AgeCommit message (Collapse)Author
2014-06-12Fix --disable-mysql and --disable-postgresqlHEADmasterGuillaume Friloux
2014-06-06Fix warnings with clang.Guillaume Friloux
2014-06-06FreeBSD exposes alloca, no need to redefine it.Guillaume Friloux
2014-05-21fix leak in error condition when escaping stringzmike
2014-05-21fix psql crash on null address during connectzmike
2014-05-21Fix a memset over a struct with size of pointer instead of struct.Guillaume Friloux
2014-05-21Updating my mail adress.Guillaume Friloux
2014-05-20completely redo build systemzmike
2014-05-20clean up build a bitzmike
2014-05-20use EINA_UNUSEDzmike
2014-03-01make esql_fake_free not inlinedzmike
this totally fucks you if you ever try to shutdown esql from a module
2014-02-09make esql_string_escape() externalzmike
2014-01-18add warning to notify users that blob types will not be freed by esqlzmike
2014-01-18remove shadow warnings from sqlite modulezmike
2014-01-18remove extra copying for binary blobs in sqlite backendzmike
adding extra copies for no reason, even if it might save the user a little work in some cases, is not a great idea
2014-01-18no wait, I'm dumb and the rest of them are wrongzmike
2014-01-18and fix sqlite blobs wtf this was written by the author of eina_valuezmike
2014-01-18fix sqlite BLOB/NULL typeszmike
2014-01-18update copyrightzmike
2013-11-14test_sqlite uses esql_query_callback_set() instead of ecore events as data ↵Jérémy Zurcher
is passed from esql_query()
2013-11-10test_sqlite: use home made assert, so that it's never bypassed by NDEBUGJérémy Zurcher
2013-11-10fix typo, complete esql_connect docJérémy Zurcher
2013-11-10esql_postgresql_error_get: fix segv improve error message retrievalJérémy Zurcher
2013-10-02Fixes a memory overlap in mysac_send_query that makes a useless memcpyGuillaume Friloux
in all cases, even when the realloc did not change the adress (which seems to be like this in most cases here). So we win some perfs.
2013-07-30Adding myself in AUTHORS since i made quite a few patches, and understand ↵Guillaume Friloux
how it works.
2013-07-22Reset the error pointers after each result set so a query error doesntGuillaume Friloux
force you to disconnect and reconnect to the MySQL server. A test program to test the patch is available here :
2013-07-22Fix a possible invalid read in esql_call_complete() where a res haveGuillaume Friloux
been freed without unreferencing it in the Esql object.
2013-05-17Changing fd_idler to fd_job to be more logic.Guillaume Friloux
Thx zmike for advice.
2013-05-17This patch modifies the way esskyuehl sends queries.Guillaume Friloux
This fixes a bug where a callback specified with esql_query_callback_set() will not be called, sending an ecore_event instead. The behavior is that sometimes, the SQL server will respond too fast, while calling esql_query_{,args,vargs}() and so our app doesnt have time to get esql_query_id and call esql_query_callback_set(). With this patch, query will be sent only once we're back in the ecore loop using an idler, so you can take your time in your functions before calling ecore_query_callback_set().
2013-05-15two kernel list pointers werent updated when reallocating the MYSAC_RES ↵Guillaume Friloux
structure, leading to reads on freed data
2013-03-28Update to last mysac, with support for multiple result sets inGuillaume Friloux
stored procedures. Yet, esql itself can only manage 1 result set (and 0 before that patch). SVN revision: 84475
2013-02-04Add support for %u, %lu, %llu in mysql_query_escape()Guillaume Friloux
SVN revision: 83598
2013-01-23As discussed with k-s, eina_value will dupe the string given by sqlite, Guillaume Friloux
not use it directly, so duping it is stupid. SVN revision: 83183
2013-01-23Fix postgre backend so blobs work.Guillaume Friloux
I cant test it because i dont have postgre, so if anyone can. SVN revision: 83170
2013-01-23Fix mysql backend to make blobs working.Guillaume Friloux
SVN revision: 83169
2013-01-23Fix sqlite backend :Guillaume Friloux
- do not access invalid memory from sqlite. - fix BLOBs. It has been needed to have an internal array of strings and blobs as sqlite will invalidate pointers at each sqlite3_step. Until now, using strings was "working" because sqlite's internal memory didnt freed/replace the data before we use it. This is luck. This is explained at : The pointers returned are valid until a type conversion occurs as described above, or until sqlite3_step() or sqlite3_reset() or sqlite3_finalize() is called. The memory space used to hold strings and BLOBs is freed automatically. Do not pass the pointers returned sqlite3_column_blob(), sqlite3_column_text(), etc. into sqlite3_free(). SVN revision: 83168
2013-01-04autotools: move to AC_CONFIG_HEADERS.Cedric BAIL
Patch by Doug Newgard <>. SVN revision: 82135
2012-12-18Re-add support for DECIMAL and NEWDECIMAL as it worksGuillaume Friloux
SVN revision: 81251
2012-12-18Revert mysac version because of ugly bug happenning sometimes. Waiting for ↵Guillaume Friloux
reply from author SVN revision: 81240
2012-12-17Updating MySAC :Guillaume Friloux
- Support for stored procedures, no matter the number of result sets - Support for functions like COUNT(), SUM()... (MYSQL_TYPE_NEWDECIMAL) Updating Esskyuehl : - Support for functions like COUNT(), SUM()... (MYSQL_TYPE_NEWDECIMAL) Esskyuehl needs to be patched (API break), for supporting more than one result set from stored procedures. SVN revision: 81068
2012-10-22MySAC lib requires a max buffer len for its internal query memory.Guillaume Friloux
Esskyuehl was fixing a 1024bytes limit, which can be low when playing with tokens (certeurope, universign...) for example. This patch is a dirty patch, a real patch would be to change MySAC to be unlimited. SVN revision: 78326
2012-08-20remove unnecessary idle timer resets in sql modulesMike Blumenkrantz
SVN revision: 75444
2012-08-17prevent multipe disconnect events on the same object, but allow cleanup ↵Mike Blumenkrantz
routines to function normally SVN revision: 75394
2012-08-17prevent multiple disconnects on the same database connectionMike Blumenkrantz
SVN revision: 75393
2012-08-17ensure that pool connect events are only emitted for overall connectMike Blumenkrantz
SVN revision: 75392
2012-08-17add bugs file for esqlMike Blumenkrantz
SVN revision: 75391
2012-08-17add 1.0s delay between database reconnect attemptsMike Blumenkrantz
SVN revision: 75389
2012-08-17move pool disconnect event serializing to regular disconnect function to ↵Mike Blumenkrantz
catch cases where all pool members have individually disconnected SVN revision: 75388
2012-08-17autoapply current database when reconnectingMike Blumenkrantz
SVN revision: 75384