added support for utf8mb4 character sets. Closes #73

Author: ifsnop

README.md

@@ -125,12 +125,14 @@ Refer to the [wiki](https://github.com/ifsnop/mysqldump-php/wiki/full-example) f
         'hex-blob' => true,
         'databases' => false,
         'add-drop-database' => false,
-        'skip-tz-utc' => false
+        'skip-tz-utc' => false,
+        'no-autocommit' => true,
+        'default-character-set' => 'utf8',
     );
 
-    $pdoSettingsDefaults = array(PDO::ATTR_PERSISTENT => true,
+    $pdoSettingsDefaults = array(
+        PDO::ATTR_PERSISTENT => true,
         PDO::ATTR_ERRMODE => PDO::ERRMODE_EXCEPTION,
-        PDO::MYSQL_ATTR_INIT_COMMAND => "SET NAMES utf8",
         PDO::MYSQL_ATTR_USE_BUFFERED_QUERY => false
     );
 
@@ -177,6 +179,14 @@ Refer to the [wiki](https://github.com/ifsnop/mysqldump-php/wiki/full-example) f
   - http://dev.mysql.com/doc/refman/5.1/en/mysqldump.html#option_mysqldump_add-drop-database
 - **skip-tz-utz**
   - http://dev.mysql.com/doc/refman/5.1/en/mysqldump.html#option_mysqldump_tz-utc
+- **no-autocommit**
+  - Option to disable autocommit (faster inserts, no problems with index keys)
+  - http://dev.mysql.com/doc/refman/4.1/en/commit.html
+- **default-character-set**
+  - utf8 (default, compatible option), utf8mb4 (for full utf8 compliance)
+  - Could be specified using the declared consts: IMysqldump\Mysqldump::UTF8 or IMysqldump\Mysqldump::UTF8MB4BZIP2
+  - http://dev.mysql.com/doc/refman/5.5/en/charset-unicode-utf8mb4.html
+  - https://mathiasbynens.be/notes/mysql-utf8mb4
 
 The following options are now enabled by default, and there is no way to disable them since
 they should always be used.
@@ -215,7 +225,6 @@ Use **SHOW GRANTS FOR user@host;** to know what privileges user has. See the fol
 
 ## TODO
 
-- Write any tests.
 - Write unit tests.
 
 ## Contributing

src/Ifsnop/Mysqldump/Mysqldump.php

@@ -41,6 +41,10 @@ class Mysqldump
     const BZIP2 = 'Bzip2';
     const NONE = 'None';
 
+    // Available connection strings
+    const UTF8 = 'utf8';
+    const UTF8MB4 = 'utf8mb4';
+
     // This can be set both on constructor or manually
     public $host;
     public $user;
@@ -85,7 +89,7 @@ public function __construct(
         $dumpSettingsDefault = array(
             'include-tables' => array(),
             'exclude-tables' => array(),
-            'compress' => 'None',
+            'compress' => Mysqldump::NONE,
             'no-data' => false,
             'add-drop-table' => false,
             'single-transaction' => true,
@@ -102,13 +106,14 @@ public function __construct(
             'add-drop-database' => false,
             'skip-tz-utz' => false,
             'no-autocommit' => true,
+            'default-character-set' => Mysqldump::UTF8,
             /* deprecated */
             'disable-foreign-keys-check' => true
         );
 
-        $pdoSettingsDefault = array(PDO::ATTR_PERSISTENT => true,
+        $pdoSettingsDefault = array(
+            PDO::ATTR_PERSISTENT => true,
             PDO::ATTR_ERRMODE => PDO::ERRMODE_EXCEPTION,
-            PDO::MYSQL_ATTR_INIT_COMMAND => "SET NAMES utf8",
             PDO::MYSQL_ATTR_USE_BUFFERED_QUERY => false
         );
 
@@ -120,6 +125,10 @@ public function __construct(
         $this->pdoSettings = self::array_replace_recursive($pdoSettingsDefault, $pdoSettings);
         $this->dumpSettings = self::array_replace_recursive($dumpSettingsDefault, $dumpSettings);
 
+        if (!isset($this->pdoSettings[PDO::MYSQL_ATTR_INIT_COMMAND])) {
+            $this->pdoSetttings[PDO::MYSQL_ATTR_INIT_COMMAND] = "SET NAMES " . $this->dumpSettings['default-character-set'];
+        }
+
         $diff = array_diff(array_keys($this->dumpSettings), array_keys($dumpSettingsDefault));
         if (count($diff)>0) {
             throw new Exception("Unexpected value in dumpSettings: (" . implode(",", $diff) . ")");
@@ -178,7 +187,7 @@ private function connect()
                         $this->pdoSettings
                     );
                     // Fix for always-unicode output
-                    $this->dbHandler->exec("SET NAMES utf8");
+                    $this->dbHandler->exec("SET NAMES " . $this->dumpSettings['default-character-set']);
                     // Store server version
                     $this->version = $this->dbHandler->getAttribute(PDO::ATTR_SERVER_VERSION);
                     break;
@@ -441,7 +450,7 @@ private function getTableStructure($tableName)
                     );
                 }
                 $this->compressManager->write(
-                    $this->typeAdapter->create_table($r)
+                    $this->typeAdapter->create_table($r, $this->dumpSettings)
                 );
                 break;
             }
@@ -915,7 +924,7 @@ public function show_create_table($tableName)
      * function create_table Get table creation code from database
      * @todo make it do something with sqlite
      */
-    public function create_table($row)
+    public function create_table($row, $dumpSettings)
     {
         return "";
     }
@@ -1163,14 +1172,14 @@ public function show_create_trigger($triggerName)
         return "SHOW CREATE TRIGGER `$triggerName`";
     }
 
-    public function create_table($row)
+    public function create_table($row, $dumpSettings)
     {
         if (!isset($row['Create Table'])) {
             throw new Exception("Error getting table code, unknown output");
         }
 
         $ret = "/*!40101 SET @saved_cs_client     = @@character_set_client */;" . PHP_EOL .
-            "/*!40101 SET character_set_client = utf8 */;" . PHP_EOL .
+            "/*!40101 SET character_set_client = " . $dumpSettings['default-character-set'] . " */;" . PHP_EOL .
             $row['Create Table'] . ";" . PHP_EOL .
             "/*!40101 SET character_set_client = @saved_cs_client */;" . PHP_EOL .
             PHP_EOL;
@@ -1462,7 +1471,7 @@ public function backup_parameters()
         $ret = "/*!40101 SET @OLD_CHARACTER_SET_CLIENT=@@CHARACTER_SET_CLIENT */;" . PHP_EOL .
             "/*!40101 SET @OLD_CHARACTER_SET_RESULTS=@@CHARACTER_SET_RESULTS */;" . PHP_EOL .
             "/*!40101 SET @OLD_COLLATION_CONNECTION=@@COLLATION_CONNECTION */;" . PHP_EOL .
-            "/*!40101 SET NAMES utf8 */;" . PHP_EOL;
+            "/*!40101 SET NAMES " . $dumpSettings['default-character-set'] . " */;" . PHP_EOL;
 
         if (false === $dumpSettings['skip-tz-utz']) {
             $ret .= "/*!40103 SET @OLD_TIME_ZONE=@@TIME_ZONE */;" . PHP_EOL .

tests/test.php

@@ -37,6 +37,19 @@
     "mysql",
     $dumpSettings);
 
-$dump->start("mysqldump-php.sql");
+$dump->start("mysqldump-php_test001.sql");
+
+
+$dumpSettings['default-character-set'] = IMysqldump\Mysqldump::UTF8MB4;
+
+$dump = new IMysqldump\Mysqldump(
+    "test002",
+    "travis",
+    "",
+    "localhost",
+    "mysql",
+    $dumpSettings);
+
+$dump->start("mysqldump-php_test002.sql");
 
 exit;

tests/test.sh

@@ -1,59 +1,87 @@
 #!/bin/bash
 
-function checksum() {
+function checksum_test001() {
 for i in 000 001 002 003 010 011 015 027 029 033 200; do
     mysql -B -e "CHECKSUM TABLE test${i}" test001 | grep -v -i checksum
 done
 }
 
+function checksum_test002() {
+for i in 201; do
+    mysql --default-character-set=utf8mb4 -B -e "CHECKSUM TABLE test${i}" test002 | grep -v -i checksum
+done
+}
+
 for i in $(seq 0 20) ; do
-    ret[i]=0
+    ret[$i]=0
 done
 
-mysql -uroot < original.sql
-ret[0]=$?
+mysql -e "CREATE USER 'travis'@'localhost' IDENTIFIED BY '';" 2> /dev/null
+mysql -e "GRANT ALL PRIVILEGES ON test001.* TO 'travis'@'localhost';" 2> /dev/null
+mysql -e "GRANT ALL PRIVILEGES ON test002.* TO 'travis'@'localhost';" 2> /dev/null
+
+mysql -uroot < test001.src.sql; ret[0]=$?
+
+mysql -uroot --default-character-set=utf8mb4 < test002.src.sql; ret[1]=$?
 
-checksum > original.checksum
+checksum_test001 > test001.src.checksum
+checksum_test002 > test002.src.checksum
 
 mysqldump -uroot test001 \
     --no-autocommit \
     --extended-insert=false \
     --hex-blob=true \
-    > mysqldump.sql
-ret[1]=$?
-
-php test.php
+    > mysqldump_test001.sql
 ret[2]=$?
 
-mysql -uroot test001 < mysqldump-php.sql
+mysqldump -uroot test002 \
+    --no-autocommit \
+    --extended-insert=false \
+    --hex-blob=true \
+    --default-character-set=utf8mb4 \
+    > mysqldump_test002.sql
 ret[3]=$?
 
-checksum > mysqldump-php.checksum
-
-cat original.sql | grep ^INSERT > original.filtered.sql
-cat mysqldump.sql | grep ^INSERT > mysqldump.filtered.sql
-cat mysqldump-php.sql | grep ^INSERT > mysqldump-php.filtered.sql
-
-diff original.filtered.sql mysqldump.filtered.sql
+php test.php
 ret[4]=$?
 
-diff original.filtered.sql mysqldump-php.filtered.sql
+mysql -uroot test001 < mysqldump-php_test001.sql
 ret[5]=$?
-
-diff original.checksum mysqldump-php.checksum
+mysql -uroot test002 < mysqldump-php_test002.sql
 ret[6]=$?
 
-rm original.checksum
-rm mysqldump-php.checksum
-rm original.filtered.sql
-rm mysqldump.filtered.sql
-rm mysqldump-php.filtered.sql
-rm mysqldump-php.sql
-rm mysqldump.sql
+checksum_test001 > mysqldump-php_test001.checksum
+checksum_test002 > mysqldump-php_test002.checksum
+
+cat test001.src.sql | grep ^INSERT > test001.filtered.sql
+cat test002.src.sql | grep ^INSERT > test002.filtered.sql
+cat mysqldump_test001.sql | grep ^INSERT > mysqldump_test001.filtered.sql
+cat mysqldump_test002.sql | grep ^INSERT > mysqldump_test002.filtered.sql
+cat mysqldump-php_test001.sql | grep ^INSERT > mysqldump-php_test001.filtered.sql
+cat mysqldump-php_test002.sql | grep ^INSERT > mysqldump-php_test002.filtered.sql
+
+diff test001.filtered.sql mysqldump_test001.filtered.sql
+ret[7]=$?
+diff test002.filtered.sql mysqldump_test002.filtered.sql
+ret[8]=$?
+
+diff test001.filtered.sql mysqldump-php_test001.filtered.sql
+ret[9]=$?
+diff test002.filtered.sql mysqldump-php_test002.filtered.sql
+ret[10]=$?
+
+diff test001.src.checksum mysqldump-php_test001.checksum
+ret[11]=$?
+diff test002.src.checksum mysqldump-php_test002.checksum
+ret[12]=$?
+
+rm *.checksum 2> /dev/null
+rm *.filtered.sql 2> /dev/null
+rm mysqldump* 2> /dev/null
 
 total=0
 for i in $(seq 0 20) ; do
-    total=(${ret[i]} + $total)
+    total=$((${ret[$i]} + $total))
 done
 
 exit $total

tests/original.sql renamed to tests/test001.src.sql

@@ -0,0 +1,13 @@
+DROP DATABASE IF EXISTS `test002`;
+CREATE DATABASE `test002`;
+USE `test002`;
+
+DROP TABLE IF EXISTS `test201`;
+CREATE TABLE `test201` (
+  `col` text COLLATE utf8mb4_unicode_ci
+) ENGINE=InnoDB DEFAULT CHARSET=utf8mb4 COLLATE=utf8mb4_unicode_ci;
+
+INSERT INTO `test201` VALUES ('áéíóú');
+INSERT INTO `test201` VALUES ('🎲');
+INSERT INTO `test201` VALUES ('🎭');
+INSERT INTO `test201` VALUES ('💩');