6cb8d74b | 17-Aug-2023 |
Andreas Gohr <andi@splitbrain.org> |
Version upped |
f71fd150 | 17-Aug-2023 |
Andreas Gohr <andi@splitbrain.org> |
don't log errors when opts table does not exist
This is to be expected on new setups and not an error. Other errors would be problematic though.
This should be the last commit related to #80 |
7467a796 | 01-Aug-2023 |
Andreas Gohr <andi@splitbrain.org> |
Version upped |
74c4ec25 | 01-Aug-2023 |
Andreas Gohr <andi@splitbrain.org> |
Set busy timeout and use WAL mode. #80
The busy timout should tell the Sqlite driver to wait to acquire locks instead of aborting immeadiately.
The Write-Ahead Logging mode should make parallel acc
Set busy timeout and use WAL mode. #80
The busy timout should tell the Sqlite driver to wait to acquire locks instead of aborting immeadiately.
The Write-Ahead Logging mode should make parallel accesses to the database more efficient. https://www.sqlite.org/wal.html
show more ...
|
64e92f7a | 26-Jul-2023 |
take <take@ruu.kr> |
translation update |
a4a13ef1 | 22-Jul-2023 |
Andreas Gohr <andi@splitbrain.org> |
Version upped |
e49844fb | 22-Jul-2023 |
Andreas Gohr <andi@splitbrain.org> |
add exception logging for #80
This should help to pinpoint the problem |
fe64ba38 | 22-Jul-2023 |
Andreas Gohr <andi@splitbrain.org> |
never set db version to 0. related to #80
There is no need to initially set the database version to 0, returning 0 is enough to run the appropriate migrations.
This will partially fix #80. The unde
never set db version to 0. related to #80
There is no need to initially set the database version to 0, returning 0 is enough to run the appropriate migrations.
This will partially fix #80. The underlying problem is probably a temporary database lock that causes an exception in getOpt(). This still needs to be addressed.
show more ...
|
fd7ab75f | 20-Jul-2023 |
Andreas Gohr <andi@splitbrain.org> |
Version upped |
c70cffc9 | 20-Jul-2023 |
Andreas Gohr <andi@splitbrain.org> |
abort migration handling early
We especially do not need to run VACUUM on every initialization. |
ecfa727d | 24-Jun-2023 |
Aleksandr Selivanov <alexgearbox@yandex.ru> |
translation update |
eec50974 | 22-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Version upped |
27eb38da | 21-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Allow passing parameters as multiple function args
This is what the old query methods allowed. Passing in an array is still somewhat cleaner, but allowing separate arguments should make porting olde
Allow passing parameters as multiple function args
This is what the old query methods allowed. Passing in an array is still somewhat cleaner, but allowing separate arguments should make porting older code a bit more convenient.
show more ...
|
9038f707 | 20-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Cleanup for the admin interface |
5a212f8e | 20-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Remove listing of the now unsupported sqlite2 databases |
ded78afb | 20-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Readd custom GROUP_CONCAT ad GROUP_CONCAT_DISTINCT #78
Since using GROUP_CONCAT with a custom separator and the DISTINCT keyword is still broken in sqlite (and might never be fixed), this readds our
Readd custom GROUP_CONCAT ad GROUP_CONCAT_DISTINCT #78
Since using GROUP_CONCAT with a custom separator and the DISTINCT keyword is still broken in sqlite (and might never be fixed), this readds our custom implementation back. Using a different name than the default allows developers to pick the native implementation when adequate.
See also https://github.com/cosmocode/sqlite/pull/73#issuecomment-1588651185
This reverts commit ff1cb7aeb9f435cab452a4ebd057f46f4dcfc245.
show more ...
|
0290deae | 15-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
log slow queries to the debug log |
33e488b3 | 13-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Added queryKeyValueList convenience method
This is a common use case in plugins |
14e7409b | 13-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
fix method calls missed in refactoring operation |
10cb854a | 13-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
renamed getDB to getPdo
This is clearer in what to expect as a return value |
a4f75120 | 12-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
do not return empty statements when splitting multiqueries |
03f14a77 | 12-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
ensure we always get PDOExceptions
PDO::prepare sometimes throws ValueErrors (and maybe other Throwables) so we convert them to PDOExceptions |
7ddaad11 | 12-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
Fix import/export handling, always split multi-statements
Even though mutliple statements can be passed to $sth->execute(), failures will only be checked in the very first statement. Failures later
Fix import/export handling, always split multi-statements
Even though mutliple statements can be passed to $sth->execute(), failures will only be checked in the very first statement. Failures later on will lead to the rest of the statements to be silently ignored rendering our rollback strategy moot.
We now always split multiline statements and run them separately (affects import and migrations).
The export was adjusted to not include the transaction handling and fix entry exports.
show more ...
|
da6f67b7 | 11-Jun-2023 |
Andreas Gohr <andi@splitbrain.org> |
check access levels only when auth is available |
74dd286d | 01-Jun-2023 |
Anna Dabrowska <dabrowska@cosmocode.de> |
Remove stale migration code
|