diff --git a/PKG-INFO b/PKG-INFO index db11821..da1d5fb 100644 --- a/PKG-INFO +++ b/PKG-INFO @@ -1,69 +1,69 @@ Metadata-Version: 2.1 Name: swh.indexer -Version: 0.0.57 +Version: 0.0.58 Summary: Software Heritage Content Indexer Home-page: https://forge.softwareheritage.org/diffusion/78/ Author: Software Heritage developers Author-email: swh-devel@inria.fr License: UNKNOWN Project-URL: Bug Reports, https://forge.softwareheritage.org/maniphest Project-URL: Funding, https://www.softwareheritage.org/donate Project-URL: Source, https://forge.softwareheritage.org/source/swh-indexer Description: swh-indexer ============ Tools to compute multiple indexes on SWH's raw contents: - content: - mimetype - ctags - language - fossology-license - metadata - revision: - metadata An indexer is in charge of: - looking up objects - extracting information from those objects - store those information in the swh-indexer db There are multiple indexers working on different object types: - content indexer: works with content sha1 hashes - revision indexer: works with revision sha1 hashes - origin indexer: works with origin identifiers Indexation procedure: - receive batch of ids - retrieve the associated data depending on object type - compute for that object some index - store the result to swh's storage Current content indexers: - mimetype (queue swh_indexer_content_mimetype): detect the encoding and mimetype - language (queue swh_indexer_content_language): detect the programming language - ctags (queue swh_indexer_content_ctags): compute tags information - fossology-license (queue swh_indexer_fossology_license): compute the license - metadata: translate file into translated_metadata dict Current revision indexers: - metadata: detects files containing metadata and retrieves translated_metadata in content_metadata table in storage or run content indexer to translate files. Platform: UNKNOWN Classifier: Programming Language :: Python :: 3 Classifier: Intended Audience :: Developers Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3) Classifier: Operating System :: OS Independent Classifier: Development Status :: 5 - Production/Stable Description-Content-Type: text/markdown Provides-Extra: testing diff --git a/swh.indexer.egg-info/PKG-INFO b/swh.indexer.egg-info/PKG-INFO index db11821..da1d5fb 100644 --- a/swh.indexer.egg-info/PKG-INFO +++ b/swh.indexer.egg-info/PKG-INFO @@ -1,69 +1,69 @@ Metadata-Version: 2.1 Name: swh.indexer -Version: 0.0.57 +Version: 0.0.58 Summary: Software Heritage Content Indexer Home-page: https://forge.softwareheritage.org/diffusion/78/ Author: Software Heritage developers Author-email: swh-devel@inria.fr License: UNKNOWN Project-URL: Bug Reports, https://forge.softwareheritage.org/maniphest Project-URL: Funding, https://www.softwareheritage.org/donate Project-URL: Source, https://forge.softwareheritage.org/source/swh-indexer Description: swh-indexer ============ Tools to compute multiple indexes on SWH's raw contents: - content: - mimetype - ctags - language - fossology-license - metadata - revision: - metadata An indexer is in charge of: - looking up objects - extracting information from those objects - store those information in the swh-indexer db There are multiple indexers working on different object types: - content indexer: works with content sha1 hashes - revision indexer: works with revision sha1 hashes - origin indexer: works with origin identifiers Indexation procedure: - receive batch of ids - retrieve the associated data depending on object type - compute for that object some index - store the result to swh's storage Current content indexers: - mimetype (queue swh_indexer_content_mimetype): detect the encoding and mimetype - language (queue swh_indexer_content_language): detect the programming language - ctags (queue swh_indexer_content_ctags): compute tags information - fossology-license (queue swh_indexer_fossology_license): compute the license - metadata: translate file into translated_metadata dict Current revision indexers: - metadata: detects files containing metadata and retrieves translated_metadata in content_metadata table in storage or run content indexer to translate files. Platform: UNKNOWN Classifier: Programming Language :: Python :: 3 Classifier: Intended Audience :: Developers Classifier: License :: OSI Approved :: GNU General Public License v3 (GPLv3) Classifier: Operating System :: OS Independent Classifier: Development Status :: 5 - Production/Stable Description-Content-Type: text/markdown Provides-Extra: testing diff --git a/swh/indexer/ctags.py b/swh/indexer/ctags.py index b6f4bb7..0679692 100644 --- a/swh/indexer/ctags.py +++ b/swh/indexer/ctags.py @@ -1,167 +1,156 @@ # Copyright (C) 2015-2017 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import click import subprocess import json from swh.model import hashutil from .language import compute_language from .indexer import ContentIndexer, DiskIndexer # Options used to compute tags __FLAGS = [ '--fields=+lnz', # +l: language # +n: line number of tag definition # +z: include the symbol's kind (function, variable, ...) '--sort=no', # sort output on tag name '--links=no', # do not follow symlinks '--output-format=json', # outputs in json ] def run_ctags(path, lang=None, ctags_command='ctags'): """Run ctags on file path with optional language. Args: path: path to the file lang: language for that path (optional) Returns: ctags' output """ optional = [] if lang: optional = ['--language-force=%s' % lang] cmd = [ctags_command] + __FLAGS + optional + [path] output = subprocess.check_output(cmd, universal_newlines=True) for symbol in output.split('\n'): if not symbol: continue js_symbol = json.loads(symbol) yield { 'name': js_symbol['name'], 'kind': js_symbol['kind'], 'line': js_symbol['line'], 'lang': js_symbol['language'], } class CtagsIndexer(ContentIndexer, DiskIndexer): CONFIG_BASE_FILENAME = 'indexer/ctags' ADDITIONAL_CONFIG = { 'workdir': ('str', '/tmp/swh/indexer.ctags'), 'tools': ('dict', { 'name': 'universal-ctags', 'version': '~git7859817b', 'configuration': { 'command_line': '''ctags --fields=+lnz --sort=no --links=no ''' '''--output-format=json ''' }, }), 'languages': ('dict', { 'ada': 'Ada', 'adl': None, 'agda': None, # ... }) } def prepare(self): super().prepare() self.working_directory = self.config['workdir'] self.language_map = self.config['languages'] self.tool = self.tools[0] def filter(self, ids): """Filter out known sha1s and return only missing ones. """ yield from self.idx_storage.content_ctags_missing(( { 'id': sha1, 'indexer_configuration_id': self.tool['id'], } for sha1 in ids )) def compute_ctags(self, path, lang): """Compute ctags on file at path with language lang. """ return run_ctags(path, lang=lang) def index(self, id, data): """Index sha1s' content and store result. Args: id (bytes): content's identifier data (bytes): raw content in bytes Returns: A dict, representing a content_mimetype, with keys: - id (bytes): content's identifier (sha1) - ctags ([dict]): ctags list of symbols """ lang = compute_language(data, log=self.log)['lang'] if not lang: return None ctags_lang = self.language_map.get(lang) if not ctags_lang: return None ctags = { 'id': id, } filename = hashutil.hash_to_hex(id) content_path = self.write_to_temp( filename=filename, data=data) result = run_ctags(content_path, lang=ctags_lang) ctags.update({ 'ctags': list(result), 'indexer_configuration_id': self.tool['id'], }) self.cleanup(content_path) return ctags def persist_index_computations(self, results, policy_update): """Persist the results in storage. Args: results ([dict]): list of content_mimetype, dict with the following keys: - id (bytes): content's identifier (sha1) - ctags ([dict]): ctags list of symbols policy_update ([str]): either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ self.idx_storage.content_ctags_add( results, conflict_update=(policy_update == 'update-dups')) - - -@click.command() -@click.option('--path', help="Path to execute index on") -def main(path): - r = list(run_ctags(path)) - print(r) - - -if __name__ == '__main__': - main() diff --git a/swh/indexer/fossology_license.py b/swh/indexer/fossology_license.py index c27bf02..37522b9 100644 --- a/swh/indexer/fossology_license.py +++ b/swh/indexer/fossology_license.py @@ -1,183 +1,172 @@ # Copyright (C) 2016-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import click import subprocess from swh.model import hashutil from .indexer import ContentIndexer, ContentRangeIndexer, DiskIndexer class MixinFossologyLicenseIndexer: """Mixin fossology license indexer. See :class:`ContentFossologyLicenseIndexer` and :class:`FossologyLicenseRangeIndexer` """ ADDITIONAL_CONFIG = { 'workdir': ('str', '/tmp/swh/indexer.fossology.license'), 'tools': ('dict', { 'name': 'nomos', 'version': '3.1.0rc2-31-ga2cbb8c', 'configuration': { 'command_line': 'nomossa ', }, }), + 'write_batch_size': ('int', 1000), } CONFIG_BASE_FILENAME = 'indexer/fossology_license' def prepare(self): super().prepare() self.working_directory = self.config['workdir'] self.tool = self.tools[0] def compute_license(self, path, log=None): """Determine license from file at path. Args: path: filepath to determine the license Returns: A dict with the following keys: - licenses ([str]): associated detected licenses to path - path (bytes): content filepath - tool (str): tool used to compute the output """ try: properties = subprocess.check_output(['nomossa', path], universal_newlines=True) if properties: res = properties.rstrip().split(' contains license(s) ') licenses = res[1].split(',') return { 'licenses': licenses, 'path': path, } except subprocess.CalledProcessError: if log: from os import path as __path log.exception('Problem during license detection for sha1 %s' % __path.basename(path)) return { 'licenses': [], 'path': path, } def index(self, id, data): """Index sha1s' content and store result. Args: id (bytes): content's identifier raw_content (bytes): raw content in bytes Returns: A dict, representing a content_license, with keys: - id (bytes): content's identifier (sha1) - license (bytes): license in bytes - path (bytes): path """ if isinstance(id, str): id = hashutil.hash_to_hex(id) content_path = self.write_to_temp( filename=id, data=data) try: properties = self.compute_license(path=content_path, log=self.log) properties.update({ 'id': id, 'indexer_configuration_id': self.tool['id'], }) finally: self.cleanup(content_path) return properties def persist_index_computations(self, results, policy_update): """Persist the results in storage. Args: results ([dict]): list of content_license, dict with the following keys: - id (bytes): content's identifier (sha1) - license (bytes): license in bytes - path (bytes): path policy_update ([str]): either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ self.idx_storage.content_fossology_license_add( results, conflict_update=(policy_update == 'update-dups')) class ContentFossologyLicenseIndexer( MixinFossologyLicenseIndexer, DiskIndexer, ContentIndexer): """Indexer in charge of: - filtering out content already indexed - reading content from objstorage per the content's id (sha1) - computing {license, encoding} from that content - store result in storage """ def filter(self, ids): """Filter out known sha1s and return only missing ones. """ yield from self.idx_storage.content_fossology_license_missing(( { 'id': sha1, 'indexer_configuration_id': self.tool['id'], } for sha1 in ids )) class FossologyLicenseRangeIndexer( MixinFossologyLicenseIndexer, DiskIndexer, ContentRangeIndexer): """FossologyLicense Range Indexer working on range of content identifiers. It: - filters out the non textual content - (optionally) filters out content already indexed (cf :callable:`range`) - reads content from objstorage per the content's id (sha1) - computes {mimetype, encoding} from that content - stores result in storage """ def indexed_contents_in_range(self, start, end): """Retrieve indexed content id within range [start, end]. Args **start** (bytes): Starting bound from range identifier **end** (bytes): End range identifier Yields: Content identifier (bytes) present in the range [start, end] """ while start: result = self.idx_storage.content_fossology_license_get_range( start, end, self.tool['id']) contents = result['ids'] for _id in contents: yield _id start = result['next'] - - -@click.command(help='Compute license for path using tool') -@click.option('--tool', default='nomossa', help="Path to tool") -@click.option('--path', required=1, help="Path to execute index on") -def main(tool, path): - indexer = ContentFossologyLicenseIndexer() - print(indexer.compute_license(tool, path)) - - -if __name__ == '__main__': - main() diff --git a/swh/indexer/mimetype.py b/swh/indexer/mimetype.py index 5652599..7dd43af 100644 --- a/swh/indexer/mimetype.py +++ b/swh/indexer/mimetype.py @@ -1,177 +1,156 @@ # Copyright (C) 2016-2018 The Software Heritage developers # See the AUTHORS file at the top-level directory of this distribution # License: GNU General Public License version 3, or any later version # See top-level LICENSE file for more information import click import magic from swh.model import hashutil -from swh.scheduler import get_scheduler from .indexer import ContentIndexer, ContentRangeIndexer def compute_mimetype_encoding(raw_content): """Determine mimetype and encoding from the raw content. Args: raw_content (bytes): content's raw data Returns: A dict with mimetype and encoding key and corresponding values (as bytes). """ r = magic.detect_from_content(raw_content) return { 'mimetype': r.mime_type.encode('utf-8'), 'encoding': r.encoding.encode('utf-8'), } class MixinMimetypeIndexer: """Mixin mimetype indexer. See :class:`ContentMimetypeIndexer` and :class:`MimetypeRangeIndexer` """ ADDITIONAL_CONFIG = { - 'scheduler': ('dict', { - 'cls': 'remote', - 'args': { - 'url': 'http://localhost:5008', - }, - }), 'tools': ('dict', { 'name': 'file', 'version': '1:5.30-1+deb9u1', 'configuration': { "type": "library", "debian-package": "python3-magic" }, }), - 'write_batch_size': ('int', 100), + 'write_batch_size': ('int', 1000), } CONFIG_BASE_FILENAME = 'indexer/mimetype' def prepare(self): super().prepare() - self.scheduler = get_scheduler(**self.config['scheduler']) self.tool = self.tools[0] def index(self, id, data): """Index sha1s' content and store result. Args: id (bytes): content's identifier data (bytes): raw content in bytes Returns: A dict, representing a content_mimetype, with keys: - id (bytes): content's identifier (sha1) - mimetype (bytes): mimetype in bytes - encoding (bytes): encoding in bytes """ try: properties = compute_mimetype_encoding(data) properties.update({ 'id': id, 'indexer_configuration_id': self.tool['id'], }) except TypeError: self.log.error('Detecting mimetype error for id %s' % ( hashutil.hash_to_hex(id), )) return None return properties def persist_index_computations(self, results, policy_update): """Persist the results in storage. Args: results ([dict]): list of content_mimetype, dict with the following keys: - id (bytes): content's identifier (sha1) - mimetype (bytes): mimetype in bytes - encoding (bytes): encoding in bytes policy_update ([str]): either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ self.idx_storage.content_mimetype_add( results, conflict_update=(policy_update == 'update-dups')) class ContentMimetypeIndexer(MixinMimetypeIndexer, ContentIndexer): """Mimetype Indexer working on list of content identifiers. It: - (optionally) filters out content already indexed (cf. :callable:`filter`) - reads content from objstorage per the content's id (sha1) - computes {mimetype, encoding} from that content - stores result in storage FIXME: - 1. Rename redundant ContentMimetypeIndexer to MimetypeIndexer - 2. Do we keep it afterwards? ~> i think this can be used with the journal """ def filter(self, ids): """Filter out known sha1s and return only missing ones. """ yield from self.idx_storage.content_mimetype_missing(( { 'id': sha1, 'indexer_configuration_id': self.tool['id'], } for sha1 in ids )) class MimetypeRangeIndexer(MixinMimetypeIndexer, ContentRangeIndexer): """Mimetype Range Indexer working on range of content identifiers. It: - (optionally) filters out content already indexed (cf :callable:`range`) - reads content from objstorage per the content's id (sha1) - computes {mimetype, encoding} from that content - stores result in storage """ def indexed_contents_in_range(self, start, end): """Retrieve indexed content id within range [start, end]. Args **start** (bytes): Starting bound from range identifier **end** (bytes): End range identifier Yields: Content identifier (bytes) present in the range [start, end] """ while start: result = self.idx_storage.content_mimetype_get_range( start, end, self.tool['id']) contents = result['ids'] for _id in contents: yield _id start = result['next'] - - -@click.command() -@click.option('--path', help="Path to execute index on") -def main(path): - with open(path, 'rb') as f: - raw_content = f.read() - - print(compute_mimetype_encoding(raw_content)) - - -if __name__ == '__main__': - main() diff --git a/version.txt b/version.txt index a44fb5a..1e222fc 100644 --- a/version.txt +++ b/version.txt @@ -1 +1 @@ -v0.0.57-0-g54b311d \ No newline at end of file +v0.0.58-0-ga8546bc \ No newline at end of file