Bring "fixes x as y" parser forward and use new parsers instead of old ones

Summary: Fixes T3872. Ref T1812. Ref T3886. Modernize the "closes x as y" string parser, and use all the new parsers instead of the old ones.

Test Plan: Made a commit full of a pile of these trigger strings, then used `scripts/repository/reparse.php --message` to reparse it. Verified that parses came back as expected using a bunch of `var_dump()`.

Reviewers: btrahan

Reviewed By: btrahan

CC: aran

Maniphest Tasks: T1812, T3872, T3886

Differential Revision: https://secure.phabricator.com/D8263

src/__phutil_library_map__.php

@@ -885,6 +885,8 @@ phutil_register_library_map(array(
     'ManiphestCreateMailReceiver' => 'applications/maniphest/mail/ManiphestCreateMailReceiver.php',
     'ManiphestCustomField' => 'applications/maniphest/field/ManiphestCustomField.php',
     'ManiphestCustomFieldNumericIndex' => 'applications/maniphest/storage/ManiphestCustomFieldNumericIndex.php',
+    'ManiphestCustomFieldStatusParser' => 'applications/maniphest/field/parser/ManiphestCustomFieldStatusParser.php',
+    'ManiphestCustomFieldStatusParserTestCase' => 'applications/maniphest/field/parser/__tests__/ManiphestCustomFieldStatusParserTestCase.php',
     'ManiphestCustomFieldStorage' => 'applications/maniphest/storage/ManiphestCustomFieldStorage.php',
     'ManiphestCustomFieldStringIndex' => 'applications/maniphest/storage/ManiphestCustomFieldStringIndex.php',
     'ManiphestDAO' => 'applications/maniphest/storage/ManiphestDAO.php',
@@ -3498,6 +3500,8 @@ phutil_register_library_map(array(
     'ManiphestCreateMailReceiver' => 'PhabricatorMailReceiver',
     'ManiphestCustomField' => 'PhabricatorCustomField',
     'ManiphestCustomFieldNumericIndex' => 'PhabricatorCustomFieldNumericIndexStorage',
+    'ManiphestCustomFieldStatusParser' => 'PhabricatorCustomFieldMonogramParser',
+    'ManiphestCustomFieldStatusParserTestCase' => 'PhabricatorTestCase',
     'ManiphestCustomFieldStorage' => 'PhabricatorCustomFieldStorage',
     'ManiphestCustomFieldStringIndex' => 'PhabricatorCustomFieldStringIndexStorage',
     'ManiphestDAO' => 'PhabricatorLiskDAO',

src/applications/differential/field/specification/DifferentialFreeformFieldSpecification.php

@@ -44,53 +44,36 @@ abstract class DifferentialFreeformFieldSpecification
       ''              => null,
     );
 
-    $prefix_regex = array();
+    $matches = id(new ManiphestCustomFieldStatusParser())
-    foreach ($prefixes as $prefix => $resolution) {
+      ->parseCorpus($message);
-      $prefix_regex[] = preg_quote($prefix, '/');
-    }
-    $prefix_regex = implode('|', $prefix_regex);
 
-    $suffix_regex = array();
+    $task_statuses = array();
-    foreach ($suffixes as $suffix => $resolution) {
+    foreach ($matches as $match) {
-      $suffix_regex[] = preg_quote($suffix, '/');
+      $prefix = phutil_utf8_strtolower($match['prefix']);
-    }
+      $suffix = phutil_utf8_strtolower($match['suffix']);
-    $suffix_regex = implode('|', $suffix_regex);
-
-    $matches = null;
-    preg_match_all(
-      "/({$prefix_regex})\s+T(\d+)\s*({$suffix_regex})/i",
-      $message,
-      $matches,
-      PREG_SET_ORDER);
-
-    $tasks_statuses = array();
-    foreach ($matches as $set) {
-      $prefix = strtolower($set[1]);
-      $task_id = (int)$set[2];
-      $suffix = strtolower($set[3]);
 
       $status = idx($suffixes, $suffix);
       if (!$status) {
         $status = idx($prefixes, $prefix);
       }
 
-      $tasks_statuses[$task_id] = $status;
+      foreach ($match['monograms'] as $task_monogram) {
+        $task_id = (int)trim($task_monogram, 'tT');
+        $task_statuses[$task_id] = $status;
+      }
     }
 
-    return $tasks_statuses;
+    return $task_statuses;
   }
 
   private function findDependentRevisions($message) {
+    $matches = id(new DifferentialCustomFieldDependsOnParser())
+      ->parseCorpus($message);
+
     $dependents = array();
-
+    foreach ($matches as $match) {
-    $matches = null;
+      foreach ($match['monograms'] as $monogram) {
-    preg_match_all(
+        $id = (int)trim($monogram, 'dD');
-      '/\b(?i:depends\s+on):?\s+D(\d+(,\s+D\d++)*)\b/',
-      $message,
-      $matches);
-
-    foreach ($matches[1] as $revisions) {
-      foreach (preg_split('/,\s+D/', $revisions) as $id) {
         $dependents[$id] = $id;
       }
     }
@@ -99,46 +82,13 @@ abstract class DifferentialFreeformFieldSpecification
   }
 
   public static function findRevertedCommits($message) {
-    $reverts = array();
+    $matches = id(new DifferentialCustomFieldRevertsParser())
-    $matches = null;
+      ->parseCorpus($message);
-
-    // NOTE: Git language is "This reverts commit X."
-    // NOTE: Mercurial language is "Backed out changeset Y".
-
-    $prefixes = array(
-      'revert'        => true,
-      'reverts'       => true,
-      'back\s*out'    => true,
-      'backs\s*out'   => true,
-      'backed\s*out'  => true,
-      'undo'          => true,
-      'undoes'        => true,
-    );
-
-    $optional = array(
-      'commit'        => true,
-      'changeset'     => true,
-      'rev'           => true,
-      'revision'      => true,
-      'change'        => true,
-      'diff'          => true,
-    );
-
-    $pre_re = implode('|', array_keys($prefixes));
-    $opt_re = implode('|', array_keys($optional));
-
-    $matches = null;
-    preg_match_all(
-      '/\b(?i:'.$pre_re.')(?:\s+(?i:'.$opt_re.'))?([rA-Z0-9a-f,\s]+)\b/',
-      $message,
-      $matches);
 
     $result = array();
-    foreach ($matches[1] as $commits) {
+    foreach ($matches as $match) {
-      $commits = preg_split('/[,\s]+/', $commits);
+      foreach ($match['monograms'] as $monogram) {
-      $commits = array_filter($commits);
+        $result[$monogram] = $monogram;
-      foreach ($commits as $commit) {
-        $result[$commit] = $commit;
       }
     }
 

src/applications/maniphest/field/parser/ManiphestCustomFieldStatusParser.php

@@ -0,0 +1,59 @@
+<?php
+
+final class ManiphestCustomFieldStatusParser
+  extends PhabricatorCustomFieldMonogramParser {
+
+  protected function getPrefixes() {
+    return array(
+      'resolve',
+      'resolves',
+      'resolved',
+      'fix',
+      'fixes',
+      'fixed',
+      'wontfix',
+      'wontfixes',
+      'wontfixed',
+      'spite',
+      'spites',
+      'spited',
+      'invalidate',
+      'invalidates',
+      'invalidated',
+      'close',
+      'closes',
+      'closed',
+      'ref',
+      'refs',
+      'references',
+      'cf.',
+    );
+  }
+
+  protected function getInfixes() {
+    return array(
+      'task',
+      'tasks',
+      'issue',
+      'issues',
+      'bug',
+      'bugs',
+    );
+  }
+
+  protected function getSuffixes() {
+    return array(
+      'as resolved',
+      'as fixed',
+      'as wontfix',
+      'as spite',
+      'out of spite',
+      'as invalid',
+    );
+  }
+
+  protected function getMonogramPattern() {
+    return '[tT]\d+';
+  }
+
+}

src/applications/maniphest/field/parser/__tests__/ManiphestCustomFieldStatusParserTestCase.php

@@ -0,0 +1,63 @@
+<?php
+
+final class ManiphestCustomFieldStatusParserTestCase
+  extends PhabricatorTestCase {
+
+  public function testParser() {
+    $map = array(
+      'quack quack quack' => array(),
+      'T123' => array(),
+      'Fixes T123' => array(
+        array(
+          'match' => 'Fixes T123',
+          'prefix' => 'Fixes',
+          'infix' => '',
+          'monograms' => array('T123'),
+          'suffix' => '',
+          'offset' => 0,
+        ),
+      ),
+      'Fixes T123, T124, and also some other bugs.' => array(
+        array(
+          'match' => 'Fixes T123, T124, ',
+          'prefix' => 'Fixes',
+          'infix' => '',
+          'monograms' => array('T123', 'T124'),
+          'suffix' => '',
+          'offset' => 0,
+        ),
+      ),
+      'Closes T1 as wontfix' => array(
+        array(
+          'match' => 'Closes T1 as wontfix',
+          'prefix' => 'Closes',
+          'infix' => '',
+          'monograms' => array('T1'),
+          'suffix' => 'as wontfix',
+          'offset' => 0,
+        ),
+      ),
+      'Fixes task T9' => array(
+        array(
+          'match' => 'Fixes task T9',
+          'prefix' => 'Fixes',
+          'infix' => 'task',
+          'monograms' => array('T9'),
+          'suffix' => '',
+          'offset' => 0,
+        ),
+      ),
+      'Fixes t2apps' => array(),
+      'fixes a bug' => array(),
+      'Prefixes T2' => array(),
+    );
+
+    foreach ($map as $input => $expect) {
+      $parser = new ManiphestCustomFieldStatusParser();
+      $output = $parser->parseCorpus($input);
+
+      $this->assertEqual($expect, $output, $input);
+    }
+  }
+
+}

src/infrastructure/customfield/parser/PhabricatorCustomFieldMonogramParser.php

@@ -47,7 +47,7 @@ abstract class PhabricatorCustomFieldMonogramParser
         'prefix' => $set[1][0],
         'infix' => $set[2][0],
         'monograms' => array_filter(preg_split('/[,\s]+/', $set[3][0])),
-        'suffix' => $set[4][0],
+        'suffix' => idx(idx($set, 4, array()), 0, ''),
         'offset' => $set[0][1],
       );
     }