source: trunk/phpgwapi/inc/adodb/docs/tips_portable_sql.htm @ 2

Revision 2, 19.4 KB checked in by niltonneto, 17 years ago (diff)

Removida todas as tags usadas pelo CVS ($Id, $Source).
Primeira versão no CVS externo.
  • Property svn:eol-style set to native
  • Property svn:executable set to *
Line 
1<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
2
3<html>
4<head>
5        <title>Tips on Writing Portable SQL for Multiple Databases for PHP</title>
6</head>
7
8<body bgcolor=white>
9<table width=100% border=0><tr><td><h2>Tips on Writing Portable SQL &nbsp;</h2></td><td>
10 <div align=right><img src="cute_icons_for_site/adodb.gif"></div></td></tr></table>
11  <p>Updated 18 Sep 2003. Added Portable Native SQL section.
12<p>
13
14 If you are writing an application that is used in multiple environments and
15  operating systems, you need to plan to support multiple databases. This article
16  is based on my experiences with multiple database systems, stretching from 4th
17  Dimension in my Mac days, to the databases I currently use, which are: Oracle,
18  FoxPro, Access, MS SQL Server and MySQL. Although most of the advice here applies
19  to using SQL with Perl, Python and other programming languages, I will focus on PHP and how
20  the <a href="http://adodb.sourceforge.net/">ADOdb</a> database abstraction library
21  offers some solutions.<p></p>
22<p>Most database vendors practice product lock-in. The best or fastest way to
23  do things is often implemented using proprietary extensions to SQL. This makes
24  it extremely hard to write portable SQL code that performs well under all conditions.
25  When the first ANSI committee got together in 1984 to standardize SQL, the database
26  vendors had such different implementations that they could only agree on the
27  core functionality of SQL. Many important application specific requirements
28  were not standardized, and after so many years since the ANSI effort began,
29  it looks as if much useful database functionality will never be standardized.
30  Even though ANSI-92 SQL has codified much more, we still have to implement portability
31  at the application level.</p>
32<h3><b>Selects</b></h3>
33<p>The SELECT statement has been standardized to a great degree. Nearly every
34  database supports the following:</p>
35<p>SELECT [cols] FROM [tables]<br>
36  &nbsp;&nbsp;[WHERE conditions]<br>
37  &nbsp; [GROUP BY cols]<br>
38  &nbsp; [HAVING conditions] <br>
39  &nbsp; [ORDER BY cols]</p>
40<p>But so many useful techniques can only be implemented by using proprietary
41  extensions. For example, when writing SQL to retrieve the first 10 rows for
42  paging, you could write...</p>
43<table width="80%" border="1" cellspacing="0" cellpadding="0" align="center">
44  <tr>
45        <td><b>Database</b></td>
46        <td><b>SQL Syntax</b></td>
47  </tr>
48  <tr>
49        <td>DB2</td>
50        <td>select * from table fetch first 10 rows only</td>
51  </tr>
52  <tr>
53        <td>Informix</td>
54        <td>select first 10 * from table</td>
55  </tr>
56  <tr>
57        <td>Microsoft SQL Server and Access</td>
58        <td>select top 10 * from table</td>
59  </tr>
60  <tr>
61        <td>MySQL and PostgreSQL</td>
62        <td>select * from table limit 10</td>
63  </tr>
64  <tr>
65        <td>Oracle 8i</td>
66        <td>select * from (select * from table) where rownum &lt;= 10</td>
67  </tr>
68</table>
69<p>This feature of getting a subset of data is so useful that in the PHP class
70  library ADOdb, we have a SelectLimit( ) function that allows you to hide the
71  implementation details within a function that will rewrite your SQL for you:</p>
72<pre>$connection-&gt;SelectLimit('select * from table', 10);
73</pre>
74<p><b>Selects: Fetch Modes</b></p>
75<p>PHP allows you to retrieve database records as arrays. You can choose to have
76  the arrays indexed by field name or number. However different low-level PHP
77  database drivers are inconsistent in their indexing efforts. ADOdb allows you
78  to determine your prefered mode. You set this by setting the variable $ADODB_FETCH_MODE
79  to either of the constants ADODB_FETCH_NUM (for numeric indexes) or ADODB_FETCH_ASSOC
80  (using field names as an associative index).</p>
81<p>The default behaviour of ADOdb varies depending on the database you are using.
82  For consistency, set the fetch mode to either ADODB_FETCH_NUM (for speed) or
83  ADODB_FETCH_ASSOC (for convenience) at the beginning of your code. </p>
84<p><b>Selects: Counting Records</b></p>
85<p>Another problem with SELECTs is that some databases do not return the number
86  of rows retrieved from a select statement. This is because the highest performance
87  databases will return records to you even before the last record has been found.
88</p>
89<p>In ADOdb, RecordCount( ) returns the number of rows returned, or will emulate
90  it by buffering the rows and returning the count after all rows have been returned.
91  This can be disabled for performance reasons when retrieving large recordsets
92  by setting the global variable $ADODB_COUNTRECS = false. This variable is checked
93  every time a query is executed, so you can selectively choose which recordsets
94  to count.</p>
95<p>If you prefer to set $ADODB_COUNTRECS = false, ADOdb still has the PO_RecordCount(
96  ) function. This will return the number of rows, or if it is not found, it will
97  return an estimate using SELECT COUNT(*):</p>
98<pre>$rs = $db-&gt;Execute(&quot;select * from table where state=$state&quot;);
99$numrows = $rs-&gt;PO_RecordCount('table', &quot;state=$state&quot;);</pre>
100<p><b>Selects: Locking</b> </p>
101<p>SELECT statements are commonly used to implement row-level locking of tables.
102  Other databases such as Oracle, Interbase, PostgreSQL and MySQL with InnoDB
103  do not require row-level locking because they use versioning to display data
104  consistent with a specific point in time.</p>
105<p>Currently, I recommend encapsulating the row-level locking in a separate function,
106  such as RowLock($table, $where):</p>
107<pre>$connection-&gt;BeginTrans( );
108$connection-&gt;RowLock($table, $where); </pre>
109<pre><font color=green># some operation</font></pre>
110<pre>if ($ok) $connection-&gt;CommitTrans( );
111else $connection-&gt;RollbackTrans( );
112</pre>
113<p><b>Selects: Outer Joins</b></p>
114<p>Not all databases support outer joins. Furthermore the syntax for outer joins
115  differs dramatically between database vendors. One portable (and possibly slower)
116  method of implementing outer joins is using UNION.</p>
117<p>For example, an ANSI-92 left outer join between two tables t1 and t2 could
118  look like:</p>
119<pre>SELECT t1.col1, t1.col2, t2.cola <br>  FROM t1 <i>LEFT JOIN</i> t2 ON t1.col = t2.col</pre>
120<p>This can be emulated using:</p>
121<pre>SELECT t1.col1, t1.col2, t2.cola FROM t1, t2 <br>     WHERE t1.col = t2.col
122   UNION ALL
123SELECT col1, col2, null FROM t1 <br>       WHERE t1.col not in (select distinct col from t2)
124</pre>
125<p>Since ADOdb 2.13, we provide some hints in the connection object as to legal
126  join variations. This is still incomplete and sometimes depends on the database
127  version you are using, but is useful as a general guideline:</p>
128<p><font face="Courier New, Courier, mono">$conn-&gt;leftOuter</font>: holds the
129  operator used for left outer joins (eg. '*='), or false if not known or not
130  available.<br>
131  <font face="Courier New, Courier, mono">$conn-&gt;rightOuter</font>: holds the
132  operator used for right outer joins (eg '=*'), or false if not known or not
133  available.<br>
134  <font face="Courier New, Courier, mono">$conn-&gt;ansiOuter</font>: boolean
135  that if true means that ANSI-92 style outer joins are supported, or false if
136  not known.</p>
137<h3><b>Inserts</b> </h3>
138<p>When you create records, you need to generate unique id's for each record.
139  There are two common techniques: (1) auto-incrementing columns and (2) sequences.
140</p>
141<p>Auto-incrementing columns are supported by MySQL, Sybase and Microsoft Access
142  and SQL Server. However most other databases do not support this feature. So
143  for portability, you have little choice but to use sequences. Sequences are
144  special functions that return a unique incrementing number every time you call
145  it, suitable to be used as database keys. In ADOdb, we use the GenID( ) function.
146  It has takes a parameter, the sequence name. Different tables can have different
147  sequences. </p>
148<pre>$id = $connection-&gt;GenID('sequence_name');<br>$connection-&gt;Execute(&quot;insert into table (id, firstname, lastname) <br>                       values ($id, $firstname, $lastname)&quot;);</pre>
149<p>For databases that do not support sequences natively, ADOdb emulates sequences
150  by creating a table for every sequence.</p>
151<h3><b>Binding</b></h3>
152<p>Binding variables in an SQL statement is another tricky feature. Binding is
153  useful because it allows pre-compilation of SQL. When inserting multiple records
154  into a database in a loop, binding can offer a 50% (or greater) speedup. However
155  many databases such as Access and MySQL do not support binding natively and
156  there is some overhead in emulating binding. Furthermore, different databases
157  (specificly Oracle!) implement binding differently. My recommendation is to
158  use binding if your database queries are too slow, but make sure you are using
159  a database that supports it like Oracle. </p>
160<p>ADOdb supports portable Prepare/Execute with:</p>
161<pre>$stmt = $db-&gt;Prepare('select * from customers where custid=? and state=?');
162$rs = $db-&gt;Execute($stmt, array($id,'New York'));</pre>
163<p>Oracle uses named bind placeholders, not "?", so to support portable binding, we have Param() that generates
164the correct placeholder (available since ADOdb 3.92):
165<pre><font color="#000000">$sql = <font color="#993300">'insert into table (col1,col2) values ('</font>.$DB-&gt;Param('a').<font color="#993300">','</font>.$DB-&gt;Param('b').<font color="#993300">')'</font>;
166<font color="#006600"># generates 'insert into table (col1,col2) values (?,?)'
167# or        'insert into table (col1,col2) values (:a,:b)</font>'
168$stmt = $DB-&gt;Prepare($sql);
169$stmt = $DB-&gt;Execute($stmt,array('one','two'));
170</font></pre>
171<a name="native"></a>
172<h2>Portable Native SQL</h2>
173<p>ADOdb provides the following functions for portably generating SQL functions
174  as strings to be merged into your SQL statements (some are only available since
175  ADOdb 3.92): </p>
176<table width="75%" border="1" align=center>
177  <tr>
178    <td width=30%><b>Function</b></td>
179    <td><b>Description</b></td>
180  </tr>
181  <tr>
182    <td>DBDate($date)</td>
183    <td>Pass in a UNIX timestamp or ISO date and it will convert it to a date
184      string formatted for INSERT/UPDATE</td>
185  </tr>
186  <tr>
187    <td>DBTimeStamp($date)</td>
188    <td>Pass in a UNIX timestamp or ISO date and it will convert it to a timestamp
189      string formatted for INSERT/UPDATE</td>
190  </tr>
191  <tr>
192    <td>SQLDate($date, $fmt)</td>
193    <td>Portably generate a date formatted using $fmt mask, for use in SELECT
194      statements.</td>
195  </tr>
196  <tr>
197    <td>OffsetDate($date, $ndays)</td>
198    <td>Portably generate a $date offset by $ndays.</td>
199  </tr>
200  <tr>
201    <td>Concat($s1, $s2, ...)</td>
202    <td>Portably concatenate strings. Alternatively, for mssql use mssqlpo driver,
203      which allows || operator.</td>
204  </tr>
205  <tr>
206    <td>IfNull($fld, $replaceNull)</td>
207    <td>Returns a string that is the equivalent of MySQL IFNULL or Oracle NVL.</td>
208  </tr>
209  <tr>
210    <td>Param($name)</td>
211    <td>Generates bind placeholders, using ? or named conventions as appropriate.</td>
212  </tr>
213  <tr><td>$db->sysDate</td><td>Property that holds the SQL function that returns today's date</td>
214</tr>
215<tr><td>$db->sysTimeStamp</td><td>Property that holds the SQL function that returns the current
216timestamp (date+time).
217</td>
218</tr>
219<tr>
220<td>$db->concat_operator</td><td>Property that holds the concatenation operator
221</td>
222</tr>
223<tr><td>$db->length</td><td>Property that holds the name of the SQL strlen function.
224</td></tr>
225
226<tr><td>$db->upperCase</td><td>Property that holds the name of the SQL strtoupper function.
227</td></tr>
228<tr><td>$db->random</td><td>Property that holds the SQL to generate a random number between 0.00 and 1.00.
229</td>
230</tr>
231<tr><td>$db->substr</td><td>Property that holds the name of the SQL substring function.
232</td></tr>
233</table>
234<p>&nbsp; </p>
235<h2>DDL and Tuning</h2>
236There are database design tools such as ERWin or Dezign that allow you to generate data definition language commands such as ALTER TABLE or CREATE INDEX from Entity-Relationship diagrams.
237<p>
238However if you prefer to use a PHP-based table creation scheme, adodb provides you with this feature. Here is the code to generate the SQL to create a table with:
239<ol>
240        <li> Auto-increment primary key 'ID', </li>
241        <li>The person's 'NAME' VARCHAR(32) NOT NULL and defaults to '', </li>
242        <li>The date and time of record creation 'CREATED', </li>
243        <li> The person's 'AGE', defaulting to 0, type NUMERIC(16). </li>
244</ol>
245<p>
246Also create a compound index consisting of 'NAME' and 'AGE':
247<pre>
248$datadict = <strong>NewDataDictionary</strong>($connection);
249$flds = "
250<font color="#660000">  ID I AUTOINCREMENT PRIMARY,
251  NAME C(32) DEFAULT '' NOTNULL,
252  CREATED T DEFTIMESTAMP,
253  AGE N(16) DEFAULT 0</font>
254";
255$sql1 = $datadict-><strong>CreateTableSQL</strong>('tabname', $flds);
256$sql2 = $datadict-><strong>CreateIndexSQL</strong>('idx_name_age', 'tabname', 'NAME,AGE');
257</pre>
258
259<h3>Data Types</h3>
260<p>Stick to a few data types that are available in most databases. Char, varchar
261  and numeric/number are supported by most databases. Most other data types (including
262  integer, boolean and float) cannot be relied on being available. I recommend
263  using char(1) or number(1) to hold booleans. </p>
264<p>Different databases have different ways of representing dates and timestamps/datetime.
265  ADOdb attempts to display all dates in ISO (YYYY-MM-DD) format. ADOdb also provides
266  DBDate( ) and DBTimeStamp( ) to convert dates to formats that are acceptable
267  to that database. Both functions accept Unix integer timestamps and date strings
268  in ISO format.</p>
269<pre>$date1 = $connection-&gt;DBDate(time( ));<br>$date2 = $connection-&gt;DBTimeStamp('2002-02-23 13:03:33');</pre>
270<p>We also provide functions to convert database dates to Unix timestamps:</p>
271<pre>$unixts = $recordset-&gt;UnixDate('#2002-02-30#'); <font color="green"># MS Access date =&gt; unix timestamp</font></pre>
272<p>The maximum length of a char/varchar field is also database specific. You can
273  only assume that field lengths of up to 250 characters are supported. This is
274  normally impractical for web based forum or content management systems. You
275  will need to be familiar with how databases handle large objects (LOBs). ADOdb
276  implements two functions, UpdateBlob( ) and UpdateClob( ) that allow you to
277  update fields holding Binary Large Objects (eg. pictures) and Character Large
278  Objects (eg. HTML articles):</p>
279<pre><font color=green># for oracle </font>
280$conn->Execute('INSERT INTO blobtable (id, blobcol) VALUES (1,empty_blob())');
281$conn->UpdateBlob('blobtable','blobcol',$blobvalue,'id=1');
282   
283<font color=green># non-oracle databases</font>
284$conn->Execute('INSERT INTO blobtable (id, blobcol) VALUES (1, null)');
285$conn->UpdateBlob('blobtable','blobcol',$blobvalue,'id=1');
286</pre>
287<p>Null handling is another area where differences can occur. This is a mine-field,
288  because 3-value logic is tricky.
289<p>In general, I avoid using nulls except for dates and default all my numeric
290  and character fields to 0 or the empty string. This maintains consistency with
291  PHP, where empty strings and zero are treated as equivalent, and avoids SQL
292  ambiguities when you use the ANY and EXISTS operators. However if your database
293  has significant amounts of missing or unknown data, using nulls might be a good
294  idea.
295  <p>
296  ADOdb also supports a portable <a href=http://phplens.com/adodb/reference.functions.concat.html#ifnull>IfNull</a> function, so you can define what to display
297  if the field contains a null.
298<h3><b>Stored Procedures</b></h3>
299<p>Stored procedures are another problem area. Some databases allow recordsets
300  to be returned in a stored procedure (Microsoft SQL Server and Sybase), and
301  others only allow output parameters to be returned. Stored procedures sometimes
302  need to be wrapped in special syntax. For example, Oracle requires such code
303  to be wrapped in an anonymous block with BEGIN and END. Also internal sql operators
304  and functions such as +, ||, TRIM( ), SUBSTR( ) or INSTR( ) vary between vendors.
305</p>
306<p>An example of how to call a stored procedure with 2 parameters and 1 return
307  value follows:</p>
308<pre>   switch ($db->databaseType) {
309        case '<font color="#993300">mssql</font>':
310          $sql = <font color="#000000"><font color="#993333">'<font color="#993300">SP_RUNSOMETHING</font>'</font></font>; break;
311        case '<font color="#993300">oci8</font>':
312          $sql =
313<font color="#993300">    </font><font color="#000000"><font color="#993300">&quot;declare RETVAL integer;begin :RETVAL := </font><font color="#000000"><font color="#993333"><font color="#993300">SP_RUNSOMETHING</font></font></font><font color="#993300">(:myid,:group);end;&quot;;
314</font>   break;</font>
315        default:
316          die('<font color="#993300">Unsupported feature</font>');
317        }
318<font color="#000000"><font color="green">      # @RETVAL = SP_RUNSOMETHING @myid,@group</font>
319        $stmt = $db-&gt;PrepareSP($sql);        <br>    $db-&gt;Parameter($stmt,$id,'<font color="#993300">myid</font>');
320        $db-&gt;Parameter($stmt,$group,'<font color="#993300">group</font>');
321        <font color="green"># true indicates output parameter<br>       </font>$db-&gt;Parameter($stmt,$ret,'<font color="#993300">RETVAL</font>',true);
322        $db-&gt;Execute($stmt); </font></pre>
323<p>As you can see, the ADOdb API is the same for both databases. But the stored
324  procedure SQL syntax is quite different between databases and is not portable,
325  so be forewarned! However sometimes you have little choice as some systems only
326  allow data to be accessed via stored procedures. This is when the ultimate portability
327  solution might be the only solution: <i>treating portable SQL as a localization
328  exercise...</i></p>
329<h3><b>SQL as a Localization Exercise</b></h3>
330<p> In general to provide real portability, you will have to treat SQL coding
331  as a localization exercise. In PHP, it has become common to define separate
332  language files for English, Russian, Korean, etc. Similarly, I would suggest
333  you have separate Sybase, Intebase, MySQL, etc files, and conditionally include
334  the SQL based on the database. For example, each MySQL SQL statement would be
335  stored in a separate variable, in a file called 'mysql-lang.inc.php'.</p>
336<pre>$sqlGetPassword = '<font color="#993300">select password from users where userid=%s</font>';
337$sqlSearchKeyword = &quot;<font color="#993300">SELECT * FROM articles WHERE match (title,body) against (%s</font>)&quot;;</pre>
338<p>In our main PHP file:</p>
339<pre><font color=green># define which database to load...</font>
340<b>$database = '<font color="#993300">mysql</font>';
341include_once(&quot;<font color="#993300">$database-lang.inc.php</font>&quot;);</b>
342
343$db = &amp;NewADOConnection($database);
344$db->PConnect(...) or die('<font color="#993300">Failed to connect to database</font>');
345
346<font color=green># search for a keyword $word</font>
347$rs = $db-&gt;Execute(sprintf($sqlSearchKeyWord,$db-&gt;qstr($word)));</pre>
348<p>Note that we quote the $word variable using the qstr( ) function. This is because
349  each database quotes strings using different conventions.</p>
350<p>
351<h3>Final Thoughts</h3>
352<p>The best way to ensure that you have portable SQL is to have your data tables designed using
353sound principles. Learn the theory of normalization and entity-relationship diagrams and model
354your data carefully. Understand how joins and indexes work and how they are used to tune performance.
355<p> Visit the following page for more references on database theory and vendors:
356  <a href="http://php.weblogs.com/sql_tutorial">http://php.weblogs.com/sql_tutorial</a>.
357  Also read this article on <a href=http://phplens.com/lens/php-book/optimizing-debugging-php.php>Optimizing PHP</a>.
358<p>
359<font size=1>(c) 2002-2003 John Lim.</font>
360
361</body>
362</html>
Note: See TracBrowser for help on using the repository browser.