Description: ------------ The following small change in the oci8.php driver should fix this. Sorry for not including a proper diff. In... $Id: oci8.php,v 1.103 2005/04/11 15:10:22 danielc Exp $ Insert the following *replacing line 1004*. //quoted table names should be passed through unchanged (expect for removing the quotes themselves if (strstr($result, '"')) { $result = str_replace('"','',$result); } else { $result = strtoupper($result); } Test script: --------------- /* in oracle: create table "Test" ( testcolumn int not null ) */ $db = DB::connect('oci8://user:pwd@tcp(hostspec)'); print_r($db->tableInfo('"Test"')); Expected result: ---------------- Array ( [0] => Array ( [table] => Test [name] => TESTCOLUMN [type] => NUMBER [len] => 22 [flags] => not_null ) ) Actual result: -------------- Array ( )

Sorry you made the mistake of using delimited identifiers. I see your suggested code uses str_replace() to remove the quotes rather than leaving them in. Why?

As is often the case, the delimited identifiers are there for "historical" reasons (namely that the database in question was ported from MSSQL where mixed case table names are allowed and tons of code uses this convention). The quotes need to be striped because the user_* views list quoted-identfiers without the quotes (at least in 10g). Oracle says [1] that quoted indentifers cannot contain a the (") character so str_replace should be safe to remove the quotes at the start and end of the $result. [1] http://oracleheva1.oracle.com/docs/cd/B19306_01/server.102/b14200/sql_elements008.htm (Rule 6)

Forgot to mention what may or may not be obivous: while user_* views don't include the quotes they do store the rest of the tablename in it's original form and that it is possible (if painful) to have "TestTable", "Testtable" and "TestTABLE" as three different tables in the same schema so we do need to perserve the case when looking up quoted table names.

Also related to quotedIdentifiers.... class DB_oci8 should override getSequenceName to allow the quote (") character. Like so perhaps... function getSequenceName($sqn) { return sprintf($this->getOption('seqname_format'), preg_replace('/[^a-z0-9_".]/i', '_', $sqn)); }

DB is in maintenance mode, new features should go into MDB2 and we can consider then if its safe to backport this.

I have decided to not change this behaviour. So it is your job to ensure that you "unquote" identifiers before passing it to tableInfo().

The problem is that it is impossible to access tableInfo on mixed case named tables currently. If you can find a better way to support that then I'm all ears.

How about something like this: $query = 'SELECT column_name, data_type, data_length, nullable'; $query.= ' FROM user_tab_columns'; $query.= ' WHERE table_name='.$db->quote(strtoupper($result), 'text'); $query.= ' OR table_name='.$db->quote($result, 'text'); $query.= ' ORDER BY column_id';

This bug has been fixed in CVS. If this was a documentation problem, the fix will appear on pear.php.net by the end of next Sunday (CET). If this was a problem with the pear.php.net website, the change should be live shortly. Otherwise, the fix will appear in the package's next release. Thank you for the report and for helping us make PEAR better.