Class: Bibliothecary::MultiParsers::Spdx

Inherits:
Object
  • Object
show all
Extended by:
Analyser::TryCache
Includes:
Analyser
Defined in:
lib/bibliothecary/multi_parsers/spdx.rb

Constant Summary collapse

WELLFORMED_LINE_REGEXP =

e.g. 'SomeText:' (allowing for leading whitespace)

/^\s*[a-zA-Z]+:/
PACKAGE_NAME_REGEXP =

e.g. 'PackageName: (allowing for excessive whitespace)

/^\s*PackageName:\s*(.*)/
PACKAGE_VERSION_REGEXP =

e.g. 'PackageVersion:' (allowing for excessive whitespace)

/^\s*PackageVersion:\s*(.*)/
PURL_REGEXP =

e.g. "ExternalRef: PACKAGE-MANAGER purl (allowing for excessive whitespace)

/^\s*ExternalRef:\s*PACKAGE[-|_]MANAGER\s*purl\s*(.*)/
NoEntries =
Class.new(StandardError)
MalformedFile =
Class.new(StandardError)

Class Method Summary collapse

Methods included from Analyser::TryCache

try_cache

Methods included from Analyser

create_analysis, create_error_analysis, included

Class Method Details

.add_entry(entries:, platform:, purl_name:, spdx_name:, purl_version:, spdx_version:, source: nil, purl_type: nil) ⇒ Object



197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 197

def self.add_entry(entries:, platform:, purl_name:, spdx_name:, purl_version:, spdx_version:, source: nil, purl_type: nil)
  package_name = purl_name || spdx_name
  package_version = purl_version || spdx_version

  return unless package_name && package_version
  return unless platform

  entries << Dependency.new(
    platform: platform ? platform.to_s : purl_type,
    name: package_name,
    requirement: package_version,
    type: "lockfile",
    source: source
  )
end

.mappingObject



31
32
33
34
35
36
37
38
39
40
41
42
43
44
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 31

def self.mapping
  {
    match_extension("spdx", case_insensitive: true) => {
      kind: "lockfile",
      parser: :parse_spdx_tag_value,
      ungroupable: true,
    },
    match_extension("spdx.json", case_insensitive: true) => {
      kind: "lockfile",
      parser: :parse_spdx_json,
      ungroupable: true,
    },
  }
end

.parse_spdx_2_json_file_contents(manifest, source) ⇒ Object



134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 134

def self.parse_spdx_2_json_file_contents(manifest, source)
  entries = Set.new
  manifest["packages"]&.each do |package|
    spdx_name = package["name"]
    spdx_version = package["versionInfo"]

    first_purl_string = package["externalRefs"]&.find { |ref| ref["referenceType"] == "purl" }&.dig("referenceLocator")
    purl = first_purl_string && PackageURL.parse(first_purl_string)
    purl_type = purl&.type
    # Use the mapped purl->bibliothecary platform, or else fall back to original platform itself.
    platform = PurlUtil::PURL_TYPE_MAPPING.fetch(purl_type, purl_type)
    purl_name = PurlUtil.full_name(purl)
    purl_version = purl&.version

    add_entry(entries: entries, platform: platform, purl_name: purl_name,
              spdx_name: spdx_name, purl_version: purl_version, spdx_version: spdx_version,
              source: source, purl_type: purl_type)
  end
  entries
end

.parse_spdx_3_json_file_contents(manifest, source) ⇒ Object



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
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 155

def self.parse_spdx_3_json_file_contents(manifest, source)
  entries = Set.new
  manifest.fetch("@graph", [])
    .select do |element|
      types = Array(element["type"] || element["@type"])
      # "software_Package" is the common one, but these may differ and include namespaces
      types.any? { |t| t.to_s.match?(/Package$/i) }
    end
    .each do |element|
      spdx_name = element["name"]
      spdx_version = element["software_packageVersion"] || element["packageVersion"] || element["versionInfo"]

      purl_string = element["externalIdentifier"]&.find do |ext_id|
        ext_id_type = ext_id["externalIdentifierType"] || ext_id["identifierType"]
        ext_id_type&.match?(/purl/i)
      end&.dig("identifier")

      purl_string ||= element.fetch("externalRef", [])
        .map { |ref| ref.fetch("locator", nil) }
        .compact
        .map(&:first)
        .find { |locator| locator.start_with?("pkg:") }

      purl_string ||= element.fetch("software_packageUrl", nil)

      next unless purl_string

      purl = PackageURL.parse(purl_string)
      purl_type = purl&.type
      # Use the mapped purl->bibliothecary platform, or else fall back to original platform itself.
      platform = PurlUtil::PURL_TYPE_MAPPING.fetch(purl_type, purl_type)
      purl_name = PurlUtil.full_name(purl)
      purl_version = purl&.version

      add_entry(entries: entries, platform: platform, purl_name: purl_name,
                spdx_name: spdx_name, purl_version: purl_version, spdx_version: spdx_version,
                source: source, purl_type: purl_type)
    end

  entries
end

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

Raises:



109
110
111
112
113
114
115
116
117
118
119
120
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 109

def self.parse_spdx_json(file_contents, options: {})
  entries = try_cache(options, options[:filename]) do
    parse_spdx_json_file_contents(
      file_contents,
      options.fetch(:filename, nil)
    )
  end

  raise NoEntries if entries.empty?

  Bibliothecary::ParserResult.new(dependencies: entries.to_a)
end

.parse_spdx_json_file_contents(file_contents, source = nil) ⇒ Object



122
123
124
125
126
127
128
129
130
131
132
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 122

def self.parse_spdx_json_file_contents(file_contents, source = nil)
  manifest = JSON.parse(file_contents)

  if manifest.key?("spdxVersion")
    parse_spdx_2_json_file_contents(manifest, source)
  elsif manifest.key?("@context")
    parse_spdx_3_json_file_contents(manifest, source)
  else
    Set.new
  end
end

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

Raises:



50
51
52
53
54
55
56
57
58
59
60
61
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 50

def self.parse_spdx_tag_value(file_contents, options: {})
  entries = try_cache(options, options[:filename]) do
    parse_spdx_tag_value_file_contents(
      file_contents,
      options.fetch(:filename, nil)
    )
  end

  raise NoEntries if entries.empty?

  Bibliothecary::ParserResult.new(dependencies: entries.to_a)
end

.parse_spdx_tag_value_file_contents(file_contents, source = nil) ⇒ Object



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
101
102
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 63

def self.parse_spdx_tag_value_file_contents(file_contents, source = nil)
  entries = Set.new
  spdx_name = spdx_version = platform = purl_name = purl_version = purl_type = nil

  file_contents.each_line do |line|
    stripped_line = line.strip
    next if skip_tag_value_line?(stripped_line)

    raise MalformedFile unless stripped_line.match?(WELLFORMED_LINE_REGEXP)

    if (match = stripped_line.match(PACKAGE_NAME_REGEXP))
      # Per the spec:
      # > A new package Information section is denoted by the package name (7.1) field.
      add_entry(entries: entries, platform: platform, purl_name: purl_name,
                spdx_name: spdx_name, purl_version: purl_version, spdx_version: spdx_version,
                source: source, purl_type: purl_type)

      # reset for this new package
      spdx_name = spdx_version = platform = purl_name = purl_version = purl_type = nil

      # capture the new package's name
      spdx_name = match[1]
    elsif (match = stripped_line.match(PACKAGE_VERSION_REGEXP))
      spdx_version = match[1]
    elsif (match = stripped_line.match(PURL_REGEXP))
      purl = PackageURL.parse(match[1])
      purl_type ||= purl&.type
      # Use the mapped purl->bibliothecary platform, or else fall back to original platform itself.
      platform = PurlUtil::PURL_TYPE_MAPPING.fetch(purl_type, purl_type)
      purl_name ||= PurlUtil.full_name(purl)
      purl_version ||= purl&.version
    end
  end

  add_entry(entries: entries, platform: platform, purl_name: purl_name,
            spdx_name: spdx_name, purl_version: purl_version, spdx_version: spdx_version,
            source: source, purl_type: purl_type)

  entries
end

.platform_nameObject



46
47
48
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 46

def self.platform_name
  raise "Spdx is a multi-parser and does not have a platform name."
end

.skip_tag_value_line?(stripped_line) ⇒ Boolean

Returns:

  • (Boolean)


104
105
106
107
# File 'lib/bibliothecary/multi_parsers/spdx.rb', line 104

def self.skip_tag_value_line?(stripped_line)
  # Ignore blank lines and comments
  stripped_line.empty? || stripped_line.start_with?("#")
end