Zip upd master #20218
Zip upd master #20218
Conversation
|
not sure about callback registering parts (ie.zip_register_ * _with_state) if we should call zend_release_fcall_info_cache on failure especially it happens only during system OOM.. wdyt @Girgias ? |
devnexen
force-pushed
the
zip_upd_master
branch
from
4a699db to
6cd1355
Compare
|
|
Ok. if that s the case that ought to be a separate bug fix. |
devnexen
force-pushed
the
zip_upd_master
branch
from
6cd1355 to
60ee5da
Compare
ext/zip/php_zip.c
Outdated
| CWD_STATE_FREE(new_state.cwd); | ||
| return 0; | ||
| efree(new_state.cwd); | ||
| return true; |
There was a problem hiding this comment.
I believe this should be false.
| } | ||
| /* }}} */ | ||
|
|
||
| # define CWD_STATE_ALLOC(l) emalloc(l) |
There was a problem hiding this comment.
I don't think such abstraction macros are necessarily bad, but making the code directly use emalloc/efree seems fine too.
ext/zip/php_zip.c
Outdated
| size_t n; | ||
| int ret; | ||
|
|
||
| char *pattern = ZSTR_VAL(spattern); |
There was a problem hiding this comment.
this could be a const pointer, no?
| case IS_LONG: | ||
| break; | ||
| case IS_STRING: | ||
| if (Z_TYPE_P(zval_file) == IS_STRING) { |
There was a problem hiding this comment.
I'm looking at the commit message. Indeed, it is weird that IS_LONG was accepted.
was IS_LONG to fetch by index eventually tough ??
devnexen
force-pushed
the
zip_upd_master
branch
from
60ee5da to
49b69b1
Compare
3 participants
No description provided.