|
From: Andreas A. <a.a...@th...> - 2001-08-26 12:09:46
|
Hi Alex,
That's how I would do it:
1:
We use a unix standard location for the shell interpreter
i.e. #!/usr/bin/php (creating a symlink isn't that hard, but only needed for
debugging purposes, see below)
2:
For the make process we introduce an new variable in
Makefile.in that is determined by a configure script i.e:
PHPCMD= "{by configure determined location}"
(after configure e.g. PHPCMD="/usr/local/bin/php -q")
in Makefiles that use command line php we just use
this syntax:
@$(PHPCMD) $(XML2PHP) ... options etcpp
pros:
- the php command is in a single place
- configure don't has to modify all php shell scripts
- Actually we could leave out the #! line and with some mods
the scripts could be "incldable" and used through the web
(maybe we need this later => builing triggerd via an
admin interface or whatever might be useful in the future).
Andi
>> 3) We should find a way to have a the interpreter
>> line in the php files more flexible through the
>> different platforms. Perl has a trick to do so
>> eval 'exec perl -S $0 "$@"'
>
> agreed, I was going to suggest a simple ./configure.sh which
> would do a grep for a php command line key.
>
> any other ideas?
|