docsterra 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
checksums.yaml ADDED
@@ -0,0 +1,7 @@
1
+ ---
2
+ SHA256:
3
+ metadata.gz: 778f499c80a748af7e5848eb9a68287b1e299a47afa54a0ff20e9de9182053a3
4
+ data.tar.gz: 8a284d3e4dcbba8921270f6211a3f5b7dde5eb5519d64e7f91f893031d2967eb
5
+ SHA512:
6
+ metadata.gz: 4dbcbab92c1caa74ddb7f3c321c21da185ed1598211e5d02b6bee9291dfb1a0bca9937531e59744ca4d7cc53d1f6dce034ec48a97b73915ec8521a589b9feb39
7
+ data.tar.gz: aa5fc647fd2545013654b6c08955c94cc70222061bba44b6025cccf814248e7de65cc1eaa20ad4d999a34b1c51aaf7932d6133daa7c5606eb3535ba616327ebe
data/CHANGELOG.md ADDED
@@ -0,0 +1,7 @@
1
+ # Changelog
2
+
3
+ ## Unreleased
4
+
5
+ ## 0.1.0 (2026-02-26)
6
+
7
+ - Initial release.
data/LICENSE.txt ADDED
@@ -0,0 +1,21 @@
1
+ The MIT License (MIT)
2
+
3
+ Copyright (c) 2026 Yudai Takada
4
+
5
+ Permission is hereby granted, free of charge, to any person obtaining a copy
6
+ of this software and associated documentation files (the "Software"), to deal
7
+ in the Software without restriction, including without limitation the rights
8
+ to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
9
+ copies of the Software, and to permit persons to whom the Software is
10
+ furnished to do so, subject to the following conditions:
11
+
12
+ The above copyright notice and this permission notice shall be included in
13
+ all copies or substantial portions of the Software.
14
+
15
+ THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
16
+ IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
17
+ FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
18
+ AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
19
+ LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
20
+ OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
21
+ THE SOFTWARE.
data/README.md ADDED
@@ -0,0 +1,133 @@
1
+ # Docsterra
2
+
3
+ `docsterra` is a Ruby gem / CLI that parses Terraform (`.tf`) files and generates infrastructure design documentation in Markdown.
4
+
5
+ Current implementation focuses on GCP Terraform resources and produces:
6
+
7
+ - resource inventory tables
8
+ - Mermaid network diagrams
9
+ - cross-product dependency diagrams
10
+ - security summaries (IAM / firewall / warnings)
11
+ - cost-related spec summaries (non-billing estimate metadata)
12
+
13
+ ## Installation
14
+
15
+ ```bash
16
+ gem install docsterra
17
+ ```
18
+
19
+ Or from a Gemfile:
20
+
21
+ ```ruby
22
+ gem "docsterra"
23
+ ```
24
+
25
+ ## Quick Start
26
+
27
+ Single Terraform root:
28
+
29
+ ```bash
30
+ docsterra generate ./terraform
31
+ ```
32
+
33
+ Multiple products (cross-product dependency analysis enabled):
34
+
35
+ ```bash
36
+ docsterra generate ./product-a/terraform ./product-b/terraform ./shared/terraform -o ./docs/infrastructure.md
37
+ ```
38
+
39
+ Dry-run summary:
40
+
41
+ ```bash
42
+ docsterra check ./terraform
43
+ ```
44
+
45
+ ## CLI Reference
46
+
47
+ ### Commands
48
+
49
+ ```bash
50
+ docsterra generate [PATHS...]
51
+ docsterra check [PATHS...]
52
+ docsterra version
53
+ ```
54
+
55
+ ### Options
56
+
57
+ - `-o`, `--output`: output markdown path (default: `./infrastructure.md`)
58
+ - `-c`, `--config`: config file path (default: `.docsterra.yml`)
59
+ - `-s`, `--sections`: comma-separated sections (`all`, `resources`, `network`, `security`, `cost`)
60
+ - `-f`, `--format`: output format (currently only `markdown`)
61
+ - `-v`, `--verbose`: verbose output
62
+ - `--ignore`: ignore glob patterns (repeatable via Thor array syntax)
63
+
64
+ ## Library Usage
65
+
66
+ ```ruby
67
+ require "docsterra"
68
+
69
+ doc = Docsterra.generate("./terraform")
70
+ doc.save("./docs/infrastructure.md")
71
+ puts doc.to_markdown
72
+
73
+ summary = Docsterra.check("./terraform")
74
+ puts summary[:resource_count]
75
+ ```
76
+
77
+ ## Configuration (`.docsterra.yml`)
78
+
79
+ ```yaml
80
+ products:
81
+ - name: "Web Application"
82
+ path: "./product-web/terraform"
83
+ - name: "Shared Infrastructure"
84
+ path: "./shared/terraform"
85
+ shared: true
86
+
87
+ output:
88
+ path: "./docs/infrastructure.md"
89
+ sections:
90
+ - resources
91
+ - network
92
+ - security
93
+ - cost
94
+
95
+ resource_attributes:
96
+ google_compute_instance:
97
+ - machine_type
98
+ - zone
99
+ - labels.env
100
+
101
+ ignore:
102
+ - "**/examples/**"
103
+ ```
104
+
105
+ ## Output Example (Sections)
106
+
107
+ - `## Overview`
108
+ - `## Cross-Product Dependencies` (Mermaid)
109
+ - `## {Product}`
110
+ - `### Resources`
111
+ - `### Network Configuration` (Mermaid)
112
+ - `### Security Settings`
113
+ - `### Cost Estimation`
114
+ - `## Appendix`
115
+
116
+ ## Limitations
117
+
118
+ - HCL parsing is best-effort. Unsupported expressions are preserved as `RawExpr` and rendered as text.
119
+ - Remote modules are not parsed (metadata only).
120
+ - Local modules are parsed recursively, but complex module evaluation/output resolution is not performed.
121
+ - No real-time GCP pricing lookup; cost output is resource spec metadata only.
122
+ - Terraform state / `terraform plan` JSON is not consumed.
123
+
124
+ ## Development
125
+
126
+ ```bash
127
+ bundle install
128
+ bundle exec rspec
129
+ ```
130
+
131
+ ## License
132
+
133
+ MIT
data/Rakefile ADDED
@@ -0,0 +1,8 @@
1
+ # frozen_string_literal: true
2
+
3
+ require "bundler/gem_tasks"
4
+ require "rspec/core/rake_task"
5
+
6
+ RSpec::Core::RakeTask.new(:spec)
7
+
8
+ task default: :spec
data/exe/docsterra ADDED
@@ -0,0 +1,6 @@
1
+ #!/usr/bin/env ruby
2
+ # frozen_string_literal: true
3
+
4
+ require "terradoc"
5
+
6
+ Terradoc::CLI.start(ARGV)
@@ -0,0 +1,104 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Docsterra
4
+ module Analyzer
5
+ class CostAnalyzer
6
+ CostItem = Struct.new(:resource_name, :resource_type, :spec, :region, :note, :resource, keyword_init: true)
7
+
8
+ def analyze(resources)
9
+ resources.filter_map do |resource|
10
+ build_cost_item(resource)
11
+ end
12
+ end
13
+
14
+ private
15
+
16
+ def build_cost_item(resource)
17
+ case resource.type
18
+ when "google_compute_instance"
19
+ CostItem.new(
20
+ resource_name: resource.attribute_text("name") || resource.name,
21
+ resource_type: resource.type,
22
+ spec: resource.attribute_text("machine_type"),
23
+ region: resource.attribute_text("zone"),
24
+ note: nil,
25
+ resource: resource
26
+ )
27
+ when "google_container_cluster"
28
+ node_type = resource.attribute_text("node_config.machine_type")
29
+ node_count = resource.attribute_text("initial_node_count")
30
+ CostItem.new(
31
+ resource_name: resource.attribute_text("name") || resource.name,
32
+ resource_type: resource.type,
33
+ spec: [node_type, node_count && "x #{node_count}"].compact.join(" "),
34
+ region: resource.attribute_text("location"),
35
+ note: "Autopilot: #{truthy_string(resource.attribute_ruby('enable_autopilot'))}",
36
+ resource: resource
37
+ )
38
+ when "google_container_node_pool"
39
+ CostItem.new(
40
+ resource_name: resource.attribute_text("name") || resource.name,
41
+ resource_type: resource.type,
42
+ spec: resource.attribute_text("node_config.machine_type"),
43
+ region: resource.attribute_text("location") || resource.attribute_text("region"),
44
+ note: node_pool_note(resource),
45
+ resource: resource
46
+ )
47
+ when "google_sql_database_instance"
48
+ CostItem.new(
49
+ resource_name: resource.attribute_text("name") || resource.name,
50
+ resource_type: resource.type,
51
+ spec: resource.attribute_text("settings.tier") || resource.attribute_text("tier"),
52
+ region: resource.attribute_text("region"),
53
+ note: "HA: #{truthy_string((resource.attribute_text('settings.availability_type') || resource.attribute_text('availability_type')).to_s == 'REGIONAL')}",
54
+ resource: resource
55
+ )
56
+ when "google_storage_bucket"
57
+ CostItem.new(
58
+ resource_name: resource.attribute_text("name") || resource.name,
59
+ resource_type: resource.type,
60
+ spec: nil,
61
+ region: resource.attribute_text("location"),
62
+ note: "class: #{resource.attribute_text('storage_class')}",
63
+ resource: resource
64
+ )
65
+ when "google_redis_instance"
66
+ CostItem.new(
67
+ resource_name: resource.attribute_text("name") || resource.name,
68
+ resource_type: resource.type,
69
+ spec: "memory_size_gb=#{resource.attribute_text('memory_size_gb')}",
70
+ region: resource.attribute_text("region"),
71
+ note: "tier: #{resource.attribute_text('tier')}",
72
+ resource: resource
73
+ )
74
+ when "google_compute_disk"
75
+ CostItem.new(
76
+ resource_name: resource.attribute_text("name") || resource.name,
77
+ resource_type: resource.type,
78
+ spec: "#{resource.attribute_text('size')}GB #{resource.attribute_text('type')}".strip,
79
+ region: resource.attribute_text("zone"),
80
+ note: nil,
81
+ resource: resource
82
+ )
83
+ end
84
+ end
85
+
86
+ def truthy_string(value)
87
+ case value
88
+ when true then "yes"
89
+ when false then "no"
90
+ else
91
+ value.nil? ? "unknown" : value.to_s
92
+ end
93
+ end
94
+
95
+ def node_pool_note(resource)
96
+ min = resource.attribute_text("autoscaling.min_node_count")
97
+ max = resource.attribute_text("autoscaling.max_node_count")
98
+ return nil if min.nil? && max.nil?
99
+
100
+ "autoscaling: #{min || '?'}-#{max || '?'}"
101
+ end
102
+ end
103
+ end
104
+ end
@@ -0,0 +1,192 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Docsterra
4
+ module Analyzer
5
+ class DependencyAnalyzer
6
+ def analyze(projects)
7
+ relationships = []
8
+ resource_index = build_resource_index(projects)
9
+ name_index = build_name_index(projects)
10
+ string_name_index = build_string_name_index(projects)
11
+ service_account_index = build_service_account_index(projects)
12
+
13
+ projects.each do |project|
14
+ relationships.concat(detect_reference_relationships(project, resource_index))
15
+ relationships.concat(detect_data_source_relationships(project, name_index))
16
+ relationships.concat(detect_shared_name_relationships(project, string_name_index))
17
+ relationships.concat(detect_service_account_relationships(project, service_account_index))
18
+ end
19
+
20
+ deduplicate(relationships)
21
+ end
22
+
23
+ private
24
+
25
+ def build_resource_index(projects)
26
+ projects.each_with_object({}) do |project, index|
27
+ project.resources.each { |resource| index[resource.identifier] = resource }
28
+ end
29
+ end
30
+
31
+ def build_name_index(projects)
32
+ projects.each_with_object(Hash.new { |hash, key| hash[key] = [] }) do |project, index|
33
+ project.resources.each do |resource|
34
+ key = [resource.type, resource.attribute_text("name") || resource.attribute_text("dataset_id") || resource.name]
35
+ index[key] << resource
36
+ end
37
+ end
38
+ end
39
+
40
+ def build_service_account_index(projects)
41
+ projects.each_with_object({}) do |project, index|
42
+ project.resources.select { |resource| resource.type == "google_service_account" }.each do |sa|
43
+ account_id = sa.attribute_text("account_id")
44
+ email = sa.attribute_text("email")
45
+ [account_id, email].compact.each { |key| index[key] = sa }
46
+ end
47
+ end
48
+ end
49
+
50
+ def build_string_name_index(projects)
51
+ projects.each_with_object(Hash.new { |hash, key| hash[key] = [] }) do |project, index|
52
+ project.resources.each do |resource|
53
+ next unless shareable_name_target?(resource)
54
+
55
+ names_for_resource(resource).each do |name|
56
+ next if name.nil? || name.empty?
57
+
58
+ index[name] << resource
59
+ end
60
+ end
61
+ end
62
+ end
63
+
64
+ def detect_reference_relationships(project, resource_index)
65
+ project.resources.flat_map do |resource|
66
+ resource.references.filter_map do |ref|
67
+ target = resource_index[reference_identifier(ref)]
68
+ next unless target
69
+ next if target.project == project
70
+
71
+ rel_type = target.project.shared? ? :shared_resource : relation_type_for_target(target)
72
+ Model::Relationship.new(
73
+ source: resource,
74
+ target: target,
75
+ type: rel_type,
76
+ detail: "Reference #{ref}"
77
+ )
78
+ end
79
+ end
80
+ end
81
+
82
+ def detect_data_source_relationships(project, name_index)
83
+ project.data_sources.filter_map do |data_source|
84
+ lookup_key = [data_source.type, data_source.attribute_text("name") || data_source.attribute_text("dataset_id")]
85
+ matches = name_index[lookup_key].reject { |resource| resource.project == project }
86
+ next if matches.empty?
87
+
88
+ target = matches.first
89
+ rel_type = target.project.shared? ? :shared_resource : relation_type_for_target(target)
90
+ Model::Relationship.new(
91
+ source: project,
92
+ target: target.project,
93
+ type: rel_type,
94
+ detail: "Data source #{data_source.type}.#{data_source.name} uses #{lookup_key[1]}"
95
+ )
96
+ end.compact
97
+ end
98
+
99
+ def detect_service_account_relationships(project, service_account_index)
100
+ project.resources.filter { |resource| resource.type.match?(/_iam_(member|binding)\z/) }.filter_map do |resource|
101
+ members = Array(resource.attribute_ruby("members")) + [resource.attribute_text("member")]
102
+ target_sa = members.compact.lazy.map { |member| find_service_account_in_member(member.to_s, service_account_index) }.find(&:itself)
103
+ next unless target_sa
104
+ next if target_sa.project == project
105
+
106
+ Model::Relationship.new(
107
+ source: resource,
108
+ target: target_sa,
109
+ type: :iam,
110
+ detail: "IAM member uses service account from #{target_sa.project.name}"
111
+ )
112
+ end
113
+ end
114
+
115
+ def detect_shared_name_relationships(project, string_name_index)
116
+ project.resources.flat_map do |resource|
117
+ flattened_attribute_strings(resource.attributes).filter_map do |text|
118
+ next unless candidate_shared_name?(text)
119
+
120
+ target = string_name_index[text].find { |candidate| candidate.project != project }
121
+ next unless target
122
+
123
+ rel_type = target.project.shared? ? :shared_resource : relation_type_for_target(target)
124
+ Model::Relationship.new(
125
+ source: resource,
126
+ target: target,
127
+ type: rel_type,
128
+ detail: "String reference uses shared resource name #{text}"
129
+ )
130
+ end
131
+ end
132
+ end
133
+
134
+ def find_service_account_in_member(member, service_account_index)
135
+ service_account_index.each do |key, resource|
136
+ return resource if key && member.include?(key)
137
+ end
138
+ nil
139
+ end
140
+
141
+ def reference_identifier(ref)
142
+ return nil unless ref
143
+
144
+ parts = ref.split(".")
145
+ return nil if parts.length < 2
146
+
147
+ parts[0, 2].join(".")
148
+ end
149
+
150
+ def relation_type_for_target(target)
151
+ target.category == :networking ? :network : :reference
152
+ end
153
+
154
+ def names_for_resource(resource)
155
+ [
156
+ resource.attribute_text("name"),
157
+ resource.attribute_text("dataset_id"),
158
+ resource.attribute_text("account_id"),
159
+ resource.name
160
+ ].compact.uniq
161
+ end
162
+
163
+ def shareable_name_target?(resource)
164
+ %i[networking iam messaging].include?(resource.category) || resource.type == "google_pubsub_topic"
165
+ end
166
+
167
+ def flattened_attribute_strings(value)
168
+ case value
169
+ when Hash
170
+ value.values.flat_map { |child| flattened_attribute_strings(child) }
171
+ when Array
172
+ value.flat_map { |child| flattened_attribute_strings(child) }
173
+ else
174
+ [Parser::ExpressionInspector.to_ruby(value)]
175
+ end.map(&:to_s).reject(&:empty?)
176
+ end
177
+
178
+ def candidate_shared_name?(text)
179
+ return false if text.start_with?("http://", "https://")
180
+ return false if text.match?(/\A\d+\z/)
181
+
182
+ text.length >= 3
183
+ end
184
+
185
+ def deduplicate(relationships)
186
+ relationships.each_with_object([]) do |relationship, unique|
187
+ unique << relationship unless unique.any? { |existing| existing.key == relationship.key }
188
+ end
189
+ end
190
+ end
191
+ end
192
+ end
@@ -0,0 +1,78 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Docsterra
4
+ module Analyzer
5
+ class NetworkAnalyzer
6
+ NETWORK_TYPES = %w[
7
+ google_compute_network
8
+ google_compute_subnetwork
9
+ google_compute_firewall
10
+ google_compute_global_address
11
+ google_compute_address
12
+ google_compute_forwarding_rule
13
+ google_compute_global_forwarding_rule
14
+ google_compute_router
15
+ google_compute_router_nat
16
+ google_dns_managed_zone
17
+ google_dns_record_set
18
+ ].freeze
19
+
20
+ LOAD_BALANCER_TYPES = %w[
21
+ google_compute_forwarding_rule
22
+ google_compute_global_forwarding_rule
23
+ ].freeze
24
+
25
+ def analyze(resources)
26
+ vpcs = resources.select { |resource| resource.type == "google_compute_network" }
27
+ subnets = resources.select { |resource| resource.type == "google_compute_subnetwork" }
28
+ firewalls = resources.select { |resource| resource.type == "google_compute_firewall" }
29
+ load_balancers = resources.select { |resource| LOAD_BALANCER_TYPES.include?(resource.type) }
30
+ endpoints = resources.reject { |resource| NETWORK_TYPES.include?(resource.type) }
31
+ links = build_links(resources, vpcs)
32
+
33
+ Model::Network.new(
34
+ vpcs: vpcs,
35
+ subnets: subnets,
36
+ firewall_rules: firewalls,
37
+ load_balancers: load_balancers,
38
+ endpoints: endpoints,
39
+ links: links
40
+ )
41
+ end
42
+
43
+ private
44
+
45
+ def build_links(resources, vpcs)
46
+ vpc_lookup = vpcs.each_with_object({}) do |vpc, index|
47
+ index[vpc.identifier] = vpc
48
+ end
49
+
50
+ resources.flat_map do |resource|
51
+ refs = Array(resource.references)
52
+ refs.filter_map do |ref|
53
+ target = target_from_reference(ref, vpc_lookup)
54
+ next unless target
55
+
56
+ {
57
+ source: resource,
58
+ target: target,
59
+ type: :network,
60
+ detail: "references #{ref}"
61
+ }
62
+ end
63
+ end.uniq
64
+ end
65
+
66
+ def target_from_reference(ref, vpc_lookup)
67
+ return nil unless ref
68
+
69
+ # ex: google_compute_network.shared.id -> google_compute_network.shared
70
+ parts = ref.split(".")
71
+ return nil if parts.length < 2
72
+
73
+ identifier = parts[0, 2].join(".")
74
+ vpc_lookup[identifier]
75
+ end
76
+ end
77
+ end
78
+ end
@@ -0,0 +1,95 @@
1
+ # frozen_string_literal: true
2
+
3
+ module Docsterra
4
+ module Analyzer
5
+ class ResourceAnalyzer
6
+ def initialize(registry: Parser::ResourceRegistry.new)
7
+ @registry = registry
8
+ end
9
+
10
+ def analyze(project)
11
+ project.resources = build_resources(project.resource_blocks, project, kind: :resource)
12
+ project.data_sources = build_resources(project.data_blocks, project, kind: :data)
13
+ project.resources
14
+ end
15
+
16
+ private
17
+
18
+ def build_resources(entries, project, kind:)
19
+ entries.map do |entry|
20
+ block = entry.block
21
+ resource_type = block.labels[0]
22
+ name = block.labels[1] || block.labels[0]
23
+ attributes = build_attribute_tree(block.body)
24
+ refs = collect_references_from_attributes(attributes)
25
+ meta = extract_meta(attributes)
26
+
27
+ Model::Resource.new(
28
+ type: resource_type,
29
+ name: name,
30
+ attributes: attributes,
31
+ project: project,
32
+ references: refs + Array(meta[:depends_on]),
33
+ meta: meta,
34
+ source_file: entry.file,
35
+ comment: block.comment,
36
+ kind: kind,
37
+ display_name: @registry.display_name_for(resource_type),
38
+ category: @registry.category_for(resource_type),
39
+ key_attributes: @registry.key_attributes_for(resource_type)
40
+ )
41
+ end
42
+ end
43
+
44
+ def build_attribute_tree(items)
45
+ tree = {}
46
+
47
+ Array(items).each do |item|
48
+ case item
49
+ when Parser::AST::Attribute
50
+ tree[item.key] = item.value
51
+ when Parser::AST::Block
52
+ block_hash = build_attribute_tree(item.body)
53
+ block_hash["__labels"] = item.labels unless item.labels.nil? || item.labels.empty?
54
+ tree[item.type] ||= []
55
+ tree[item.type] << block_hash
56
+ end
57
+ end
58
+
59
+ tree
60
+ end
61
+
62
+ def collect_references_from_attributes(value)
63
+ case value
64
+ when Hash
65
+ value.values.flat_map { |child| collect_references_from_attributes(child) }.uniq
66
+ when Array
67
+ value.flat_map { |child| collect_references_from_attributes(child) }.uniq
68
+ else
69
+ Parser::ExpressionInspector.collect_references(value)
70
+ end
71
+ end
72
+
73
+ def extract_meta(attributes)
74
+ {
75
+ count: attributes["count"],
76
+ for_each: attributes["for_each"],
77
+ depends_on: normalize_depends_on(attributes["depends_on"])
78
+ }.compact
79
+ end
80
+
81
+ def normalize_depends_on(node)
82
+ return [] if node.nil?
83
+
84
+ case node
85
+ when Parser::AST::ListExpr
86
+ node.elements.map { |element| Parser::ExpressionInspector.to_text(element) }.compact
87
+ when Parser::AST::RawExpr
88
+ [node.text]
89
+ else
90
+ [Parser::ExpressionInspector.to_text(node)]
91
+ end
92
+ end
93
+ end
94
+ end
95
+ end