diff --git a/swh/indexer/ctags.py b/swh/indexer/ctags.py index b29e4c7..1e32bf3 100644 --- a/swh/indexer/ctags.py +++ b/swh/indexer/ctags.py @@ -1,151 +1,154 @@ -# Copyright (C) 2015-2017 The Software Heritage developers +# Copyright (C) 2015-2020 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 subprocess import json +from typing import Dict, List + from swh.model import hashutil from .indexer import ContentIndexer, write_to_temp # 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 compute_language(content, log=None): raise NotImplementedError( 'Language detection was unreliable, so it is currently disabled. ' 'See https://forge.softwareheritage.org/D1455') 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) Yields: dict: 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): 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'] 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 index(self, id, data): """Index sha1s' content and store result. Args: id (bytes): content's identifier data (bytes): raw content in bytes Returns: dict: 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) with write_to_temp( filename=filename, data=data, working_directory=self.working_directory) as content_path: result = run_ctags(content_path, lang=ctags_lang) ctags.update({ 'ctags': list(result), 'indexer_configuration_id': self.tool['id'], }) return ctags - def persist_index_computations(self, results, policy_update): + def persist_index_computations( + self, results: List[Dict], policy_update: str) -> Dict: """Persist the results in storage. Args: - results ([dict]): list of content_mimetype, dict with the + results: 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 + policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ - self.idx_storage.content_ctags_add( + return self.idx_storage.content_ctags_add( results, conflict_update=(policy_update == 'update-dups')) diff --git a/swh/indexer/fossology_license.py b/swh/indexer/fossology_license.py index ed9a431..7e00bc4 100644 --- a/swh/indexer/fossology_license.py +++ b/swh/indexer/fossology_license.py @@ -1,173 +1,181 @@ -# Copyright (C) 2016-2018 The Software Heritage developers +# Copyright (C) 2016-2020 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 logging import subprocess -from typing import Optional +from typing import Any, Dict, List, Optional from swh.model import hashutil from .indexer import ContentIndexer, ContentRangeIndexer, write_to_temp -def compute_license(path, log=None): +logger = logging.getLogger(__name__) + + +def compute_license(path): """Determine license from file at path. Args: path: filepath to determine the license Returns: dict: A dict with the following keys: - licenses ([str]): associated detected licenses to path - path (bytes): content filepath """ try: properties = subprocess.check_output(['nomossa', path], universal_newlines=True) if properties: res = properties.rstrip().split(' contains license(s) ') licenses = res[1].split(',') else: licenses = [] 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)) + from os import path as __path + logger.exception('Problem during license detection for sha1 %s' % + __path.basename(path)) return { 'licenses': [], 'path': path, } class MixinFossologyLicenseIndexer: """Mixin fossology license indexer. See :class:`FossologyLicenseIndexer` 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' # type: Optional[str] + tool: Any + idx_storage: Any def prepare(self): super().prepare() self.working_directory = self.config['workdir'] - def index(self, id, data): + def index(self, id: bytes, data: Optional[bytes] = None, + **kwargs) -> Dict[str, Any]: """Index sha1s' content and store result. Args: id (bytes): content's identifier raw_content (bytes): associated raw content to content id Returns: dict: A dict, representing a content_license, with keys: - id (bytes): content's identifier (sha1) - license (bytes): license in bytes - path (bytes): path - indexer_configuration_id (int): tool used to compute the output """ assert isinstance(id, bytes) + assert data is not None with write_to_temp( filename=hashutil.hash_to_hex(id), # use the id as pathname data=data, working_directory=self.working_directory) as content_path: - properties = compute_license(path=content_path, log=self.log) + properties = compute_license(path=content_path) properties.update({ 'id': id, 'indexer_configuration_id': self.tool['id'], }) return properties - def persist_index_computations(self, results, policy_update): + def persist_index_computations( + self, results: List[Dict], policy_update: str) -> Dict: """Persist the results in storage. Args: - results ([dict]): list of content_license, dict with the + results: 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 + policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ - self.idx_storage.content_fossology_license_add( + return self.idx_storage.content_fossology_license_add( results, conflict_update=(policy_update == 'update-dups')) class FossologyLicenseIndexer( MixinFossologyLicenseIndexer, 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, ContentRangeIndexer): """FossologyLicense Range Indexer working on range of content identifiers. - filters out the non textual content - (optionally) filters out content already indexed (cf :meth:`.indexed_contents_in_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 Returns: dict: a dict with keys: - **ids** [bytes]: iterable of content ids within the range. - **next** (Optional[bytes]): The next range of sha1 starts at this sha1 if any """ return self.idx_storage.content_fossology_license_get_range( start, end, self.tool['id']) diff --git a/swh/indexer/indexer.py b/swh/indexer/indexer.py index 75b65c8..fb5fa8a 100644 --- a/swh/indexer/indexer.py +++ b/swh/indexer/indexer.py @@ -1,596 +1,594 @@ # Copyright (C) 2016-2020 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 abc import os import logging import shutil import tempfile from contextlib import contextmanager -from typing import Any, Dict, Tuple, Generator, Union, List -from typing import Set +from typing import ( + Any, Dict, Iterator, List, Optional, Set, Tuple, Union +) from swh.scheduler import CONFIG as SWH_CONFIG from swh.storage import get_storage from swh.core.config import SWHConfig from swh.objstorage import get_objstorage from swh.objstorage.exc import ObjNotFoundError from swh.indexer.storage import get_indexer_storage, INDEXER_CFG_KEY from swh.model import hashutil from swh.core import utils @contextmanager def write_to_temp( - filename: str, data: bytes, working_directory: str -) -> Generator[str, None, None]: + filename: str, data: bytes, working_directory: str) -> Iterator[str]: """Write the sha1's content in a temporary file. Args: filename: one of sha1's many filenames data: the sha1's content to write in temporary file working_directory: the directory into which the file is written Returns: The path to the temporary file created. That file is filled in with the raw content's data. """ os.makedirs(working_directory, exist_ok=True) temp_dir = tempfile.mkdtemp(dir=working_directory) content_path = os.path.join(temp_dir, filename) with open(content_path, 'wb') as f: f.write(data) yield content_path shutil.rmtree(temp_dir) class BaseIndexer(SWHConfig, metaclass=abc.ABCMeta): """Base class for indexers to inherit from. The main entry point is the :func:`run` function which is in charge of triggering the computations on the batch dict/ids received. Indexers can: - filter out ids whose data has already been indexed. - retrieve ids data from storage or objstorage - index this data depending on the object and store the result in storage. To implement a new object type indexer, inherit from the BaseIndexer and implement indexing: :meth:`~BaseIndexer.run`: object_ids are different depending on object. For example: sha1 for content, sha1_git for revision, directory, release, and id for origin To implement a new concrete indexer, inherit from the object level classes: :class:`ContentIndexer`, :class:`RevisionIndexer`, :class:`OriginIndexer`. Then you need to implement the following functions: :meth:`~BaseIndexer.filter`: filter out data already indexed (in storage). :meth:`~BaseIndexer.index_object`: compute index on id with data (retrieved from the storage or the objstorage by the id key) and return the resulting index computation. :meth:`~BaseIndexer.persist_index_computations`: persist the results of multiple index computations in the storage. The new indexer implementation can also override the following functions: :meth:`~BaseIndexer.prepare`: Configuration preparation for the indexer. When overriding, this must call the `super().prepare()` instruction. :meth:`~BaseIndexer.check`: Configuration check for the indexer. When overriding, this must call the `super().check()` instruction. :meth:`~BaseIndexer.register_tools`: This should return a dict of the tool(s) to use when indexing or filtering. """ results: List[Dict] CONFIG = 'indexer/base' DEFAULT_CONFIG = { INDEXER_CFG_KEY: ('dict', { 'cls': 'remote', 'args': { 'url': 'http://localhost:5007/' } }), 'storage': ('dict', { 'cls': 'remote', 'args': { 'url': 'http://localhost:5002/', } }), 'objstorage': ('dict', { 'cls': 'remote', 'args': { 'url': 'http://localhost:5003/', } }) } ADDITIONAL_CONFIG = {} # type: Dict[str, Tuple[str, Any]] USE_TOOLS = True catch_exceptions = True """Prevents exceptions in `index()` from raising too high. Set to False in tests to properly catch all exceptions.""" scheduler: Any def __init__(self, config=None, **kw) -> None: """Prepare and check that the indexer is ready to run. """ super().__init__() if config is not None: self.config = config elif SWH_CONFIG: self.config = SWH_CONFIG.copy() else: config_keys = ('base_filename', 'config_filename', 'additional_configs', 'global_config') config_args = {k: v for k, v in kw.items() if k in config_keys} if self.ADDITIONAL_CONFIG: config_args.setdefault('additional_configs', []).append( self.ADDITIONAL_CONFIG) self.config = self.parse_config_file(**config_args) self.prepare() self.check() self.log.debug('%s: config=%s', self, self.config) def prepare(self) -> None: """Prepare the indexer's needed runtime configuration. Without this step, the indexer cannot possibly run. """ config_storage = self.config.get('storage') if config_storage: self.storage = get_storage(**config_storage) objstorage = self.config['objstorage'] self.objstorage = get_objstorage(objstorage['cls'], objstorage['args']) idx_storage = self.config[INDEXER_CFG_KEY] self.idx_storage = get_indexer_storage(**idx_storage) _log = logging.getLogger('requests.packages.urllib3.connectionpool') _log.setLevel(logging.WARN) self.log = logging.getLogger('swh.indexer') if self.USE_TOOLS: self.tools = list(self.register_tools( self.config.get('tools', []))) self.results = [] @property def tool(self) -> Dict: return self.tools[0] def check(self) -> None: """Check the indexer's configuration is ok before proceeding. If ok, does nothing. If not raise error. """ if self.USE_TOOLS and not self.tools: raise ValueError('Tools %s is unknown, cannot continue' % self.tools) def _prepare_tool(self, tool: Dict[str, Any]) -> Dict[str, Any]: """Prepare the tool dict to be compliant with the storage api. """ return {'tool_%s' % key: value for key, value in tool.items()} def register_tools( self, tools: Union[Dict[str, Any], List[Dict[str, Any]]] ) -> List[Dict[str, Any]]: """Permit to register tools to the storage. Add a sensible default which can be overridden if not sufficient. (For now, all indexers use only one tool) Expects the self.config['tools'] property to be set with one or more tools. Args: tools: Either a dict or a list of dict. Returns: list: List of dicts with additional id key. Raises: ValueError: if not a list nor a dict. """ if isinstance(tools, list): tools = list(map(self._prepare_tool, tools)) elif isinstance(tools, dict): tools = [self._prepare_tool(tools)] else: raise ValueError('Configuration tool(s) must be a dict or list!') if tools: return self.idx_storage.indexer_configuration_add(tools) else: return [] - def index( - self, id: bytes, data: bytes - ) -> Dict[str, Any]: + def index(self, id: bytes, data: Optional[bytes] = None, + **kwargs) -> Dict[str, Any]: """Index computation for the id and associated raw data. Args: id: identifier data: id's data from storage or objstorage depending on object type Returns: dict: a dict that makes sense for the :meth:`.persist_index_computations` method. """ raise NotImplementedError() - def filter(self, ids: List[bytes]) -> Generator[bytes, None, None]: + def filter(self, ids: List[bytes]) -> Iterator[bytes]: """Filter missing ids for that particular indexer. Args: ids: list of ids Yields: iterator of missing ids """ yield from ids @abc.abstractmethod - def persist_index_computations(self, results, policy_update): + def persist_index_computations(self, results, policy_update) -> Dict: """Persist the computation resulting from the index. Args: results ([result]): List of results. One result is the result of the index function. policy_update ([str]): either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them Returns: - None + a summary dict of what has been inserted in the storage """ - pass - - @abc.abstractmethod - def run(self, ids, policy_update, **kwargs): - """Given a list of ids: - - - retrieves the data from the storage - - executes the indexing computations - - stores the results (according to policy_update) - - Args: - ids ([bytes]): id's identifier list - policy_update (str): either 'update-dups' or 'ignore-dups' to - respectively update duplicates or ignore them - **kwargs: passed to the `index` method - - """ - pass + return {} class ContentIndexer(BaseIndexer): """A content indexer working on a list of ids directly. To work on indexer range, use the :class:`ContentRangeIndexer` instead. Note: :class:`ContentIndexer` is not an instantiable object. To use it, one should inherit from this class and override the methods mentioned in the :class:`BaseIndexer` class. """ - - def run(self, ids, policy_update, **kwargs): + def run(self, ids: Union[List[bytes], bytes, str], policy_update: str, + **kwargs) -> Dict: """Given a list of ids: - retrieve the content from the storage - execute the indexing computations - store the results (according to policy_update) Args: ids (Iterable[Union[bytes, str]]): sha1's identifier list policy_update (str): either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them **kwargs: passed to the `index` method + Returns: + A summary Dict of the task's status + """ - ids = [hashutil.hash_to_bytes(id_) if isinstance(id_, str) else id_ - for id_ in ids] + status = 'uneventful' + sha1s = [hashutil.hash_to_bytes(id_) if isinstance(id_, str) else id_ + for id_ in ids] results = [] + summary: Dict = {} try: - for sha1 in ids: + for sha1 in sha1s: try: raw_content = self.objstorage.get(sha1) except ObjNotFoundError: self.log.warning('Content %s not found in objstorage' % hashutil.hash_to_hex(sha1)) continue res = self.index(sha1, raw_content, **kwargs) if res: # If no results, skip it results.append(res) - - self.persist_index_computations(results, policy_update) + status = 'eventful' + summary = self.persist_index_computations(results, policy_update) self.results = results except Exception: if not self.catch_exceptions: raise self.log.exception( 'Problem when reading contents metadata.') + status = 'failed' + finally: + summary['status'] = status + return summary class ContentRangeIndexer(BaseIndexer): """A content range indexer. This expects as input a range of ids to index. To work on a list of ids, use the :class:`ContentIndexer` instead. Note: :class:`ContentRangeIndexer` is not an instantiable object. To use it, one should inherit from this class and override the methods mentioned in the :class:`BaseIndexer` class. """ @abc.abstractmethod def indexed_contents_in_range( self, start: bytes, end: bytes ) -> Any: """Retrieve indexed contents within range [start, end]. Args: start: Starting bound from range identifier end: End range identifier Yields: bytes: Content identifier present in the range ``[start, end]`` """ pass def _list_contents_to_index( self, start: bytes, end: bytes, indexed: Set[bytes] - ) -> Generator[bytes, None, None]: + ) -> Iterator[bytes]: """Compute from storage the new contents to index in the range [start, end]. The already indexed contents are skipped. Args: start: Starting bound from range identifier end: End range identifier indexed: Set of content already indexed. Yields: bytes: Identifier of contents to index. """ if not isinstance(start, bytes) or not isinstance(end, bytes): raise TypeError('identifiers must be bytes, not %r and %r.' % (start, end)) while start: result = self.storage.content_get_range(start, end) contents = result['contents'] for c in contents: _id = hashutil.hash_to_bytes(c['sha1']) if _id in indexed: continue yield _id start = result['next'] def _index_contents( self, start: bytes, end: bytes, indexed: Set[bytes], **kwargs: Any - ) -> Generator[Dict, None, None]: + ) -> Iterator[Dict]: """Index the contents from within range [start, end] Args: start: Starting bound from range identifier end: End range identifier indexed: Set of content already indexed. Yields: dict: Data indexed to persist using the indexer storage """ for sha1 in self._list_contents_to_index(start, end, indexed): try: raw_content = self.objstorage.get(sha1) except ObjNotFoundError: self.log.warning('Content %s not found in objstorage' % hashutil.hash_to_hex(sha1)) continue - res = self.index(sha1, raw_content, **kwargs) # type: ignore + res = self.index(sha1, raw_content, **kwargs) if res: if not isinstance(res['id'], bytes): raise TypeError( '%r.index should return ids as bytes, not %r' % (self.__class__.__name__, res['id'])) yield res def _index_with_skipping_already_done( - self, start: bytes, end: bytes - ) -> Generator[Dict, None, None]: + self, start: bytes, end: bytes) -> Iterator[Dict]: """Index not already indexed contents in range [start, end]. Args: start: Starting range identifier end: Ending range identifier Yields: dict: Content identifier present in the range ``[start, end]`` which are not already indexed. """ while start: indexed_page = self.indexed_contents_in_range(start, end) contents = indexed_page['ids'] _end = contents[-1] if contents else end yield from self._index_contents( start, _end, contents) start = indexed_page['next'] - def run(self, start, end, skip_existing=True, **kwargs): + def run(self, start: Union[bytes, str], end: Union[bytes, str], + skip_existing: bool = True, **kwargs) -> Dict: """Given a range of content ids, compute the indexing computations on the contents within. Either the indexer is incremental (filter out existing computed data) or not (compute everything from scratch). Args: - start (Union[bytes, str]): Starting range identifier - end (Union[bytes, str]): Ending range identifier - skip_existing (bool): Skip existing indexed data + start: Starting range identifier + end: Ending range identifier + skip_existing: Skip existing indexed data (default) or not **kwargs: passed to the `index` method Returns: - bool: True if data was indexed, False otherwise. + A dict with the task's status """ - with_indexed_data = False + status = 'uneventful' + summary: Dict = {} try: - if isinstance(start, str): - start = hashutil.hash_to_bytes(start) - if isinstance(end, str): - end = hashutil.hash_to_bytes(end) + range_start = hashutil.hash_to_bytes(start) \ + if isinstance(start, str) else start + range_end = hashutil.hash_to_bytes(end) \ + if isinstance(end, str) else end if skip_existing: - gen = self._index_with_skipping_already_done(start, end) + gen = self._index_with_skipping_already_done( + range_start, range_end) else: - gen = self._index_contents(start, end, indexed=[]) - - for results in utils.grouper(gen, - n=self.config['write_batch_size']): - self.persist_index_computations( - results, policy_update='update-dups') - with_indexed_data = True + gen = self._index_contents( + range_start, range_end, indexed=set([])) + + for contents in utils.grouper( + gen, n=self.config['write_batch_size']): + res = self.persist_index_computations( + contents, policy_update='update-dups') + summary['content_mimetype:add'] += res.get( + 'content_mimetype:add') + status = 'eventful' except Exception: if not self.catch_exceptions: raise self.log.exception( 'Problem when computing metadata.') + status = 'failed' finally: - return with_indexed_data + summary['status'] = status + return summary class OriginIndexer(BaseIndexer): """An object type indexer, inherits from the :class:`BaseIndexer` and implements Origin indexing using the run method Note: the :class:`OriginIndexer` is not an instantiable object. To use it in another context one should inherit from this class and override the methods mentioned in the :class:`BaseIndexer` class. """ - def run(self, origin_urls, policy_update='update-dups', - next_step=None, **kwargs): + def run(self, origin_urls: List[str], + policy_update: str = 'update-dups', **kwargs) -> Dict: """Given a list of origin urls: - retrieve origins from storage - execute the indexing computations - store the results (according to policy_update) Args: - origin_urls ([str]): list of origin urls. - policy_update (str): either 'update-dups' or 'ignore-dups' to + origin_urls: list of origin urls. + policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates (default) or ignore them - parse_ids (bool): Do we need to parse id or not (default) **kwargs: passed to the `index` method """ results = self.index_list(origin_urls, **kwargs) - - self.persist_index_computations(results, policy_update) + summary = self.persist_index_computations(results, policy_update) self.results = results + return summary def index_list(self, origins: List[Any], **kwargs: Any) -> List[Dict]: results = [] for origin in origins: try: res = self.index(origin, **kwargs) if res: # If no results, skip it results.append(res) except Exception: if not self.catch_exceptions: raise self.log.exception( 'Problem when processing origin %s', origin['id']) return results class RevisionIndexer(BaseIndexer): """An object type indexer, inherits from the :class:`BaseIndexer` and implements Revision indexing using the run method Note: the :class:`RevisionIndexer` is not an instantiable object. To use it in another context one should inherit from this class and override the methods mentioned in the :class:`BaseIndexer` class. """ - def run(self, ids, policy_update): + def run(self, ids: Union[str, bytes], policy_update: str) -> Dict: """Given a list of sha1_gits: - retrieve revisions from storage - execute the indexing computations - store the results (according to policy_update) Args: - ids ([bytes or str]): sha1_git's identifier list - policy_update (str): either 'update-dups' or 'ignore-dups' to + ids: sha1_git's identifier list + policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ results = [] - ids = [hashutil.hash_to_bytes(id_) if isinstance(id_, str) else id_ - for id_ in ids] - revs = self.storage.revision_get(ids) + revs = self.storage.revision_get( + hashutil.hash_to_bytes(id_) if isinstance(id_, str) else id_ + for id_ in ids) for rev in revs: if not rev: self.log.warning('Revisions %s not found in storage' % list(map(hashutil.hash_to_hex, ids))) continue try: res = self.index(rev) if res: # If no results, skip it results.append(res) except Exception: if not self.catch_exceptions: raise self.log.exception( 'Problem when processing revision') - self.persist_index_computations(results, policy_update) + summary = self.persist_index_computations(results, policy_update) self.results = results + return summary diff --git a/swh/indexer/metadata.py b/swh/indexer/metadata.py index ae9a990..50c31b6 100644 --- a/swh/indexer/metadata.py +++ b/swh/indexer/metadata.py @@ -1,368 +1,377 @@ # Copyright (C) 2017-2020 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 from copy import deepcopy -from typing import Any, List, Dict, Tuple, Callable, Generator +from typing import Any, Callable, Dict, Iterator, List, Tuple from swh.core.utils import grouper from swh.indexer.codemeta import merge_documents from swh.indexer.indexer import ContentIndexer, RevisionIndexer, OriginIndexer from swh.indexer.origin_head import OriginHeadIndexer from swh.indexer.metadata_dictionary import MAPPINGS from swh.indexer.metadata_detector import detect_metadata from swh.indexer.storage import INDEXER_CFG_KEY from swh.model import hashutil REVISION_GET_BATCH_SIZE = 10 ORIGIN_GET_BATCH_SIZE = 10 def call_with_batches( f: Callable[[List[Dict[str, Any]]], Dict['str', Any]], args: List[Dict[str, str]], batch_size: int -) -> Generator[str, None, None]: +) -> Iterator[str]: """Calls a function with batches of args, and concatenates the results. """ groups = grouper(args, batch_size) for group in groups: yield from f(list(group)) class ContentMetadataIndexer(ContentIndexer): """Content-level indexer This indexer is in charge of: - filtering out content already indexed in content_metadata - reading content from objstorage with the content's id sha1 - computing metadata by given context - using the metadata_dictionary as the 'swh-metadata-translator' tool - store result in content_metadata table """ def filter(self, ids): """Filter out known sha1s and return only missing ones. """ yield from self.idx_storage.content_metadata_missing(( { 'id': sha1, 'indexer_configuration_id': self.tool['id'], } for sha1 in ids )) def index(self, id, data, log_suffix='unknown revision'): """Index sha1s' content and store result. Args: id (bytes): content's identifier data (bytes): raw content in bytes Returns: dict: dictionary representing a content_metadata. If the translation wasn't successful the metadata keys will be returned as None """ result = { 'id': id, 'indexer_configuration_id': self.tool['id'], 'metadata': None } try: mapping_name = self.tool['tool_configuration']['context'] log_suffix += ', content_id=%s' % hashutil.hash_to_hex(id) result['metadata'] = \ MAPPINGS[mapping_name](log_suffix).translate(data) except Exception: self.log.exception( "Problem during metadata translation " "for content %s" % hashutil.hash_to_hex(id)) if result['metadata'] is None: return None return result def persist_index_computations( self, results: List[Dict], policy_update: str - ) -> None: + ) -> Dict: """Persist the results in storage. Args: results: list of content_metadata, dict with the following keys: - id (bytes): content's identifier (sha1) - metadata (jsonb): detected metadata policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ - self.idx_storage.content_metadata_add( + return self.idx_storage.content_metadata_add( results, conflict_update=(policy_update == 'update-dups')) class RevisionMetadataIndexer(RevisionIndexer): """Revision-level indexer This indexer is in charge of: - filtering revisions already indexed in revision_intrinsic_metadata table with defined computation tool - retrieve all entry_files in root directory - use metadata_detector for file_names containing metadata - compute metadata translation if necessary and possible (depends on tool) - send sha1s to content indexing if possible - store the results for revision """ ADDITIONAL_CONFIG = { 'tools': ('dict', { 'name': 'swh-metadata-detector', 'version': '0.0.2', 'configuration': { }, }), } def filter(self, sha1_gits): """Filter out known sha1s and return only missing ones. """ yield from self.idx_storage.revision_intrinsic_metadata_missing(( { 'id': sha1_git, 'indexer_configuration_id': self.tool['id'], } for sha1_git in sha1_gits )) def index(self, rev): """Index rev by processing it and organizing result. use metadata_detector to iterate on filenames - if one filename detected -> sends file to content indexer - if multiple file detected -> translation needed at revision level Args: rev (dict): revision artifact from storage Returns: dict: dictionary representing a revision_intrinsic_metadata, with keys: - id (str): rev's identifier (sha1_git) - indexer_configuration_id (bytes): tool used - metadata: dict of retrieved metadata """ result = { 'id': rev['id'], 'indexer_configuration_id': self.tool['id'], 'mappings': None, 'metadata': None } try: root_dir = rev['directory'] dir_ls = list(self.storage.directory_ls(root_dir, recursive=False)) if [entry['type'] for entry in dir_ls] == ['dir']: # If the root is just a single directory, recurse into it # eg. PyPI packages, GNU tarballs subdir = dir_ls[0]['target'] dir_ls = self.storage.directory_ls(subdir, recursive=False) files = [entry for entry in dir_ls if entry['type'] == 'file'] detected_files = detect_metadata(files) (mappings, metadata) = self.translate_revision_intrinsic_metadata( detected_files, log_suffix='revision=%s' % hashutil.hash_to_hex(rev['id'])) result['mappings'] = mappings result['metadata'] = metadata except Exception as e: self.log.exception( 'Problem when indexing rev: %r', e) return result def persist_index_computations( self, results: List[Dict], policy_update: str - ) -> None: + ) -> Dict: """Persist the results in storage. Args: results: 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: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ # TODO: add functions in storage to keep data in # revision_intrinsic_metadata - self.idx_storage.revision_intrinsic_metadata_add( + return self.idx_storage.revision_intrinsic_metadata_add( results, conflict_update=(policy_update == 'update-dups')) def translate_revision_intrinsic_metadata( self, detected_files: Dict[str, List[Any]], log_suffix: str ) -> Tuple[List[Any], List[Any]]: """ Determine plan of action to translate metadata when containing one or multiple detected files: Args: detected_files: dictionary mapping context names (e.g., "npm", "authors") to list of sha1 Returns: (List[str], dict): list of mappings used and dict with translated metadata according to the CodeMeta vocabulary """ used_mappings = [MAPPINGS[context].name for context in detected_files] metadata = [] tool = { 'name': 'swh-metadata-translator', 'version': '0.0.2', 'configuration': { }, } # TODO: iterate on each context, on each file # -> get raw_contents # -> translate each content config = { k: self.config[k] for k in [INDEXER_CFG_KEY, 'objstorage', 'storage'] } config['tools'] = [tool] for context in detected_files.keys(): cfg = deepcopy(config) cfg['tools'][0]['configuration']['context'] = context c_metadata_indexer = ContentMetadataIndexer(config=cfg) # sha1s that are in content_metadata table sha1s_in_storage = [] metadata_generator = self.idx_storage.content_metadata_get( detected_files[context]) for c in metadata_generator: # extracting metadata sha1 = c['id'] sha1s_in_storage.append(sha1) local_metadata = c['metadata'] # local metadata is aggregated if local_metadata: metadata.append(local_metadata) sha1s_filtered = [item for item in detected_files[context] if item not in sha1s_in_storage] if sha1s_filtered: # content indexing try: c_metadata_indexer.run(sha1s_filtered, policy_update='ignore-dups', log_suffix=log_suffix) # on the fly possibility: for result in c_metadata_indexer.results: local_metadata = result['metadata'] metadata.append(local_metadata) except Exception: self.log.exception( "Exception while indexing metadata on contents") metadata = merge_documents(metadata) return (used_mappings, metadata) class OriginMetadataIndexer(OriginIndexer): ADDITIONAL_CONFIG = RevisionMetadataIndexer.ADDITIONAL_CONFIG USE_TOOLS = False def __init__(self, config=None, **kwargs) -> None: super().__init__(config=config, **kwargs) self.origin_head_indexer = OriginHeadIndexer(config=config) self.revision_metadata_indexer = RevisionMetadataIndexer(config=config) def index_list(self, origin_urls): head_rev_ids = [] origins_with_head = [] origins = list(call_with_batches( self.storage.origin_get, [{'url': url} for url in origin_urls], ORIGIN_GET_BATCH_SIZE)) for origin in origins: if origin is None: continue head_result = self.origin_head_indexer.index(origin['url']) if head_result: origins_with_head.append(origin) head_rev_ids.append(head_result['revision_id']) head_revs = list(call_with_batches( self.storage.revision_get, head_rev_ids, REVISION_GET_BATCH_SIZE)) assert len(head_revs) == len(head_rev_ids) results = [] for (origin, rev) in zip(origins_with_head, head_revs): if not rev: self.log.warning('Missing head revision of origin %r', origin['url']) continue rev_metadata = self.revision_metadata_indexer.index(rev) orig_metadata = { 'from_revision': rev_metadata['id'], 'id': origin['url'], 'metadata': rev_metadata['metadata'], 'mappings': rev_metadata['mappings'], 'indexer_configuration_id': rev_metadata['indexer_configuration_id'], } results.append((orig_metadata, rev_metadata)) return results def persist_index_computations( self, results: List[Dict], policy_update: str - ) -> None: + ) -> Dict: conflict_update = (policy_update == 'update-dups') # Deduplicate revisions rev_metadata: List[Any] = [] orig_metadata: List[Any] = [] revs_to_delete: List[Any] = [] origs_to_delete: List[Any] = [] + summary: Dict = {} for (orig_item, rev_item) in results: assert rev_item['metadata'] == orig_item['metadata'] if not rev_item['metadata'] or \ rev_item['metadata'].keys() <= {'@context'}: # If we didn't find any metadata, don't store a DB record # (and delete existing ones, if any) if rev_item not in revs_to_delete: revs_to_delete.append(rev_item) if orig_item not in origs_to_delete: origs_to_delete.append(orig_item) else: if rev_item not in rev_metadata: rev_metadata.append(rev_item) if orig_item not in orig_metadata: orig_metadata.append(orig_item) if rev_metadata: - self.idx_storage.revision_intrinsic_metadata_add( + summary_rev = self.idx_storage.revision_intrinsic_metadata_add( rev_metadata, conflict_update=conflict_update) + summary.update(summary_rev) if orig_metadata: - self.idx_storage.origin_intrinsic_metadata_add( + summary_ori = self.idx_storage.origin_intrinsic_metadata_add( orig_metadata, conflict_update=conflict_update) + summary.update(summary_ori) # revs_to_delete should always be empty unless we changed a mapping # to detect less files or less content. # However, origs_to_delete may be empty whenever an upstream deletes # a metadata file. if origs_to_delete: - self.idx_storage.origin_intrinsic_metadata_delete(origs_to_delete) + summary_ori = self.idx_storage.origin_intrinsic_metadata_delete( + origs_to_delete) + summary.update(summary_ori) if revs_to_delete: - self.idx_storage.revision_intrinsic_metadata_delete(revs_to_delete) + summary_rev = self.idx_storage.revision_intrinsic_metadata_delete( + revs_to_delete) + summary.update(summary_ori) + + return summary diff --git a/swh/indexer/mimetype.py b/swh/indexer/mimetype.py index 0d391bb..fd8b86e 100644 --- a/swh/indexer/mimetype.py +++ b/swh/indexer/mimetype.py @@ -1,152 +1,154 @@ # Copyright (C) 2016-2020 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 from typing import Optional, Dict, Any, List import magic from .indexer import ContentIndexer, ContentRangeIndexer if not hasattr(magic.Magic, 'from_buffer'): raise ImportError( 'Expected "import magic" to import python-magic, but file_magic ' 'was imported instead.') def compute_mimetype_encoding(raw_content: bytes) -> Dict[str, bytes]: """Determine mimetype and encoding from the raw content. Args: raw_content: content's raw data Returns: dict: mimetype and encoding key and corresponding values. """ m = magic.Magic(mime=True, mime_encoding=True) res = m.from_buffer(raw_content) (mimetype, encoding) = res.split('; charset=') return { 'mimetype': mimetype, 'encoding': encoding, } class MixinMimetypeIndexer: """Mixin mimetype indexer. See :class:`MimetypeIndexer` and :class:`MimetypeRangeIndexer` """ tool: Dict[str, Any] idx_storage: Any ADDITIONAL_CONFIG = { 'tools': ('dict', { 'name': 'file', 'version': '1:5.30-1+deb9u1', 'configuration': { "type": "library", "debian-package": "python3-magic" }, }), 'write_batch_size': ('int', 1000), } CONFIG_BASE_FILENAME = 'indexer/mimetype' # type: Optional[str] - def index(self, id: bytes, data: bytes) -> Dict[str, Any]: + def index(self, id: bytes, data: Optional[bytes] = None, + **kwargs) -> Dict[str, Any]: """Index sha1s' content and store result. Args: id: content's identifier data: raw content in bytes Returns: dict: content's mimetype; dict keys being - id: content's identifier (sha1) - mimetype: mimetype in bytes - encoding: encoding in bytes """ + assert data is not None properties = compute_mimetype_encoding(data) properties.update({ 'id': id, 'indexer_configuration_id': self.tool['id'], }) return properties def persist_index_computations( - self, results: List[Dict], policy_update: List[str] - ) -> None: + self, results: List[Dict], policy_update: str + ) -> Dict: """Persist the results in storage. Args: results: list of content's mimetype dicts (see :meth:`.index`) policy_update: either 'update-dups' or 'ignore-dups' to respectively update duplicates or ignore them """ - self.idx_storage.content_mimetype_add( + return self.idx_storage.content_mimetype_add( results, conflict_update=(policy_update == 'update-dups')) class MimetypeIndexer(MixinMimetypeIndexer, ContentIndexer): """Mimetype Indexer working on list of content identifiers. It: - (optionally) filters out content already indexed (cf. :meth:`.filter`) - reads content from objstorage per the content's id (sha1) - computes {mimetype, encoding} from that content - stores result in storage """ 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 :meth:`.indexed_contents_in_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: bytes, end: bytes ) -> Dict[str, Optional[bytes]]: """Retrieve indexed content id within range [start, end]. Args: start: Starting bound from range identifier end: End range identifier Returns: dict: a dict with keys: - ids: iterable of content ids within the range. - next: The next range of sha1 starts at this sha1 if any """ return self.idx_storage.content_mimetype_get_range( start, end, self.tool['id']) diff --git a/swh/indexer/origin_head.py b/swh/indexer/origin_head.py index 707d643..abfbb02 100644 --- a/swh/indexer/origin_head.py +++ b/swh/indexer/origin_head.py @@ -1,158 +1,158 @@ # Copyright (C) 2018-2020 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 from typing import List, Tuple, Any, Dict, Union import re import click import logging from swh.indexer.indexer import OriginIndexer class OriginHeadIndexer(OriginIndexer): """Origin-level indexer. This indexer is in charge of looking up the revision that acts as the "head" of an origin. In git, this is usually the commit pointed to by the 'master' branch.""" USE_TOOLS = False def persist_index_computations( self, results: Any, policy_update: str - ) -> None: + ) -> Dict: """Do nothing. The indexer's results are not persistent, they should only be piped to another indexer.""" - pass + return {} # Dispatch def index(self, origin_url): latest_visit = self.storage.origin_visit_get_latest( origin_url, allowed_statuses=['full'], require_snapshot=True) if latest_visit is None: return None latest_snapshot = self.storage.snapshot_get(latest_visit['snapshot']) method = getattr( self, '_try_get_%s_head' % latest_visit['type'], self._try_get_head_generic) rev_id = method(latest_snapshot) if rev_id is not None: return { 'origin_url': origin_url, 'revision_id': rev_id, } # could not find a head revision return None # Tarballs _archive_filename_re = re.compile( rb'^' rb'(?P.*)[-_]' rb'(?P[0-9]+(\.[0-9])*)' rb'(?P[-+][a-zA-Z0-9.~]+?)?' rb'(?P(\.[a-zA-Z0-9]+)+)' rb'$') @classmethod def _parse_version( cls: Any, filename: str ) -> Tuple[Union[float, int], ...]: """Extracts the release version from an archive filename, to get an ordering whose maximum is likely to be the last version of the software >>> OriginHeadIndexer._parse_version(b'foo') (-inf,) >>> OriginHeadIndexer._parse_version(b'foo.tar.gz') (-inf,) >>> OriginHeadIndexer._parse_version(b'gnu-hello-0.0.1.tar.gz') (0, 0, 1, 0) >>> OriginHeadIndexer._parse_version(b'gnu-hello-0.0.1-beta2.tar.gz') (0, 0, 1, -1, 'beta2') >>> OriginHeadIndexer._parse_version(b'gnu-hello-0.0.1+foobar.tar.gz') (0, 0, 1, 1, 'foobar') """ res = cls._archive_filename_re.match(filename) if res is None: return (float('-infinity'),) version = [int(n) for n in res.group('version').decode().split('.')] if res.group('preversion') is None: version.append(0) else: preversion = res.group('preversion').decode() if preversion.startswith('-'): version.append(-1) version.append(preversion[1:]) elif preversion.startswith('+'): version.append(1) version.append(preversion[1:]) else: assert False, res.group('preversion') return tuple(version) def _try_get_ftp_head(self, snapshot: Dict[str, Any]) -> Any: archive_names = list(snapshot['branches']) max_archive_name = max(archive_names, key=self._parse_version) r = self._try_resolve_target(snapshot['branches'], max_archive_name) return r # Generic def _try_get_head_generic( self, snapshot: Dict[str, Any] ) -> Any: # Works on 'deposit', 'pypi', and VCSs. try: branches = snapshot['branches'] except KeyError: return None else: return ( self._try_resolve_target(branches, b'HEAD') or self._try_resolve_target(branches, b'master') ) def _try_resolve_target(self, branches: Dict, target_name: bytes) -> Any: try: target = branches[target_name] if target is None: return None while target['target_type'] == 'alias': target = branches[target['target']] if target is None: return None if target['target_type'] == 'revision': return target['target'] elif target['target_type'] == 'content': return None # TODO elif target['target_type'] == 'directory': return None # TODO elif target['target_type'] == 'release': return None # TODO else: assert False except KeyError: return None @click.command() @click.option('--origins', '-i', help='Origins to lookup, in the "type+url" format', multiple=True) def main(origins: List[str]) -> None: rev_metadata_indexer = OriginHeadIndexer() rev_metadata_indexer.run(origins) if __name__ == '__main__': logging.basicConfig(level=logging.INFO) main() diff --git a/swh/indexer/rehash.py b/swh/indexer/rehash.py index 038b63e..bec65ec 100644 --- a/swh/indexer/rehash.py +++ b/swh/indexer/rehash.py @@ -1,177 +1,190 @@ # Copyright (C) 2017-2020 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 logging import itertools from collections import defaultdict from typing import Dict, Any, Tuple, List, Generator from swh.core import utils from swh.core.config import SWHConfig from swh.model import hashutil from swh.objstorage import get_objstorage from swh.objstorage.exc import ObjNotFoundError from swh.storage import get_storage class RecomputeChecksums(SWHConfig): """Class in charge of (re)computing content's hashes. Hashes to compute are defined across 2 configuration options: compute_checksums ([str]) list of hash algorithms that py:func:`swh.model.hashutil.MultiHash.from_data` function should be able to deal with. For variable-length checksums, a desired checksum length should also be provided. Their format is : e.g: blake2:512 recompute_checksums (bool) a boolean to notify that we also want to recompute potential existing hashes specified in compute_checksums. Default to False. """ DEFAULT_CONFIG = { # The storage to read from or update metadata to 'storage': ('dict', { 'cls': 'remote', 'args': { 'url': 'http://localhost:5002/' }, }), # The objstorage to read contents' data from 'objstorage': ('dict', { 'cls': 'pathslicing', 'args': { 'root': '/srv/softwareheritage/objects', 'slicing': '0:2/2:4/4:6', }, }), # the set of checksums that should be computed. # Examples: 'sha1_git', 'blake2b512', 'blake2s256' 'compute_checksums': ( 'list[str]', []), # whether checksums that already exist in the DB should be # recomputed/updated or left untouched 'recompute_checksums': ('bool', False), # Number of contents to retrieve blobs at the same time 'batch_size_retrieve_content': ('int', 10), # Number of contents to update at the same time 'batch_size_update': ('int', 100), } CONFIG_BASE_FILENAME = 'indexer/rehash' def __init__(self) -> None: self.config = self.parse_config_file() self.storage = get_storage(**self.config['storage']) self.objstorage = get_objstorage(**self.config['objstorage']) self.compute_checksums = self.config['compute_checksums'] self.recompute_checksums = self.config[ 'recompute_checksums'] self.batch_size_retrieve_content = self.config[ 'batch_size_retrieve_content'] self.batch_size_update = self.config[ 'batch_size_update'] self.log = logging.getLogger('swh.indexer.rehash') if not self.compute_checksums: raise ValueError('Checksums list should not be empty.') def _read_content_ids( self, contents: List[Dict[str, Any]] ) -> Generator[bytes, Any, None]: """Read the content identifiers from the contents. """ for c in contents: h = c['sha1'] if isinstance(h, str): h = hashutil.hash_to_bytes(h) yield h def get_new_contents_metadata( self, all_contents: List[Dict[str, Any]] ) -> Generator[Tuple[Dict[str, Any], List[Any]], Any, None]: """Retrieve raw contents and compute new checksums on the contents. Unknown or corrupted contents are skipped. Args: all_contents: List of contents as dictionary with the necessary primary keys Yields: tuple: tuple of (content to update, list of checksums computed) """ content_ids = self._read_content_ids(all_contents) for contents in utils.grouper(content_ids, self.batch_size_retrieve_content): contents_iter = itertools.tee(contents, 2) try: content_metadata = self.storage.content_get_metadata( [s for s in contents_iter[0]]) except Exception: self.log.exception( 'Problem when reading contents metadata.') continue for content in content_metadata: # Recompute checksums provided in compute_checksums options if self.recompute_checksums: checksums_to_compute = list(self.compute_checksums) else: # Compute checksums provided in compute_checksums # options not already defined for that content checksums_to_compute = [h for h in self.compute_checksums if not content.get(h)] if not checksums_to_compute: # Nothing to recompute continue try: raw_content = self.objstorage.get(content['sha1']) except ObjNotFoundError: self.log.warning('Content %s not found in objstorage!' % content['sha1']) continue content_hashes = hashutil.MultiHash.from_data( raw_content, hash_names=checksums_to_compute).digest() content.update(content_hashes) yield content, checksums_to_compute - def run(self, contents: List[Dict[str, Any]]) -> None: + def run(self, contents: List[Dict[str, Any]]) -> Dict: """Given a list of content: - (re)compute a given set of checksums on contents available in our object storage - update those contents with the new metadata - Args: - contents: contents as dictionary with necessary keys. - key present in such dictionary should be the ones defined in - the 'primary_key' option. + Args: + contents: contents as dictionary with necessary keys. + key present in such dictionary should be the ones defined in + the 'primary_key' option. + + Returns: + A summary dict with key 'status', task' status and 'count' the + number of updated contents. """ + status = 'uneventful' + count = 0 for data in utils.grouper( self.get_new_contents_metadata(contents), self.batch_size_update): groups: Dict[str, List[Any]] = defaultdict(list) for content, keys_to_update in data: keys = ','.join(keys_to_update) groups[keys].append(content) for keys_to_update, contents in groups.items(): keys = keys_to_update.split(',') try: self.storage.content_update(contents, keys=keys) + count += len(contents) + status = 'eventful' except Exception: self.log.exception('Problem during update.') continue + + return { + 'status': status, + 'count': count, + } diff --git a/swh/indexer/tasks.py b/swh/indexer/tasks.py index dc47146..0e4331f 100644 --- a/swh/indexer/tasks.py +++ b/swh/indexer/tasks.py @@ -1,57 +1,50 @@ -# Copyright (C) 2016-2019 The Software Heritage developers +# Copyright (C) 2016-2020 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 from celery import current_app as app from .mimetype import MimetypeIndexer, MimetypeRangeIndexer from .ctags import CtagsIndexer from .fossology_license import ( FossologyLicenseIndexer, FossologyLicenseRangeIndexer ) from .rehash import RecomputeChecksums from .metadata import OriginMetadataIndexer @app.task(name=__name__ + '.OriginMetadata') def origin_metadata(*args, **kwargs): - results = OriginMetadataIndexer().run(*args, **kwargs) - return getattr(results, 'results', results) + return OriginMetadataIndexer().run(*args, **kwargs) @app.task(name=__name__ + '.Ctags') def ctags(*args, **kwargs): - results = CtagsIndexer().run(*args, **kwargs) - return getattr(results, 'results', results) + return CtagsIndexer().run(*args, **kwargs) @app.task(name=__name__ + '.ContentFossologyLicense') def fossology_license(*args, **kwargs): - results = FossologyLicenseIndexer().run(*args, **kwargs) - return getattr(results, 'results', results) + return FossologyLicenseIndexer().run(*args, **kwargs) @app.task(name=__name__ + '.RecomputeChecksums') def recompute_checksums(*args, **kwargs): - results = RecomputeChecksums().run(*args, **kwargs) - return getattr(results, 'results', results) + return RecomputeChecksums().run(*args, **kwargs) @app.task(name=__name__ + '.ContentMimetype') def mimetype(*args, **kwargs): - results = MimetypeIndexer().run(*args, **kwargs) - return {'status': 'eventful' if results else 'uneventful'} + return MimetypeIndexer().run(*args, **kwargs) @app.task(name=__name__ + '.ContentRangeMimetype') def range_mimetype(*args, **kwargs): - results = MimetypeRangeIndexer().run(*args, **kwargs) - return {'status': 'eventful' if results else 'uneventful'} + return MimetypeRangeIndexer().run(*args, **kwargs) @app.task(name=__name__ + '.ContentRangeFossologyLicense') def range_license(*args, **kwargs): - results = FossologyLicenseRangeIndexer().run(*args, **kwargs) - return {'status': 'eventful' if results else 'uneventful'} + return FossologyLicenseRangeIndexer().run(*args, **kwargs) diff --git a/swh/indexer/tests/test_fossology_license.py b/swh/indexer/tests/test_fossology_license.py index 8b15b90..90bfa40 100644 --- a/swh/indexer/tests/test_fossology_license.py +++ b/swh/indexer/tests/test_fossology_license.py @@ -1,181 +1,181 @@ # Copyright (C) 2017-2018 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 unittest import pytest from unittest.mock import patch from typing import Any, Dict from swh.indexer import fossology_license from swh.indexer.fossology_license import ( FossologyLicenseIndexer, FossologyLicenseRangeIndexer, compute_license ) from swh.indexer.tests.utils import ( SHA1_TO_LICENSES, CommonContentIndexerTest, CommonContentIndexerRangeTest, BASE_TEST_CONFIG, fill_storage, fill_obj_storage, filter_dict, ) class BasicTest(unittest.TestCase): @patch("swh.indexer.fossology_license.subprocess") def test_compute_license(self, mock_subprocess): """Computing licenses from a raw content should return results """ for path, intermediary_result, output in [ (b'some/path', None, []), (b'some/path/2', [], []), (b'other/path', ' contains license(s) GPL,AGPL', ['GPL', 'AGPL'])]: mock_subprocess.check_output.return_value = intermediary_result - actual_result = compute_license(path, log=None) + actual_result = compute_license(path) self.assertEqual(actual_result, { 'licenses': output, 'path': path, }) -def mock_compute_license(path, log=None): +def mock_compute_license(path): """path is the content identifier """ if isinstance(id, bytes): path = path.decode('utf-8') # path is something like /tmp/tmpXXX/ so we keep only the sha1 part path = path.split('/')[-1] return { 'licenses': SHA1_TO_LICENSES.get(path) } CONFIG = { **BASE_TEST_CONFIG, 'workdir': '/tmp', 'tools': { 'name': 'nomos', 'version': '3.1.0rc2-31-ga2cbb8c', 'configuration': { 'command_line': 'nomossa ', }, }, } # type: Dict[str, Any] RANGE_CONFIG = dict(list(CONFIG.items()) + [('write_batch_size', 100)]) class TestFossologyLicenseIndexer(CommonContentIndexerTest, unittest.TestCase): """Language indexer test scenarios: - Known sha1s in the input list have their data indexed - Unknown sha1 in the input list are not indexed """ def get_indexer_results(self, ids): yield from self.idx_storage.content_fossology_license_get(ids) def setUp(self): super().setUp() # replace actual license computation with a mock self.orig_compute_license = fossology_license.compute_license fossology_license.compute_license = mock_compute_license self.indexer = FossologyLicenseIndexer(CONFIG) self.indexer.catch_exceptions = False self.idx_storage = self.indexer.idx_storage fill_storage(self.indexer.storage) fill_obj_storage(self.indexer.objstorage) self.id0 = '01c9379dfc33803963d07c1ccc748d3fe4c96bb5' self.id1 = '688a5ef812c53907562fe379d4b3851e69c7cb15' self.id2 = 'da39a3ee5e6b4b0d3255bfef95601890afd80709' # empty content tool = {k.replace('tool_', ''): v for (k, v) in self.indexer.tool.items()} # then self.expected_results = { self.id0: { 'tool': tool, 'licenses': SHA1_TO_LICENSES[self.id0], }, self.id1: { 'tool': tool, 'licenses': SHA1_TO_LICENSES[self.id1], }, self.id2: { 'tool': tool, 'licenses': SHA1_TO_LICENSES[self.id2], } } def tearDown(self): super().tearDown() fossology_license.compute_license = self.orig_compute_license class TestFossologyLicenseRangeIndexer( CommonContentIndexerRangeTest, unittest.TestCase): """Range Fossology License Indexer tests. - new data within range are indexed - no data outside a range are indexed - with filtering existing indexed data prior to compute new index - without filtering existing indexed data prior to compute new index """ def setUp(self): super().setUp() # replace actual license computation with a mock self.orig_compute_license = fossology_license.compute_license fossology_license.compute_license = mock_compute_license self.indexer = FossologyLicenseRangeIndexer(config=RANGE_CONFIG) self.indexer.catch_exceptions = False fill_storage(self.indexer.storage) fill_obj_storage(self.indexer.objstorage) self.id0 = '01c9379dfc33803963d07c1ccc748d3fe4c96bb5' self.id1 = '02fb2c89e14f7fab46701478c83779c7beb7b069' self.id2 = '103bc087db1d26afc3a0283f38663d081e9b01e6' tool_id = self.indexer.tool['id'] self.expected_results = { self.id0: { 'id': self.id0, 'indexer_configuration_id': tool_id, 'licenses': SHA1_TO_LICENSES[self.id0] }, self.id1: { 'id': self.id1, 'indexer_configuration_id': tool_id, 'licenses': SHA1_TO_LICENSES[self.id1] }, self.id2: { 'id': self.id2, 'indexer_configuration_id': tool_id, 'licenses': SHA1_TO_LICENSES[self.id2] } } def tearDown(self): super().tearDown() fossology_license.compute_license = self.orig_compute_license def test_fossology_w_no_tool(): with pytest.raises(ValueError): FossologyLicenseIndexer(config=filter_dict(CONFIG, 'tools')) def test_fossology_range_w_no_tool(): with pytest.raises(ValueError): FossologyLicenseRangeIndexer(config=filter_dict(RANGE_CONFIG, 'tools')) diff --git a/swh/indexer/tests/utils.py b/swh/indexer/tests/utils.py index fee5c52..084460d 100644 --- a/swh/indexer/tests/utils.py +++ b/swh/indexer/tests/utils.py @@ -1,766 +1,765 @@ # Copyright (C) 2017-2019 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 abc import datetime import functools import random from typing import Dict, Any import unittest from hypothesis import strategies from swh.model import hashutil from swh.model.hashutil import hash_to_bytes, hash_to_hex from swh.indexer.storage import INDEXER_CFG_KEY BASE_TEST_CONFIG: Dict[str, Dict[str, Any]] = { 'storage': { 'cls': 'pipeline', 'steps': [ {'cls': 'validate'}, {'cls': 'memory'}, ] }, 'objstorage': { 'cls': 'memory', 'args': { }, }, INDEXER_CFG_KEY: { 'cls': 'memory', 'args': { }, }, } ORIGIN_VISITS = [ { 'type': 'git', 'url': 'https://github.com/SoftwareHeritage/swh-storage'}, { 'type': 'ftp', 'url': 'rsync://ftp.gnu.org/gnu/3dldf'}, { 'type': 'deposit', 'url': 'https://forge.softwareheritage.org/source/jesuisgpl/'}, { 'type': 'pypi', 'url': 'https://pypi.org/project/limnoria/'}, { 'type': 'svn', 'url': 'http://0-512-md.googlecode.com/svn/'}, { 'type': 'git', 'url': 'https://github.com/librariesio/yarn-parser'}, { 'type': 'git', 'url': 'https://github.com/librariesio/yarn-parser.git'}, ] SNAPSHOTS = [ { 'origin': 'https://github.com/SoftwareHeritage/swh-storage', 'branches': { b'refs/heads/add-revision-origin-cache': { 'target': b'L[\xce\x1c\x88\x8eF\t\xf1"\x19\x1e\xfb\xc0' b's\xe7/\xe9l\x1e', 'target_type': 'revision'}, b'refs/head/master': { 'target': b'8K\x12\x00d\x03\xcc\xe4]bS\xe3\x8f{\xd7}' b'\xac\xefrm', 'target_type': 'revision'}, b'HEAD': { 'target': b'refs/head/master', 'target_type': 'alias'}, b'refs/tags/v0.0.103': { 'target': b'\xb6"Im{\xfdLb\xb0\x94N\xea\x96m\x13x\x88+' b'\x0f\xdd', 'target_type': 'release'}, }}, { 'origin': 'rsync://ftp.gnu.org/gnu/3dldf', 'branches': { b'3DLDF-1.1.4.tar.gz': { 'target': b'dJ\xfb\x1c\x91\xf4\x82B%]6\xa2\x90|\xd3\xfc' b'"G\x99\x11', 'target_type': 'revision'}, b'3DLDF-2.0.2.tar.gz': { 'target': b'\xb6\x0e\xe7\x9e9\xac\xaa\x19\x9e=' b'\xd1\xc5\x00\\\xc6\xfc\xe0\xa6\xb4V', 'target_type': 'revision'}, b'3DLDF-2.0.3-examples.tar.gz': { 'target': b'!H\x19\xc0\xee\x82-\x12F1\xbd\x97' b'\xfe\xadZ\x80\x80\xc1\x83\xff', 'target_type': 'revision'}, b'3DLDF-2.0.3.tar.gz': { 'target': b'\x8e\xa9\x8e/\xea}\x9feF\xf4\x9f\xfd\xee' b'\xcc\x1a\xb4`\x8c\x8by', 'target_type': 'revision'}, b'3DLDF-2.0.tar.gz': { 'target': b'F6*\xff(?\x19a\xef\xb6\xc2\x1fv$S\xe3G' b'\xd3\xd1m', 'target_type': 'revision'} }}, { 'origin': 'https://forge.softwareheritage.org/source/jesuisgpl/', 'branches': { b'master': { 'target': b'\xe7n\xa4\x9c\x9f\xfb\xb7\xf76\x11\x08{' b'\xa6\xe9\x99\xb1\x9e]q\xeb', 'target_type': 'revision'} }, 'id': b"h\xc0\xd2a\x04\xd4~'\x8d\xd6\xbe\x07\xeda\xfa\xfbV" b"\x1d\r "}, { 'origin': 'https://pypi.org/project/limnoria/', 'branches': { b'HEAD': { 'target': b'releases/2018.09.09', 'target_type': 'alias'}, b'releases/2018.09.01': { 'target': b'<\xee1(\xe8\x8d_\xc1\xc9\xa6rT\xf1\x1d' b'\xbb\xdfF\xfdw\xcf', 'target_type': 'revision'}, b'releases/2018.09.09': { 'target': b'\x83\xb9\xb6\xc7\x05\xb1%\xd0\xfem\xd8k' b'A\x10\x9d\xc5\xfa2\xf8t', 'target_type': 'revision'}}, 'id': b'{\xda\x8e\x84\x7fX\xff\x92\x80^\x93V\x18\xa3\xfay' b'\x12\x9e\xd6\xb3'}, { 'origin': 'http://0-512-md.googlecode.com/svn/', 'branches': { b'master': { 'target': b'\xe4?r\xe1,\x88\xab\xec\xe7\x9a\x87\xb8' b'\xc9\xad#.\x1bw=\x18', 'target_type': 'revision'}}, 'id': b'\xa1\xa2\x8c\n\xb3\x87\xa8\xf9\xe0a\x8c\xb7' b'\x05\xea\xb8\x1f\xc4H\xf4s'}, { 'origin': 'https://github.com/librariesio/yarn-parser', 'branches': { b'HEAD': { 'target': hash_to_bytes( '8dbb6aeb036e7fd80664eb8bfd1507881af1ba9f'), 'target_type': 'revision'}}}, { 'origin': 'https://github.com/librariesio/yarn-parser.git', 'branches': { b'HEAD': { 'target': hash_to_bytes( '8dbb6aeb036e7fd80664eb8bfd1507881af1ba9f'), 'target_type': 'revision'}}}, ] REVISIONS = [{ 'id': hash_to_bytes('8dbb6aeb036e7fd80664eb8bfd1507881af1ba9f'), 'message': 'Improve search functionality', 'author': { 'name': b'Andrew Nesbitt', 'fullname': b'Andrew Nesbitt ', 'email': b'andrewnez@gmail.com' }, 'committer': { 'name': b'Andrew Nesbitt', 'fullname': b'Andrew Nesbitt ', 'email': b'andrewnez@gmail.com' }, 'committer_date': { 'negative_utc': None, 'offset': 120, 'timestamp': { 'microseconds': 0, 'seconds': 1380883849 } }, 'type': 'git', 'synthetic': False, 'date': { 'negative_utc': False, 'timestamp': { 'seconds': 1487596456, 'microseconds': 0 }, 'offset': 0 }, 'directory': b'10' }] DIRECTORY_ID = b'10' DIRECTORY_ENTRIES = [{ 'name': b'index.js', 'type': 'file', 'target': b'abc', 'perms': 33188, }, { 'name': b'package.json', 'type': 'file', 'target': b'cde', 'perms': 33188, }, { 'name': b'.github', 'type': 'dir', 'target': b'11', 'perms': 16384, } ] SHA1_TO_LICENSES = { '01c9379dfc33803963d07c1ccc748d3fe4c96bb5': ['GPL'], '02fb2c89e14f7fab46701478c83779c7beb7b069': ['Apache2.0'], '103bc087db1d26afc3a0283f38663d081e9b01e6': ['MIT'], '688a5ef812c53907562fe379d4b3851e69c7cb15': ['AGPL'], 'da39a3ee5e6b4b0d3255bfef95601890afd80709': [], } SHA1_TO_CTAGS = { '01c9379dfc33803963d07c1ccc748d3fe4c96bb5': [{ 'name': 'foo', 'kind': 'str', 'line': 10, 'lang': 'bar', }], 'd4c647f0fc257591cc9ba1722484229780d1c607': [{ 'name': 'let', 'kind': 'int', 'line': 100, 'lang': 'haskell', }], '688a5ef812c53907562fe379d4b3851e69c7cb15': [{ 'name': 'symbol', 'kind': 'float', 'line': 99, 'lang': 'python', }], } OBJ_STORAGE_DATA = { '01c9379dfc33803963d07c1ccc748d3fe4c96bb5': b'this is some text', '688a5ef812c53907562fe379d4b3851e69c7cb15': b'another text', '8986af901dd2043044ce8f0d8fc039153641cf17': b'yet another text', '02fb2c89e14f7fab46701478c83779c7beb7b069': b""" import unittest import logging from swh.indexer.mimetype import MimetypeIndexer from swh.indexer.tests.test_utils import MockObjStorage class MockStorage(): def content_mimetype_add(self, mimetypes): self.state = mimetypes self.conflict_update = conflict_update def indexer_configuration_add(self, tools): return [{ 'id': 10, }] """, '103bc087db1d26afc3a0283f38663d081e9b01e6': b""" #ifndef __AVL__ #define __AVL__ typedef struct _avl_tree avl_tree; typedef struct _data_t { int content; } data_t; """, '93666f74f1cf635c8c8ac118879da6ec5623c410': b""" (should 'pygments (recognize 'lisp 'easily)) """, '26a9f72a7c87cc9205725cfd879f514ff4f3d8d5': b""" { "name": "test_metadata", "version": "0.0.1", "description": "Simple package.json test for indexer", "repository": { "type": "git", "url": "https://github.com/moranegg/metadata_test" } } """, 'd4c647f0fc257591cc9ba1722484229780d1c607': b""" { "version": "5.0.3", "name": "npm", "description": "a package manager for JavaScript", "keywords": [ "install", "modules", "package manager", "package.json" ], "preferGlobal": true, "config": { "publishtest": false }, "homepage": "https://docs.npmjs.com/", "author": "Isaac Z. Schlueter (http://blog.izs.me)", "repository": { "type": "git", "url": "https://github.com/npm/npm" }, "bugs": { "url": "https://github.com/npm/npm/issues" }, "dependencies": { "JSONStream": "~1.3.1", "abbrev": "~1.1.0", "ansi-regex": "~2.1.1", "ansicolors": "~0.3.2", "ansistyles": "~0.1.3" }, "devDependencies": { "tacks": "~1.2.6", "tap": "~10.3.2" }, "license": "Artistic-2.0" } """, 'a7ab314d8a11d2c93e3dcf528ca294e7b431c449': b""" """, 'da39a3ee5e6b4b0d3255bfef95601890afd80709': b'', # 626364 hash_to_hex(b'bcd'): b'unimportant content for bcd', # 636465 hash_to_hex(b'cde'): b""" { "name": "yarn-parser", "version": "1.0.0", "description": "Tiny web service for parsing yarn.lock files", "main": "index.js", "scripts": { "start": "node index.js", "test": "mocha" }, "engines": { "node": "9.8.0" }, "repository": { "type": "git", "url": "git+https://github.com/librariesio/yarn-parser.git" }, "keywords": [ "yarn", "parse", "lock", "dependencies" ], "author": "Andrew Nesbitt", "license": "AGPL-3.0", "bugs": { "url": "https://github.com/librariesio/yarn-parser/issues" }, "homepage": "https://github.com/librariesio/yarn-parser#readme", "dependencies": { "@yarnpkg/lockfile": "^1.0.0", "body-parser": "^1.15.2", "express": "^4.14.0" }, "devDependencies": { "chai": "^4.1.2", "mocha": "^5.2.0", "request": "^2.87.0", "test": "^0.6.0" } } """ } YARN_PARSER_METADATA = { '@context': 'https://doi.org/10.5063/schema/codemeta-2.0', 'url': 'https://github.com/librariesio/yarn-parser#readme', 'codeRepository': 'git+git+https://github.com/librariesio/yarn-parser.git', 'author': [{ 'type': 'Person', 'name': 'Andrew Nesbitt' }], 'license': 'https://spdx.org/licenses/AGPL-3.0', 'version': '1.0.0', 'description': "Tiny web service for parsing yarn.lock files", 'issueTracker': 'https://github.com/librariesio/yarn-parser/issues', 'name': 'yarn-parser', 'keywords': ['yarn', 'parse', 'lock', 'dependencies'], 'type': 'SoftwareSourceCode', } json_dict_keys = strategies.one_of( strategies.characters(), strategies.just('type'), strategies.just('url'), strategies.just('name'), strategies.just('email'), strategies.just('@id'), strategies.just('@context'), strategies.just('repository'), strategies.just('license'), strategies.just('repositories'), strategies.just('licenses'), ) """Hypothesis strategy that generates strings, with an emphasis on those that are often used as dictionary keys in metadata files.""" generic_json_document = strategies.recursive( strategies.none() | strategies.booleans() | strategies.floats() | strategies.characters(), lambda children: ( strategies.lists(children, min_size=1) | strategies.dictionaries(json_dict_keys, children, min_size=1) ) ) """Hypothesis strategy that generates possible values for values of JSON metadata files.""" def json_document_strategy(keys=None): """Generates an hypothesis strategy that generates metadata files for a JSON-based format that uses the given keys.""" if keys is None: keys = strategies.characters() else: keys = strategies.one_of(map(strategies.just, keys)) return strategies.dictionaries(keys, generic_json_document, min_size=1) def _tree_to_xml(root, xmlns, data): def encode(s): "Skips unpaired surrogates generated by json_document_strategy" return s.encode('utf8', 'replace') def to_xml(data, indent=b' '): if data is None: return b'' elif isinstance(data, (bool, str, int, float)): return indent + encode(str(data)) elif isinstance(data, list): return b'\n'.join(to_xml(v, indent=indent) for v in data) elif isinstance(data, dict): lines = [] for (key, value) in data.items(): lines.append(indent + encode('<{}>'.format(key))) lines.append(to_xml(value, indent=indent+b' ')) lines.append(indent + encode(''.format(key))) return b'\n'.join(lines) else: raise TypeError(data) return b'\n'.join([ '<{} xmlns="{}">'.format(root, xmlns).encode(), to_xml(data), ''.format(root).encode(), ]) class TreeToXmlTest(unittest.TestCase): def test_leaves(self): self.assertEqual( _tree_to_xml('root', 'http://example.com', None), b'\n\n' ) self.assertEqual( _tree_to_xml('root', 'http://example.com', True), b'\n True\n' ) self.assertEqual( _tree_to_xml('root', 'http://example.com', 'abc'), b'\n abc\n' ) self.assertEqual( _tree_to_xml('root', 'http://example.com', 42), b'\n 42\n' ) self.assertEqual( _tree_to_xml('root', 'http://example.com', 3.14), b'\n 3.14\n' ) def test_dict(self): self.assertIn( _tree_to_xml('root', 'http://example.com', { 'foo': 'bar', 'baz': 'qux' }), [ b'\n' b' \n bar\n \n' b' \n qux\n \n' b'', b'\n' b' \n qux\n \n' b' \n bar\n \n' b'' ] ) def test_list(self): self.assertEqual( _tree_to_xml('root', 'http://example.com', [ {'foo': 'bar'}, {'foo': 'baz'}, ]), b'\n' b' \n bar\n \n' b' \n baz\n \n' b'' ) def xml_document_strategy(keys, root, xmlns): """Generates an hypothesis strategy that generates metadata files for an XML format that uses the given keys.""" return strategies.builds( functools.partial(_tree_to_xml, root, xmlns), json_document_strategy(keys)) def filter_dict(d, keys): 'return a copy of the dict with keys deleted' if not isinstance(keys, (list, tuple)): keys = (keys, ) return dict((k, v) for (k, v) in d.items() if k not in keys) def fill_obj_storage(obj_storage): """Add some content in an object storage.""" for (obj_id, content) in OBJ_STORAGE_DATA.items(): obj_storage.add(content, obj_id=hash_to_bytes(obj_id)) def fill_storage(storage): visit_types = {} for visit in ORIGIN_VISITS: storage.origin_add_one({'url': visit['url']}) visit_types[visit['url']] = visit['type'] for snap in SNAPSHOTS: origin_url = snap['origin'] visit = storage.origin_visit_add( origin=origin_url, date=datetime.datetime.now(), type=visit_types[origin_url]) snap_id = snap.get('id') or \ bytes([random.randint(0, 255) for _ in range(32)]) storage.snapshot_add([{ 'id': snap_id, 'branches': snap['branches'] }]) storage.origin_visit_update( origin_url, visit['visit'], status='full', snapshot=snap_id) storage.revision_add(REVISIONS) contents = [] for (obj_id, content) in OBJ_STORAGE_DATA.items(): content_hashes = hashutil.MultiHash.from_data(content).digest() contents.append({ 'data': content, 'length': len(content), 'status': 'visible', 'sha1': hash_to_bytes(obj_id), 'sha1_git': hash_to_bytes(obj_id), 'sha256': content_hashes['sha256'], 'blake2s256': content_hashes['blake2s256'] }) storage.content_add(contents) storage.directory_add([{ 'id': DIRECTORY_ID, 'entries': DIRECTORY_ENTRIES, }]) class CommonContentIndexerTest(metaclass=abc.ABCMeta): legacy_get_format = False """True if and only if the tested indexer uses the legacy format. see: https://forge.softwareheritage.org/T1433 """ def get_indexer_results(self, ids): """Override this for indexers that don't have a mock storage.""" return self.indexer.idx_storage.state def assert_legacy_results_ok(self, sha1s, expected_results=None): # XXX old format, remove this when all endpoints are # updated to the new one # see: https://forge.softwareheritage.org/T1433 sha1s = [sha1 if isinstance(sha1, bytes) else hash_to_bytes(sha1) for sha1 in sha1s] actual_results = list(self.get_indexer_results(sha1s)) if expected_results is None: expected_results = self.expected_results self.assertEqual(len(expected_results), len(actual_results), (expected_results, actual_results)) for indexed_data in actual_results: _id = indexed_data['id'] expected_data = expected_results[hashutil.hash_to_hex(_id)].copy() expected_data['id'] = _id self.assertEqual(indexed_data, expected_data) def assert_results_ok(self, sha1s, expected_results=None): if self.legacy_get_format: self.assert_legacy_results_ok(sha1s, expected_results) return sha1s = [sha1 if isinstance(sha1, bytes) else hash_to_bytes(sha1) for sha1 in sha1s] actual_results = list(self.get_indexer_results(sha1s)) if expected_results is None: expected_results = self.expected_results self.assertEqual(len(expected_results), len(actual_results), (expected_results, actual_results)) for indexed_data in actual_results: (_id, indexed_data) = list(indexed_data.items())[0] expected_data = expected_results[hashutil.hash_to_hex(_id)].copy() expected_data = [expected_data] self.assertEqual(indexed_data, expected_data) def test_index(self): """Known sha1 have their data indexed """ sha1s = [self.id0, self.id1, self.id2] # when self.indexer.run(sha1s, policy_update='update-dups') self.assert_results_ok(sha1s) # 2nd pass self.indexer.run(sha1s, policy_update='ignore-dups') self.assert_results_ok(sha1s) def test_index_one_unknown_sha1(self): """Unknown sha1 are not indexed""" sha1s = [self.id1, '799a5ef812c53907562fe379d4b3851e69c7cb15', # unknown '800a5ef812c53907562fe379d4b3851e69c7cb15'] # unknown # when self.indexer.run(sha1s, policy_update='update-dups') # then expected_results = { k: v for k, v in self.expected_results.items() if k in sha1s } self.assert_results_ok(sha1s, expected_results) class CommonContentIndexerRangeTest: """Allows to factorize tests on range indexer. """ def setUp(self): self.contents = sorted(OBJ_STORAGE_DATA) def assert_results_ok(self, start, end, actual_results, expected_results=None): if expected_results is None: expected_results = self.expected_results actual_results = list(actual_results) for indexed_data in actual_results: _id = indexed_data['id'] assert isinstance(_id, bytes) indexed_data = indexed_data.copy() indexed_data['id'] = hash_to_hex(indexed_data['id']) self.assertEqual(indexed_data, expected_results[hash_to_hex(_id)]) self.assertTrue(start <= _id <= end) _tool_id = indexed_data['indexer_configuration_id'] self.assertEqual(_tool_id, self.indexer.tool['id']) def test__index_contents(self): """Indexing contents without existing data results in indexed data """ _start, _end = [self.contents[0], self.contents[2]] # output hex ids start, end = map(hashutil.hash_to_bytes, (_start, _end)) # given actual_results = list(self.indexer._index_contents( start, end, indexed={})) self.assert_results_ok(start, end, actual_results) def test__index_contents_with_indexed_data(self): """Indexing contents with existing data results in less indexed data """ _start, _end = [self.contents[0], self.contents[2]] # output hex ids start, end = map(hashutil.hash_to_bytes, (_start, _end)) data_indexed = [self.id0, self.id2] # given actual_results = self.indexer._index_contents( start, end, indexed=set(map(hash_to_bytes, data_indexed))) # craft the expected results expected_results = self.expected_results.copy() for already_indexed_key in data_indexed: expected_results.pop(already_indexed_key) self.assert_results_ok( start, end, actual_results, expected_results) def test_generate_content_get(self): """Optimal indexing should result in indexed data """ _start, _end = [self.contents[0], self.contents[2]] # output hex ids start, end = map(hashutil.hash_to_bytes, (_start, _end)) # given actual_results = self.indexer.run(start, end) # then self.assertTrue(actual_results) def test_generate_content_get_input_as_bytes(self): """Optimal indexing should result in indexed data Input are in bytes here. """ _start, _end = [self.contents[0], self.contents[2]] # output hex ids start, end = map(hashutil.hash_to_bytes, (_start, _end)) # given - actual_results = self.indexer.run( # checks the bytes input this time + actual_results = self.indexer.run( start, end, skip_existing=False) # no already indexed data so same result as prior test # then - self.assertTrue(actual_results) + self.assertEquals(actual_results, {'status': 'uneventful'}) def test_generate_content_get_no_result(self): """No result indexed returns False""" _start, _end = ['0000000000000000000000000000000000000000', '0000000000000000000000000000000000000001'] start, end = map(hashutil.hash_to_bytes, (_start, _end)) # given - actual_results = self.indexer.run( - start, end, incremental=False) + actual_results = self.indexer.run(start, end, incremental=False) # then - self.assertFalse(actual_results) + self.assertEquals(actual_results, {'status': 'uneventful'})