diff --git a/swh/loader/dir/converters.py b/swh/loader/dir/converters.py index 034ea54..dd5173a 100644 --- a/swh/loader/dir/converters.py +++ b/swh/loader/dir/converters.py @@ -1,148 +1,148 @@ # Copyright (C) 2015 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 """Convert dir objects to dictionaries suitable for swh.storage""" import datetime from swh.loader.dir.git.git import GitType from swh.loader.dir.git import git, utils def to_datetime(ts): """Convert a timestamp to utc datetime. """ return datetime.datetime.utcfromtimestamp(ts).replace( tzinfo=datetime.timezone.utc) def format_to_minutes(offset_str): """Convert a git string timezone format string (e.g +0200, -0310) to minutes. Args: offset_str: a string representing an offset. Returns: A positive or negative number of minutes of such input """ sign = offset_str[0] hours = int(offset_str[1:3]) minutes = int(offset_str[3:]) + (hours * 60) return minutes if sign == '+' else -1 * minutes def blob_to_content(obj, log=None, max_content_size=None, origin_id=None): if 'data' not in obj: filepath = obj['path'] - content_raw, length = utils._read_raw(filepath) + content_raw = open(filepath, 'rb').read() obj.update({'data': content_raw, - 'length': length}) + 'length': len(content_raw)}) return _blob_to_content(obj, log, max_content_size, origin_id) def _blob_to_content(obj, log=None, max_content_size=None, origin_id=None): """Convert to a compliant swh content. """ size = obj['length'] ret = { 'sha1': obj['sha1'], 'sha256': obj['sha256'], 'sha1_git': obj['sha1_git'], 'length': size, 'perms': obj['perms'].value, 'type': obj['type'].value, } if max_content_size and size > max_content_size: if log: log.info('Skipping content %s, too large (%s > %s)' % (utils.hash_to_hex(obj['sha1_git']), size, max_content_size)) ret.update({'status': 'absent', 'reason': 'Content too large', 'origin': origin_id}) return ret ret.update({ 'status': 'visible', 'data': obj['data'], }) return ret # Map of type to swh types _entry_type_map = { GitType.TREE: 'dir', GitType.BLOB: 'file', GitType.COMM: 'rev', } def tree_to_directory(tree, objects, log=None): """Format a tree as a directory """ entries = [] for entry in objects[tree['path']]: entries.append({ 'type': _entry_type_map[entry['type']], 'perms': int(entry['perms'].value), 'name': entry['name'], 'target': entry['sha1_git'] }) return { 'id': tree['sha1_git'], 'entries': entries } def commit_to_revision(commit, objects, log=None): """Format a commit as a revision. """ upper_directory = objects[git.ROOT_TREE_KEY][0] return { 'id': commit['sha1_git'], 'date': to_datetime(commit['author_date']), 'date_offset': format_to_minutes(commit['author_offset']), 'committer_date': to_datetime(commit['committer_date']), 'committer_date_offset': format_to_minutes(commit['committer_offset']), 'type': commit['type'], 'directory': upper_directory['sha1_git'], 'message': commit['message'], 'author_name': commit['author_name'], 'author_email': commit['author_email'], 'committer_name': commit['committer_name'], 'committer_email': commit['committer_email'], 'synthetic': True, 'metadata': commit['metadata'], 'parents': [], } def annotated_tag_to_release(release, log=None): """Format a swh release. """ return { 'id': release['sha1_git'], 'revision': release['revision'], 'name': release['name'], 'comment': release['comment'], 'date': to_datetime(release['date']), 'date_offset': format_to_minutes(release['offset']), 'author_name': release['author_name'], 'author_email': release['author_email'], 'synthetic': True, } diff --git a/swh/loader/dir/git/utils.py b/swh/loader/dir/git/utils.py index 04afe0d..b1f8037 100644 --- a/swh/loader/dir/git/utils.py +++ b/swh/loader/dir/git/utils.py @@ -1,121 +1,98 @@ # Copyright (C) 2015 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 hashlib import os from io import BytesIO from swh.core import hashutil hashfile = hashutil.hashfile hash_to_hex = hashutil.hash_to_hex hex_to_hash = hashutil.hex_to_hash def _new_hash(header_type, length): """Initialize a digest object (as returned by python's hashlib) for the git sha1 algorithm. This is in charge of pre-computing the needed header for git. Args: header_type: a git sha1 type ('blob', 'tree', 'commit', 'tag') length: Length of content to hash. Could be None if when hashing with sha1 and sha256 Returns: A digest object Raises: ValueError if header_type is not one of 'blob', 'commit', 'tree', 'tag' """ h = hashlib.new('sha1') if header_type not in ('blob', 'commit', 'tree', 'tag'): raise ValueError( 'Only supported types are blob, commit, tree, tag') h.update(('%s %d\0' % (header_type, length)).encode('ascii')) return h def _hash_file_obj(f, header_type, length): """hash (git sha1) the content of a file-like object f with header_type and length. Returns: A dictionary with 'sha1_git' as key and value the computed sha1_git. Raises: ValueError if header_type is not one of 'blob', 'commit', 'tree', 'tag' """ h = _new_hash(header_type, length) while True: chunk = f.read(hashutil.HASH_BLOCK_SIZE) if not chunk: break h.update(chunk) return {'sha1_git': h.digest()} def hashdata(data, header_type): """Hash data as git sha1 with header_type. Returns: A dictionary with 'sha1_git' as key and value the computed sha1_git. Raises: ValueError if header_type is not one of 'blob', 'commit', 'tree', 'tag' """ buf = BytesIO(data) return _hash_file_obj(buf, header_type, len(data)) -def _read_raw(filepath): - """Read filepath's raw content and returns it. - - Args: - filepath: absolute path to an existing file. - - Returns: - raw content in bytes + its length - - """ - content_raw = b'' - length = 0 - with open(filepath, 'rb') as f: - while True: - chunk = f.read(hashutil.HASH_BLOCK_SIZE) - if not chunk: - break - content_raw += chunk - length += len(chunk) - - return content_raw, length - - def hashlink(linkpath): """Compute hashes for a link. Args: linkpath: the absolute path name to a symbolic link. Returns: dictionary with sha1_git as key and the actual binary sha1 as value. """ raw_data = os.readlink(linkpath) hashes = hashutil.hashdata(raw_data) hashes.update({ 'data': raw_data, 'length': len(raw_data) }) return hashes