Consolidate repository identity resolution and detection code

Summary: Ref T13444. Send all repository identity/detection through a new "DiffusionRepositoryIdentityEngine" which handles resolution and detection updates in one place.

Test Plan:
  - Ran `bin/repository reparse --message ...`, saw author/committer identity updates.
  - Added "goose@example.com" to my email addresses, ran daemons, saw the identity relationship get picked up.
  - Ran `bin/repository rebuild-identities ...`, saw sensible rebuilds.

Maniphest Tasks: T13444

Differential Revision: https://secure.phabricator.com/D20910

src/__phutil_library_map__.php

@@ -985,6 +985,7 @@ phutil_register_library_map(array(
     'DiffusionRepositoryFunctionDatasource' => 'applications/diffusion/typeahead/DiffusionRepositoryFunctionDatasource.php',
     'DiffusionRepositoryHistoryManagementPanel' => 'applications/diffusion/management/DiffusionRepositoryHistoryManagementPanel.php',
     'DiffusionRepositoryIdentityEditor' => 'applications/diffusion/editor/DiffusionRepositoryIdentityEditor.php',
+    'DiffusionRepositoryIdentityEngine' => 'applications/diffusion/identity/DiffusionRepositoryIdentityEngine.php',
     'DiffusionRepositoryIdentitySearchEngine' => 'applications/diffusion/query/DiffusionRepositoryIdentitySearchEngine.php',
     'DiffusionRepositoryLimitsManagementPanel' => 'applications/diffusion/management/DiffusionRepositoryLimitsManagementPanel.php',
     'DiffusionRepositoryListController' => 'applications/diffusion/controller/DiffusionRepositoryListController.php',
@@ -6966,6 +6967,7 @@ phutil_register_library_map(array(
     'DiffusionRepositoryFunctionDatasource' => 'PhabricatorTypeaheadCompositeDatasource',
     'DiffusionRepositoryHistoryManagementPanel' => 'DiffusionRepositoryManagementPanel',
     'DiffusionRepositoryIdentityEditor' => 'PhabricatorApplicationTransactionEditor',
+    'DiffusionRepositoryIdentityEngine' => 'Phobject',
     'DiffusionRepositoryIdentitySearchEngine' => 'PhabricatorApplicationSearchEngine',
     'DiffusionRepositoryLimitsManagementPanel' => 'DiffusionRepositoryManagementPanel',
     'DiffusionRepositoryListController' => 'DiffusionController',

src/applications/diffusion/identity/DiffusionRepositoryIdentityEngine.php

@@ -0,0 +1,80 @@
+<?php
+
+final class DiffusionRepositoryIdentityEngine
+  extends Phobject {
+
+  private $viewer;
+  private $sourcePHID;
+
+  public function setViewer(PhabricatorUser $viewer) {
+    $this->viewer = $viewer;
+    return $this;
+  }
+
+  public function getViewer() {
+    return $this->viewer;
+  }
+
+  public function setSourcePHID($source_phid) {
+    $this->sourcePHID = $source_phid;
+    return $this;
+  }
+
+  public function getSourcePHID() {
+    if (!$this->sourcePHID) {
+      throw new PhutilInvalidStateException('setSourcePHID');
+    }
+
+    return $this->sourcePHID;
+  }
+
+  public function newResolvedIdentity($raw_identity) {
+    $identity = $this->loadRawIdentity($raw_identity);
+
+    if (!$identity) {
+      $identity = $this->newIdentity($raw_identity);
+    }
+
+    return $this->updateIdentity($identity);
+  }
+
+  public function newUpdatedIdentity(PhabricatorRepositoryIdentity $identity) {
+    return $this->updateIdentity($identity);
+  }
+
+  private function loadRawIdentity($raw_identity) {
+    $viewer = $this->getViewer();
+
+    return id(new PhabricatorRepositoryIdentityQuery())
+      ->setViewer($viewer)
+      ->withIdentityNames(array($raw_identity))
+      ->executeOne();
+  }
+
+  private function newIdentity($raw_identity) {
+    $source_phid = $this->getSourcePHID();
+
+    return id(new PhabricatorRepositoryIdentity())
+      ->setAuthorPHID($source_phid)
+      ->setIdentityName($raw_identity);
+  }
+
+  private function resolveIdentity(PhabricatorRepositoryIdentity $identity) {
+    $raw_identity = $identity->getIdentityName();
+
+    return id(new DiffusionResolveUserQuery())
+      ->withName($raw_identity)
+      ->execute();
+  }
+
+  private function updateIdentity(PhabricatorRepositoryIdentity $identity) {
+    $resolved_phid = $this->resolveIdentity($identity);
+
+    $identity
+      ->setAutomaticGuessedUserPHID($resolved_phid)
+      ->save();
+
+    return $identity;
+  }
+
+}

src/applications/repository/management/PhabricatorRepositoryManagementRebuildIdentitiesWorkflow.php

@@ -3,6 +3,8 @@
 final class PhabricatorRepositoryManagementRebuildIdentitiesWorkflow
   extends PhabricatorRepositoryManagementWorkflow {
 
+  private $identityCache = array();
+
   protected function didConstruct() {
     $this
       ->setName('rebuild-identities')
@@ -94,31 +96,21 @@ final class PhabricatorRepositoryManagementRebuildIdentitiesWorkflow
   }
 
   private function getIdentityForCommit(
-    PhabricatorRepositoryCommit $commit, $identity_name) {
+    PhabricatorRepositoryCommit $commit,
+    $raw_identity) {
 
-    static $seen = array();
-    $identity_key = PhabricatorHash::digestForIndex($identity_name);
-    if (empty($seen[$identity_key])) {
-      try {
-        $user_phid = id(new DiffusionResolveUserQuery())
-          ->withName($identity_name)
-          ->execute();
+    if (!isset($this->identityCache[$raw_identity])) {
+      $viewer = $this->getViewer();
 
-        $identity = id(new PhabricatorRepositoryIdentity())
-          ->setAuthorPHID($commit->getPHID())
-          ->setIdentityName($identity_name)
-          ->setAutomaticGuessedUserPHID($user_phid)
-          ->save();
-      } catch (AphrontDuplicateKeyQueryException $ex) {
-          // Somehow this identity already exists?
-        $identity = id(new PhabricatorRepositoryIdentityQuery())
-          ->setViewer(PhabricatorUser::getOmnipotentUser())
-          ->withIdentityNames(array($identity_name))
-          ->executeOne();
-      }
-      $seen[$identity_key] = $identity;
+      $identity = id(new DiffusionRepositoryIdentityEngine())
+        ->setViewer($viewer)
+        ->setSourcePHID($commit->getPHID())
+        ->newResolvedIdentity($raw_identity);
+
+      $this->identityCache[$raw_identity] = $identity;
     }
 
-    return $seen[$identity_key];
+    return $this->identityCache[$raw_identity];
   }
+
 }

src/applications/repository/worker/PhabricatorRepositoryIdentityChangeWorker.php

@@ -1,7 +1,7 @@
 <?php
 
 final class PhabricatorRepositoryIdentityChangeWorker
-extends PhabricatorWorker {
+  extends PhabricatorWorker {
 
   protected function doWork() {
     $viewer = PhabricatorUser::getOmnipotentUser();
@@ -15,18 +15,20 @@ extends PhabricatorWorker {
       ->executeOne();
 
     $emails = id(new PhabricatorUserEmail())->loadAllWhere(
-      'userPHID = %s ORDER BY address',
+      'userPHID = %s',
       $user->getPHID());
 
+    $identity_engine = id(new DiffusionRepositoryIdentityEngine())
+      ->setViewer($viewer);
+
     foreach ($emails as $email) {
       $identities = id(new PhabricatorRepositoryIdentityQuery())
         ->setViewer($viewer)
-        ->withEmailAddresses($email->getAddress())
+        ->withEmailAddresses(array($email->getAddress()))
         ->execute();
 
       foreach ($identities as $identity) {
-        $identity->setAutomaticGuessedUserPHID($user->getPHID())
-          ->save();
+        $identity_engine->newUpdatedIdentity($identity);
       }
     }
   }

src/applications/repository/worker/commitmessageparser/PhabricatorRepositoryCommitMessageParserWorker.php

@@ -65,37 +65,20 @@ abstract class PhabricatorRepositoryCommitMessageParserWorker
     $message = $ref->getMessage();
     $committer = $ref->getCommitter();
     $hashes = $ref->getHashes();
+    $has_committer = (bool)strlen($committer);
 
-    $author_identity = id(new PhabricatorRepositoryIdentityQuery())
-      ->setViewer(PhabricatorUser::getOmnipotentUser())
-      ->withIdentityNames(array($author))
-      ->executeOne();
+    $viewer = PhabricatorUser::getOmnipotentUser();
 
-    if (!$author_identity) {
-      $author_identity = id(new PhabricatorRepositoryIdentity())
-        ->setAuthorPHID($commit->getPHID())
-        ->setIdentityName($author)
-        ->setAutomaticGuessedUserPHID(
-          $this->resolveUserPHID($commit, $author))
-        ->save();
-    }
+    $identity_engine = id(new DiffusionRepositoryIdentityEngine())
+      ->setViewer($viewer)
+      ->setSourcePHID($commit->getPHID());
 
+    $author_identity = $identity_engine->newResolvedIdentity($author);
+
+    if ($has_committer) {
+      $committer_identity = $identity_engine->newResolvedIdentity($committer);
+    } else {
       $committer_identity = null;
-
-    if ($committer) {
-      $committer_identity = id(new PhabricatorRepositoryIdentityQuery())
-        ->setViewer(PhabricatorUser::getOmnipotentUser())
-        ->withIdentityNames(array($committer))
-        ->executeOne();
-
-      if (!$committer_identity) {
-        $committer_identity = id(new PhabricatorRepositoryIdentity())
-          ->setAuthorPHID($commit->getPHID())
-          ->setIdentityName($committer)
-          ->setAutomaticGuessedUserPHID(
-            $this->resolveUserPHID($commit, $committer))
-          ->save();
-      }
     }
 
     $data = id(new PhabricatorRepositoryCommitData())->loadOneWhere(
@@ -117,11 +100,11 @@ abstract class PhabricatorRepositoryCommitMessageParserWorker
       'authorIdentityPHID', $author_identity->getPHID());
     $data->setCommitDetail(
       'authorPHID',
-      $this->resolveUserPHID($commit, $author));
+      $author_identity->getCurrentEffectiveUserPHID());
 
     $data->setCommitMessage($message);
 
-    if (strlen($committer)) {
+    if ($has_committer) {
       $data->setCommitDetail('committer', $committer);
 
       $data->setCommitDetail('committerName', $ref->getCommitterName());
@@ -129,7 +112,8 @@ abstract class PhabricatorRepositoryCommitMessageParserWorker
 
       $data->setCommitDetail(
         'committerPHID',
-        $this->resolveUserPHID($commit, $committer));
+        $committer_identity->getCurrentEffectiveUserPHID());
+
       $data->setCommitDetail(
         'committerIdentityPHID', $committer_identity->getPHID());
 
@@ -177,15 +161,6 @@ abstract class PhabricatorRepositoryCommitMessageParserWorker
       PhabricatorRepositoryCommit::IMPORTED_MESSAGE);
   }
 
-  private function resolveUserPHID(
-    PhabricatorRepositoryCommit $commit,
-    $user_name) {
-
-    return id(new DiffusionResolveUserQuery())
-      ->withName($user_name)
-      ->execute();
-  }
-
   private function closeRevisions(
     PhabricatorUser $actor,
     DiffusionCommitRef $ref,