def test_xml_parser_repomd_repo01_nowarningcb(self): repomd = cr.Repomd() cr.xml_parse_repomd(REPO_01_REPOMD, repomd) self.assertEqual(repomd.revision, "1334667230") self.assertEqual(repomd.repo_tags, []) self.assertEqual(repomd.distro_tags, []) self.assertEqual(repomd.content_tags, []) self.assertEqual(len(repomd.records), 3)
def third_method(): """Parsing main metadata types (primary, filelists, other) at the same time. This approach significantly reduces memory footprint because we don't need to keep all the packages in memory, user can handle them one by one. The API reflects xml_parse_primary/filelists/other except that it handles all of them at the same time. """ def warningcb(warning_type, message): print("PARSER WARNING: %s" % message) return True repomd = cr.Repomd() cr.xml_parse_repomd(os.path.join(REPO_PATH, "repodata/repomd.xml"), repomd, warningcb) primary_xml_path = None filelists_xml_path = None other_xml_path = None for record in repomd.records: if record.type == "primary": primary_xml_path = os.path.join(REPO_PATH, record.location_href) elif record.type == "filelists": filelists_xml_path = os.path.join(REPO_PATH, record.location_href) elif record.type == "other": other_xml_path = os.path.join(REPO_PATH, record.location_href) # # Main XML metadata parsing (primary, filelists, other) # def pkgcb(pkg): # Called when whole package entry from all 3 metadata xml files is parsed print_package_info(pkg) cr.xml_parse_main_metadata_together(primary_xml_path, filelists_xml_path, other_xml_path, None, pkgcb, warningcb, False)
def streaming_iterator(): """Parsing main metadata types (primary, filelists, other) at the same time. This approach significantly reduces memory footprint because we don't need to keep all the packages in memory, user can handle them one by one. This is the most flexible method, and the recommended one if you need all of the RPM metadata. If you only need to parse one file it might not be the most efficient. """ def warningcb(warning_type, message): print("PARSER WARNING: %s" % message) return True repomd = cr.Repomd() cr.xml_parse_repomd(os.path.join(REPO_PATH, "repodata/repomd.xml"), repomd, warningcb) primary_xml_path = None filelists_xml_path = None other_xml_path = None for record in repomd.records: if record.type == "primary": primary_xml_path = os.path.join(REPO_PATH, record.location_href) elif record.type == "filelists": filelists_xml_path = os.path.join(REPO_PATH, record.location_href) elif record.type == "other": other_xml_path = os.path.join(REPO_PATH, record.location_href) # # Main XML metadata parsing (primary, filelists, other) # package_iterator = cr.PackageIterator(primary_path=primary_xml_path, filelists_path=filelists_xml_path, other_path=other_xml_path, warningcb=warningcb) for pkg in package_iterator: # Called when whole package entry from all 3 metadata xml files is parsed print_package_info(pkg)
def test_xml_parser_repomd_repo01(self): warnings = [] def warningcb(warn_type, msg): warnings.append((warn_type, msg)) repomd = cr.Repomd() cr.xml_parse_repomd(REPO_01_REPOMD, repomd, warningcb) self.assertEqual(warnings, []) self.assertEqual(repomd.revision, "1334667230") self.assertEqual(repomd.repo_tags, []) self.assertEqual(repomd.distro_tags, []) self.assertEqual(repomd.content_tags, []) self.assertEqual(len(repomd.records), 3) self.assertEqual(repomd.records[0].type, "filelists") self.assertEqual(repomd.records[0].location_real, None) self.assertEqual( repomd.records[0].location_href, "repodata/c7db035d0e6f1b2e883a7fa3229e2d2be70c05a8b8d2b57dbb5f9c1a67483b6c-filelists.xml.gz" ) self.assertEqual( repomd.records[0].checksum, "c7db035d0e6f1b2e883a7fa3229e2d2be70c05a8b8d2b57dbb5f9c1a67483b6c") self.assertEqual(repomd.records[0].checksum_type, "sha256") self.assertEqual( repomd.records[0].checksum_open, "85bc611be5d81ac8da2fe01e98ef741d243d1518fcc46ada70660020803fbf09") self.assertEqual(repomd.records[0].checksum_open_type, "sha256") self.assertEqual(repomd.records[0].timestamp, 1334667230) self.assertEqual(repomd.records[0].size, 273) self.assertEqual(repomd.records[0].size_open, 389) self.assertEqual(repomd.records[0].db_ver, 0) self.assertEqual(repomd.records[1].type, "other") self.assertEqual(repomd.records[1].location_real, None) self.assertEqual( repomd.records[1].location_href, "repodata/b752a73d9efd4006d740f943db5fb7c2dd77a8324bd99da92e86bd55a2c126ef-other.xml.gz" ) self.assertEqual( repomd.records[1].checksum, "b752a73d9efd4006d740f943db5fb7c2dd77a8324bd99da92e86bd55a2c126ef") self.assertEqual(repomd.records[1].checksum_type, "sha256") self.assertEqual( repomd.records[1].checksum_open, "da6096c924349af0c326224a33be0cdb26897fbe3d25477ac217261652449445") self.assertEqual(repomd.records[1].checksum_open_type, "sha256") self.assertEqual(repomd.records[1].timestamp, 1334667230) self.assertEqual(repomd.records[1].size, 332) self.assertEqual(repomd.records[1].size_open, 530) self.assertEqual(repomd.records[1].db_ver, 0) self.assertEqual(repomd.records[2].type, "primary") self.assertEqual(repomd.records[2].location_real, None) self.assertEqual( repomd.records[2].location_href, "repodata/6c662d665c24de9a0f62c17d8fa50622307739d7376f0d19097ca96c6d7f5e3e-primary.xml.gz" ) self.assertEqual( repomd.records[2].checksum, "6c662d665c24de9a0f62c17d8fa50622307739d7376f0d19097ca96c6d7f5e3e") self.assertEqual(repomd.records[2].checksum_type, "sha256") self.assertEqual( repomd.records[2].checksum_open, "0fc6cadf97d515e87491d24dc9712d8ddaf2226a21ae7f131ff42d71a877c496") self.assertEqual(repomd.records[2].checksum_open_type, "sha256") self.assertEqual(repomd.records[2].timestamp, 1334667230) self.assertEqual(repomd.records[2].size, 782) self.assertEqual(repomd.records[2].size_open, 2085) self.assertEqual(repomd.records[2].db_ver, 0)
def test_xml_parser_repomd_repo01(self): warnings = [] def warningcb(warn_type, msg): warnings.append((warn_type, msg)) repomd = cr.Repomd() cr.xml_parse_repomd(REPO_01_REPOMD, repomd, warningcb) self.assertEqual(warnings, []) self.assertEqual(repomd.revision, "1334667230") self.assertEqual(repomd.repo_tags, []) self.assertEqual(repomd.distro_tags, []) self.assertEqual(repomd.content_tags, []) self.assertEqual(len(repomd.records), 3) self.assertEqual(repomd.records[0].type, "filelists") self.assertEqual(repomd.records[0].location_real, None) self.assertEqual(repomd.records[0].location_href, "repodata/c7db035d0e6f1b2e883a7fa3229e2d2be70c05a8b8d2b57dbb5f9c1a67483b6c-filelists.xml.gz") self.assertEqual(repomd.records[0].checksum, "c7db035d0e6f1b2e883a7fa3229e2d2be70c05a8b8d2b57dbb5f9c1a67483b6c") self.assertEqual(repomd.records[0].checksum_type, "sha256") self.assertEqual(repomd.records[0].checksum_open, "85bc611be5d81ac8da2fe01e98ef741d243d1518fcc46ada70660020803fbf09") self.assertEqual(repomd.records[0].checksum_open_type, "sha256") self.assertEqual(repomd.records[0].timestamp, 1334667230) self.assertEqual(repomd.records[0].size, 273) self.assertEqual(repomd.records[0].size_open, 389) self.assertEqual(repomd.records[0].db_ver, 0) self.assertEqual(repomd.records[1].type, "other") self.assertEqual(repomd.records[1].location_real, None) self.assertEqual(repomd.records[1].location_href, "repodata/b752a73d9efd4006d740f943db5fb7c2dd77a8324bd99da92e86bd55a2c126ef-other.xml.gz") self.assertEqual(repomd.records[1].checksum, "b752a73d9efd4006d740f943db5fb7c2dd77a8324bd99da92e86bd55a2c126ef") self.assertEqual(repomd.records[1].checksum_type, "sha256") self.assertEqual(repomd.records[1].checksum_open, "da6096c924349af0c326224a33be0cdb26897fbe3d25477ac217261652449445") self.assertEqual(repomd.records[1].checksum_open_type, "sha256") self.assertEqual(repomd.records[1].timestamp, 1334667230) self.assertEqual(repomd.records[1].size, 332) self.assertEqual(repomd.records[1].size_open, 530) self.assertEqual(repomd.records[1].db_ver, 0) self.assertEqual(repomd.records[2].type, "primary") self.assertEqual(repomd.records[2].location_real, None) self.assertEqual(repomd.records[2].location_href, "repodata/6c662d665c24de9a0f62c17d8fa50622307739d7376f0d19097ca96c6d7f5e3e-primary.xml.gz") self.assertEqual(repomd.records[2].checksum, "6c662d665c24de9a0f62c17d8fa50622307739d7376f0d19097ca96c6d7f5e3e") self.assertEqual(repomd.records[2].checksum_type, "sha256") self.assertEqual(repomd.records[2].checksum_open, "0fc6cadf97d515e87491d24dc9712d8ddaf2226a21ae7f131ff42d71a877c496") self.assertEqual(repomd.records[2].checksum_open_type, "sha256") self.assertEqual(repomd.records[2].timestamp, 1334667230) self.assertEqual(repomd.records[2].size, 782) self.assertEqual(repomd.records[2].size_open, 2085) self.assertEqual(repomd.records[2].db_ver, 0)
def second_method(): """Prefered method for repodata parsing. Important callbacks for repodata parsing: newpkgcb -------- Via newpkgcb (Package callback) you could directly affect if the current package element shoud be parsed or not. This decision could be based on three values that are available as attributtes in the <package> element. This values are: - pkgId (package checksum) - name (package name) - arch (package architecture) (Note: This is applicable only for filelists.xml and other.xml, primary.xml doesn't contain this information in <package> element) If newpkgcb returns a package object, the parsed data will be loaded to this package object. If it returns a None, package element is skiped. This could help you to reduce a memory requirements because non wanted packages could be skiped without need to store them into the memory. If no newpkgcb is specified, default callback returning a new package object is used. pkgcb ----- Callback called when a <package> element parsing is done. Its argument is a package object that has been previously returned by the newpkgcb. This function should return True if parsing should continue or False if parsing should be interrupted. Note: Both callbacks are optional, BUT at least one MUST be used (newpkgcb or pkgcb)! warningcb --------- Warning callbacks is called when a non-fatal oddity of prased XML is detected. If True is returned, parsing continues. If return value is False, parsing is terminated. This callback is optional. """ primary_xml_path = None filelists_xml_path = None other_xml_path = None # # repomd.xml parsing # # Parse repomd.xml to get paths (1. Method - Repomd object based) # Pros: Easy to use repomd = cr.Repomd(os.path.join(REPO_PATH, "repodata/repomd.xml")) # Parse repomd.xml (2. Method - Parser based) # Pros: Warning callback could be specified def warningcb(warning_type, message): """Optional callback for warnings about wierd stuff and formatting in XML. :param warning_type: Integer value. One from the XML_WARNING_* constants. :param message: String message. """ print "PARSER WARNING: %s" % message return True repomd2 = cr.Repomd() cr.xml_parse_repomd(os.path.join(REPO_PATH, "repodata/repomd.xml"), repomd2, warningcb) # Get stuff we need # (repomd or repomd2 could be used, both have the same values) for record in repomd.records: if record.type == "primary": primary_xml_path = record.location_href elif record.type == "filelists": filelists_xml_path = record.location_href elif record.type == "other": other_xml_path = record.location_href # # Main XML metadata parsing (primary, filelists, other) # packages = {} def pkgcb(pkg): # Called when whole package entry in xml is parsed packages[pkg.pkgId] = pkg def newpkgcb(pkgId, name, arch): # Called when new package entry is encountered # And only opening <package> element is parsed # This function has to return a package to which # parsed data will be added or None if this package # should be skiped. return packages.get(pkgId, None) # Option do_files tells primary parser to skip <file> element of package. # If you plan to parse filelists.xml after the primary.xml, always # set do_files to False. cr.xml_parse_primary(os.path.join(REPO_PATH, primary_xml_path), pkgcb=pkgcb, do_files=False, warningcb=warningcb) cr.xml_parse_filelists(os.path.join(REPO_PATH, filelists_xml_path), newpkgcb=newpkgcb, warningcb=warningcb) cr.xml_parse_other(os.path.join(REPO_PATH, other_xml_path), newpkgcb=newpkgcb, warningcb=warningcb) for pkg in packages.itervalues(): print_package_info(pkg)
def oneshot_callback(): """Parse one file at a time into a set of packages. Use of this method is discouraged. newpkgcb -------- Via newpkgcb (Package callback) you could directly affect if the current package element should be parsed or not. This decision could be based on three values that are available as attributtes in the <package> element. This values are: - pkgId (package checksum) - name (package name) - arch (package architecture) (Note: This is applicable only for filelists.xml and other.xml, primary.xml doesn't contain this information in <package> element) If newpkgcb returns a package object, the parsed data will be loaded to this package object. If it returns a None, package element is skiped. This could help you to reduce a memory requirements because non wanted packages could be skiped without need to store them into the memory. If no newpkgcb is specified, default callback returning a new package object is used. pkgcb ----- Callback called when a <package> element parsing is done. Its argument is a package object that has been previously returned by the newpkgcb. This function should return True if parsing should continue or False if parsing should be interrupted. Note: Both callbacks are optional, BUT at least one MUST be used (newpkgcb or pkgcb)! warningcb --------- Warning callbacks is called when a non-fatal oddity of prased XML is detected. If True is returned, parsing continues. If return value is False, parsing is terminated. This callback is optional. """ primary_xml_path = None filelists_xml_path = None other_xml_path = None # # repomd.xml parsing # # Parse repomd.xml to get paths (1. Method - Repomd object based) # Pros: Easy to use repomd = cr.Repomd(os.path.join(REPO_PATH, "repodata/repomd.xml")) # Parse repomd.xml (2. Method - Parser based) # Pros: Warning callback could be specified def warningcb(warning_type, message): """Optional callback for warnings about wierd stuff and formatting in XML. :param warning_type: Integer value. One from the XML_WARNING_* constants. :param message: String message. """ print("PARSER WARNING: %s" % message) return True repomd2 = cr.Repomd() cr.xml_parse_repomd(os.path.join(REPO_PATH, "repodata/repomd.xml"), repomd2, warningcb) # Get stuff we need # (repomd or repomd2 could be used, both have the same values) for record in repomd.records: if record.type == "primary": primary_xml_path = record.location_href elif record.type == "filelists": filelists_xml_path = record.location_href elif record.type == "other": other_xml_path = record.location_href # # Main XML metadata parsing (primary, filelists, other) # packages = {} def pkgcb(pkg): # Called when whole package entry in xml is parsed packages[pkg.pkgId] = pkg def newpkgcb(pkgId, name, arch): # Called when new package entry is encountered # And only opening <package> element is parsed # This function has to return a package to which # parsed data will be added or None if this package # should be skiped. return packages.get(pkgId, None) # Option do_files tells primary parser to skip <file> element of package. # If you plan to parse filelists.xml after the primary.xml, always # set do_files to False. cr.xml_parse_primary(os.path.join(REPO_PATH, primary_xml_path), pkgcb=pkgcb, do_files=False, warningcb=warningcb) cr.xml_parse_filelists(os.path.join(REPO_PATH, filelists_xml_path), newpkgcb=newpkgcb, warningcb=warningcb) cr.xml_parse_other(os.path.join(REPO_PATH, other_xml_path), newpkgcb=newpkgcb, warningcb=warningcb) for pkg in packages.values(): print_package_info(pkg)
def _load_cached_updateinfo(self): """ Load the cached updateinfo.xml from '../{tag}.repocache/repodata' """ seen_ids = set() from_cache = set() existing_ids = set() # Parse the updateinfo out of the repomd updateinfo = None repomd_xml = os.path.join(self.cached_repodata, 'repomd.xml') repomd = cr.Repomd() cr.xml_parse_repomd(repomd_xml, repomd) for record in repomd.records: if record.type == 'updateinfo': updateinfo = os.path.join(os.path.dirname( os.path.dirname(self.cached_repodata)), record.location_href) break assert updateinfo, 'Unable to find updateinfo' # Load the metadata with createrepo_c log.info('Loading cached updateinfo: %s', updateinfo) uinfo = cr.UpdateInfo(updateinfo) # Determine which updates are present in the cache for update in uinfo.updates: existing_ids.add(update.id) # Generate metadata for any new builds for update in self.updates: seen_ids.add(update.alias) if not update.alias: self.missing_ids.append(update.title) continue if update.alias in existing_ids: notice = None for value in uinfo.updates: if value.title == update.title: notice = value break if not notice: log.warn('%s ID in cache but notice cannot be found', update.title) self.add_update(update) continue if notice.updated_date: if notice.updated_date < update.date_modified: log.debug('Update modified, generating new notice: %s' % update.title) self.add_update(update) else: log.debug('Loading updated %s from cache' % update.title) from_cache.add(update.alias) elif update.date_modified: log.debug('Update modified, generating new notice: %s' % update.title) self.add_update(update) else: log.debug('Loading %s from cache' % update.title) from_cache.add(update.alias) else: log.debug('Adding new update notice: %s' % update.title) self.add_update(update) # Add all relevant notices from the cache to this document for notice in uinfo.updates: if notice.id in from_cache: log.debug('Keeping existing notice: %s', notice.title) self.uinfo.append(notice) else: # Keep all security notices in the stable repo if self.request is not UpdateRequest.testing: if notice.type == 'security': if notice.id not in seen_ids: log.debug('Keeping existing security notice: %s', notice.title) self.uinfo.append(notice) else: log.debug('%s already added?', notice.title) else: log.debug('Purging cached stable notice %s', notice.title) else: log.debug('Purging cached testing update %s', notice.title)