location, 'rb', encoding='UTF-8') as fh:
xml = fh.read()
except UnicodeDecodeError as _a:
xml = analysis.unicode_text(location)
else:
xml = text
xml = xml[xml.find('<project'):]
xml = STRIP_NAMESPACE_RE.sub('<project>', xml, 1)
parser = etree.XMLParser(
recover=True,
remove_comments=True,
remove_pis=True,
remove_blank_text=True, resolve_entities=False
)
self._xml = etree.fromstring(xml, parser=parser)
# FXIME: we do not use a client for now. there are pending issues at pymaven to address this
self._client = None
self.model_version = self._get_attribute('modelVersion')
if not self.model_version:
# for version 3
self.model_version = self._get_attribute('pomVersion')
self.group_id = self._get_attribute('groupId')
self.artifact_id = self._get_attribute('artifactId')
self.version = self._get_attribute('version')
self.classifier = self._get_attribute('classifier')
self.packaging = self._get_attribute('packaging') or 'jar'
self.name = self._get_attribute('name')
self.description = self._get_attribute('description')
self.inception_year = self._get_attribute('inceptionYear')
self.url = self._get_attribute('url')
self.organization_name = self._get_attribute('organization/name')
self.organization_url = self._get_attribute('organization/url')
self.licenses = list(self._find_licenses())
self.developers = list(self._find_parties('developers/developer'))
self.contributors = list(self._find_parties('contributors/contributor'))
self.mailing_lists = list(self._find_mailing_lists())
self.scm = self._find_scm()
self.issue_management = self._find_issue_management()
self.ci_management = self._find_ci_management()
self.distribution_management = self._find_distribution_management()
self.repositories = list(self._find_repositories('repositories/repository'))
self.plugin_repositories = list(self._find_repositories('pluginRepositories/pluginRepository'))
self.modules = self._get_attributes_list('modules/module')
# FIXME: this attribute should be collected with the parent but
# is not retrieved yet by pymaven it points to the relative path
# where to find the full parent POM
self.parent_relative_path = self._get_attribute('relativePath') # or '../pom.xml'
# FIXME: Other types that are not collected for now (or
# indirectly through dependencies management) include: build,
# reporting, profiles, etc
# dynamic attributes
self._parent = None
self._dep_mgmt = None
self._dependencies = None
self._properties = None