String manipulation.

[Nuno Paquete]

Hi,

I've got a php script that receives parts of URLs from another script and
transforms it to be HTTP compliant.
For example, if I receive a parameter like "dir=some directory/" I have to
convert it to "dir=some%20directory/".
I make this with this 2 lines of code:

$str = explode(" ",$_GET["dir"]);
fputs($fp,"GET /share/" . implode("%20",$str) . " HTTP/1.0\r\n");

Everything was working fine, until I have created a directory called "C++".
I don't know why but the code above convert it to "C ".
It should keep it unchanged, because what the code do is searching for SPACE
characters ' ' only and change it to "%20". I was not expecting that it
would change '+' characters.
What is wrong with the code?

Thanks in advance,
Nuno Paquete

[Berislav Lopac]

Nuno Paquete wrote:

Hi,

I've got a php script that receives parts of URLs from another script
and transforms it to be HTTP compliant.
For example, if I receive a parameter like "dir=some directory/" I
have to convert it to "dir=some%20directory/".
I make this with this 2 lines of code:

$str = explode(" ",$_GET["dir"]);
fputs($fp,"GET /share/" . implode("%20",$str) . " HTTP/1.0\r\n");

You're killing flies with a shotgun. What you really need is
http://www.php.net/manual/en/function.urlencode.php

Berislav

--
If the Internet is a Marx Brothers movie, and Web, e-mail, and IRC are
Groucho, Chico, and Harpo, then Usenet is Zeppo.

[Alvaro G Vicario]

*** Nuno Paquete wrote/escribió (Fri, 20 Aug 2004 11:53:01 +0100):

For example, if I receive a parameter like "dir=some directory/" I have to
convert it to "dir=some%20directory/".

Two options:

echo urlencode('dir=some directory/'); // prints 'dir=some+directory/'
echo rawurlencode('dir=some directory/'); // prints 'dir=some%20directory/'

I'm unsure though about what standards say about encoding spaces in URLs
(i.e, which is better, "+" or "%20"?)

--
-- Álvaro G. Vicario - Burgos, Spain
-- Questions sent to my mailbox will be billed ;-)
--

[John Dunlop]

Alvaro G Vicario wrote:

[ ... ]

echo urlencode('dir=some directory/'); // prints 'dir=some+directory/'
echo rawurlencode('dir=some directory/'); // prints 'dir=some%20directory/'
But remember that equals signs and solidi are percent-
encoded too. Your examples output 'dir%3Dsome+directory%2F'
and 'dir%3Dsome%20directory%2F' respectively.
I'm unsure though about what standards say about encoding spaces in URLs
(i.e, which is better, "+" or "%20"?)

I had misunderstood this, but it was explained to me a while
ago. Hopefully now I have it right. ;o)

The plus sign is reserved in query components, which means
the semantics of the URI change if it's replaced by its
percent-encoding. How it's interpreted is up to the program
handling it. The URI standard doesn't assign any meaning to
plus signs; whereas %20 always means a US-ASCII space.

HAGW Álvaro!

--
Jock

[Nuno Paquete]

Alvaro G Vicario wrote:

*** Nuno Paquete wrote/escribió (Fri, 20 Aug 2004 11:53:01 +0100):

For example, if I receive a parameter like "dir=some directory/" I have
to convert it to "dir=some%20directory/".

Two options:

echo urlencode('dir=some directory/'); // prints 'dir=some+directory/'
echo rawurlencode('dir=some directory/'); // prints
'dir=some%20directory/'

I'm unsure though about what standards say about encoding spaces in URLs
(i.e, which is better, "+" or "%20"?)

Hi.
The functions you described both encode everything, including slashs (/),
changing it for %2F. I want to keep '/' unchanged.
What I think that is strange is that I want to manipulate a string that was
supposed to be already encoded, because when the script is asked from
another page, it is accessed from a link like this: "script.php?dir=some
20directory" but then, inside script.php $_GET["dir"] becomes "some
directory". Why?
So, I need to encode it again (understandable), but I don't want to enconde
'/' characters.

Any suggestions?

Regards.

[John Dunlop]

Nuno Paquete wrote:

[ ... ]

What I think that is strange is that I want to manipulate a string that was
supposed to be already encoded, because when the script is asked from
another page, it is accessed from a link like this: "script.php?dir=some
20directory" but then, inside script.php $_GET["dir"] becomes "some
directory". Why?
Because %20 means a space; %20 is the percent-encoding of a
US-ASCII space character; see RFC2396 sec. 2.4. If you
intended $_GET['dir'] to be 'some%20directory', the percent
sign must itself be percent-encoded as %25, giving a query
component of 'dir=some%2520directory'.
So, I need to encode it again (understandable), but I don't want to enconde
'/' characters.

Any suggestions?

How did you do it before?

--
Jock

[Nuno Paquete]

> How did you do it before?

I get the result of the web server. This is a script that deals with HTTP
connections. I establish the connection, I "speak" HTTP with the server
(GET /dir/ HTTP/1.0) and receive the directory content (Index Directory).
Then I manipulate the result, and this result already contains the url
encoded.
But now I need to POST the (already encoded) url and strangely I cant get
the encoded result on the other side. It seems that it is decoded in the
middle, because "some%20directory" becomes "some directory".
Any suggestions?

Thanks again.

[Nuno Paquete]

Nuno Paquete wrote:

How did you do it before?

I get the result of the web server. This is a script that deals with HTTP
connections. I establish the connection, I "speak" HTTP with the server
(GET /dir/ HTTP/1.0) and receive the directory content (Index Directory).
Then I manipulate the result, and this result already contains the url
encoded.
But now I need to POST the (already encoded) url and strangely I cant get
the encoded result on the other side. It seems that it is decoded in the
middle, because "some%20directory" becomes "some directory".
Any suggestions?

Thanks again.

It's done.
Instead of make links with GET method, I did make a form and then used POST
method.
It's working perfectly, and I don't need to use urlencode() function.

Best regards,
Nuno Paquete

[Alvaro G Vicario]

*** Nuno Paquete wrote/escribió (Sat, 21 Aug 2004 14:18:45 +0100):

The functions you described both encode everything, including slashs (/),
changing it for %2F. I want to keep '/' unchanged.

If the slash is part of the parameter you must encode it.

Also, as John said, it also encodes the equal size, which should be left
outside the function call.

--
-- Álvaro G. Vicario - Burgos, Spain
-- Questions sent to my mailbox will be billed ;-)
--

[Alvaro G Vicario]

*** Alvaro G Vicario wrote/escribió (Mon, 23 Aug 2004 08:19:59 +0200):

Also, as John said, it also encodes the equal size

sign, not size. Sorry for the typos...
--
-- Álvaro G. Vicario - Burgos, Spain
-- Questions sent to my mailbox will be billed ;-)
--