iev 0.3.9 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 8ad06b28a53df39790f6b46c5e1ca4cec92e294d3331b772e1977be2759ad53f
-  data.tar.gz: 688d3afc08ea75f5a49f24c3b13264414871015c5e795a0e14b8a75b280d35f2
+  metadata.gz: 382780f439ab289a11d44053962bb7dbf043688fd68cb2328b8746429c19c6d3
+  data.tar.gz: 72151b973c479bfc82e61b3d0da4648f11ac54e9344b8c71e3bbf5d91a3a3c3b
 SHA512:
-  metadata.gz: 19cb09cf27f99308cd6c67778c7668896656f153b7ae89b114fed7219786d9bd56dcd3444c669dec1961aa66c0be84601012b45829f5fef21da9c819b1f838b1
-  data.tar.gz: 4e6c952bfebc8a4da93f1573634520f7ec6f868641a71980fe9ba2c5e2abe784fddea50fc330c74ed44636938ce881bb8726aba2d011db59a4720d8c56762508
+  metadata.gz: d36fad5c1d853d39ffa5d8d67e98ecaedd3c40206237706ccb177ff14957ebea9616420bd41878b84ac34cdab1e9c5c7890143024b0a3070fb6c8a994d769518
+  data.tar.gz: 8c5e7f0a9510fd562fc92db4ba225e5f9b7580e0e703525ec062b3d404ab57c804da824f94af0e3047738033577aaae8394970d47c349b1ac6b1979f7fbfa02d

data/.github/workflows/rake.yml

@@ -2,8 +2,6 @@ name: rake
 
 permissions:
   contents: read
-  actions: write         # Add this if you need to approve or trigger workflows
-
 
 on:
   push:

data/.github/workflows/release.yml

@@ -2,6 +2,8 @@ name: release
 
 permissions:
   contents: write
+  packages: write
+  id-token: write
 
 on:
   workflow_dispatch:
@@ -22,4 +24,4 @@ jobs:
       next_version: ${{ github.event.inputs.next_version }}
     secrets:
       rubygems-api-key: ${{ secrets.GLOSSARIST_CI_RUBYGEMS_API_KEY }}
-      pat_token: ${{ secrets.GITHUB_TOKEN }}
+      pat_token: ${{ secrets.GLOSSARIST_CI_PAT_TOKEN }}

data/.gitignore

@@ -14,4 +14,6 @@
 .rspec_status
 .vscode/
 
-concepts/
+/concepts/
+Gemfile.lock
+TODO*

data/CLAUDE.md

@@ -0,0 +1,50 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Build and Test Commands
+
+- Run all tests: `bundle exec rake` or `bundle exec rspec`
+- Run a single test file: `bundle exec rspec spec/iev/term_builder_spec.rb`
+- Run a single test example: `bundle exec rspec spec/iev/term_builder_spec.rb:42`
+- Lint: `bundle exec rubocop`
+- CI uses the shared `metanorma/ci` generic-rake workflow
+
+## Architecture
+
+This is a Ruby gem (`iev`) for working with the International Electrotechnical Vocabulary (IEV) from IEC Electropedia. It is part of the Glossarist ecosystem and depends on the `glossarist` gem for concept modeling.
+
+### Two Main Data Flows
+
+**1. Fetching terms (API usage):** `Iev.get(code, lang)` or `Iev.fetch_concept(code)`
+- `DataSource` checks a local path (`IEV_DATA_PATH` env var) for YAML concept files, then falls back to fetching from a remote GitHub repo (`glossarist/glossarist-data-iev`), with file-based caching in `IEV_CACHE_DIR` (defaults to system tmpdir).
+- `Db` wraps `DataSource` with a two-tier cache (global + local) using `DbCache`, which stores versioned XML files on disk.
+
+**2. Converting Excel exports (CLI usage):** via `exe/iev` (Thor-based)
+- `xlsx2yaml`: Excel → `DbWriter` (in-memory SQLite) → `TermBuilder` (row-by-row) → `Glossarist::LocalizedConcept` objects → YAML concept files
+- `xlsx2db`: Excel → `DbWriter` → SQLite file
+- `db2yaml`: SQLite → `TermBuilder` → YAML concept files
+
+### Key Modules
+
+- `TermBuilder` — the core converter that turns a spreadsheet row into a `Glossarist::LocalizedConcept`. Handles definition splitting (notes/examples extraction), term designation parsing, and source parsing.
+- `SourceParser` — parses the SOURCE column from IEV exports, normalizing references (CEI→IEC, UIT→ITU, etc.) and extracting ref/clause/relationship using extensive regex matching.
+- `TermAttrsParser` — parses the TERMATTRIBUTE field (gender, plurality, part of speech, geographical area, abbreviations).
+- `SupersessionParser` — parses the REPLACES field for deprecated term relationships.
+- `Converter::MathmlToAsciimath` — converts MathML markup to AsciiMath using Plurimath.
+- `Utilities` — HTML processing: converts IEV cross-references (`<a href=IEV...>`) to `{{term, IEV:code}}` format, handles figures, images, bold tags, and newline normalization.
+
+### Configuration
+
+`Iev.configure` yields a `Config` object with:
+- `data_path` — local path to YAML concept files (env: `IEV_DATA_PATH`)
+- `cache_dir` — cache directory (env: `IEV_CACHE_DIR`, default: system tmpdir)
+- `remote_base_url` — base URL for remote concept YAML fetching
+
+## Key Conventions
+
+- Ruby >= 3.1.0 required
+- `plurimath` and `unitsml` are optional runtime dependencies — loaded with `rescue LoadError`, so the `DataSource`/`Db` APIs work without them
+- The IEV Excel export format is specific to IEC-internal use; column structure is documented in README.adoc
+- Language codes: the spreadsheet uses ISO 639-1 (2-char like "en"), internally converted to ISO 639-2/3 (3-char like "eng") via `Iso639Code` and `DataConversions`
+- `DataConversions` is a refinement (`using DataConversions`) that adds `.sanitize` and `.decode_html` methods to String

data/Gemfile

@@ -2,8 +2,11 @@
 
 source "https://rubygems.org"
 
+gem "benchmark"
 gem "canon"
 gem "openssl"
+gem "w3c_api", "~> 0.2.0"
+gem "lutaml-model", github: "lutaml/lutaml-model", branch: "main"
 gem "rake"
 gem "rspec"
 gem "rubocop"

data/README.adoc

@@ -6,6 +6,13 @@ image:https://github.com/glossarist/iev/workflows/rake/badge.svg["Build Status",
 image:https://img.shields.io/github/issues-pr-raw/glossarist/iev.svg["Pull Requests", link="https://github.com/glossarist/iev/pulls"]
 image:https://img.shields.io/github/commits-since/glossarist/iev/latest.svg["Commits since latest",link="https://github.com/glossarist/iev/releases"]
 
+[WARNING]
+--
+As of 2026-02-10, Electropedia is behind AWS WAF, which blocks simple HTTP clients.
+Scraping now uses Ferrum (headless Chrome via DevTools Protocol) to handle the WAF challenge.
+Chrome/Chromium must be installed for scraping to work.
+--
+
 
 == Purpose
 
@@ -54,24 +61,26 @@ $ gem install iev
 
 The gem comes with the `iev` executable, which provides the following commands:
 
-`iev xlsx2yaml FILE`::
-Converts Excel IEV exports to YAMLs.
+`iev export FILE -o OUTPUT_DIR`::
+Exports IEV data to Glossarist YAML format.
+Supports both Excel (`.xlsx`/`.xls`) and SQLite (`.sqlite3`/`.sqlite`/`.db`) input files.
+Format is detected automatically from the file extension.
 
 `iev xlsx2db FILE`::
 Imports Excel to SQLite database.
 
-`iev db2yaml DB_FILE`:::
-Exports SQLite to IEV YAMLs.
+`iev fetch CODE`::
+Fetches a single IEV concept and outputs YAML to stdout.
 
 WARNING: The IEV XLSX export files can only be obtained from the IEC
 Electropedia administrator.
 
 
-=== Fetching IEV terms from Electropedia
+=== Fetching IEV terms from cached data
 
 [source, ruby]
 ----
-# Get term
+# Get term designation (from local YAML or GitHub remote)
 
 Iev.get("103-01-02", "en")
 => "functional"
@@ -83,26 +92,67 @@ Iev.get("111-11-11", "en")
 # If language not found
 Iev.get("103-01-02", "eee")
 => nil
+
+# Fetch full concept data (all languages)
+Iev.fetch_concept("103-01-02")
+=> { "id" => "103-01-02", "data" => { ... } }
+
+# Fetch localized term data
+Iev.fetch_term("103-01-02", "en")
+=> { "term" => "functional", ... }
+----
+
+=== Scraping IEV terms from Electropedia
+
+Requires Chrome/Chromium installed. Uses Ferrum (headless Chrome) to bypass AWS WAF.
+
+[source, ruby]
+----
+# Scrape concept data directly from electropedia.org
+Iev.scrape_concept("103-01-02")
+=> { "id" => "103-01-02", "data" => { "identifier" => "103-01-02", "localized_concepts" => { "eng" => { ... }, ... } } }
+
+# Custom browser options (e.g., headless mode, window size)
+scraper = Iev::Scraper.new(browser_opts: { headless: true, window_size: [1280, 800] })
+concept = scraper.fetch_concept("103-01-02")
 ----
 
 
-=== Converting IEV Excel exports to a Glossarist dataset
+=== Converting IEV data to a Glossarist dataset
 
-Run the following command:
+The `export` command converts an IEV Excel export or SQLite database into
+Glossarist YAML concept files:
 
 [source,sh]
 ----
-$ iev xlsx2yaml [termbase.xlsx]
+# From an Excel export
+$ iev export termbase.xlsx -o /path/to/output
+
+# From a SQLite database
+$ iev export termbase.sqlite3 -o /path/to/output
+
+# With filters
+$ iev export termbase.xlsx -o /output --only-concepts "103-%" --only-languages "en,fr"
 ----
 
-Where,
+The output directory will contain a `concepts/` subdirectory with Glossarist
+concept and localized concept YAML files.
 
-* the Glossarist termbase YAML files will be created at `[termbase.yaml]` at
-  the current working directory;
+You can also use the `Iev::Exporter` class programmatically:
 
-* ConceptYAML files for each concept (each concept identified by a unique "`Term-ID`")
-  are created under the `concepts/` directory (based on the current working directory).
-  i.e. `concepts/concept-338.yaml`.
+[source,ruby]
+----
+# Export from Excel
+Iev::Exporter.new("termbase.xlsx", output_dir: "/path/to/output").export
+
+# Export from SQLite with filters
+collection = Iev::Exporter.new("termbase.sqlite3",
+  output_dir: "/path/to/output",
+  only_concepts: "103-%",
+  only_languages: "en,fr",
+).export
+# collection is a Glossarist::ManagedConceptCollection
+----
 
 
 == Structure of the IEV Excel export

data/exe/iev

@@ -2,6 +2,17 @@
 # frozen_string_literal: true
 
 require_relative "../lib/iev"
+
+# CLI dependencies — loaded eagerly for the conversion pipeline
+require "benchmark"
+require "creek"
+require "glossarist"
+require "nokogiri"
+require "relaton"
+require "relaton_bib"
+require "sequel"
+require "thor"
+
 require_relative "../lib/iev/cli"
 
 Iev::Cli.start(ARGV)

data/iev.gemspec

@@ -19,16 +19,17 @@ Gem::Specification.new do |spec|
   spec.bindir = "exe"
   spec.executables = spec.files.grep(%r{^exe/}) { |f| File.basename(f) }
   spec.require_paths = ["lib"]
-  spec.required_ruby_version = Gem::Requirement.new(">= 3.1.0")
+  spec.required_ruby_version = Gem::Requirement.new(">= 3.2.0")
 
   spec.add_dependency "creek", "~> 2.6"
   spec.add_dependency "glossarist", ">= 2.3.0"
-  spec.add_dependency "mechanize", "~> 2.10"
-  spec.add_dependency "nokogiri", "~> 1.17"
+  spec.add_dependency "ferrum", "~> 0.15"
+  spec.add_dependency "nokogiri", "~> 1.19"
   spec.add_dependency "plurimath"
+  spec.add_dependency "lutaml-model", "~> 0.8.0"
   spec.add_dependency "relaton", "~> 1.18"
   spec.add_dependency "sequel", "~> 5.40"
-  spec.add_dependency "sqlite3", "~> 1.7.0"
+  spec.add_dependency "sqlite3", "~> 1.7"
   spec.add_dependency "thor", "~> 1.0"
   spec.add_dependency "unitsml"
 end

data/lib/iev/cli/command.rb

@@ -8,102 +8,145 @@ module Iev
     class Command < Thor
       include CommandHelper
 
+      desc "export FILE", "Export IEV data to Glossarist YAML format"
+      long_desc <<~DESC
+        Exports IEV data from an Excel (.xlsx/.xls) or SQLite (.sqlite3/.sqlite/.db)
+        file to Glossarist YAML concept files.
+
+        The input format is detected automatically from the file extension.
+      DESC
+      option :output, desc: "Output directory", aliases: :o, default: Dir.pwd
+      option :only_concepts,
+             desc: "Only process concepts with IEVREF matching this pattern " \
+                   "(SQL LIKE wildcards: % and _)"
+      option :only_languages,
+             desc: "Only export these languages, skip concepts which aren't " \
+                   "translated to any of them (comma-separated list, language " \
+                   "codes must be as in spreadsheet)"
+      option :progress, type: :boolean,
+                        desc: "Enables or disables progress indicator. By default disabled " \
+                              "when 'CI' environment variable is set and enabled otherwise"
+      option :profile, type: :boolean, default: false,
+                       desc: "Generates profiler reports for this program, requires ruby-prof"
+      option :debug_term_attributes, type: :boolean, default: false,
+                                     desc: "Enables debug messages about term attributes recognition"
+      option :debug_sources, type: :boolean, default: false,
+                             desc: "Enables debug messages about authoritative sources recognition"
+      option :debug_relaton, type: :boolean, default: false,
+                             desc: "Enables debug messages about Relaton integration"
+      def export(file)
+        handle_generic_options(options)
+
+        Iev::Exporter.new(
+          file,
+          output_dir: options[:output],
+          only_concepts: options[:only_concepts],
+          only_languages: options[:only_languages],
+        ).export
+
+        info "Done!"
+      end
+
       desc "xlsx2yaml FILE", "Converts Excel IEV exports to YAMLs."
+      option :output, desc: "Output directory", aliases: :o, default: Dir.pwd
+      option :only_concepts,
+             desc: "Only process concepts with IEVREF matching this argument, " \
+                   "'%' and '_' wildcards are supported and have meaning as in SQL " \
+                   "LIKE operator"
+      option :only_languages,
+             desc: "Only export these languages, skip concepts which aren't " \
+                   "translated to any of them (comma-separated list, language " \
+                   "codes must be as in spreadsheet)"
+      option :progress, type: :boolean,
+                        desc: "Enables or disables progress indicator. By default disabled " \
+                              "when 'CI' environment variable is set and enabled otherwise"
+      option :profile, type: :boolean, default: false,
+                       desc: "Generates profiler reports for this program, requires ruby-prof"
+      option :debug_term_attributes, type: :boolean, default: false
+      option :debug_sources, type: :boolean, default: false
+      option :debug_relaton, type: :boolean, default: false
       def xlsx2yaml(file)
         handle_generic_options(options)
-        db = Sequel.sqlite
-        DbWriter.new(db).import_spreadsheet(file)
-        ds = filter_dataset(db, options)
-        collection = build_collection_from_dataset(ds)
-        save_collection_to_files(collection, options[:output])
+
+        Iev::Exporter.new(
+          file,
+          output_dir: options[:output],
+          only_concepts: options[:only_concepts],
+          only_languages: options[:only_languages],
+        ).export
+
+        summary
+      end
+
+      desc "db2yaml DB_FILE", "Exports SQLite to IEV YAMLs."
+      option :output, desc: "Output directory", aliases: :o, default: Dir.pwd
+      option :only_concepts,
+             desc: "Only process concepts with IEVREF matching this argument, " \
+                   "'%' and '_' wildcards are supported and have meaning as in SQL " \
+                   "LIKE operator"
+      option :only_languages,
+             desc: "Only export these languages, skip concepts which aren't " \
+                   "translated to any of them (comma-separated list, language " \
+                   "codes must be as in spreadsheet)"
+      option :progress, type: :boolean,
+                        desc: "Enables or disables progress indicator. By default disabled " \
+                              "when 'CI' environment variable is set and enabled otherwise"
+      option :profile, type: :boolean, default: false,
+                       desc: "Generates profiler reports for this program, requires ruby-prof"
+      option :debug_term_attributes, type: :boolean, default: false
+      option :debug_sources, type: :boolean, default: false
+      option :debug_relaton, type: :boolean, default: false
+      def db2yaml(dbfile)
+        handle_generic_options(options)
+
+        Iev::Exporter.new(
+          dbfile,
+          output_dir: options[:output],
+          only_concepts: options[:only_concepts],
+          only_languages: options[:only_languages],
+        ).export
+
         summary
       end
 
       desc "xlsx2db FILE", "Imports Excel to SQLite database."
+      option :output, desc: "Output file", aliases: :o,
+                      default: File.join(Dir.pwd, "concepts.sqlite3")
+      option :progress, type: :boolean,
+                        desc: "Enables or disables progress indicator. By default disabled " \
+                              "when 'CI' environment variable is set and enabled otherwise"
+      option :profile, type: :boolean, default: false,
+                       desc: "Generates profiler reports for this program, requires ruby-prof"
       def xlsx2db(file)
         handle_generic_options(options)
-        # Instantiating an in-memory db and dumping it later is faster than
-        # just working on file db.
         db = Sequel.sqlite
         DbWriter.new(db).import_spreadsheet(file)
         save_db_to_file(db, options[:output])
         summary
       end
 
-      desc "db2yaml DB_FILE", "Exports SQLite to IEV YAMLs."
-      def db2yaml(dbfile)
-        handle_generic_options(options)
-        db = Sequel.sqlite(dbfile)
-        ds = filter_dataset(db, options)
-        collection = build_collection_from_dataset(ds)
-        save_collection_to_files(collection, options[:output])
-        summary
+      desc "fetch CODE", "Fetch an IEV concept and output YAML to stdout."
+      option :scrape, type: :boolean, default: false,
+                      desc: "Scrape from Electropedia instead of using cached data"
+      def fetch(code)
+        raw = if options[:scrape]
+                Scraper.new.fetch_concept(code)
+              else
+                DataSource.fetch_concept(code)
+              end
+
+        unless raw
+          warn "IEV: concept #{code} not found."
+          exit 1
+        end
+
+        concept = build_concept_from_raw(code, raw)
+        print_concept_grouped_yaml(concept)
       end
 
       def self.exit_on_failure?
         true
       end
-
-      # Options must be declared at the bottom because Thor must have commands
-      # defined in advance.
-
-      def self.shared_option(name, methods:, **kwargs)
-        [*methods].each { |m| option name, for: m, **kwargs }
-      end
-
-      shared_option :only_concepts,
-                    desc: "Only process concepts with IEVREF matching this argument, " \
-                          "'%' and '_' wildcards are supported and have meaning as in SQL " \
-                          "LIKE operator",
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :only_languages,
-                    desc: "Only export these languages, skip concepts which aren't " \
-                          "translated to any of them (comma-separated list, language " \
-                          "codes must be as in spreadsheet)",
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :output,
-                    desc: "Output directory",
-                    aliases: :o,
-                    default: Dir.pwd,
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :output,
-                    desc: "Output file",
-                    aliases: :o,
-                    default: File.join(Dir.pwd, "concepts.sqlite3"),
-                    methods: :xlsx2db
-
-      shared_option :progress,
-                    type: :boolean,
-                    desc: "Enables or disables progress indicator. By default disabled " \
-                          "when 'CI' environment variable is set and enabled otherwise",
-                    methods: %i[xlsx2yaml xlsx2db db2yaml]
-
-      shared_option :debug_term_attributes,
-                    desc: "Enables debug messages about term attributes recognition",
-                    type: :boolean,
-                    default: false,
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :debug_sources,
-                    desc: "Enables debug messages about authoritative sources recognition",
-                    type: :boolean,
-                    default: false,
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :debug_relaton,
-                    desc: "Enables debug messages about Relaton integration",
-                    type: :boolean,
-                    default: false,
-                    methods: %i[xlsx2yaml db2yaml]
-
-      shared_option :profile,
-                    desc: "Generates profiler reports for this program, requires ruby-prof",
-                    type: :boolean,
-                    default: false,
-                    methods: %i[xlsx2yaml xlsx2db db2yaml]
     end
   end
 end

data/lib/iev/cli/command_helper.rb

@@ -10,17 +10,6 @@ module Iev
 
       protected
 
-      def save_collection_to_files(collection, output_dir)
-        Profiler.measure("writing-yamls") do
-          info "Writing concepts to files..."
-          path = File.expand_path("./concepts", output_dir)
-          FileUtils.mkdir_p(path)
-          collection.save_to_files(path)
-        end
-      end
-
-      # NOTE: Implementation examples here:
-      # https://www.rubydoc.info/github/luislavena/sqlite3-ruby/SQLite3/Backup
       def save_db_to_file(src_db, dbfile)
         info "Saving database to a file..."
         src_db.synchronize do |src_conn|
@@ -35,13 +24,6 @@ module Iev
         info "Done!"
       end
 
-      def collection_file_path(file, output_dir)
-        output_dir.join(Pathname.new(file).basename.sub_ext(".yaml"))
-      end
-
-      # Handles various generic options, e.g. detailed debug switches.
-      # Assigns some global variables accordingly, so these settings are
-      # available throughout the program.
       def handle_generic_options(options)
         $IEV_PROFILE = options[:profile]
         $IEV_PROGRESS = options.fetch(:progress, !ENV["CI"])
@@ -54,31 +36,68 @@ module Iev
         end
       end
 
-      def filter_dataset(db, options)
-        query = db[:concepts]
+      def build_concept_from_raw(code, raw)
+        concept = Glossarist::ManagedConcept.of_yaml(
+          "id" => code,
+          "data" => { "id" => code },
+        )
 
-        if options[:only_concepts]
-          query = query.where(Sequel.ilike(:ievref,
-                                           options[:only_concepts]))
+        localized = extract_localized(raw)
+        localized.each do |lang, entry|
+          l10n = build_localized_concept(code, lang, entry)
+          concept.add_l10n(l10n)
         end
 
-        query = query.where(language: options[:only_languages].split(",")) if options[:only_languages]
+        concept
+      end
+
+      def extract_localized(raw)
+        # Scraper format: raw["data"]["localized_concepts"] => {lang => {term, definition}}
+        data = raw["data"]
+        if data && data["localized_concepts"]
+          return data["localized_concepts"]
+        end
+
+        # DataSource format: raw itself, keys are lang codes
+        raw.each_with_object({}) do |(k, v), h|
+          h[k] = v if v.is_a?(Hash) && v["terms"]
+        end
+      end
+
+      def build_localized_concept(code, lang, entry)
+        terms = if entry["terms"]
+                  entry["terms"].map { |t| Glossarist::Designation::Expression.new(**t.transform_keys(&:to_sym)) }
+                else
+                  [Glossarist::Designation::Expression.new(
+                    designation: entry["term"],
+                    normative_status: "preferred",
+                  )]
+                end
+
+        cd = Glossarist::ConceptData.new
+        cd.id = code
+        cd.language_code = lang
+        cd.terms = terms
+
+        definition = entry["definition"]
+        if definition
+          content = definition.is_a?(String) ? definition : definition
+          cd.definition = [Glossarist::DetailedDefinition.new(content: content)]
+        end
 
-        query
+        l10n = Glossarist::LocalizedConcept.new
+        l10n.data = cd
+        l10n.id = code
+        l10n
       end
 
-      def build_collection_from_dataset(dataset)
-        Profiler.measure("building-collection") do
-          Glossarist::ManagedConceptCollection.new.tap do |concept_collection|
-            dataset.each do |row|
-              term = TermBuilder.build_from(row)
-              next unless term
-
-              concept = concept_collection.fetch_or_initialize(term.id)
-              concept.add_l10n(term)
-            end
-          end
+      def print_concept_grouped_yaml(concept)
+        content = []
+        content << concept.to_yaml
+        concept.localized_concepts.each_key do |lang|
+          content << concept.localization(lang).to_yaml
         end
+        puts content.join("\n")
       end
     end
   end

data/lib/iev/config.rb

@@ -0,0 +1,31 @@
+# frozen_string_literal: true
+
+require "tmpdir"
+
+module Iev
+  class Config
+    DEFAULT_REMOTE_BASE_URL = "https://raw.githubusercontent.com/glossarist/glossarist-data-iev/main/concepts"
+
+    attr_accessor :data_path, :cache_dir, :remote_base_url
+
+    def initialize
+      @data_path = ENV["IEV_DATA_PATH"]
+      @cache_dir = ENV["IEV_CACHE_DIR"] || File.join(Dir.tmpdir, "iev-cache")
+      @remote_base_url = DEFAULT_REMOTE_BASE_URL
+    end
+  end
+
+  class << self
+    def config
+      @config ||= Config.new
+    end
+
+    def configure
+      yield(config) if block_given?
+    end
+
+    def reset_config!
+      @config = nil
+    end
+  end
+end