Add an "{anchor #xyz}" rule to Remarkup

Summary: Ref T13410. Fixes T4280. Allows you to put a named anchor into a document explicitly.

Test Plan: Used `{anchor ...}` in Remarkup, used location bar to jump to anchors.

Maniphest Tasks: T13410, T4280

Differential Revision: https://secure.phabricator.com/D20825

src/__phutil_library_map__.php

@@ -5598,6 +5598,7 @@ phutil_register_library_map(array(
     'PhutilQsprintfInterface' => 'infrastructure/storage/xsprintf/PhutilQsprintfInterface.php',
     'PhutilQueryString' => 'infrastructure/storage/xsprintf/PhutilQueryString.php',
     'PhutilRealNameContextFreeGrammar' => 'infrastructure/lipsum/PhutilRealNameContextFreeGrammar.php',
+    'PhutilRemarkupAnchorRule' => 'infrastructure/markup/markuprule/PhutilRemarkupAnchorRule.php',
     'PhutilRemarkupBlockInterpreter' => 'infrastructure/markup/blockrule/PhutilRemarkupBlockInterpreter.php',
     'PhutilRemarkupBlockRule' => 'infrastructure/markup/blockrule/PhutilRemarkupBlockRule.php',
     'PhutilRemarkupBlockStorage' => 'infrastructure/markup/PhutilRemarkupBlockStorage.php',
@@ -12391,6 +12392,7 @@ phutil_register_library_map(array(
     'PhutilPhabricatorAuthAdapter' => 'PhutilOAuthAuthAdapter',
     'PhutilQueryString' => 'Phobject',
     'PhutilRealNameContextFreeGrammar' => 'PhutilContextFreeGrammar',
+    'PhutilRemarkupAnchorRule' => 'PhutilRemarkupRule',
     'PhutilRemarkupBlockInterpreter' => 'Phobject',
     'PhutilRemarkupBlockRule' => 'Phobject',
     'PhutilRemarkupBlockStorage' => 'Phobject',

src/docs/user/userguide/remarkup.diviner

@@ -715,6 +715,18 @@ Press {key down down-right right LP} to activate the hadoken technique.
 > Press {key down down-right right LP} to activate the hadoken technique.
 
 
+Anchors
+========
+
+You can use `{anchor #xyz}` to create a document anchor and later link to
+it directly with `#xyz` in the URI.
+
+Headers also automatically create named anchors.
+
+If you navigate to `#xyz` in your browser location bar, the page will scroll
+to the first anchor with "xyz" as a prefix of the anchor name.
+
+
 = Fullscreen Mode =
 
 Remarkup editors provide a fullscreen composition mode. This can make it easier

src/infrastructure/markup/PhabricatorMarkupEngine.php

@@ -539,6 +539,7 @@ final class PhabricatorMarkupEngine extends Phobject {
     $rules[] = new PhutilRemarkupDelRule();
     $rules[] = new PhutilRemarkupUnderlineRule();
     $rules[] = new PhutilRemarkupHighlightRule();
+    $rules[] = new PhutilRemarkupAnchorRule();
 
     foreach (self::loadCustomInlineRules() as $rule) {
       $rules[] = clone $rule;

src/infrastructure/markup/blockrule/PhutilRemarkupHeaderBlockRule.php

@@ -162,12 +162,7 @@ final class PhutilRemarkupHeaderBlockRule extends PhutilRemarkupBlockRule {
 
   public static function getAnchorNameFromHeaderText($text) {
     $anchor = phutil_utf8_strtolower($text);
-
+    $anchor = PhutilRemarkupAnchorRule::normalizeAnchor($anchor);
-    // Replace all latin characters which are not "a-z" or "0-9" with "-".
-    // Preserve other characters, since non-latin letters and emoji work
-    // fine in anchors.
-    $anchor = preg_replace('/[\x00-\x2F\x3A-\x60\x7B-\x7F]+/', '-', $anchor);
-    $anchor = trim($anchor, '-');
 
     // Truncate the fragment to something reasonable.
     $anchor = id(new PhutilUTF8StringTruncator())

src/infrastructure/markup/markuprule/PhutilRemarkupAnchorRule.php

@@ -0,0 +1,69 @@
+<?php
+
+final class PhutilRemarkupAnchorRule extends PhutilRemarkupRule {
+
+  public function getPriority() {
+    return 200.0;
+  }
+
+  public function apply($text) {
+    return preg_replace_callback(
+      '/{anchor\s+#([^\s}]+)}/s',
+      array($this, 'markupAnchor'),
+      $text);
+  }
+
+  protected function markupAnchor(array $matches) {
+    $engine = $this->getEngine();
+
+    if ($engine->isTextMode()) {
+      return null;
+    }
+
+    if ($engine->isHTMLMailMode()) {
+      return null;
+    }
+
+    if ($engine->isAnchorMode()) {
+      return null;
+    }
+
+    if (!$this->isFlatText($matches[0])) {
+      return $matches[0];
+    }
+
+    if (!self::isValidAnchorName($matches[1])) {
+      return $matches[0];
+    }
+
+    $tag_view = phutil_tag(
+      'a',
+      array(
+        'name' => $matches[1],
+      ),
+      '');
+
+    return $this->getEngine()->storeText($tag_view);
+  }
+
+  public static function isValidAnchorName($anchor_name) {
+    $normal_anchor = self::normalizeAnchor($anchor_name);
+
+    if ($normal_anchor === $anchor_name) {
+      return true;
+    }
+
+    return false;
+  }
+
+  public static function normalizeAnchor($anchor) {
+    // Replace all latin characters which are not "a-z" or "0-9" with "-".
+    // Preserve other characters, since non-latin letters and emoji work
+    // fine in anchors.
+    $anchor = preg_replace('/[\x00-\x2F\x3A-\x60\x7B-\x7F]+/', '-', $anchor);
+    $anchor = trim($anchor, '-');
+
+    return $anchor;
+  }
+
+}