httpd-modules-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sorin Manolache <sor...@gmail.com>
Subject Re: path problem
Date Sun, 22 Mar 2009 21:19:40 GMT
On Sun, Mar 22, 2009 at 22:05, Alexander Hjalmarsson <hjalle@sgh.se> wrote:
> -----Original Message-----
> From: Sorin Manolache [mailto:sorinm@gmail.com]
> Sent: den 22 mars 2009 19:57
> To: modules-dev@httpd.apache.org
> Subject: Re: path problem
>
> On Sun, Mar 22, 2009 at 11:24, Alexander Hjalmarsson <hjalle@sgh.se> wrote:
>> Hi!
>>
>> I want to get the path in a module I've created but I have some kind of
>> problem with it. It just works like the ordinary "pwd", and in 30% of the
>> cases it works like it should, but in the rest, it doesn't return anything
>> at all.
>>
>> if ((buffad = (char *)malloc((size_t)sizead)) != NULL)
>>   ptrad = getcwd(buffad, (size_t)sizead);
>>
>> That is the call that is made, and buffed is a char *, as well as ptrad.
>> sizead is a long.
>>
>> Can this be a problem with permissions or something? The only cases that
> it
>> do work for is when theres a more advanced script that is called, such as
> a
>> forum or a blog (phpbb or wordpress in example). For normal, static html
>> pages, it just returns "/" instead of the real path.
>>
>> I've checked all the paths by writing each request to a file so I manually
>> can check them. I can't see why it should have anything to do with
>> permissions, but since it does work in some cases but not all, that's my
>> first thought.
>>
>> I would love all kind of help here! Feel free to ask anything, I'll try to
>> answer as good as I can!
>>
>> I've tried on all kind of forums, irc-networks and mailing lists with no
>> luck, but I guess this is the "right" place :)
>
> Hello,
>
> getcwd returns the working directory of the apache process. It has
> nothing to do with modules.
>
> A process may work in any directory. If, after it starts running, the
> process performs a chdir(whatever_existing_dir) system call, all
> subsequent calls to getcwd will return whatever_existing_dir. You can
> see the working directory of a process with process ID x by running
>
> ls -l /proc/x/cwd
>
> Which path do you want to get? The path of the served file?
>
> S
>
> --
> A: Because it reverses the logical flow of conversation.
> Q: Why is top-posting frowned upon?
> A: Top-posting.
> Q: What is the most annoying thing in e-mail?
> --------
>
> I'm sorry if I wasn't clear with my question! I want the path of the served
> file and the problem is that it doesn't always returns the correct path,
> sometimes it just returns "/".
>
> In example, if the served file is domain.com/index.html, the path should be
> /www/domain/a/b/c/abc/htdocs/index.html, but then without the index.html,
> since the directory path is just /www/domain/a/b/c/abc/htdocs/. That is the
> structure I have now, but in most cases it doesn't work. As I said before,
> it looks like it works more often with forum software and blog software then
> static content or non-complex php-scripts. I would really love to solve
> this, so thank you for answering so far:)
>
>

Please read again my previous answer why getcwd is not the solution.
Also read the man page of getcwd.

I think r->filename should be what you're looking for, where r is a
pointer to the request_rec of the currently served request. Pass
r->filename to dirname() (man 3 dirname) in order to extract the
directory name from the path of the served file.

Sorin


-- 
A: Because it reverses the logical flow of conversation.
Q: Why is top-posting frowned upon?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

Mime
View raw message