Let's solve both. pgsql patch is untested and quite frankly I do not care, but I still have added %u support and I hope it works.
| Comment | File | Size | Author |
|---|---|---|---|
| #12 | database_query_load.patch | 6.41 KB | bdragon |
| #7 | pg_menu.png | 164.96 KB | agentrickard |
| #6 | database_query_load.patch | 5.52 KB | chx |
| #1 | database_query_load_0.patch | 4.77 KB | chx |
| database_query_load.patch | 4.76 KB | chx |
Comments
Comment #1
chx commentedDowngrading and adding %u tothe regexp too.
Comment #2
dries commentedCare to provide a bit more background/context? How does this relate to ip2long?
Comment #3
chx commenteds integers > 2147483647 overflow with %d, see also http://drupal.org/node/134292
Comment #4
dries commentedAh, right! :-)
Comment #5
catchno longer applies.
Comment #6
chx commentedOK, rerolled. The patch adds %u to the useable placeholders.
Comment #7
agentrickardThe patch creates multiple issues with the menu system -- causing all pages to return NOT FOUND. Trying to decipher exactly what.
Screenshot is of testsite home page after applying the patch. Using HEAD and pgSQL 8.2.3.
Comment #8
agentrickardInitial finding: db_query_load_values() fails to return a value, but I still have odd behavior. Investigating.
Changes line 474.
Comment #9
bdragon commentedUm... db_query_load_values is not returning $query.
Comment #10
agentrickardFor reference, here's the similar function from database-mysql-common.inc:
Comment #11
agentrickardFixing the return value for db_query_load_values allows this statement to work correctly.
Where {test} is defined as:
pgSQL 8.2.3
Changes in database_pgsql.inc:
Comment #12
bdragon commentedPatch and an additional comment because I had to scratch my head for a second to figure out what was going on.
Comment #13
agentrickardPatch works as desired; %u syntax functions correctly in pgSQL. RTBC.
Comment #14
gábor hojtsyBeing an API change, I hand this over to Dries for consideration.
Comment #15
dries commentedThis issue combines two things: it adds support for longs _and_ it rewrite part of the database API for performance's sake. This should be split in two issues.
Comment #16
dpearcefl commentedDoes this issue exist in current D6?
Comment #17
dpearcefl commentedComment #18
dpearcefl commented