On Wed, Dec 24, 2014 at 7:53 PM, Xinchen Hui <larue...@php.net> wrote: > On Wed, Dec 24, 2014 at 3:57 PM, Pierre Joye <pierre....@gmail.com> wrote: >> >> On Dec 24, 2014 2:38 PM, "Stanislav Malyshev" <smalys...@gmail.com> wrote: >>> >>> Hi! >>> >>> > But: return 0 and return FAILURE... which is simpler? >>> >>> It's equally simple to write, but FAILURE of course is way simpler to >>> understand when read. >> >> I totally agree. >> >> I do not care much about the value of failure or success but I am tired to >> have to read the code to see if it is 0, 1, or -1 on failure. >> >> The kind of uniformization I would like to see for the php internals APIs. >> >> About the argument for the lack of info in function signature: >> >> A simple typedef will solve it, for the good: >> >> status php_foo(); >> >> Or something along this line. >> >> Yes, it will mean yet another large set of changes for ext developers. But >> at this point, it may be a good time to do it. > hmm, okey > > so, make the functions which use SUCCESS/FAILURE return php_success type? > > and maybe also typedef php_success php_status?
more like: #include <stdio.h> /* From http://lxr.php.net/xref/PHP_TRUNK/Zend/zend_types.h#48 */ typedef enum { SUCCESS = 0, FAILURE = -1, /* this MUST stay a negative number, or it may affect functions! */ } ZEND_RESULT_CODE; typedef enum ZEND_RESULT_CODE status; status func (status s) { printf("%s\n", s == FAILURE ? "FAILURE" : "SUCCESS"); return s; } int main (void) { status s; s = SUCCESS; func (s); s = FAILURE; func(s); return 0; } according to the current way FAILURE/SUCCESS are defined. Cheers, -- Pierre @pierrejoye | http://www.libgd.org -- PHP Internals - PHP Runtime Development Mailing List To unsubscribe, visit: http://www.php.net/unsub.php
