Class: Bibliothecary::Parsers::Maven

Inherits:
Object
  • Object
show all
Includes:
Analyser
Defined in:
lib/bibliothecary/parsers/maven.rb

Constant Summary collapse

GRADLE_TYPE_REGEX =

e.g. “annotationProcessor - Annotation processors and their dependencies for source set ‘main’.”

/^(\w+)/
GRADLE_DEP_REGEX =

e.g. “| \— com.google.guava:guava:23.5-jre (*)”

/(\+---|\\---){1}/
GRADLE_PROJECT_REGEX =

Dependencies that are on-disk projects, eg: e.g. “-– project :api:my-internal-project” e.g. “+— my-group:my-alias:1.2.3 -> project :client (*)”

/project :(\S+)?/
GRADLE_LINE_ENDING_REGEX =

line ending legend: © means a dependency constraint, (n) means not resolved, or (*) means resolved previously, e.g. org.springframework.boot:spring-boot-starter-web:2.1.0.M3 (*) e.g. the “(n)” in “+— my-group:my-name:1.2.3 (n)”

/(\((c|n|\*)\))$/
GRADLE_DEPENDENCY_METHODS =

Builtin methods: docs.gradle.org/current/userguide/java_plugin.html#tab:configurations Deprecated methods: docs.gradle.org/current/userguide/upgrading_version_6.html#sec:configuration_removal

%w(api compile compileClasspath compileOnly compileOnlyApi implementation runtime runtimeClasspath runtimeOnly testCompile testCompileOnly testImplementation testRuntime testRuntimeOnly)
GRADLE_VERSION_REGEX =

Intentionally overly-simplified regexes to scrape deps from build.gradle (Groovy) and build.gradle.kts (Kotlin) files. To be truly useful bibliothecary would need full Groovy / Kotlin parsers that speaks Gradle, because the Groovy and Kotlin DSLs have many dynamic ways of declaring dependencies.

/[\w.-]+/
GRADLE_VAR_INTERPOLATION_REGEX =

e.g. ‘$myVersion’

/\$\w+/
GRADLE_CODE_INTERPOLATION_REGEX =

e.g. ‘$"version"’

/\$\{.*\}/
GRADLE_GAV_REGEX =

e.g. “group:artifactId:1.2.3”

/([\w.-]+)\:([\w.-]+)(?:\:(#{GRADLE_VERSION_REGEX}|#{GRADLE_VAR_INTERPOLATION_REGEX}|#{GRADLE_CODE_INTERPOLATION_REGEX}))?/
GRADLE_GROOVY_SIMPLE_REGEX = 
/(#{GRADLE_DEPENDENCY_METHODS.join('|')})\s*\(?\s*['"]#{GRADLE_GAV_REGEX}['"]/m
GRADLE_KOTLIN_SIMPLE_REGEX = 
/(#{GRADLE_DEPENDENCY_METHODS.join('|')})\s*\(\s*"#{GRADLE_GAV_REGEX}"/m
MAVEN_PROPERTY_REGEX = 
/\$\{(.+?)\}/
MAX_DEPTH = 
5
SBT_TYPE_REGEX =

e.g. “[info] test:”

/^\[info\]\s+([-\w]+):$/
SBT_DEP_REGEX =

e.g. “[info] org.typelevel:spire-util_2.12”

/^\[info\]\s+(.+)$/
SBT_VERSION_REGEX =

e.g. “[info] - 1.7.5”

/^\[info\]\s+-\s+(.+)$/
SBT_FIELD_REGEX =

e.g. “[info] homepage: www.slf4j.org”

/^\[info\]\s+([^:]+):\s+(.+)$/
SBT_IGNORE_REGEX =

e.g. “[info] ”

/^\[info\]\s*$/

Class Method Summary collapse

Methods included from Analyser

create_analysis, create_error_analysis, included

Class Method Details

.extract_pom_dep_info(xml, dependency, name, parent_properties = {}) ⇒ Object

TODO: it might be worth renaming parent_properties to parent_elements so that more can be inherited from the parent pom than just <properties> here (see maven.apache.org/pom.html#inheritance)



366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
 
# File 'lib/bibliothecary/parsers/maven.rb', line 366

def self.extract_pom_dep_info(xml, dependency, name, parent_properties = {})
  field = dependency.locate(name).first
  return nil if field.nil?

  value = field.nodes.first
  value = value.value if value.is_a?(Ox::CData)
  # whitespace in dependency tags should be ignored
  value = value&.strip
  match = value&.match(MAVEN_PROPERTY_REGEX)
  if match
    return extract_property(xml, match[1], value, parent_properties)
  else
    return value
  end
end

.extract_pom_info(xml, location, parent_properties = {}) ⇒ Object 



359
360
361
 
# File 'lib/bibliothecary/parsers/maven.rb', line 359

def self.extract_pom_info(xml, location, parent_properties = {})
  extract_pom_dep_info(xml, xml, location, parent_properties)
end

.extract_property(xml, property_name, value, parent_properties = {}, depth = 0) ⇒ Object 



386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
 
# File 'lib/bibliothecary/parsers/maven.rb', line 386

def self.extract_property(xml, property_name, value, parent_properties = {}, depth = 0)
  prop_value = property_value(xml, property_name, parent_properties)
  return value unless prop_value
  # don't resolve more than 5 levels deep to avoid potential circular references

  resolved_value = replace_value_with_prop(value, prop_value, property_name)
  # check to see if we just resolved to another property name
  match = resolved_value.match(MAVEN_PROPERTY_REGEX)
  if match && depth < MAX_DEPTH
    depth += 1
    return extract_property(xml, match[1], resolved_value, parent_properties, depth)
  else
    return resolved_value
  end
end

.gradle_dependency_name(group, name) ⇒ Object 



345
346
347
348
349
350
351
352
353
354
355
356
357
 
# File 'lib/bibliothecary/parsers/maven.rb', line 345

def self.gradle_dependency_name(group, name)
  if group.empty? && name.include?(":")
    group, name = name.split(":", 2)
  end

  # Strip comments, and single/doublequotes
  [group, name].map do |part|
    part
      .gsub(/\s*\/\/.*$/, "") # Comments
      .gsub(/^["']/, "") # Beginning single/doublequotes
      .gsub(/["']$/, "") # Ending single/doublequotes
  end.join(":")
end

.ivy_report?(file_contents) ⇒ Boolean

Returns:

  • (Boolean) 


118
119
120
121
122
123
124
125
126
127
 
# File 'lib/bibliothecary/parsers/maven.rb', line 118

def self.ivy_report?(file_contents)
  doc = Ox.parse file_contents
  root = doc&.locate("ivy-report")&.first
  return !root.nil?
rescue Exception # rubocop:disable Lint/RescueException
  # We rescue exception here since native libs can throw a non-StandardError
  # We don't want to throw errors during the matching phase, only during
  # parsing after we match.
  false
end

.mapping ⇒ Object 



60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
 
# File 'lib/bibliothecary/parsers/maven.rb', line 60

def self.mapping
  {
    match_filename("ivy.xml", case_insensitive: true) => {
      kind: 'manifest',
      parser: :parse_ivy_manifest
    },
    match_filename("pom.xml", case_insensitive: true) => {
      kind: 'manifest',
      parser: :parse_standalone_pom_manifest
    },
    match_filename("build.gradle", case_insensitive: true) => {
      kind: 'manifest',
      parser: :parse_gradle
    },
    match_filename("build.gradle.kts", case_insensitive: true) => {
      kind: 'manifest',
      parser: :parse_gradle_kts
    },
    match_extension(".xml", case_insensitive: true) => {
      content_matcher: :ivy_report?,
      kind: 'lockfile',
      parser: :parse_ivy_report
    },
    match_filename("gradle-dependencies-q.txt", case_insensitive: true) => {
      kind: 'lockfile',
      parser: :parse_gradle_resolved
    },
    match_filename("maven-resolved-dependencies.txt", case_insensitive: true) => {
      kind: 'lockfile',
      parser: :parse_maven_resolved
    },
    match_filename("sbt-update-full.txt", case_insensitive: true) => {
      kind: 'lockfile',
      parser: :parse_sbt_update_full
    },
    match_filename("maven-dependency-tree.txt", case_insensitive: true) => {
      kind: 'lockfile',
      parser: :parse_maven_tree
    }
  }
end

.parse_gradle(file_contents, options: {}) ⇒ Object 



319
320
321
322
323
324
325
326
327
328
329
330
 
# File 'lib/bibliothecary/parsers/maven.rb', line 319

def self.parse_gradle(file_contents, options: {})
  file_contents
  .scan(GRADLE_GROOVY_SIMPLE_REGEX)                                                # match 'implementation "group:artifactId:version"'
  .reject { |(_type, group, artifactId, _version)| group.nil? || artifactId.nil? } # remove any matches with missing group/artifactId
  .map { |(type, group, artifactId, version)|
    {
      name: [group, artifactId].join(":"),
      requirement: version || "*",
      type: type
    }
  }
end

.parse_gradle_kts(file_contents, options: {}) ⇒ Object 



332
333
334
335
336
337
338
339
340
341
342
343
 
# File 'lib/bibliothecary/parsers/maven.rb', line 332

def self.parse_gradle_kts(file_contents, options: {})
  file_contents
    .scan(GRADLE_KOTLIN_SIMPLE_REGEX)                                                # match 'implementation("group:artifactId:version")'
    .reject { |(_type, group, artifactId, _version)| group.nil? || artifactId.nil? } # remove any matches with missing group/artifactId
    .map { |(type, group, artifactId, version)|
      {
        name: [group, artifactId].join(":"),
        requirement: version || "*",
        type: type
      }
    }
end

.parse_gradle_resolved(file_contents, options: {}) ⇒ Object 



154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
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/bibliothecary/parsers/maven.rb', line 154

def self.parse_gradle_resolved(file_contents, options: {})
  current_type = nil

  file_contents.split("\n").map do |line|
    current_type_match = GRADLE_TYPE_REGEX.match(line)
    current_type = current_type_match.captures[0] if current_type_match

    gradle_dep_match = GRADLE_DEP_REGEX.match(line)
    next unless gradle_dep_match

    split = gradle_dep_match.captures[0]

    # gradle can import on-disk projects and deps will be listed under them, e.g. `+--- project :test:integration`,
    # so we treat these projects as "internal" deps with requirement of "1.0.0"
    if (project_match = line.match(GRADLE_PROJECT_REGEX))
      # an empty project name is self-referential (i.e. a cycle), and we don't need to track the manifest's project itself, e.g. "+--- project :"
      next if project_match[1].nil?

      # project names can have colons (e.g. for gradle projects in subfolders), which breaks maven artifact naming assumptions, so just replace them with hyphens.
      project_name = project_match[1].gsub(/:/, "-")
      line = line.sub(GRADLE_PROJECT_REGEX, "internal:#{project_name}:1.0.0")
    end

    dep = line
      .split(split)[1]
      .sub(GRADLE_LINE_ENDING_REGEX, "")
      .sub(/ FAILED$/, "") # dependency could not be resolved (but still may have a version)
      .sub(" -> ", ":") # handle version arrow syntax
      .strip
      .split(":")

    # A testImplementation line can look like this so just skip those
    # \--- org.springframework.security:spring-security-test (n)
    next unless dep.length >= 3

    if dep.count == 6
      # get name from renamed package resolution "org:name:version -> renamed_org:name:version"
      {
        original_name: dep[0,2].join(":"),
        original_requirement: dep[2],
        name: dep[-3..-2].join(":"),
        requirement: dep[-1],
        type: current_type
      }
    elsif dep.count == 5
      # get name from renamed package resolution "org:name -> renamed_org:name:version"
      {
        original_name: dep[0,2].join(":"),
        original_requirement: "*",
        name: dep[-3..-2].join(":"),
        requirement: dep[-1],
        type: current_type
      }
    else
      # get name from version conflict resolution ("org:name:version -> version") and no-resolution ("org:name:version")
      {
        name: dep[0..1].join(":"),
        requirement: dep[-1],
        type: current_type
      }
    end
  end
    .compact
    .uniq { |item| item.values_at(:name, :requirement, :type, :original_name, :original_requirement) }
end

.parse_ivy_manifest(file_contents, options: {}) ⇒ Object 



106
107
108
109
110
111
112
113
114
115
116
 
# File 'lib/bibliothecary/parsers/maven.rb', line 106

def self.parse_ivy_manifest(file_contents, options: {})
  manifest = Ox.parse file_contents
  manifest.dependencies.locate('dependency').map do |dependency|
    attrs = dependency.attributes
    {
      name: "#{attrs[:org]}:#{attrs[:name]}",
      requirement: attrs[:rev],
      type: 'runtime'
    }
  end
end

.parse_ivy_report(file_contents, options: {}) ⇒ Object 



129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
 
# File 'lib/bibliothecary/parsers/maven.rb', line 129

def self.parse_ivy_report(file_contents, options: {})
  doc = Ox.parse file_contents
  root = doc.locate("ivy-report").first
  raise "ivy-report document does not have ivy-report at the root" if root.nil?
  info = doc.locate("ivy-report/info").first
  raise "ivy-report document lacks <info> element" if info.nil?
  type = info.attributes[:conf]
  type = "unknown" if type.nil?
  modules = doc.locate("ivy-report/dependencies/module")
  modules.map do |mod|
    attrs = mod.attributes
    org = attrs[:organisation]
    name = attrs[:name]
    version = mod.locate('revision').first&.attributes[:name]

    next nil if org.nil? or name.nil? or version.nil?

    {
      name: "#{org}:#{name}",
      requirement: version,
      type: type
    }
  end.compact
end

.parse_maven_resolved(file_contents, options: {}) ⇒ Object 



220
221
222
223
224
225
226
 
# File 'lib/bibliothecary/parsers/maven.rb', line 220

def self.parse_maven_resolved(file_contents, options: {})
  Strings::ANSI.sanitize(file_contents)
    .split("\n")
    .map(&method(:parse_resolved_dep_line))
    .compact
    .uniq
end

.parse_maven_tree(file_contents, options: {}) ⇒ Object 



228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
 
# File 'lib/bibliothecary/parsers/maven.rb', line 228

def self.parse_maven_tree(file_contents, options: {})
  captures = Strings::ANSI.sanitize(file_contents)
    .gsub(/\r\n?/, "\n")
    .scan(/^\[INFO\](?:(?:\+-)|\||(?:\\-)|\s)+((?:[\w\.-]+:)+[\w\.\-${}]+)/)
    .flatten
    .uniq

  deps = captures.map do |item|
    parts = item.split(":")
    case parts.count
    when 4
      version = parts[-1]
      type = parts[-2]
    when 5..6
      version, type = parts[-2..]
    end
    {
      name: parts[0..1].join(":"),
      requirement: version,
      type: type
    }
  end

  # First dep line will be the package itself (unless we're only analyzing a single line)
  package = deps[0]
  deps.size < 2 ? deps : deps[1..-1].reject { |d| d[:name] == package[:name] && d[:requirement] == package[:requirement] }
end

.parse_pom_manifest(file_contents, parent_properties = {}, options: {}) ⇒ Object

parent_properties is used by Libraries: github.com/librariesio/libraries.io/blob/e970925aade2596a03268b6e1be785eba8502c62/app/models/package_manager/maven.rb#L129



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
310
311
312
313
314
315
316
317
 
# File 'lib/bibliothecary/parsers/maven.rb', line 277

def self.parse_pom_manifest(file_contents, parent_properties = {}, options: {})
  manifest = Ox.parse file_contents
  xml = manifest.respond_to?('project') ? manifest.project : manifest
  [].tap do |deps|
    # <dependencyManagement> is a namespace to specify artifact configuration (e.g. version), but it doesn't
    # actually add dependencies to your project. Grab these and keep them for reference while parsing <dependencies>
    # Ref: https://maven.apache.org/pom.html#Dependency_Management
    # Ref: https://maven.apache.org/guides/introduction/introduction-to-dependency-mechanism.html#transitive-dependencies
    dependencyManagement = xml.locate("dependencyManagement/dependencies/dependency").map do |dep|
      {
        groupId: extract_pom_dep_info(xml, dep, "groupId", parent_properties),
        artifactId: extract_pom_dep_info(xml, dep, "artifactId", parent_properties),
        version: extract_pom_dep_info(xml, dep, "version", parent_properties),
        scope: extract_pom_dep_info(xml, dep, "scope", parent_properties),
      }
    end
    # <dependencies> is the namespace that will add dependencies to your project.
    xml.locate("dependencies/dependency").each do |dep|
      groupId = extract_pom_dep_info(xml, dep, 'groupId', parent_properties)
      artifactId = extract_pom_dep_info(xml, dep, 'artifactId', parent_properties)
      version = extract_pom_dep_info(xml, dep, 'version', parent_properties)
      scope = extract_pom_dep_info(xml, dep, 'scope', parent_properties)

      # Use any dep configurations from <dependencyManagement> as fallbacks
      if (depConfig = dependencyManagement.find { |d| d[:groupId] == groupId && d[:artifactId] == artifactId })
        version ||= depConfig[:version]
        scope ||= depConfig[:scope]
      end

      dep_hash = {
        name: "#{groupId}:#{artifactId}",
        requirement: version,
        type: scope || 'runtime',
      }
      # optional field is, itself, optional, and will be either "true" or "false"
      optional = extract_pom_dep_info(xml, dep, 'optional', parent_properties)
      dep_hash[:optional] = optional == "true" unless optional.nil?
      deps.push(dep_hash)
    end
  end
end

.parse_resolved_dep_line(line) ⇒ Object 



256
257
258
259
260
261
262
263
264
265
266
267
268
269
 
# File 'lib/bibliothecary/parsers/maven.rb', line 256

def self.parse_resolved_dep_line(line)
  # filter out anything that doesn't look like a
  # resolved dep line
  return unless line[/  .*:[^-]+-- /]

  dep_parts = line.strip.split(":")
  return unless dep_parts.length == 5
  # org.springframework.boot:spring-boot-starter-web:jar:2.0.3.RELEASE:compile -- module spring.boot.starter.web [auto]
  {
    name: dep_parts[0, 2].join(":"),
    requirement: dep_parts[3],
    type: dep_parts[4].split("--").first.strip
  }
end

.parse_sbt_deps(type, lines) ⇒ Object 



465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
 
# File 'lib/bibliothecary/parsers/maven.rb', line 465

def self.parse_sbt_deps(type, lines)
  deps = []
  while lines.any? and not SBT_TYPE_REGEX.match(lines[0])
    line = lines.shift

    next if SBT_IGNORE_REGEX.match(line)

    dep_match = SBT_DEP_REGEX.match(line)
    if dep_match
      versions = parse_sbt_versions(type, dep_match.captures[0], lines)
      deps.concat(versions)
    else
      lines.unshift(line)
      break
    end
  end

  deps
end

.parse_sbt_update_full(file_contents, options: {}) ⇒ Object 



428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
 
# File 'lib/bibliothecary/parsers/maven.rb', line 428

def self.parse_sbt_update_full(file_contents, options: {})
  all_deps = []
  type = nil
  lines = file_contents.split("\n")
  while lines.any?
    line = lines.shift

    type_match = SBT_TYPE_REGEX.match(line)
    next unless type_match
    type = type_match.captures[0]

    deps = parse_sbt_deps(type, lines)
    all_deps.concat(deps)
  end

  # strip out evicted dependencies
  all_deps.select! do |dep|
    dep[:fields]["evicted"] != "true"
  end

  # in the future, we could use "callers" in the fields to
  # decide which deps are direct root deps and which are
  # pulled in by another dep.  The direct deps have the sbt
  # project name as a caller.

  # clean out any duplicates (I'm pretty sure sbt will have done this for
  # us so this is paranoia, basically)
  squished = all_deps.compact.uniq {|item| [item[:name], item[:requirement], item[:type]]}

  # get rid of the fields
  squished.each do |dep|
    dep.delete(:fields)
  end

  return squished
end

.parse_sbt_version(type, name, version, lines) ⇒ Object 



502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
 
# File 'lib/bibliothecary/parsers/maven.rb', line 502

def self.parse_sbt_version(type, name, version, lines)
  fields = {}
  while lines.any? and not SBT_TYPE_REGEX.match(lines[0])
    line = lines.shift

    field_match = SBT_FIELD_REGEX.match(line)
    if field_match
      fields[field_match.captures[0]] = field_match.captures[1]
    else
      lines.unshift(line)
      break
    end
  end

  {
    name: name,
    requirement: version,
    type: type,
    # we post-process using some of these fields and then delete them again
    fields: fields
  }
end

.parse_sbt_versions(type, name, lines) ⇒ Object 



485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
 
# File 'lib/bibliothecary/parsers/maven.rb', line 485

def self.parse_sbt_versions(type, name, lines)
  versions = []
  while lines.any? and not SBT_TYPE_REGEX.match(lines[0])
    line = lines.shift

    version_match = SBT_VERSION_REGEX.match(line)
    if version_match
      versions.push(parse_sbt_version(type, name, version_match.captures[0], lines))
    else
      lines.unshift(line)
      break
    end
  end

  versions
end

.parse_standalone_pom_manifest(file_contents, options: {}) ⇒ Object 



271
272
273
 
# File 'lib/bibliothecary/parsers/maven.rb', line 271

def self.parse_standalone_pom_manifest(file_contents, options: {})
  parse_pom_manifest(file_contents, {}, options: options)
end

.property_value(xml, property_name, parent_properties) ⇒ Object 



402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
 
# File 'lib/bibliothecary/parsers/maven.rb', line 402

def self.property_value(xml, property_name, parent_properties)
  # the xml root is <project> so lookup the non property name in the xml
  # this converts ${project/group.id} -> ${group/id}
  non_prop_name = property_name.gsub(".", "/").gsub("project/", "")
  return "${#{property_name}}" if !xml.respond_to?("properties") && parent_properties.empty? && xml.locate(non_prop_name).empty?

  prop_field = xml.properties.locate(property_name).first if xml.respond_to?("properties")
  parent_prop = parent_properties[property_name] ||                 # e.g. "${foo}"
    parent_properties[property_name.sub(/^project\./, '')] ||       # e.g. "${project.foo}"
    parent_properties[property_name.sub(/^project\.parent\./, '')]  # e.g. "${project.parent.foo}"

  if prop_field
    prop_field.nodes.first
  elsif parent_prop
    parent_prop
  elsif xml.locate(non_prop_name).first
    # see if the value to look up is a field under the project
    # examples are ${project.groupId} or ${project.version}
    xml.locate(non_prop_name).first.nodes.first
  elsif xml.locate("parent/#{non_prop_name}").first
    # see if the value to look up is a field under the project parent
    # examples are ${project.groupId} or ${project.version}
    xml.locate("parent/#{non_prop_name}").first.nodes.first
  end
end

.replace_value_with_prop(original_value, property_value, property_name) ⇒ Object 



382
383
384
 
# File 'lib/bibliothecary/parsers/maven.rb', line 382

def self.replace_value_with_prop(original_value, property_value, property_name)
  original_value.gsub("${#{property_name}}", property_value)
end