refactor: Update globe button to modify lang in URL paths 🌋

.github/workflows/ci.yml

@@ -37,11 +37,12 @@ jobs:
     - name: Check broken links
       shell: bash
       continue-on-error: false
-      # Exclude checking locales for each link 'lang=*'
+      # Exclude the following: non-en locales, and 3-letter keyboard redirects (/lao, /ipa)
       run: |
         set +e
         set +o pipefail
-        npx broken-link-checker http://localhost:8053/_test --recursive --ordered ---host-requests 50 -e --filter-level 3 --exclude '*/donate' --exclude '*lang=*' | tee blc.log
+        npx broken-link-checker http://localhost:8053/_test --recursive --ordered ---host-requests 50 -e --filter-level 3 \
+          --exclude '*/donate' --exclude '*/de/*' --exclude '*/es/*' --exclude '*/fr/*' --exclude '*/ipa/*' --exclude '*/km/*' --exclude '*lang=*' --exclude '*/lao/*' | tee blc.log
         echo "BLC_RESULT=${PIPESTATUS[0]}" >> "$GITHUB_ENV"
 
     - name: Report on broken links

.htaccess

@@ -203,75 +203,82 @@ RewriteCond "%{DOCUMENT_ROOT}/cdn/$1$2" -f
 RewriteRule "^cdn/(.+\.)(gif|js|png|svg)$" "/cdn/$1$2" [END]
 
 
+#
+# Intentionally not using regex from Validation::validate_bcp47 because we only need lang-script-region triplet
+# ^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)
+# [NC] for case-insensitive
+# Long regex for "BCP-47" codes = $1, rest of URL $7 onward
+#
+
 #
 # keyboards Install | Download | Share | bare | .json -->
 #
 
 # /keyboards/install/[id] to /keyboards/install.php
-RewriteRule "^(en)/keyboards/install/([^/]+)$" "/_content/keyboards/install.php?id=$2&lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/install/([^/]+)$" "/_content/keyboards/install.php?id=$7&lang=$1" [NC,END,QSA]
 
 # /keyboards/download/[id] to /keyboards/keyboard.php
 # This formerly redirected to a download, but we no longer need it; keep it for
 # legacy links
-RewriteRule "^(en)/keyboards/download/([^/]+)$" "/_content/keyboards/keyboard.php?id=$2&lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/download/([^/]+)$" "/_content/keyboards/keyboard.php?id=$7&lang=$1" [NC,END,QSA]
 
 # /keyboards/share/[id] to /keyboards/share.php
 # if the keyboard exists in the repo, then share.php will redirect to /keyboards/<id>
-RewriteRule "^(en)/keyboards/share/([^/]+)$" "/_content/keyboards/share.php?id=$2&lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/share/([^/]+)$" "/_content/keyboards/share.php?id=$7&lang=$1" [NC,END,QSA]
 
 # /keyboards/{id}.json to /keyboards/keyboard.json.php
-RewriteRule "^(en)/keyboards/(?!keyboard.json)(.*)\.json$" "/_content/keyboards/keyboard.json.php?id=$2&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/(?!keyboard.json)(.*)\.json$" "/_content/keyboards/keyboard.json.php?id=$7&lang=$1" [NC,END]
 
 # /keyboards/{id} to /keyboards/keyboard.php
-RewriteRule "^(en)/keyboards/(?!index\.php|install|keyboard|session|share)([^/]+)$" "/_content/keyboards/keyboard.php?id=$2&lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/(?!index\.php|install|keyboard|session|share)([^/]+)$" "/_content/keyboards/keyboard.php?id=$7&lang=$1" [NC,END,QSA]
 
 #
 # keyboards search
 #
 
 # /keyboards/languages to /keyboards/index.php
-RewriteRule "^(en)/keyboards/languages/(.*)" "/_content/keyboards/index.php?q=l:id:$2&lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/languages/(.*)" "/_content/keyboards/index.php?q=l:id:$7&lang=$1" [NC,END,QSA]
 
 # /keyboards/download to /keyboards/download.php
-RewriteRule "^(en)/keyboards/download(.php)?" "/_content/keyboards/download.php?lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/download(.php)?" "/_content/keyboards/download.php?lang=$1" [NC,END,QSA]
 
 # /keyboards/legacy to /keyboards/keyboard.php
-RewriteRule "^(en)/keyboards/legacy/(.*)" "/_content/keyboards/keyboard.php?legacy=$2&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/legacy/(.*)" "/_content/keyboards/keyboard.php?legacy=$7&lang=$1" [NC,END]
 
 # /keyboards/countries to /keyboards/index.php
-RewriteRule "^(en)/keyboards/countries/(.*)" "/_content/keyboards/index.php?q=c:id:$2&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/keyboards/countries/(.*)" "/_content/keyboards/index.php?q=c:id:$7&lang=$1" [NC,END]
 
 
 #
 # contact/exception
 #
 
 # /contact/exception to /contact/exception.php
-RewriteRule "^(en)/contact/exception(.php)?" "/_content/contact/exception.php?lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/contact/exception(.php)?" "/_content/contact/exception.php?lang=$1" [NC,END,QSA]
 
-RewriteRule "^(en)/contact/exception?id=(.+)$" "/_content/contact/exception?id=$2&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/contact/exception?id=(.+)$" "/_content/contact/exception?id=$7&lang=$1" [NC,END]
 
 #
 # ios
 #
-RewriteRule "^(en)/(iphone|ipad)(/.*)?" "/_content/iphone-and-ipad/$3?lang=$1" [END,QSA]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(iphone|ipad)(/.*)?" "/_content/iphone-and-ipad/$8?lang=$1" [NC,END,QSA]
 
 #
 # downloads/releases
 #
 
 # releases-tier/download
 # note: the tier is currently ignored
-RewriteRule "^(en)/downloads/releases/(alpha|beta|stable)/(.+)$" "/_content/downloads/releases/_version_downloads.php?tier=$2&version=$3&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/downloads/releases/(alpha|beta|stable)/(.+)$" "/_content/downloads/releases/_version_downloads.php?tier=$7&version=$8&lang=$1" [NC,END]
 
 # releases-download
-RewriteRule "^(en)/downloads/releases/(.+)$" "/_content/downloads/releases/_version_downloads.php?version=$2&lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/downloads/releases/(.+)$" "/_content/downloads/releases/_version_downloads.php?version=$7&lang=$1" [NC,END]
 
 #
 # Assets don't use lang param
 #
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2" -f
-RewriteRule "^(en)/(.+)$" "/_content/$2" [END]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7" -f
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(.+)$" "/_content/$7$8" [NC,END]
 
 # TODO: do we want en/lao --> en/keyboards/basic_kbdlao? as well as /lao --> ...
 
@@ -281,32 +288,30 @@ RewriteRule "^(en)/(.+)$" "/_content/$2" [END]
 
 # TODO: mdhost currently in a different path than help.keyman
 
-# TODO: en --> should be a bcp-47 type 'lang' match (e.g. ab[-x][-y]) ignore case, keep it a fairly loose expression [a-z][a-z][a-z]?(-...)?
-
 ##############################################
 
 # Rewrite file to file.md
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2.md" -f
-RewriteRule "^(en)/(.+)$" "/_includes/includes/md/mdhost.php?file=_content/$2.md&lang=$1" [END]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7.md" -f
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(.+)$" "/_includes/includes/md/mdhost.php?file=_content/$7.md&lang=$1" [NC,END]
 
 # Rewrite file to file.php
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2.php" -f
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2.md" !-f
-RewriteRule "^(en)/(.+)$" "/_content/$2.php?lang=$1" [END]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7.php" -f
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7.md" !-f
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(.+)$" "/_content/$7.php?lang=$1" [NC,END]
 
 # Rewrite folder/ to folder/index.md
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2/index.md" -f
-RewriteRule "^(en)/(.+)/$" "/_includes/includes/md/mdhost.php?file=_content/$2/index.md&lang=$1" [END]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7/index.md" -f
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(.+)/$" "/_includes/includes/md/mdhost.php?file=_content/$7/index.md&lang=$1" [NC,END]
 
 # Rewrite folder/ to folder/index.php
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2/index.php" -f
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2/index.md" !-f
-RewriteRule "^(en)/(.+)/$" "/_content/$2/index.php?lang=$1" [END]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7/index.php" -f
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7/index.md" !-f
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/(.+)/$" "/_content/$7/index.php?lang=$1" [NC,END]
 
 # Root page
-RewriteRule "^(en)/$" "/_content/index.php?lang=$1" [END]
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)/$" "/_content/index.php?lang=$1" [NC,END]
 
 # Finally, append the terminating slash for folders, given it is no longer
 # done automatically because we put DirectorySlash off
-RewriteCond "%{DOCUMENT_ROOT}/_content/$2$3" -d
-RewriteRule "^(en)(.*)([^/])?$" "$1$2$3/" [R,QSA]
+RewriteCond "%{DOCUMENT_ROOT}/_content/$7$8" -d
+RewriteRule "^(([a-z]{2,3})(-([A-Za-z]{4}))?(-([a-z]{2}|[0-9]{3}))?)(.*)([^/])?$" "$1$7$8/" [NC,R,QSA]

_includes/2020/templates/Menu.php

@@ -3,7 +3,9 @@
 
   namespace Keyman\Site\com\keyman\templates;
 
+  use Keyman\Site\com\keyman\Locale;
   use Keyman\Site\com\keyman\Util;
+  use Keyman\Site\com\keyman\Validation;
   use Keyman\Site\Common\KeymanVersion;
   use Keyman\Site\Common\KeymanHosts;
 
@@ -34,40 +36,48 @@ public static function render(array $fields): void {
     }
 
     /**
-     * Generate the URL with query to change the UI language
+     * Modify link of the current URL for a given UI language
      * @param language - language tag to use
+     * @return string - modified URL
      */
     private static function change_ui_language($language): string {
       // Parse the current URI for populating the UI dropdown
       $url = isset($_SERVER['REQUEST_URI']) ? $_SERVER['REQUEST_URI'] : '/';
       $parts = parse_url($url);
 
-      if (!empty($parts['query'])) {
-        parse_str($parts['query'], $queryParams);
-      } else {
-        $queryParams = [];
+      $path = '';
+      // Replace language if current language in path is valid BCP-47
+      // Note: Validate::validate_bcp47 differs from regex in .htaccess
+      if (!empty($parts['path'])) {
+        $path = explode("/", $parts['path']);
+        if ($path[1] != null ) {
+           /* TODO: add validation back && class_exists('\\Keyman\\Site\\com\\keyman\\Validation') &&
+            \Keyman\Site\com\keyman\Validation::validate_bcp47($path[1]) != null*/
+          $path[1] = $language;
+        } else {
+          // original URL didn't have a valid BCP-47 so inert it
+          array_splice($path, 1, 0, $language);
+        }
       }
 
-      // Set the language query
-      $queryParams['lang'] = $language;
-      $query = http_build_query($queryParams);
-
-      return $parts['path'] . "?" . $query;
+      // Rebuild the path
+      $newPath = implode("/", $path);
+      if (!empty($parts['query'])) {
+        // Append query
+        $newPath .= "?" . $parts['query'];
+      }
+      return $newPath;
     }
 
     /**
      * Generate links that correspond to the UI options
-     * As UI languages get added, we'll need to update this.
      */
     private static function render_ui_list() {
       echo "<ul>\n                <!-- Just use autonyms -->\n";
-      $linkArray = array(
-        'en' => array(Menu::change_ui_language('en'), 'English'),
-        'de' => array(Menu::change_ui_language('de'), 'Deutsch'),
-        'es' => array(Menu::change_ui_language('es'), 'Español'),
-        'fr' => array(Menu::change_ui_language('fr'), 'Français'),
-        'km' => array(Menu::change_ui_language('km'), 'ខ្មែរ')
-      );
+      $linkArray = array();
+      foreach(DISPLAY_NAMES as $id => $name) {
+        $linkArray[$id] = array(Menu::change_ui_language($id), $name);
+      }
 
       foreach($linkArray as $id) {
 echo <<<END
@@ -79,25 +89,18 @@ private static function render_ui_list() {
 
     /**
      * Render the globe dropdown for changing the UI language
-     * Limitation: Currently only visible on pages that use localized strings
      * @param divID - <div> ID to handle 3 cases:
      * ui-language (default) Desktop globe hover
      * ui-language1 - Desktop globe hover
      * phone - Mobile list
      */
     private static function render_globe_dropdown($divID = "ui-language"): void {
-      global $page_is_using_locale;
-      if (!isset($page_is_using_locale) || !$page_is_using_locale) {
-        // only render on pages that use localized strings
-        return;
-      }
-
       // Phone layout
       $globeClass = '';
       if ($divID === "phone") {
 ?>
 <div class="phone-menu-item">
-            <h3><span><img src="<?php echo Util::cdn("img/globe.png"); ?>" alt="UI globe dropdown" /></span> Keyboard Search UI</h3>
+            <h3><span><img src="<?php echo Util::cdn("img/globe.png"); ?>" alt="UI globe dropdown" /></span> Display in:</h3>
             <?= Menu::render_ui_list(); ?>
         </div>
       <?php
@@ -204,7 +207,7 @@ private static function render_top_menu(object $fields): void {
         <img id="header-bottom" src="<?php echo Util::cdn("img/headerbar.png"); ?>" alt='Header bottom' />
         <div id="help">
 
-          <span id='free'>Keyman is <a href='/free'>free and open source</a></span>
+          <span id='free'>Keyman is <a href='/<?=$fields->lang?>/free'>free and open source</a></span>
 
           <form action="/<?=$fields->lang?>/search/" method="get" role="search">
             <div class="search-wrap">

_includes/locale/Locale.php

@@ -10,6 +10,14 @@
 
   use \Keyman\Site\Common\KeymanHosts;
 
+  // As UI languages get added, we'll need to update this.
+  define('DISPLAY_NAMES', [
+    'en' => 'English',
+    'de' => 'Deutsch',
+    'es' => 'Español',
+    'fr' => 'Français',
+    'km' => 'ខ្មែរ (Khmer)']);
+
   class Locale {
     public const DEFAULT_LOCALE = 'en';