Xataface 1.1 Released
Archived from the Xataface News forum.
shannah — Mon Jan 05, 2009 8:24 pm
Xataface 1.1 has been released. I includes a number of bug fixes and performance improvements so it is recommended for all users using 1.0.x.
Upgrading should be a simple matter of replacing your old Xataface folder with this new one. Applications should work fine unchanged.
dratman — Mon Jan 12, 2009 5:13 pm
shannah wrote:Xataface 1.1 has been released. I includes a number of bug fixes and performance improvements so it is recommended for all users using 1.0.x.
Upgrading should be a simple matter of replacing your old Xataface folder with this new one. Applications should work fine unchanged.
Steve,
I noticed on Freshmeat you wrote that the SQL parsing had been extended. Is there any documentation of what can (or what can’t) be parsed?
Thank you.
Ralph
shannah — Mon Jan 12, 2009 5:19 pm
I haven’t documented it in detail. But the parser used to have problems with the following:
- Expressions in the select field list.
e.g. select a, 10+5 as b from foo
This is now fixed.
- Subqueries in the select list.
e.g. select a, (select max(b) from c) as d from foo
This is now fixed.
- Group by clauses can now contain expressions.
These changes were made primarily to improve the support for managing multilingual content in the database, but they may affect/improve other areas also.
The changes were made to the lib/SQL/Parser.php and lib/SQL/Compiler.php classes primarily, which, once upon a time were from the PEAR class repository; but they have been overhauled so much that they barely resemble the original parsers.
Note also that these changes were for version 1.1 to 1.1.1 . Upgrading from 1.0.x to 1.1 involves wide variety of changes (you can see the changes.txt file with the distribution for an abbreviated list).
dratman — Mon Jan 12, 2009 6:03 pm
Thanks, Steve, for your quick reply. I will look at those sources and see if I can understand, in a general way, how the parser works, and maybe even get a glimpse of how you use the output of the parser. An interesting topic, because to me it suggests a move at least a little bit into the semantics of a query.
Ralph