Add extensive alpha3 languages from ISO639-3 (#71)

.github/workflows/update-spec-language-extensive.yml

@@ -0,0 +1,33 @@
+name: Daily Scripts spec update
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: '55 17 * * *'
+
+jobs:
+  update-specs:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+      - name: Install dependencies
+        run: composer install
+      - name: Run standard update command
+        run: composer update-spec-language-extensive
+      - name: Run CSFixer
+        run: composer cs:fix
+      - name: Create Pull Request
+        id: cpr
+        uses: peter-evans/create-pull-request@v4
+        with:
+          commit-message: Automatic Language Extensive spec update from upstream changes
+          delete-branch: true
+          title: 'Automatic Language Extensive spec update'
+          branch: update-spec/language-extensive
+          body: |
+            This PR makes sure the content of this package is updated with upstream changes in the specs.
+          labels: |
+            spec-update
+            automated pr
+          reviewers: prinsfrank

composer.json

@@ -42,6 +42,7 @@
         "update-spec-currency": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\CurrencyMapping",
         "update-spec-http-methods": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\HttpMethodMapping",
         "update-spec-http-status-codes": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\HttpStatusCodeMapping",
+        "update-spec-language-extensive": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\LanguageExtensiveMapping",
         "update-spec-language": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\LanguageMapping",
         "update-spec-scripts": "@update-spec PrinsFrank\\Standards\\Dev\\DataSource\\Mapping\\ScriptMapping",
         "detect-drivers": "vendor/bin/bdi detect drivers"

dev/DataSource/DataSourceMappingProvider.php

@@ -3,6 +3,7 @@
 
 namespace PrinsFrank\Standards\Dev\DataSource;
 
+use PrinsFrank\Standards\Dev\DataSource\Mapping\LanguageExtensiveMapping;
 use PrinsFrank\Standards\Dev\DataSource\Mapping\CountryMapping;
 use PrinsFrank\Standards\Dev\DataSource\Mapping\CurrencyMapping;
 use PrinsFrank\Standards\Dev\DataSource\Mapping\HttpMethodMapping;
@@ -24,6 +25,7 @@ public function provide(): array
             CurrencyMapping::class,
             HttpMethodMapping::class,
             HttpStatusCodeMapping::class,
+            LanguageExtensiveMapping::class,
             LanguageMapping::class,
             ScriptMapping::class,
         ];

dev/DataSource/Mapping/LanguageExtensiveMapping.php

@@ -0,0 +1,70 @@
+<?php
+declare(strict_types=1);
+
+namespace PrinsFrank\Standards\Dev\DataSource\Mapping;
+
+use PrinsFrank\Standards\Dev\DataSource\Sorting\SortingInterface;
+use PrinsFrank\Standards\Dev\DataSource\Sorting\ValueWithDeprecatedTagsSeparateSorting;
+use PrinsFrank\Standards\Dev\DataTarget\EnumCase;
+use PrinsFrank\Standards\Dev\DataTarget\EnumFile;
+use PrinsFrank\Standards\Language\LanguageAlpha3Extensive;
+use Symfony\Component\Panther\Client;
+use Symfony\Component\Panther\DomCrawler\Crawler;
+
+/**
+ * @template TDataSet of object{Id: string, Part2B: string, Part2T: string, Part1: string, Scope: string, LanguageType: string, RefName: string, Comment: string}
+ * @implements Mapping<TDataSet>
+ */
+class LanguageExtensiveMapping implements Mapping
+{
+    public static function url(): string
+    {
+        return 'https://iso639-3.sil.org/sites/iso639-3/files/downloads/iso-639-3_Latin1.tab';
+    }
+
+    public static function toDataSet(Client $client, Crawler $crawler): array
+    {
+        $dataSet = [];
+        foreach (explode("\n", $client->getPageSource()) as $lineNumber => $line) {
+            if ($lineNumber === 0) {
+                continue;
+            }
+
+            $columns = explode("\t", $line);
+
+            $record               = (object) [];
+            $record->Id           = $columns[0];
+            $record->Part2B       = $columns[1];
+            $record->Part2T       = $columns[2];
+            $record->Part1        = $columns[3];
+            $record->Scope        = $columns[4];
+            $record->LanguageType = $columns[5];
+            $record->RefName      = $columns[6];
+            $record->Comment      = $columns[7];
+
+            /** @var TDataSet $record */
+            $dataSet[] = $record;
+        }
+
+        return $dataSet;
+    }
+
+    public static function toEnumMapping(array $dataSet): array
+    {
+        $languageExtensive = new EnumFile(LanguageAlpha3Extensive::class);
+        foreach ($dataSet as $dataItem) {
+            if (in_array($dataItem->RefName, ['Fa D\'ambu', 'C\'lela'], true)) {
+                continue;
+            }
+
+            $languageExtensive->addCase(new EnumCase($dataItem->RefName, $dataItem->Id));
+        }
+
+        return [$languageExtensive];
+    }
+
+    public static function getSorting(): SortingInterface
+    {
+        return new ValueWithDeprecatedTagsSeparateSorting();
+    }
+}

dev/DataTarget/NameNormalizer.php

@@ -15,7 +15,7 @@ public static function normalize(string $key): string
             throw new TransliterationException();
         }
 
-        $key = str_replace([' ', ';', ',', '(', ')', '-', '.', '\'', '*', '[', ']'], '_', $key);
+        $key = str_replace([' ', ';', ',', '(', ')', '-', '.', '\'', '/', '|', '=', '!', '?', '*', '[', ']', '~'], '_', $key);
 
         return trim(str_replace(['__', '__'], ['_', '_'], $key), '_');
     }