src/applications/diffusion/query/lowlevel/DiffusionLowLevelCommitQuery.php

   1 <?php
   2
   3 /**
   4  * Populate a @{class:DiffusionCommitRef} with information about a specific
   5  * commit in a repository. This is a low-level query which talks directly to
   6  * the underlying VCS.
   7  */
   8 final class DiffusionLowLevelCommitQuery
   9   extends DiffusionLowLevelQuery {
  10
  11   private $identifier;
  12
  13   public function withIdentifier($identifier) {
  14     $this->identifier = $identifier;
  15     return $this;
  16   }
  17
  18   protected function executeQuery() {
  19     if (!strlen($this->identifier)) {
  20       throw new PhutilInvalidStateException('withIdentifier');
  21     }
  22
  23     $type = $this->getRepository()->getVersionControlSystem();
  24     switch ($type) {
  25       case PhabricatorRepositoryType::REPOSITORY_TYPE_GIT:
  26         $result = $this->loadGitCommitRef();
  27         break;
  28       case PhabricatorRepositoryType::REPOSITORY_TYPE_MERCURIAL:
  29         $result = $this->loadMercurialCommitRef();
  30         break;
  31       case PhabricatorRepositoryType::REPOSITORY_TYPE_SVN:
  32         $result = $this->loadSubversionCommitRef();
  33         break;
  34       default:
  35         throw new Exception(pht('Unsupported repository type "%s"!', $type));
  36     }
  37
  38     return $result;
  39   }
  40
  41   private function loadGitCommitRef() {
  42     $repository = $this->getRepository();
  43
  44     // See T5028. The "%B" (raw body) mode is not present in very old versions
  45     // of Git. Use "%s" and "%b" ("subject" and "wrapped body") as an
  46     // approximation.
  47
  48     $git_binary = PhutilBinaryAnalyzer::getForBinary('git');
  49     $git_version = $git_binary->getBinaryVersion();
  50     if (version_compare($git_version, '1.7.2', '>=')) {
  51       $body_format = '%B';
  52       $split_body = false;
  53     } else {
  54       $body_format = '%s%x00%b';
  55       $split_body = true;
  56     }
  57
  58     $argv = array();
  59
  60     $argv[] = '-n';
  61     $argv[] = '1';
  62
  63     $argv[] = '--encoding=UTF-8';
  64
  65     $argv[] = sprintf(
  66       '--format=%s',
  67       implode(
  68         '%x00',
  69         array(
  70           '%e',
  71           '%cn',
  72           '%ce',
  73           '%an',
  74           '%ae',
  75           '%T',
  76           '%at',
  77           $body_format,
  78
  79           // The "git log" output includes a trailing newline. We want to
  80           // faithfully capture only the exact text of the commit message,
  81           // so include an explicit terminator: this makes sure the exact
  82           // body text is surrounded by "\0" characters.
  83           '~',
  84         )));
  85
  86     // Even though we pass --encoding here, git doesn't always succeed, so
  87     // we try a little harder, since git *does* tell us what the actual encoding
  88     // is correctly (unless it doesn't; encoding is sometimes empty).
  89     list($info) = $repository->execxLocalCommand(
  90       'log -n 1 %Ls %s --',
  91       $argv,
  92       gitsprintf('%s', $this->identifier));
  93
  94     $parts = explode("\0", $info);
  95     $encoding = array_shift($parts);
  96
  97     foreach ($parts as $key => $part) {
  98       if ($encoding) {
  99         $part = phutil_utf8_convert($part, 'UTF-8', $encoding);
 100       }
 101       $parts[$key] = phutil_utf8ize($part);
 102       if (!strlen($parts[$key])) {
 103         $parts[$key] = null;
 104       }
 105     }
 106
 107     $hashes = array(
 108       id(new DiffusionCommitHash())
 109         ->setHashType(ArcanistDifferentialRevisionHash::HASH_GIT_COMMIT)
 110         ->setHashValue($this->identifier),
 111       id(new DiffusionCommitHash())
 112         ->setHashType(ArcanistDifferentialRevisionHash::HASH_GIT_TREE)
 113         ->setHashValue($parts[4]),
 114     );
 115
 116     $author_epoch = (int)$parts[5];
 117     if (!$author_epoch) {
 118       $author_epoch = null;
 119     }
 120
 121     if ($split_body) {
 122       // Here, the body is: "subject", "\0", "wrapped body". Stitch the
 123       // pieces back together by putting a newline between them if both
 124       // parts are nonempty.
 125
 126       $head = $parts[6];
 127       $tail = $parts[7];
 128
 129       if (strlen($head) && strlen($tail)) {
 130         $body = $head."\n\n".$tail;
 131       } else if (strlen($head)) {
 132         $body = $head;
 133       } else if (strlen($tail)) {
 134         $body = $tail;
 135       } else {
 136         $body = '';
 137       }
 138     } else {
 139       // Here, the body is the raw unwrapped body.
 140       $body = $parts[6];
 141     }
 142
 143     return id(new DiffusionCommitRef())
 144       ->setCommitterName($parts[0])
 145       ->setCommitterEmail($parts[1])
 146       ->setAuthorName($parts[2])
 147       ->setAuthorEmail($parts[3])
 148       ->setHashes($hashes)
 149       ->setAuthorEpoch($author_epoch)
 150       ->setMessage($body);
 151   }
 152
 153   private function loadMercurialCommitRef() {
 154     $repository = $this->getRepository();
 155
 156     list($stdout) = $repository->execxLocalCommand(
 157       'log --template %s --rev %s',
 158       '{author}\\n{desc}',
 159       hgsprintf('%s', $this->identifier));
 160
 161     list($author, $message) = explode("\n", $stdout, 2);
 162
 163     $author = phutil_utf8ize($author);
 164     $message = phutil_utf8ize($message);
 165
 166     list($author_name, $author_email) = $this->splitUserIdentifier($author);
 167
 168     $hashes = array(
 169       id(new DiffusionCommitHash())
 170         ->setHashType(ArcanistDifferentialRevisionHash::HASH_MERCURIAL_COMMIT)
 171         ->setHashValue($this->identifier),
 172     );
 173
 174     return id(new DiffusionCommitRef())
 175       ->setAuthorName($author_name)
 176       ->setAuthorEmail($author_email)
 177       ->setMessage($message)
 178       ->setHashes($hashes);
 179   }
 180
 181   private function loadSubversionCommitRef() {
 182     $repository = $this->getRepository();
 183
 184     list($xml) = $repository->execxRemoteCommand(
 185       'log --xml --limit 1 %s',
 186       $repository->getSubversionPathURI(null, $this->identifier));
 187
 188     // Subversion may send us back commit messages which won't parse because
 189     // they have non UTF-8 garbage in them. Slam them into valid UTF-8.
 190     $xml = phutil_utf8ize($xml);
 191     $log = new SimpleXMLElement($xml);
 192     $entry = $log->logentry[0];
 193
 194     $author = (string)$entry->author;
 195     $message = (string)$entry->msg;
 196
 197     list($author_name, $author_email) = $this->splitUserIdentifier($author);
 198
 199     // No hashes in Subversion.
 200     $hashes = array();
 201
 202     return id(new DiffusionCommitRef())
 203       ->setAuthorName($author_name)
 204       ->setAuthorEmail($author_email)
 205       ->setMessage($message)
 206       ->setHashes($hashes);
 207   }
 208
 209   private function splitUserIdentifier($user) {
 210     $email = new PhutilEmailAddress($user);
 211
 212     if ($email->getDisplayName() || $email->getDomainName()) {
 213       $user_name = $email->getDisplayName();
 214       $user_email = $email->getAddress();
 215     } else {
 216       $user_name = $email->getAddress();
 217       $user_email = null;
 218     }
 219
 220     return array($user_name, $user_email);
 221   }
 222
 223 }