Module: Datadog::CI::Git::LocalRepository

Defined in:
lib/datadog/ci/git/local_repository.rb

Class Method Summary collapse

Class Method Details

.base_commit_sha(base_branch: nil) ⇒ Object

On best effort basis determines the git sha of the most likely base branch for the current PR.



360
361
362
363
364
365
366
367
368
 
# File 'lib/datadog/ci/git/local_repository.rb', line 360

def self.base_commit_sha(base_branch: nil)
  Telemetry.git_command(Ext::Telemetry::Command::BASE_COMMIT_SHA)

  BaseBranchShaDetector.base_branch_sha(base_branch)
rescue => e
  Telemetry.track_error(e, Ext::Telemetry::Command::BASE_COMMIT_SHA)
  log_failure(e, "git base ref")
  nil
end

.current_folder_name ⇒ Object 



78
79
80
 
# File 'lib/datadog/ci/git/local_repository.rb', line 78

def self.current_folder_name
  File.basename(root)
end

.filter_invalid_commits(commits) ⇒ Object 



377
378
379
 
# File 'lib/datadog/ci/git/local_repository.rb', line 377

def self.filter_invalid_commits(commits)
  commits.filter { |commit| Utils::Git.valid_commit_sha?(commit) }
end

.get_changed_files_from_diff(base_commit) ⇒ Object

Returns a Set of normalized file paths changed since the given base_commit. If base_commit is nil, returns nil. On error, returns nil.



313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
 
# File 'lib/datadog/ci/git/local_repository.rb', line 313

def self.get_changed_files_from_diff(base_commit)
  return nil if base_commit.nil?

  Datadog.logger.debug { "calculating git diff from base_commit: #{base_commit}" }

  Telemetry.git_command(Ext::Telemetry::Command::DIFF)

  begin
    # 1. Run the git diff command

    # @type var output: String?
    output = nil
    duration_ms = Core::Utils::Time.measure(:float_millisecond) do
      output = CLI.exec_git_command(["diff", "-U0", "--word-diff=porcelain", base_commit, "HEAD"], timeout: CLI::LONG_TIMEOUT)
    end
    Telemetry.git_command_ms(Ext::Telemetry::Command::DIFF, duration_ms)

    Datadog.logger.debug { "git diff output: #{output}" }

    return nil if output.nil?

    # 2. Parse the output to extract which files changed
    changed_files = Set.new
    output.each_line do |line|
      # Match lines like: diff --git a/foo/bar.rb b/foo/bar.rb
      # This captures git changes on file level
      match = /^diff --git a\/(?<file>.+?) b\//.match(line)
      if match && match[:file]
        changed_file = match[:file]
        # Normalize to repo root
        normalized_changed_file = relative_to_root(changed_file)
        changed_files << normalized_changed_file unless normalized_changed_file.nil? || normalized_changed_file.empty?

        Datadog.logger.debug { "matched changed_file: #{changed_file} from line: #{line}" }
        Datadog.logger.debug { "normalized_changed_file: #{normalized_changed_file}" }
      end
    end
    changed_files
  rescue => e
    Telemetry.track_error(e, Ext::Telemetry::Command::DIFF)
    log_failure(e, "get changed files from diff")
    nil
  end
end

.get_upstream_branch ⇒ Object 



370
371
372
373
374
375
 
# File 'lib/datadog/ci/git/local_repository.rb', line 370

def self.get_upstream_branch
  CLI.exec_git_command(["rev-parse", "--abbrev-ref", "--symbolic-full-name", "@{upstream}"])
rescue => e
  Datadog.logger.debug { "Error getting upstream: #{e}" }
  nil
end

.git_branch ⇒ Object 



113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
 
# File 'lib/datadog/ci/git/local_repository.rb', line 113

def self.git_branch
  Telemetry.git_command(Ext::Telemetry::Command::GET_BRANCH)
  # @type var res: String?
  res = nil

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    res = CLI.exec_git_command(["rev-parse", "--abbrev-ref", "HEAD"])
  end

  Telemetry.git_command_ms(Ext::Telemetry::Command::GET_BRANCH, duration_ms)
  res
rescue => e
  log_failure(e, "git branch")
  Telemetry.track_error(e, Ext::Telemetry::Command::GET_BRANCH)
  nil
end

.git_commit_message ⇒ Object 



137
138
139
140
141
142
 
# File 'lib/datadog/ci/git/local_repository.rb', line 137

def self.git_commit_message
  CLI.exec_git_command(["log", "-n", "1", "--format=%B"])
rescue => e
  log_failure(e, "git commit message")
  nil
end

.git_commit_sha ⇒ Object 



106
107
108
109
110
111
 
# File 'lib/datadog/ci/git/local_repository.rb', line 106

def self.git_commit_sha
  CLI.exec_git_command(["rev-parse", "HEAD"])
rescue => e
  log_failure(e, "git commit sha")
  nil
end

.git_commit_users ⇒ Object 



144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
 
# File 'lib/datadog/ci/git/local_repository.rb', line 144

def self.git_commit_users
  # Get committer and author information in one command.
  output = CLI.exec_git_command(["show", "-s", "--format=%an\t%ae\t%at\t%cn\t%ce\t%ct"])
  unless output
    Datadog.logger.debug(
      "Unable to read git commit users: git command output is nil"
    )
    nil_user = NilUser.new
    return [nil_user, nil_user]
  end

  author_name, author_email, author_timestamp,
    committer_name, committer_email, committer_timestamp = output.split("\t").each(&:strip!)

  author = User.new(author_name, author_email, author_timestamp)
  committer = User.new(committer_name, committer_email, committer_timestamp)

  [author, committer]
rescue => e
  log_failure(e, "git commit users")

  nil_user = NilUser.new
  [nil_user, nil_user]
end

.git_commits ⇒ Object

returns maximum of 1000 latest commits in the last month



170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
 
# File 'lib/datadog/ci/git/local_repository.rb', line 170

def self.git_commits
  Telemetry.git_command(Ext::Telemetry::Command::GET_LOCAL_COMMITS)

  # @type var output: String?
  output = nil

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    output = CLI.exec_git_command(["log", "--format=%H", "-n", "1000", "--since=\"1 month ago\""], timeout: CLI::LONG_TIMEOUT)
  end

  Telemetry.git_command_ms(Ext::Telemetry::Command::GET_LOCAL_COMMITS, duration_ms)

  return [] if output.nil?

  output.split("\n")
rescue => e
  log_failure(e, "git commits")
  Telemetry.track_error(e, Ext::Telemetry::Command::GET_LOCAL_COMMITS)
  []
end

.git_commits_rev_list(included_commits:, excluded_commits:) ⇒ Object 



191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
 
# File 'lib/datadog/ci/git/local_repository.rb', line 191

def self.git_commits_rev_list(included_commits:, excluded_commits:)
  Telemetry.git_command(Ext::Telemetry::Command::GET_OBJECTS)
  included_commits_list = filter_invalid_commits(included_commits)
  excluded_commits_list = filter_invalid_commits(excluded_commits).map { |sha| "^#{sha}" }

  # @type var res: String?
  res = nil

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    cmd = [
      "rev-list",
      "--objects",
      "--no-object-names",
      "--filter=blob:none",
      "--since=\"1 month ago\""
    ] + excluded_commits_list + included_commits_list

    res = CLI.exec_git_command(cmd, timeout: CLI::LONG_TIMEOUT)
  end

  Telemetry.git_command_ms(Ext::Telemetry::Command::GET_OBJECTS, duration_ms)

  res
rescue => e
  log_failure(e, "git commits rev list")
  Telemetry.track_error(e, Ext::Telemetry::Command::GET_OBJECTS)
  nil
end

.git_generate_packfiles(included_commits:, excluded_commits:, path:) ⇒ Object 



220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
 
# File 'lib/datadog/ci/git/local_repository.rb', line 220

def self.git_generate_packfiles(included_commits:, excluded_commits:, path:)
  return nil unless File.exist?(path)

  commit_tree = git_commits_rev_list(included_commits: included_commits, excluded_commits: excluded_commits)
  return nil if commit_tree.nil?

  basename = SecureRandom.hex(4)

  Telemetry.git_command(Ext::Telemetry::Command::PACK_OBJECTS)

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    CLI.exec_git_command(
      ["pack-objects", "--compression=9", "--max-pack-size=3m", "#{path}/#{basename}"],
      stdin: commit_tree,
      timeout: CLI::LONG_TIMEOUT
    )
  end
  Telemetry.git_command_ms(Ext::Telemetry::Command::PACK_OBJECTS, duration_ms)

  basename
rescue => e
  log_failure(e, "git generate packfiles")
  Telemetry.track_error(e, Ext::Telemetry::Command::PACK_OBJECTS)
  nil
end

.git_repository_url ⇒ Object 



82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
 
# File 'lib/datadog/ci/git/local_repository.rb', line 82

def self.git_repository_url
  Telemetry.git_command(Ext::Telemetry::Command::GET_REPOSITORY)
  # @type var res: String?
  res = nil

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    res = CLI.exec_git_command(["ls-remote", "--get-url"])
  end

  Telemetry.git_command_ms(Ext::Telemetry::Command::GET_REPOSITORY, duration_ms)
  res
rescue => e
  log_failure(e, "git repository url")
  Telemetry.track_error(e, Ext::Telemetry::Command::GET_REPOSITORY)
  nil
end

.git_root ⇒ Object 



99
100
101
102
103
104
 
# File 'lib/datadog/ci/git/local_repository.rb', line 99

def self.git_root
  CLI.exec_git_command(["rev-parse", "--show-toplevel"])
rescue => e
  log_failure(e, "git root path")
  nil
end

.git_shallow_clone? ⇒ Boolean

Returns:

  • (Boolean) 


246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
 
# File 'lib/datadog/ci/git/local_repository.rb', line 246

def self.git_shallow_clone?
  Telemetry.git_command(Ext::Telemetry::Command::CHECK_SHALLOW)
  res = false

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    res = CLI.exec_git_command(["rev-parse", "--is-shallow-repository"]) == "true"
  end
  Telemetry.git_command_ms(Ext::Telemetry::Command::CHECK_SHALLOW, duration_ms)

  res
rescue => e
  log_failure(e, "git shallow clone")
  Telemetry.track_error(e, Ext::Telemetry::Command::CHECK_SHALLOW)
  false
end

.git_tag ⇒ Object 



130
131
132
133
134
135
 
# File 'lib/datadog/ci/git/local_repository.rb', line 130

def self.git_tag
  CLI.exec_git_command(["tag", "--points-at", "HEAD"])
rescue => e
  log_failure(e, "git tag")
  nil
end

.git_unshallow ⇒ Object 



262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
 
# File 'lib/datadog/ci/git/local_repository.rb', line 262

def self.git_unshallow
  Telemetry.git_command(Ext::Telemetry::Command::UNSHALLOW)
  # @type var res: String?
  res = nil

  default_remote = CLI.exec_git_command(["config", "--default", "origin", "--get", "clone.defaultRemoteName"])&.strip
  head_commit = git_commit_sha
  upstream_branch = get_upstream_branch

  # Build array of remotes to try, filtering out nil values
  unshallow_remotes = []
  unshallow_remotes << head_commit if head_commit
  unshallow_remotes << upstream_branch if upstream_branch
  unshallow_remotes << nil # Ensure the loop runs at least once, even if no valid remotes are available. This acts as a fallback mechanism.

  duration_ms = Core::Utils::Time.measure(:float_millisecond) do
    unshallow_remotes.each do |remote|
      unshallowing_errored = false

      res =
        begin
          # @type var cmd: Array[String]
          cmd = [
            "fetch",
            "--shallow-since=\"1 month ago\"",
            "--update-shallow",
            "--filter=blob:none",
            "--recurse-submodules=no",
            default_remote
          ]
          cmd << remote if remote

          CLI.exec_git_command(cmd, timeout: CLI::UNSHALLOW_TIMEOUT)
        rescue => e
          log_failure(e, "git unshallow")
          Telemetry.track_error(e, Ext::Telemetry::Command::UNSHALLOW)
          unshallowing_errored = true
          nil
        end

      # If the command succeeded, break and return the result
      break [] if res && !unshallowing_errored
    end
  end

  Telemetry.git_command_ms(Ext::Telemetry::Command::UNSHALLOW, duration_ms)
  res
end

.log_failure(e, action) ⇒ Object 



381
382
383
384
385
 
# File 'lib/datadog/ci/git/local_repository.rb', line 381

def self.log_failure(e, action)
  Datadog.logger.debug(
    "Unable to perform #{action}: #{e.class.name} #{e.message} at #{Array(e.backtrace).first}"
  )
end

.relative_to_root(path) ⇒ Object

ATTENTION: this function is running in a hot path and must be optimized for performance



26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
 
# File 'lib/datadog/ci/git/local_repository.rb', line 26

def self.relative_to_root(path)
  return "" if path.nil?

  root_path = root
  return path if root_path.nil?

  if File.absolute_path?(path)
    # prefix_index is where the root path ends in the given path
    prefix_index = root_path.size

    # impossible case - absolute paths are returned from code coverage tool that always checks
    # that root is a prefix of the path
    return "" if path.size < prefix_index

    prefix_index += 1 if path[prefix_index] == File::SEPARATOR
    res = path[prefix_index..]
  else
    # prefix_to_root is a difference between the root path and the given path
    if @prefix_to_root == ""
      return path
    elsif @prefix_to_root
      return File.join(@prefix_to_root, path)
    end

    pathname = Pathname.new(File.expand_path(path))
    root_path = Pathname.new(root_path)

    # relative_path_from is an expensive function
    res = pathname.relative_path_from(root_path).to_s

    unless defined?(@prefix_to_root)
      @prefix_to_root = res.gsub(path, "") if res.end_with?(path)
    end
  end

  res || ""
end

.repository_name ⇒ Object 



64
65
66
67
68
69
70
71
72
73
74
75
76
 
# File 'lib/datadog/ci/git/local_repository.rb', line 64

def self.repository_name
  return @repository_name if defined?(@repository_name)

  git_remote_url = git_repository_url

  # return git repository name from remote url without .git extension
  last_path_segment = git_remote_url.split("/").last if git_remote_url
  @repository_name = last_path_segment.gsub(".git", "") if last_path_segment
  @repository_name ||= current_folder_name
rescue => e
  log_failure(e, "git repository name")
  @repository_name = current_folder_name
end

.root ⇒ Object 



18
19
20
21
22
 
# File 'lib/datadog/ci/git/local_repository.rb', line 18

def self.root
  return @root if defined?(@root)

  @root = git_root || Dir.pwd
end