Add PhabricatorHash::digestForIndex()

Summary: Does this seem reasonable? It's a bit more compact than digest() (6 bits / byte instead of 4 bits / byte) and 72 bits of entropy @ 12 bytes instead of 128 bits of entropy @ 32 bytes. I feel like it's important to preserve the printability, though, and this seemed like a fairly good balance of concerns.

Test Plan: unit tests

Reviewers: vrana

Reviewed By: vrana

CC: aran, yemao932

Differential Revision: https://secure.phabricator.com/D4253

src/__phutil_library_map__.php

@@ -813,6 +813,7 @@ phutil_register_library_map(array(
     'PhabricatorGlobalUploadTargetView' => 'applications/files/view/PhabricatorGlobalUploadTargetView.php',
     'PhabricatorHandleObjectSelectorDataView' => 'applications/phid/handle/view/PhabricatorHandleObjectSelectorDataView.php',
     'PhabricatorHash' => 'infrastructure/util/PhabricatorHash.php',
+    'PhabricatorHashTestCase' => 'infrastructure/util/__tests__/PhabricatorHashTestCase.php',
     'PhabricatorHeaderView' => 'view/layout/PhabricatorHeaderView.php',
     'PhabricatorHelpController' => 'applications/help/controller/PhabricatorHelpController.php',
     'PhabricatorHelpKeyboardShortcutController' => 'applications/help/controller/PhabricatorHelpKeyboardShortcutController.php',
@@ -2098,6 +2099,7 @@ phutil_register_library_map(array(
     'PhabricatorGarbageCollectorDaemon' => 'PhabricatorDaemon',
     'PhabricatorGlobalLock' => 'PhutilLock',
     'PhabricatorGlobalUploadTargetView' => 'AphrontView',
+    'PhabricatorHashTestCase' => 'PhabricatorTestCase',
     'PhabricatorHeaderView' => 'AphrontView',
     'PhabricatorHelpController' => 'PhabricatorController',
     'PhabricatorHelpKeyboardShortcutController' => 'PhabricatorHelpController',

src/infrastructure/util/PhabricatorHash.php

@@ -2,6 +2,13 @@
 
 final class PhabricatorHash {
 
+
+  /**
+   * Digest a string for general use, including use which relates to security.
+   *
+   * @param   string  Input string.
+   * @return  string  32-byte hexidecimal SHA1+HMAC hash.
+   */
   public static function digest($string) {
     $key = PhabricatorEnv::getEnvConfig('security.hmac-key');
     if (!$key) {
@@ -12,4 +19,39 @@ final class PhabricatorHash {
     return hash_hmac('sha1', $string, $key);
   }
 
+
+  /**
+   * Digest a string for use in, e.g., a MySQL index. This produces a short
+   * (12-byte), case-sensitive alphanumeric string with 72 bits of entropy,
+   * which is generally safe in most contexts (notably, URLs).
+   *
+   * This method emphasizes compactness, and should not be used for security
+   * related hashing (for general purpose hashing, see @{method:digest}).
+   *
+   * @param   string  Input string.
+   * @return  string  12-byte, case-sensitive alphanumeric hash of the string
+   *                  which
+   */
+  public static function digestForIndex($string) {
+    $hash = sha1($string, $raw_output = true);
+
+    static $map;
+    if ($map === null) {
+      $map = "0123456789".
+             "abcdefghij".
+             "klmnopqrst".
+             "uvwxyzABCD".
+             "EFGHIJKLMN".
+             "OPQRSTUVWX".
+             "YZ._";
+    }
+
+    $result = '';
+    for ($ii = 0; $ii < 12; $ii++) {
+      $result .= $map[(ord($hash[$ii]) & 0x3F)];
+    }
+
+    return $result;
+  }
+
 }

src/infrastructure/util/__tests__/PhabricatorHashTestCase.php

@@ -0,0 +1,41 @@
+<?php
+
+final class PhabricatorHashTestCase extends PhabricatorTestCase {
+
+  public function testHashForIndex() {
+    $map = array(
+      'dog' => 'Aliif7Qjd5ct',
+      'cat' => 'toudDsue3Uv8',
+      'rat' => 'RswaKgTjqOuj',
+      'bat' => 'rAkJKyX4YdYm',
+    );
+
+    foreach ($map as $input => $expect) {
+      $this->assertEqual(
+        $expect,
+        PhabricatorHash::digestForIndex($input),
+        "Input: {$input}");
+    }
+
+    // Test that the encoding produces 6 bits of entropy per byte.
+    $entropy = array(
+      'dog', 'cat', 'rat', 'bat', 'dig', 'fig', 'cot',
+      'cut', 'fog', 'rig', 'rug', 'dug', 'mat', 'pat',
+      'eat', 'tar', 'pot',
+    );
+
+    $seen = array();
+    foreach ($entropy as $input) {
+      $chars = preg_split('//', PhabricatorHash::digestForIndex($input));
+      foreach ($chars as $char) {
+        $seen[$char] = true;
+      }
+    }
+
+    $this->assertEqual(
+      (1 << 6),
+      count($seen),
+      "Distinct characters in hash of: {$input}");
+  }
+
+}