Skip to content

Loading…

DDC-2238: doctrine:schema:update partially broken #2936

Closed
doctrinebot opened this Issue · 8 comments

2 participants

@doctrinebot

Jira issue originally created by user tom:

the app/console doctrine:schema:update command generates a seemingly random number of statements like these:

ALTER TABLE geodata ALTER humidity SET ;
ALTER TABLE geodata ALTER lake SET ;
ALTER TABLE message ALTER translate SET ;

which are obvious invalid SQL commands. The mappings are fine, validate and the application works just fine. Here's an example from the mapping files including two of the above statements:





I am using doctrine2-spatial as an extension for GIS information. This problem shows up both in entities using spatial data and entities not using spatial data.

I'll gladly help debug this, as right now I can't update my dev database with --force, I need to use --dump-sql and filter out the invalid lines.

@doctrinebot

Comment created by @beberlei:

Can you dump the SchemaDiff/ColumnDiff instances that are returned from lib/Doctrine/DBAL/Schema/Comparator.php?

@doctrinebot

Comment created by tom:

requested dump of SchemaDiff

@doctrinebot

Comment created by tom:

added in the file schemadiff.txt - I added this in Schematool.php - getUpdateSchemaSql():

    $comparator = new Comparator();
    $schemaDiff = $comparator->compare($fromSchema, $toSchema);
    \Doctrine\Common\Util\Debug::dump($schemaDiff, 6);

if you need other output, just tell me what entity to dump and where. I'll be happy to help.

The file also contains the buggy update statements it creates towards the end. There's a few non-crucial bugs included, where it alters the geospatial columns to themselves (i.e. river.course already is a geometry/linestring). I don't worry about those because they don't break anything.

@doctrinebot

Comment created by tom:

Is there any update on this? I'm still having this issue, with many different entities, always the same problem, for example:

ALTER TABLE event ALTER priority SET ;

Which is an integer field on an entity that doesn't have any GIS elements, so I'm not even sure if it's caused by that anymore.

@doctrinebot

Comment created by @beberlei:

[~tom] I added a fix for PostgreSQL today, can you verify again if this works? Its included in the 2.3 branch.

@doctrinebot

Issue was closed with resolution "Duplicate"

@beberlei beberlei was assigned by doctrinebot
@doctrinebot doctrinebot closed this
@doctrinebot doctrinebot added the Bug label
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.