def parse_pom(location, check_is_pom=False):
"""
Return a MavenPom object from the Maven POM file at location.
"""
pom = _get_mavenpom(location, check_is_pom)
if not pom:
return {}
return pom.to_dict()
def parse(location=None, text=None, check_is_pom=True, extra_properties=None):
"""
Return a MavenPomPackage or None.
Parse a pom file at `location` or using the provided `text` (one or
the other but not both).
Check if the location is a POM if `check_is_pom` is True.
When resolving the POM, use an optional `extra_properties` mapping
of name/value pairs to resolve properties.
"""
pom = get_maven_pom(location, text, check_is_pom, extra_properties)
if not pom:
return
if TRACE:
logger.debug('parse: pom:.to_dict()\n{}'.format(pformat(pom.to_dict())))
version = pom.version
# pymaven whart
if version == 'latest.release':
version = None
qualifiers = {}
classifier = pom.classifier
if classifier:
qualifiers['classifier'] = classifier
packaging = pom.packaging
if packaging:
extension = get_extension(packaging)
if extension and extension not in ('jar', 'pom'):
# we use type as in the PURL spec: this is a problematic field with
# complex defeinition in Maven
qualifiers['type'] = extension
declared_license = pom.licenses
source_packages = []
# TODO: what does this mean????
if not classifier and all([pom.group_id, pom.artifact_id, version]):
spurl = PackageURL(
type=MavenPomPackage.default_type,
namespace=pom.group_id,
name=pom.artifact_id,
version=version,
# we hardcode the source qualifier for now...
qualifiers=dict(classifier='sources'))
source_packages = [spurl.to_string()]
pname = pom.name or ''
pdesc = pom.description or ''
if pname == pdesc:
description = pname
else:
description = [d for d in (pname, pdesc) if d]
description = '\n'.join(description)
issue_mngt = pom.issue_management or {}
bug_tracking_url = issue_mngt.get('url')
scm = pom.scm or {}
vcs_url, code_view_url = build_vcs_and_code_view_urls(scm)
# FIXME: there are still other data to map in a Package
package = MavenPomPackage(
namespace=pom.group_id,
name=pom.artifact_id,
version=version,
qualifiers=qualifiers or None,
description=description or None,
homepage_url=pom.url or None,
declared_license=declared_license or None,
parties=get_parties(pom),
dependencies=get_dependencies(pom),
source_packages=source_packages,
bug_tracking_url=bug_tracking_url,
vcs_url=vcs_url,
code_view_url=code_view_url,
)
return package
def parse(
location,
datasource_id,
package_type,
primary_language,
base_url='https://repo1.maven.org/maven2',
):
"""
Yield Packagedata objects from parsing a Maven pom file at `location` or
using the provided `text` (one or the other but not both).
"""
pom = get_maven_pom(location=location)
if not pom:
return
if TRACE:
ptd = pformat(pom.to_dict())
logger.debug(f'PomXmlHandler.parse: pom:.to_dict()\n{ptd}')
version = pom.version
# pymaven whart
if version == 'latest.release':
version = None
qualifiers = {}
classifier = pom.classifier
if classifier:
qualifiers['classifier'] = classifier
packaging = pom.packaging
if packaging:
extension = get_extension(packaging)
if extension and extension not in ('jar', 'pom'):
# we use type as in the PURL spec: this is a problematic field with
# complex defeinition in Maven
qualifiers['type'] = extension
declared_license = pom.licenses
group_id = pom.group_id
artifact_id = pom.artifact_id
# craft a source package purl for the main binary
source_packages = []
is_main_binary_jar = not classifier and all(
[group_id, artifact_id, version])
if is_main_binary_jar:
spurl = PackageURL(
type=package_type,
namespace=group_id,
name=artifact_id,
version=version,
# we hardcode the source qualifier for now...
qualifiers=dict(classifier='sources'))
source_packages = [spurl.to_string()]
pname = pom.name or ''
pdesc = pom.description or ''
if pname == pdesc:
description = pname
else:
description = [d for d in (pname, pdesc) if d]
description = '\n'.join(description)
issue_mngt = pom.issue_management or {}
bug_tracking_url = issue_mngt.get('url')
scm = pom.scm or {}
urls = build_vcs_and_code_view_urls(scm)
urls.update(
get_urls(
namespace=group_id,
name=artifact_id,
version=version,
qualifiers=qualifiers,
base_url=base_url,
))
# FIXME: there are still other data to map in a PackageData
package_data = models.PackageData(
datasource_id=datasource_id,
type=package_type,
primary_language=primary_language,
namespace=group_id,
name=artifact_id,
version=version,
qualifiers=qualifiers or None,
description=description or None,
homepage_url=pom.url or None,
declared_license=declared_license or None,
parties=get_parties(pom),
dependencies=get_dependencies(pom),
source_packages=source_packages,
bug_tracking_url=bug_tracking_url,
**urls,
)
if not package_data.license_expression and package_data.declared_license:
package_data.license_expression = models.compute_normalized_license(
package_data.declared_license)
yield package_data
