Posted by chx on May 16, 2009 at 3:09pm
Jump to:
| Project: | Oracle driver |
| Version: | 7.x-1.x-dev |
| Component: | Code |
| Category: | bug report |
| Priority: | normal |
| Assigned: | aaaristo |
| Status: | closed (fixed) |
Issue Summary
Yay, an Oracle driver! However, I do not see any blob handling so all the text field are constricted in length to 4096. Also, it would be great if you would report on the project page how far this driver goes -- does Drupal 7 install, how many tests pass.
Comments
#1
chx, the blob handling is all there...
It passes al the Database test cases and i have tried it with all core modules enabled.
#2
#3
YAY there is, indeed, blob handling. Have you tried to run more than Database tests? LIke, all tests?
#4
No actually i don't. I tried to run a couple of tests but i've seen some minor things i should fix. I think i'am going to test the whole suite in a couple of weeks, but i don't know if is the right time: i mean 7.x is -dev... so i'am worried about debuggin the 7.x code instead of the oracle driver code...
For example, it seems that the DatabaseConnection->lastInsertId($sequence_name) method always return the last inserted id of any db_prefix, for example on MySQL:
if you insert the BID of BLOCKS table with $db_prefix == DUMMY and then call lastInsertId it will return N.
if the you recall it without $db_prefix it will return N again, even if db_prefix is null or empty.
Is this a "wanted" way of work or a "side effect". In any of this two cases i have to reproduce the same functionality on the Oracle driver for compatibility, so...
WDYT? Is a good idea to run the whole suite now?
#5
Well the whole suite passes on MySQL (guaranteed by testing infra) and Josh Waihi told me a few days ago that postgresql is passing too. I do not get your problem with last insert id, the db prefix is not really supposed to change in between two function calls just like that...? Care to show some code? Moved the issue to what we are talking now, btw.
BTW: congrats on Oracle driver.
#6
Thanks chx,
now the lastinsertId problem is fixed it was my misunderstanding of some tricky things on sequence name generation.
So actually i'am going to test the whole suite.
#7
#8
it seams to pass al the suite... :)
(openid does not work either on mysql on my testbed so... )
and
File space used tests: 4 passes, 6 fails, and 0 exceptions
seems not to be related with the database driver so..
again it seems to work.
#9
#10
found the files problem...
#11
was a & -> BITAND translation problem fixed
#12
The OpenID problem was dued to the lack of php-bcmath.i386 library.