Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.11/site-packages/dulwich/reflog.py: 20%

1# reflog.py -- Parsing and writing reflog files 

2# Copyright (C) 2015 Jelmer Vernooij and others. 

3# 

4# SPDX-License-Identifier: Apache-2.0 OR GPL-2.0-or-later 

5# Dulwich is dual-licensed under the Apache License, Version 2.0 and the GNU 

6# General Public License as published by the Free Software Foundation; version 2.0 

7# or (at your option) any later version. You can redistribute it and/or 

8# modify it under the terms of either of these two licenses. 

9# 

10# Unless required by applicable law or agreed to in writing, software 

11# distributed under the License is distributed on an "AS IS" BASIS, 

12# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

13# See the License for the specific language governing permissions and 

14# limitations under the License. 

15# 

16# You should have received a copy of the licenses; if not, see 

17# <http://www.gnu.org/licenses/> for a copy of the GNU General Public License 

18# and <http://www.apache.org/licenses/LICENSE-2.0> for a copy of the Apache 

19# License, Version 2.0. 

20# 

21 

22"""Utilities for reading and generating reflogs.""" 

23 

24__all__ = [ 

25 "drop_reflog_entry", 

26 "expire_reflog", 

27 "format_reflog_line", 

28 "iter_reflogs", 

29 "parse_reflog_line", 

30 "parse_reflog_spec", 

31 "read_reflog", 

32] 

33 

34import collections 

35import os 

36from collections.abc import Callable, Generator 

37from pathlib import Path 

38from typing import IO, BinaryIO 

39 

40from .file import _GitFile 

41from .objects import ZERO_SHA, format_timezone, parse_timezone 

42 

43Entry = collections.namedtuple( 

44 "Entry", 

45 ["old_sha", "new_sha", "committer", "timestamp", "timezone", "message"], 

46) 

47 

48 

49def parse_reflog_spec(refspec: str | bytes) -> tuple[bytes, int]: 

50 """Parse a reflog specification like 'HEAD@{1}' or 'refs/heads/master@{2}'. 

51 

52 Args: 

53 refspec: Reflog specification (e.g., 'HEAD@{1}', 'master@{0}') 

54 

55 Returns: 

56 Tuple of (ref_name, index) where index is in Git reflog order (0 = newest) 

57 

58 Raises: 

59 ValueError: If the refspec is not a valid reflog specification 

60 """ 

61 if isinstance(refspec, str): 

62 refspec = refspec.encode("utf-8") 

63 

64 if b"@{" not in refspec: 

65 raise ValueError( 

66 f"Invalid reflog spec: {refspec!r}. Expected format: ref@{{n}}" 

67 ) 

68 

69 ref, rest = refspec.split(b"@{", 1) 

70 if not rest.endswith(b"}"): 

71 raise ValueError( 

72 f"Invalid reflog spec: {refspec!r}. Expected format: ref@{{n}}" 

73 ) 

74 

75 index_str = rest[:-1] 

76 if not index_str.isdigit(): 

77 raise ValueError( 

78 f"Invalid reflog index: {index_str!r}. Expected integer in ref@{{n}}" 

79 ) 

80 

81 # Use HEAD if no ref specified (e.g., "@{1}") 

82 if not ref: 

83 ref = b"HEAD" 

84 

85 return ref, int(index_str) 

86 

87 

88def format_reflog_line( 

89 old_sha: bytes | None, 

90 new_sha: bytes, 

91 committer: bytes, 

92 timestamp: int | float, 

93 timezone: int, 

94 message: bytes, 

95) -> bytes: 

96 """Generate a single reflog line. 

97 

98 Args: 

99 old_sha: Old Commit SHA 

100 new_sha: New Commit SHA 

101 committer: Committer name and e-mail 

102 timestamp: Timestamp 

103 timezone: Timezone 

104 message: Message 

105 """ 

106 if old_sha is None: 

107 old_sha = ZERO_SHA 

108 # A reflog entry is a single line and the message is the last, tab-separated 

109 # field. Collapse embedded whitespace (in particular newlines) to single 

110 # spaces so a multi-line or crafted message cannot split into extra physical 

111 # lines and forge additional reflog entries. Matches Git's copy_reflog_msg. 

112 message = b" ".join(message.split()) 

113 return ( 

114 old_sha 

115 + b" " 

116 + new_sha 

117 + b" " 

118 + committer 

119 + b" " 

120 + str(int(timestamp)).encode("ascii") 

121 + b" " 

122 + format_timezone(timezone) 

123 + b"\t" 

124 + message 

125 ) 

126 

127 

128def parse_reflog_line(line: bytes) -> Entry: 

129 """Parse a reflog line. 

130 

131 Args: 

132 line: Line to parse 

133 Returns: Tuple of (old_sha, new_sha, committer, timestamp, timezone, 

134 message) 

135 """ 

136 (begin, message) = line.split(b"\t", 1) 

137 (old_sha, new_sha, rest) = begin.split(b" ", 2) 

138 (committer, timestamp_str, timezone_str) = rest.rsplit(b" ", 2) 

139 return Entry( 

140 old_sha, 

141 new_sha, 

142 committer, 

143 int(timestamp_str), 

144 parse_timezone(timezone_str)[0], 

145 message, 

146 ) 

147 

148 

149def read_reflog( 

150 f: BinaryIO | IO[bytes] | _GitFile, 

151) -> Generator[Entry, None, None]: 

152 """Read reflog. 

153 

154 Args: 

155 f: File-like object 

156 Returns: Iterator over Entry objects 

157 """ 

158 for line in f: 

159 yield parse_reflog_line(line.rstrip(b"\n")) 

160 

161 

162def drop_reflog_entry(f: BinaryIO, index: int, rewrite: bool = False) -> None: 

163 """Drop the specified reflog entry. 

164 

165 Args: 

166 f: File-like object 

167 index: Reflog entry index (in Git reflog reverse 0-indexed order) 

168 rewrite: If a reflog entry's predecessor is removed, set its 

169 old SHA to the new SHA of the entry that now precedes it 

170 """ 

171 if index < 0: 

172 raise ValueError(f"Invalid reflog index {index}") 

173 

174 log = [] 

175 offset = f.tell() 

176 for line in f: 

177 log.append((offset, parse_reflog_line(line.rstrip(b"\n")))) 

178 offset = f.tell() 

179 

180 inverse_index = len(log) - index - 1 

181 write_offset = log[inverse_index][0] 

182 f.seek(write_offset) 

183 

184 if index == 0: 

185 f.truncate() 

186 return 

187 

188 del log[inverse_index] 

189 if rewrite and index > 0 and log: 

190 if inverse_index == 0: 

191 previous_new = ZERO_SHA 

192 else: 

193 previous_new = log[inverse_index - 1][1].new_sha 

194 offset, entry = log[inverse_index] 

195 log[inverse_index] = ( 

196 offset, 

197 Entry( 

198 previous_new, 

199 entry.new_sha, 

200 entry.committer, 

201 entry.timestamp, 

202 entry.timezone, 

203 entry.message, 

204 ), 

205 ) 

206 

207 for _, entry in log[inverse_index:]: 

208 f.write( 

209 format_reflog_line( 

210 entry.old_sha, 

211 entry.new_sha, 

212 entry.committer, 

213 entry.timestamp, 

214 entry.timezone, 

215 entry.message, 

216 ) 

217 + b"\n" 

218 ) 

219 f.truncate() 

220 

221 

222def expire_reflog( 

223 f: BinaryIO, 

224 expire_time: int | None = None, 

225 expire_unreachable_time: int | None = None, 

226 reachable_checker: Callable[[bytes], bool] | None = None, 

227) -> int: 

228 """Expire reflog entries based on age and reachability. 

229 

230 Args: 

231 f: File-like object for the reflog 

232 expire_time: Expire entries older than this timestamp (seconds since epoch). 

233 If None, entries are not expired based on age alone. 

234 expire_unreachable_time: Expire unreachable entries older than this 

235 timestamp. If None, unreachable entries are not expired. 

236 reachable_checker: Optional callable that takes a SHA and returns True 

237 if the commit is reachable. If None, all entries are considered 

238 reachable. 

239 

240 Returns: 

241 Number of entries expired 

242 """ 

243 if expire_time is None and expire_unreachable_time is None: 

244 return 0 

245 

246 entries = [] 

247 offset = f.tell() 

248 for line in f: 

249 entries.append((offset, parse_reflog_line(line.rstrip(b"\n")))) 

250 offset = f.tell() 

251 

252 # Filter entries that should be kept 

253 kept_entries = [] 

254 expired_count = 0 

255 

256 for offset, entry in entries: 

257 should_expire = False 

258 

259 # Check if entry is reachable 

260 is_reachable = True 

261 if reachable_checker is not None: 

262 is_reachable = reachable_checker(entry.new_sha) 

263 

264 # Apply expiration rules 

265 # Check the appropriate expiration time based on reachability 

266 if is_reachable: 

267 if expire_time is not None and entry.timestamp < expire_time: 

268 should_expire = True 

269 else: 

270 if ( 

271 expire_unreachable_time is not None 

272 and entry.timestamp < expire_unreachable_time 

273 ): 

274 should_expire = True 

275 

276 if should_expire: 

277 expired_count += 1 

278 else: 

279 kept_entries.append((offset, entry)) 

280 

281 # Write back the kept entries 

282 if expired_count > 0: 

283 f.seek(0) 

284 for _, entry in kept_entries: 

285 f.write( 

286 format_reflog_line( 

287 entry.old_sha, 

288 entry.new_sha, 

289 entry.committer, 

290 entry.timestamp, 

291 entry.timezone, 

292 entry.message, 

293 ) 

294 + b"\n" 

295 ) 

296 f.truncate() 

297 

298 return expired_count 

299 

300 

301def iter_reflogs(logs_dir: str) -> Generator[bytes, None, None]: 

302 """Iterate over all reflogs in a repository. 

303 

304 Args: 

305 logs_dir: Path to the logs directory (e.g., .git/logs) 

306 

307 Yields: 

308 Reference names (as bytes) that have reflogs 

309 """ 

310 if not os.path.exists(logs_dir): 

311 return 

312 

313 logs_path = Path(logs_dir) 

314 for log_file in logs_path.rglob("*"): 

315 if log_file.is_file(): 

316 # Get the ref name by removing the logs_dir prefix 

317 ref_name = str(log_file.relative_to(logs_path)) 

318 # Convert path separators to / for refs 

319 ref_name = ref_name.replace(os.sep, "/") 

320 yield ref_name.encode("utf-8")