relaton-xsf 1.20.0 → 2.0.0.pre.alpha.1

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6ae057ad0c2acaaf2543cd58bf877e787df3c9b402fb5ab18deea706426a4665
-  data.tar.gz: 7e2267c02b7e2e94c3dc49f11b94757e021cee1e0b05cc38af70f6488fa0d851
+  metadata.gz: f37b079bdc2af9ac10f413caecc791b69360a948628e0bdbf11a19e32648f53b
+  data.tar.gz: c2d4606232e12b41bd8b6673f0ae67ff101b37487ecb1e36247ed0521f66d8bc
 SHA512:
-  metadata.gz: d85e7fab9e5e36b81bf349244cc9cb289a3a5609bc6bf153800da9a3f687f3dd1ba38eb8d55c56db1916abf22dea97b3dde5609218eab5d1ce638e224c93b498
-  data.tar.gz: 7987e50e1904656ac8f2fa0491faf0efc4cf19594a60822c7af81330d04581a1696b8b34b4316b8f2cdfb82d36b0aeed543daed5c9745edbc736927feb521457
+  metadata.gz: 4f35da54cefed765b7b810959bbbb44ae982b89eed78516f17822090f6ac00a7767ab8dbb3707de3553e7d92e7fa034441a930eec7db0fc9a887e1cb4bb29f55
+  data.tar.gz: 32607532cadab4c5be0a4d391502e5c2b0621ae1c4f3787ab95d4096de2ff5d91a181cf62f7cd88a8de90372a0ae221651398448f4790ce75f23f12db2ec1576

data/.rubocop.yml

@@ -7,6 +7,6 @@ require: rubocop-rails
 inherit_from:
   - https://raw.githubusercontent.com/riboseinc/oss-guides/master/ci/rubocop.yml
 AllCops:
-  TargetRubyVersion: 2.7
+  TargetRubyVersion: 3.2
 Rails:
   Enabled: false

data/CLAUDE.md

@@ -0,0 +1,58 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## What is relaton-xsf?
+
+A Ruby gem for bibliographic retrieval of XMPP XEP (XMPP Extension Protocol) specifications. Part of the Relaton family of gems. Fetches data from https://xmpp.org/extensions/refs/ and the relaton-data-xsf GitHub repository.
+
+## Commands
+
+- `bundle exec rspec` — run all tests
+- `bundle exec rspec spec/relaton/xsf/processor_spec.rb` — run a single spec file
+- `bundle exec rubocop` — lint
+- `bundle exec rubocop -a` — lint with auto-fix
+- `bin/console` — interactive console with gem loaded
+
+## Architecture
+
+Namespace: `Relaton::Xsf` (under `lib/relaton/xsf/`). Branch `lutaml-integration` uses the new nested namespace (not the old `RelatonXsf`).
+
+Key classes and their base classes from relaton-core:
+
+| Class | Base | Role |
+|---|---|---|
+| `Processor` | `Relaton::Core::Processor` | Plugin entry point for relaton registry |
+| `Bibliography` | Module (extends self) | Search & get interface (`search`, `get`) |
+| `HitCollection` | `Relaton::Core::HitCollection` | Collection of search results |
+| `Hit` | `Relaton::Core::Hit` | Single result; lazy-loads YAML from GitHub |
+| `DataFetcher` | `Relaton::Core::DataFetcher` | Crawls xmpp.org, parses BibXML, saves docs |
+| `Item` / `Bibitem` / `Bibdata` | `Relaton::Bib::Item` | Bibliographic item models (lutaml-model based) |
+
+Data flow: `Processor#get` → `Bibliography.get` → `HitCollection.search` → `Hit#item` → fetches YAML → `Relaton::Bib::Item.from_yaml`
+
+DataFetcher flow: Crawls `https://xmpp.org/extensions/refs/`, parses each XML ref via `Relaton::Bib::Converter::BibXml.to_item`, sets `ext.flavor = "xsf"`, saves to disk.
+
+Constants: `INDEXFILE = "index-v1"`, `GHDATA_URL` points to relaton-data-xsf `data-v2` branch.
+
+## Testing
+
+- RSpec with VCR cassettes (`spec/vcr_cassettes/`) for HTTP interactions
+- WebMock disables all external network connections
+- Fixtures in `spec/fixtures/` (item.yaml, bibdata.xml, bibitem.xml)
+- Round-trip tests verify YAML→Item→YAML and XML→Item→XML fidelity
+- `DataFetcher` is lazily required — specs that test it must `require "relaton/xsf/data_fetcher"` explicitly
+- Same for `Processor` — `require "relaton/xsf/processor"`
+
+## Key dependencies
+
+- `relaton-core` — abstract base classes (Processor, HitCollection, Hit, DataFetcher)
+- `relaton-bib` — bibliographic models, XML/YAML serialization (lutaml-model based)
+- `relaton-index` — index management for quick document lookups
+- `mechanize` — HTTP fetching and HTML parsing
+
+## Style
+
+- RuboCop with relaton shared config (inherits from riboseinc/oss-guides)
+- Target Ruby version: 3.1
+- Logging via `Relaton::Xsf::Util` (extends `Relaton::Bib::Util`, PROGNAME = "relaton-xsf")

data/README.adoc

@@ -1,4 +1,4 @@
-= Relaton-XSF: bibliographic retrieval of XMPP XEP specifications
+= Relaton::XSF: bibliographic retrieval of XMPP XEP specifications
 
 image:https://img.shields.io/gem/v/relaton-xsf.svg["Gem Version", link="https://rubygems.org/gems/relaton-xsf"]
 image:https://github.com/relaton/relaton-xsf/workflows/macos/badge.svg["Build Status (macOS)", link="https://github.com/relaton/relaton-xsf/actions?workflow=macos"]
@@ -42,14 +42,14 @@ Or install it yourself as:
 
 [source,ruby]
 ----
-require 'relaton_xsf'
+require 'relaton/xsf'
 => true
 
-hit_collection = RelatonXsf::Bibliography.search("XEP 0001")
-=> <RelatonXsf::HitCollection:0x00000000019780 @ref=XEP 0001 @fetched=false>
+hit_collection = Relaton::Xsf::Bibliography.search("XEP 0001")
+=> <Relaton::Xsf::HitCollection:0x000000000013c0 @ref=XEP 0001 @fetched=false>
 
-item = hit_collection[0].fetch
-=> #<RelatonXsf::BibliographicItem:0x000000011167a518
+item = hit_collection[0].item
+=> #<Relaton::Bib::ItemData:0x0000000124d510b8
  ...
 ----
 
@@ -57,9 +57,12 @@ item = hit_collection[0].fetch
 [source,ruby]
 ----
 item.to_xml
-=> "<bibitem id="XEP0001" type="standard" schema-version="v1.2.9">
-      <fetched>2023-07-18</fetched>
-      <title format="text/plain" language="en" script="Latn">XMPP Extension Protocols</title>
+=> "<bibitem id="XEP0001" type="standard" schema-version="v1.4.1">
+      <fetched>2026-03-04</fetched>
+      <title language="en" script="Latn">XMPP Extension Protocols</title>
+      <uri type="src">http://xmpp.org/extensions/xep-0001.html</uri>
+      <uri type="HTML">http://xmpp.org/extensions/xep-0001.html</uri>
+      <docidentifier type="XEP" primary="true">XEP 0001</docidentifier>
       ...
     <bibitem>"
 ----
@@ -67,12 +70,16 @@ With argument `bibdata: true` it outputs XML wrapped by `bibdata` element and ad
 [source,ruby]
 ----
 item.to_xml bibdata: true
-=> "<bibdata type="standard" schema-version="v1.2.9">
-      <fetched>2023-07-18</fetched>
-      <title format="text/plain" language="en" script="Latn">XMPP Extension Protocols</title>
+=> "<bibdata type="standard" schema-version="v1.4.1">
+      <fetched>2026-03-04</fetched>
+      <title language="en" script="Latn">XMPP Extension Protocols</title>
+      <uri type="src">http://xmpp.org/extensions/xep-0001.html</uri>
+      <uri type="HTML">http://xmpp.org/extensions/xep-0001.html</uri>
+      <docidentifier type="XEP" primary="true">XEP 0001</docidentifier>
       ...
       <ext>
         <doctype>rfc</doctype>
+        <flavor>xsf</flavor>
       </ext>
   </bibdata>"
 ----
@@ -80,35 +87,34 @@ item.to_xml bibdata: true
 === Get document by reference
 [source,ruby]
 ----
-item = RelatonXsf::Bibliography.get "XEP 0001"
-[relaton-xsf] (XEP 0001) Fetching from Relaton repository ...
-[relaton-xsf] (XEP 0001) Found `XEP 0001`
-=> #<RelatonXsf::BibliographicItem:0x000000011275cd18
+item = Relaton::Xsf::Bibliography.get "XEP 0001"
+[relaton-xsf] INFO: (XEP 0001) Fetching from Relaton repository ...
+[relaton-xsf] INFO: (XEP 0001) Found: `XEP 0001`
+=> #<Relaton::Bib::ItemData:0x0000000125036f58
 ...
 
-item.docidentifier.first.id
+item.docidentifier.first.content
 => "XEP 0001"
 ----
 
-=== Typed links
+=== Typed source links
 
-XSF publications have `src` type link.
+XSF publications have `src` type source link.
 
 [source,ruby]
 ----
-item.link
-=> [#<RelatonBib::TypedUri:0x0000000113ad5ca0
-  @content=#<Addressable::URI:0xcc24 URI:http://xmpp.org/extensions/xep-0001.html>,
-  @language=nil,
-  @script=nil,
-  @type="src">]
+item.source[0].type
+=> "src"
+
+item.source[0].content
+=> "http://xmpp.org/extensions/xep-0001.html"
 ----
 
 === Fetch data
 
 This gem uses the https://xmpp.org/extensions/refs/ dataset as a data source.
 
-The method `RelatonXsf::DataFetcher.fetch(output: "data", format: "yaml")` fetches all the documents from the dataset and saves them to the `./data` folder in YAML format.
+The method `Relaton::Xsf::DataFetcher.fetch(output: "data", format: "yaml")` fetches all the documents from the dataset and saves them to the `./data` folder in YAML format.
 Arguments:
 
 - `output` - folder to save documents (default './data').
@@ -116,7 +122,9 @@ Arguments:
 
 [source,ruby]
 ----
-RelatonXsf::DataFetcher.fetch
+require 'relaton/xsf/data_fetcher'
+
+Relaton::Xsf::DataFetcher.fetch
 Started at: 2021-09-01 18:01:01 +0200
 Stopped at: 2021-09-01 18:01:43 +0200
 Done in: 42 sec.
@@ -125,7 +133,7 @@ Done in: 42 sec.
 
 === Logging
 
-RelatonXsf uses the relaton-logger gem for logging. By default, it logs to STDOUT. To change the log levels and add other loggers, read the https://github.com/relaton/relaton-logger#usage[relaton-logger] documentation.
+Relaton::Xsf uses the relaton-logger gem for logging. By default, it logs to STDOUT. To change the log levels and add other loggers, read the https://github.com/relaton/relaton-logger#usage[relaton-logger] documentation.
 
 == Development
 

data/lib/relaton/xsf/bibdata.rb

@@ -0,0 +1,7 @@
+module Relaton
+  module Xsf
+    class Bibdata < Item
+      include Bib::BibdataShared
+    end
+  end
+end

data/lib/relaton/xsf/bibitem.rb

@@ -0,0 +1,7 @@
+module Relaton
+  module Xsf
+    class Bibitem < Item
+      include Bib::BibitemShared
+    end
+  end
+end

data/lib/relaton/xsf/bibliography.rb

@@ -0,0 +1,24 @@
+module Relaton
+  module Xsf
+    module Bibliography
+      extend self
+
+      def search(ref)
+        HitCollection.new(ref).search
+      end
+
+      def get(code, _year = nil, _opts = {})
+        Util.info "Fetching from Relaton repository ...", key: code
+        result = search(code)
+        if result.empty?
+          Util.info "Not found.", key: code
+          return
+        end
+
+        bib = result.first.item
+        Util.info "Found: `#{bib.docidentifier.first.content}`", key: code
+        bib
+      end
+    end
+  end
+end

data/lib/relaton/xsf/data_fetcher.rb

@@ -0,0 +1,56 @@
+require "relaton/core"
+
+module Relaton
+  module Xsf
+    class DataFetcher < Relaton::Core::DataFetcher
+      def index
+        @index ||= Relaton::Index.find_or_create :xsf, file: "#{INDEXFILE}.yaml"
+      end
+
+      def fetch(_source = nil)
+        agent = Mechanize.new
+        resp = agent.get "https://xmpp.org/extensions/refs/"
+        resp.xpath("//a[contains(@href, 'XEP-')]").each do |link|
+          doc = agent.get link[:href]
+          bib = Relaton::Bib::Converter::BibXml.to_item doc.body
+          save_doc bib
+        rescue StandardError => e
+          Util.warn "Failed to parse #{link[:href]}: #{e.message}"
+        end
+        index.save
+      end
+
+      def save_doc(bib)
+        return unless bib
+
+        bib.ext ||= Relaton::Bib::Ext.new
+        bib.ext.flavor = "xsf"
+
+        docid = bib.docidentifier.detect(&:primary) || bib.docidentifier.first
+        id = docid&.content
+        return unless id
+
+        file = output_file id
+        if @files.include? file
+          Util.warn "File #{file} already exists"
+        else
+          @files << file
+        end
+        File.write file, serialize(bib), encoding: "UTF-8"
+        index.add_or_update id, file
+      end
+
+      def to_yaml(bib)
+        bib.to_yaml
+      end
+
+      def to_xml(bib)
+        bib.to_xml bibdata: true
+      end
+
+      def to_bibxml(bib)
+        bib.to_rfcxml
+      end
+    end
+  end
+end

data/lib/relaton/xsf/hit.rb

@@ -0,0 +1,17 @@
+module Relaton
+  module Xsf
+    class Hit < Relaton::Core::Hit
+      def item
+        return @doc if @doc
+
+        agent = Mechanize.new
+        resp = agent.get hit[:url]
+        hash = YAML.safe_load resp.body
+        hash["fetched"] = Date.today.to_s
+        @doc = Relaton::Bib::Item.from_yaml hash.to_yaml
+      rescue StandardError => e
+        raise Relaton::RequestError, e.message
+      end
+    end
+  end
+end

data/lib/relaton/xsf/hit_collection.rb

@@ -0,0 +1,24 @@
+module Relaton
+  module Xsf
+    class HitCollection < Relaton::Core::HitCollection
+      GHDATA_URL = "https://raw.githubusercontent.com/relaton/relaton-data-xsf/data-v2/".freeze
+
+      def search
+        @array = index.search(ref).sort_by { |hit| hit[:id] }.map do |row|
+          Hit.new url: "#{GHDATA_URL}#{row[:file]}"
+        end
+        self
+      rescue StandardError => e
+        raise Relaton::RequestError, e.message
+      end
+
+      def index
+        @index ||= Relaton::Index.find_or_create(
+          :xsf,
+          url: "#{GHDATA_URL}#{INDEXFILE}.zip",
+          file: "#{INDEXFILE}.yaml",
+        )
+      end
+    end
+  end
+end

data/lib/relaton/xsf/item.rb

@@ -0,0 +1,7 @@
+module Relaton
+  module Xsf
+    class Item < Bib::Item
+      model Bib::ItemData
+    end
+  end
+end

data/lib/relaton/xsf/processor.rb

@@ -0,0 +1,50 @@
+# frozen_string_literal: true
+require "relaton/core/processor"
+
+module Relaton
+  module Xsf
+    class Processor < Relaton::Core::Processor
+      attr_reader :idtype
+
+      def initialize # rubocop:disable Lint/MissingSuper
+        @short = :relaton_xsf
+        @prefix = "XEP"
+        @defaultprefix = %r{^XEP\s}
+        @idtype = "XEP"
+        @datasets = %w[xep-xmpp]
+      end
+
+      def get(code, date, opts)
+        require_relative "../xsf"
+        Relaton::Xsf::Bibliography.get(code, date, opts)
+      end
+
+      def fetch_data(_source, opts)
+        require_relative "data_fetcher"
+        Relaton::Xsf::DataFetcher.fetch(**opts)
+      end
+
+      def from_xml(xml)
+        require_relative "../xsf"
+        Relaton::Bib::Item.from_xml(xml)
+      end
+
+      def from_yaml(yaml)
+        require_relative "../xsf"
+        Relaton::Bib::Item.from_yaml(yaml)
+      end
+
+      def grammar_hash
+        require_relative "../xsf"
+        Relaton::Xsf.grammar_hash
+      end
+
+      def remove_index_file
+        require_relative "../xsf"
+        Relaton::Index.find_or_create(
+          :xsf, url: true, file: "#{INDEXFILE}.yaml"
+        ).remove_file
+      end
+    end
+  end
+end

data/lib/relaton/xsf/util.rb

@@ -0,0 +1,9 @@
+module Relaton
+  module Xsf
+    module Util
+      extend Relaton::Bib::Util
+
+      PROGNAME = "relaton-xsf".freeze
+    end
+  end
+end

data/lib/relaton/xsf/version.rb

@@ -0,0 +1,7 @@
+# frozen_string_literal: true
+
+module Relaton
+  module Xsf
+    VERSION = "2.0.0-alpha.1"
+  end
+end

data/lib/relaton/xsf.rb

@@ -0,0 +1,29 @@
+# frozen_string_literal: true
+
+require "mechanize"
+require "relaton/index"
+require "relaton/bib"
+require_relative "xsf/version"
+require_relative "xsf/util"
+require_relative "xsf/item"
+require_relative "xsf/bibitem"
+require_relative "xsf/bibdata"
+require_relative "xsf/hit"
+require_relative "xsf/hit_collection"
+require_relative "xsf/bibliography"
+
+module Relaton
+  module Xsf
+    INDEXFILE = "index-v1"
+
+    class Error < StandardError; end
+
+    # Your code goes here...
+    def self.grammar_hash
+      # gem_path = File.expand_path "..", __dir__
+      # grammars_path = File.join gem_path, "grammars", "*"
+      # grammars = Dir[grammars_path].sort.map { |gp| File.read gp }.join
+      Digest::MD5.hexdigest Relaton::Bib::VERSION # grammars
+    end
+  end
+end

metadata

@@ -1,14 +1,13 @@
 --- !ruby/object:Gem::Specification
 name: relaton-xsf
 version: !ruby/object:Gem::Version
-  version: 1.20.0
+  version: 2.0.0.pre.alpha.1
 platform: ruby
 authors:
 - Ribose Inc.
-autorequire:
 bindir: exe
 cert_chain: []
-date: 2024-12-13 00:00:00.000000000 Z
+date: 1980-01-02 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: mechanize
@@ -30,14 +29,28 @@ dependencies:
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.20.0
+        version: 2.0.0.pre.alpha.4
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - "~>"
       - !ruby/object:Gem::Version
-        version: 1.20.0
+        version: 2.0.0.pre.alpha.4
+- !ruby/object:Gem::Dependency
+  name: relaton-core
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.0.9
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: 0.0.9
 - !ruby/object:Gem::Dependency
   name: relaton-index
   requirement: !ruby/object:Gem::Requirement
@@ -52,8 +65,8 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: 0.2.0
-description: 'RelatonIso: retrieve ISO Standards for bibliographic use using the IsoBibliographicItem
-  model'
+description: 'Relaton::Xsf: retrieve ISO Standards for bibliographic use using the
+  IsoBibliographicItem model'
 email:
 - open.source@ribose.com
 executables: []
@@ -62,29 +75,28 @@ extra_rdoc_files: []
 files:
 - ".rspec"
 - ".rubocop.yml"
+- CLAUDE.md
 - Gemfile
 - LICENSE.txt
 - README.adoc
 - Rakefile
-- lib/relaton_xsf.rb
-- lib/relaton_xsf/bibliographic_item.rb
-- lib/relaton_xsf/bibliography.rb
-- lib/relaton_xsf/bibxml_parser.rb
-- lib/relaton_xsf/data_fetcher.rb
-- lib/relaton_xsf/hash_converter.rb
-- lib/relaton_xsf/hit.rb
-- lib/relaton_xsf/hit_collection.rb
-- lib/relaton_xsf/processor.rb
-- lib/relaton_xsf/util.rb
-- lib/relaton_xsf/version.rb
-- lib/relaton_xsf/xml_parser.rb
+- lib/relaton/xsf.rb
+- lib/relaton/xsf/bibdata.rb
+- lib/relaton/xsf/bibitem.rb
+- lib/relaton/xsf/bibliography.rb
+- lib/relaton/xsf/data_fetcher.rb
+- lib/relaton/xsf/hit.rb
+- lib/relaton/xsf/hit_collection.rb
+- lib/relaton/xsf/item.rb
+- lib/relaton/xsf/processor.rb
+- lib/relaton/xsf/util.rb
+- lib/relaton/xsf/version.rb
 - sig/relaton_xsf.rbs
 homepage: https://github.com/relaton/relaton-xsf
 licenses:
 - BSD-2-Clause
 metadata:
   homepage_uri: https://github.com/relaton/relaton-xsf
-post_install_message:
 rdoc_options: []
 require_paths:
 - lib
@@ -92,16 +104,15 @@ required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
-      version: 2.7.0
+      version: 3.2.0
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
   - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.27
-signing_key:
+rubygems_version: 3.6.9
 specification_version: 4
-summary: 'RelatonIso: retrieve ISO Standards for bibliographic use using the IsoBibliographicItem
+summary: 'Relaton::Xsf: retrieve ISO Standards for bibliographic use using the IsoBibliographicItem
   model'
 test_files: []

data/lib/relaton_xsf/bibliographic_item.rb

@@ -1,4 +0,0 @@
-module RelatonXsf
-  class BibliographicItem < RelatonBib::BibliographicItem
-  end
-end

data/lib/relaton_xsf/bibliography.rb

@@ -1,22 +0,0 @@
-module RelatonXsf
-  module Bibliography
-    extend self
-
-    def search(ref)
-      HitCollection.new(ref).search
-    end
-
-    def get(code, _year = nil, _opts = {})
-      Util.info "Fetching from Relaton repository ...", key: code
-      result = search(code)
-      if result.empty?
-        Util.info "Not found.", key: code
-        return
-      end
-
-      bib = result.first.fetch
-      Util.info "Found: `#{bib.docidentifier.first.id}`", key: code
-      bib
-    end
-  end
-end

data/lib/relaton_xsf/bibxml_parser.rb

@@ -1,12 +0,0 @@
-module RelatonXsf
-  module BibXMLParser
-    include RelatonBib::BibXMLParser
-    extend self
-
-    # @param item_hash [Hash]
-    # @return [RelatonXsf::BibliographicItem]
-    def bib_item(item_hash)
-      BibliographicItem.new(**item_hash)
-    end
-  end
-end

data/lib/relaton_xsf/data_fetcher.rb

@@ -1,57 +0,0 @@
-module RelatonXsf
-  class DataFetcher
-    # @param output [String]
-    # @param format [String]
-    def initialize(output, format)
-      @output = output
-      @format = format
-      @ext = format.sub(/^bib/, "")
-      @files = []
-    end
-
-    def self.fetch(output: "data", format: "yaml")
-      warn "fetching data to #{output} in #{format} format"
-      t1 = Time.now
-      warn "start at #{t1}"
-      FileUtils.mkdir_p output
-      new(output, format).fetch
-      t2 = Time.now
-      t = t2 - t1
-      warn "finished at #{t2} (#{t.round} seconds)"
-    end
-
-    def index
-      @index ||= Relaton::Index.find_or_create :xsf, file: "index-v1.yaml"
-    end
-
-    def fetch
-      agent = Mechanize.new
-      resp = agent.get "https://xmpp.org/extensions/refs/"
-      resp.xpath("//a[contains(@href, 'XEP-')]").each do |link|
-        doc = agent.get link[:href]
-        bib = BibXMLParser.parse doc.body
-        write_doc bib
-      end
-      index.save
-    end
-
-    def write_doc(bib)
-      id = bib.docidentifier.find(&:primary).id
-      file = File.join @output, "#{id.gsub(' ', '-').downcase}.#{@ext}"
-      if @files.include? file
-        Util.warn "#{file} already exists"
-      end
-      File.write file, serialize(bib), encoding: "UTF-8"
-      @files << file
-      index.add_or_update id, file
-    end
-
-    def serialize(bib)
-      case @format
-      when "yaml" then bib.to_hash.to_yaml
-      when "xml" then bib.to_xml bibdata: true
-      else bib.send "to_#{@format}"
-      end
-    end
-  end
-end

data/lib/relaton_xsf/hash_converter.rb

@@ -1,11 +0,0 @@
-module RelatonXsf
-  module HashConverter
-    include RelatonBib::HashConverter
-    extend self
-    # @param item_hash [Hash]
-    # @return [RelatonBib::BibliographicItem]
-    def bib_item(item_hash)
-      BibliographicItem.new(**item_hash)
-    end
-  end
-end

data/lib/relaton_xsf/hit.rb

@@ -1,15 +0,0 @@
-module RelatonXsf
-  class Hit < RelatonBib::Hit
-    def fetch
-      return @doc if @doc
-
-      agent = Mechanize.new
-      resp = agent.get hit[:url]
-      hash = YAML.safe_load resp.body
-      hash["fetched"] = Date.today.to_s
-      @doc = BibliographicItem.from_hash hash
-    rescue StandardError => e
-      raise RelatonBib::RequestError, e.message
-    end
-  end
-end

data/lib/relaton_xsf/hit_collection.rb

@@ -1,19 +0,0 @@
-module RelatonXsf
-  class HitCollection < RelatonBib::HitCollection
-    INDEX_FILE = "index-v1.yaml".freeze
-    GHDATA_URL = "https://raw.githubusercontent.com/relaton/relaton-data-xsf/main/".freeze
-
-    def search
-      @array = index.search(text).sort_by { |hit| hit[:id] }.map do |row|
-        Hit.new url: "#{GHDATA_URL}#{row[:file]}"
-      end
-      self
-    rescue StandardError => e
-      raise RelatonBib::RequestError, e.message
-    end
-
-    def index
-      @index ||= Relaton::Index.find_or_create :xsf, url: "#{GHDATA_URL}index-v1.zip", file: INDEX_FILE
-    end
-  end
-end

data/lib/relaton_xsf/processor.rb

@@ -1,61 +0,0 @@
-require "relaton/processor"
-
-module RelatonXsf
-  class Processor < Relaton::Processor
-    attr_reader :idtype
-
-    def initialize # rubocop:disable Lint/MissingSuper
-      @short = :relaton_xsf
-      @prefix = "XEP"
-      @defaultprefix = %r{^XEP\s}
-      @idtype = "XEP"
-      @datasets = %w[xep-xmpp]
-    end
-
-    # @param code [String]
-    # @param date [String, nil] year
-    # @param opts [Hash]
-    # @return [RelatonXsf::BibliographicItem]
-    def get(code, date, opts)
-      ::RelatonXsf::Bibliography.get(code, date, opts)
-    end
-
-    #
-    # Fetch all the documents from http://xml2rfc.tools.ietf.org/public/rfc/bibxml-3gpp-new/
-    #
-    # @param [String] source source name
-    # @param [Hash] opts
-    # @option opts [String] :output directory to output documents
-    # @option opts [String] :format
-    #
-    def fetch_data(_source, opts)
-      ::RelatonXsf::DataFetcher.fetch(**opts)
-    end
-
-    # @param xml [String]
-    # @return [RelatonXsf::BibliographicItem]
-    def from_xml(xml)
-      ::RelatonXsf::XMLParser.from_xml xml
-    end
-
-    # @param hash [Hash]
-    # @return [RelatonXsf::BibliographicItem]
-    def hash_to_bib(hash)
-      item_hash = ::RelatonXsf::HashConverter.hash_to_bib(hash)
-      ::RelatonXsf::BibliographicItem.new(**item_hash)
-    end
-
-    # Returns hash of XML grammar
-    # @return [String]
-    def grammar_hash
-      @grammar_hash ||= ::RelatonBib.grammar_hash
-    end
-
-    #
-    # Remove index file
-    #
-    def remove_index_file
-      Relaton::Index.find_or_create(:xsf, url: true, file: HitCollection::INDEX_FILE).remove_file
-    end
-  end
-end

data/lib/relaton_xsf/util.rb

@@ -1,6 +0,0 @@
-module RelatonXsf
-  module Util
-    extend RelatonBib::Util
-    PROGNAME = "relaton-xsf".freeze
-  end
-end

data/lib/relaton_xsf/version.rb

@@ -1,5 +0,0 @@
-# frozen_string_literal: true
-
-module RelatonXsf
-  VERSION = "1.20.0"
-end

data/lib/relaton_xsf/xml_parser.rb

@@ -1,14 +0,0 @@
-module RelatonXsf
-  class XMLParser < RelatonBib::XMLParser
-    #
-    # Create bibliographic item
-    #
-    # @param item_hash [Hash] bibliographic item hash
-    #
-    # @return [RelatonXsf::BibliographicItem] bibliographic item
-    #
-    def self.bib_item(item_hash)
-      BibliographicItem.new(**item_hash)
-    end
-  end
-end

data/lib/relaton_xsf.rb

@@ -1,20 +0,0 @@
-# frozen_string_literal: true
-
-require "mechanize"
-require "relaton/index"
-require "relaton_bib"
-require_relative "relaton_xsf/version"
-require_relative "relaton_xsf/util"
-require_relative "relaton_xsf/bibliographic_item"
-require_relative "relaton_xsf/bibliography"
-require_relative "relaton_xsf/hit_collection"
-require_relative "relaton_xsf/hit"
-require_relative "relaton_xsf/xml_parser"
-require_relative "relaton_xsf/bibxml_parser"
-require_relative "relaton_xsf/hash_converter"
-require_relative "relaton_xsf/data_fetcher"
-
-module RelatonXsf
-  class Error < StandardError; end
-  # Your code goes here...
-end