trouble using variable in fopen

[Bill]

For my personal use I am accessing a railway timetable page, parsing its
contents and then sending brief relevant information as the subject line of
an email to be read on a mobile phone.
The problem comes when I try to parse a second page, the url of which, I
have derived from an initial page.
I cannot go directly to the second (target) page but from the initial page
derive the url and put it in a string called $strTarget

In one particular instance $strTarget reads thus:
http://www.livedepartureboards.co.uk...&J=1370626&R=0

If I hard code this string like this
$fp =
fopen("http://www.livedepartu reboards.co.uk/ldb/train.aspx?T=FR NCMB++&J=137062 6&R=0",
'r');
then the target page opens ok and conveys the timetable information.
However if, as I have to do, I code it using the variable $strTarget like
this
$fp = fopen($strTarge t, 'r');
then the page that opens is always the timetable site's default page for the
message,
"Details for the requested train are no longer available."

I have checked and double checked to make sure this is indeed what is
happening.
Can anyone explain what is going on here?
Thanks
Bill

[CJ Llewellyn]

"Bill" <bi**@spamsucks .com> wrote in message
news:4L******** ***********@new s.xtra.co.nz...

For my personal use I am accessing a railway timetable page, parsing its
contents and then sending brief relevant information as the subject line of an email to be read on a mobile phone.
The problem comes when I try to parse a second page, the url of which, I
have derived from an initial page.
I cannot go directly to the second (target) page but from the initial page
derive the url and put it in a string called $strTarget

In one particular instance $strTarget reads thus:
http://www.livedepartureboards.co.uk...&J=1370626&R=0
If I hard code this string like this
$fp =
fopen("http://www.livedepartu reboards.co.uk/ldb/train.aspx?T=FR NCMB++&J=1370
626&R=0", 'r');
then the target page opens ok and conveys the timetable information.
However if, as I have to do, I code it using the variable $strTarget like
this
echo "[$strTarget]<br>\n";
$fp = fopen($strTarge t, 'r');

Your parsing routine cannot be working as well as you are expecting it to
work.

[Bill]

"CJ Llewellyn" <in*****@exampl e.con> wrote in message
news:co******** **@slavica.ukpo st.com...

"Bill" <bi**@spamsucks .com> wrote in message
news:4L******** ***********@new s.xtra.co.nz...

For my personal use I am accessing a railway timetable page, parsing its
contents and then sending brief relevant information as the subject line

of

an email to be read on a mobile phone.
The problem comes when I try to parse a second page, the url of which, I
have derived from an initial page.
I cannot go directly to the second (target) page but from the initial
page
derive the url and put it in a string called $strTarget

In one particular instance $strTarget reads thus:

http://www.livedepartureboards.co.uk...&J=1370626&R=0

If I hard code this string like this
$fp =

fopen("http://www.livedepartu reboards.co.uk/ldb/train.aspx?T=FR NCMB++&J=1370
626&R=0",

'r');
then the target page opens ok and conveys the timetable information.
However if, as I have to do, I code it using the variable $strTarget like
this

echo "[$strTarget]<br>\n";

$fp = fopen($strTarge t, 'r');

Your parsing routine cannot be working as well as you are expecting it to
work.

It was echo $strTarget that gave me
http://www.livedepartureboards.co.uk...70626&R=0which I then used to hard code the function

[Janwillem Borleffs]

Bill wrote:

http://www.livedepartureboards.co.uk...&J=1370626&R=0
which I then used to hard code the function

Even when I paste the URL in my browser, I'm getting an error.
JW

[Bill]

"Janwillem Borleffs" <jw@jwscripts.c om> wrote in message
news:41******** **************@ news.euronet.nl ...

Bill wrote:

http://www.livedepartureboards.co.uk...&J=1370626&R=0
which I then used to hard code the function

Even when I paste the URL in my browser, I'm getting an error.
JW

That is because the url is for a page giving details of a specific train.
The train it referred to has run long ago.so you now get the message that
the page is no longer available.
However at the time I ran the test I assure you that the string when hard
coded opened the required page
but when it was contained in a variable it opened a default "details no
longer available" page.
I have checked and double checked this.
Immediately after the variable routed me to the default page I could paste
the url into my browser and get the page I wanted, and also get it
programmaticall y by hard coding the string.

Thank you for taking the time to check out my problem.
In general terms I am asking:
Is there any way php treats a variable differently from a hard coded string
when used in a function?

[Janwillem Borleffs]

Bill wrote:

Thank you for taking the time to check out my problem.
In general terms I am asking:
Is there any way php treats a variable differently from a hard coded
string when used in a function?

No, but the problem can be related to livedeparturebo ards.co.uk expecting
browser specific headers, which are not send with the fopen function.

When these headers aren't received, their script can be instructed to abort
the operation.

Try to use fsockopen instead of fopen, which is capable of sending headers
with the aid of fputs. The following example sends an user agent header
along the request:

$host = 'ivedeparturebo ards.co.uk';
$post = 80;
$path = '/ldb/train.aspx?T=FR NCMB++&J=137062 6&R=0';
$fp = fsockopen($host , $port);
fputs($fp, "GET $path HTTP/1.0\r\n");
fputs($fp, "Host: $host\r\n");
fputs($fp, "User-Agent: {$_SERVER['HTTP_USER_AGEN T']}\r\n");
fputs($fp, "\r\n");

fpassthru($fp);
fclose($fp);
JW

[Bill]

> Try to use fsockopen instead of fopen, which is capable of sending headers

with the aid of fputs. The following example sends an user agent header
along the request:

$host = 'ivedeparturebo ards.co.uk';
$post = 80;
$path = '/ldb/train.aspx?T=FR NCMB++&J=137062 6&R=0';
$fp = fsockopen($host , $port);
fputs($fp, "GET $path HTTP/1.0\r\n");
fputs($fp, "Host: $host\r\n");
fputs($fp, "User-Agent: {$_SERVER['HTTP_USER_AGEN T']}\r\n");
fputs($fp, "\r\n");

fpassthru($fp);
fclose($fp);
JW

Thanks for this detailed help.
I am afraid I now have problems with fsockopen()
When I try the above with $host = 'livedepartureb oards.co.uk'; I get

Warning: fsockopen() [function.fsocko pen]: unable to connect to
livedeparturebo ards.co.uk:0 (Failed to parse address
"livedepartureb oards.co.uk")

I have tried numerous variations for $host but they all give me the above
error or

Warning: fsockopen() [function.fsocko pen]: php_network_get addresses:
getaddrinfo failed: No address associated with hostname

In case it is relevant stream_get_tran sports() returns : Array ( [0] =>
tcp [1] => udp [2] => unix [3] => udg )

[Janwillem Borleffs]

Bill wrote:

Thanks for this detailed help.
I am afraid I now have problems with fsockopen()
When I try the above with $host = 'livedepartureb oards.co.uk'; I get

Warning: fsockopen() [function.fsocko pen]: unable to connect to
livedeparturebo ards.co.uk:0 (Failed to parse address
"livedepartureb oards.co.uk")

That's because the first two lines contain errors. Change them into the
following:

$host = 'www.livedepart ureboards.co.uk ';
$port = 80;

However, when you run the code, you will notice an object moved message. The
following code also deals with this:

<?php

$host = 'www.livedepart ureboards.co.uk ';
$port = 80;
$path = '/ldb/train.aspx?T=FR NCMB++&J=137062 6&R=0';
$break = false;

while (true) {
$fp = fsockopen($host , $port);
fputs($fp, "GET $path HTTP/1.0\r\n");
fputs($fp, "Host: $host\r\n");
fputs($fp, "User-Agent: {$_SERVER['HTTP_USER_AGEN T']}\r\n");
fputs($fp, "\r\n");

// Get the response code
$response = (int) strstr(fgets($f p, 1024), " ");

if (($response != 301) && ($response != 302)) {
fpassthru($fp);
$break = true;
} else {
while (!preg_match("/^Location:(.+)$/i",
fgets($fp, 1024), $header));
$path = trim($header[1]);
}
fclose($fp);
if ($break) break;
}

?>

Mind you, this code is provided "as is" and does not guarantee flawless
behaviour under all circumstances!
JW

[Bill]

"Janwillem Borleffs" <jw@jwscripts.c om> wrote in message
news:41******** *************** @news.euronet.n l...
....

$host = 'www.livedepart ureboards.co.uk ';
$port = 80;
$path = '/ldb/train.aspx?T=FR NCMB++&J=137062 6&R=0'; .... Mind you, this code is provided "as is" and does not guarantee flawless
behaviour under all circumstances!

Thank you for your patient help.I am learning a lot from it and the code
does now work
However I am still getting the same problem when I try to use a variable
rather than hard code the value in $path.
Here is the code for a page you can try to see what I mean:
I hope my comments are clear.

<?php
$fp = fopen("http://www.livedepartu reboards.co.uk/ldb/summary.aspx?T= FNC",
'r');
if (stream_set_tim eout($fp, 5) == false){
echo "<p>&nbsp;</p><p><font color=\"red\">E rror: Connection Timed
Out</font></p>";
fclose($fp);
}elseif (!$fp){
echo "<p><font color=\"red\">E rror: Could Not Connect</font></p>";
fclose($fp);
}else{
$events = fgets($fp);
while ($events){
$portsmouth = strpos($events, 'Portsmouth');
if($portsmouth ){
$strBegins =strpos($events , 'train' );
$strEnds =strpos($events , '>Port' );
$strWanted = substr ( $events, $strBegins,$str Ends -
$strBegins -1);
$strTarget ="/ldb/".$strWante d;
}
$events = fgets($fp);
}
fclose($fp);
}

echo "strTarget is ".$strTarget."< br><br>";
echo " the url we want to go to is
www.livedepartu reboards.co.uk" .$strTarget." <br><br>";
echo " you can paste this into your browser to confirm<br>";
echo "If you promptly hard code the path variable with ".$strTarge t." you
will get the proper page:<br><br> ";
echo "However the following is the content of the page accessed by using the
strTarget variable in the path <br><br><br>" ;

$host = 'www.livedepart ureboards.co.uk ';
$port = 80;

//comment out the path you are not using
//$path = 'PASTE HERE';
$path = $strTarget;
$break = false;

while (true) {
$fp = fsockopen($host , $port);
fputs($fp, "GET $path HTTP/1.0\r\n");
fputs($fp, "Host: $host\r\n");
fputs($fp, "User-Agent: {$_SERVER['HTTP_USER_AGEN T']}\r\n");
fputs($fp, "\r\n");

// Get the response code
$response = (int) strstr(fgets($f p, 1024), " ");

if (($response != 301) && ($response != 302)) {
fpassthru($fp);
$break = true;
} else {
while (!preg_match("/^Location:(.+)$/i",
fgets($fp, 1024), $header));
$path = trim($header[1]);
}
fclose($fp);
if ($break) break;
}
?>

[Janwillem Borleffs]

Bill wrote:

Thank you for your patient help.I am learning a lot from it and the code
does now work
However I am still getting the same problem when I try to use a variable
rather than hard code the value in $path.
Here is the code for a page you can try to see what I mean:
I hope my comments are clear.

I see what's happening. In order to be XHTML complient, the page you are
retrieving contains links like the following:

<td><a href="train.asp x?T=FRNCMB++&am p;J=1400907&amp ;R=0">London
Waterloo</a></td>

So, instead of '&', it uses the entity ('&amp;'). This is okay when
clicking the link in your browser, which will automatically convert the
entity to the implied character.

PHP doesn't have this mechanism, so you will have to tell it explicitly
to convert entities. One way of doing this is by calling str_replace, e.g.:

$path = str_replace("&a mp;", "&", $strTarget);
JW