regular expression inquiry

[vito]

I'm processing the following sequence with length more than 100k

1 cagatgctga taaaaaagtg tgttcctcat agcatttatt taattgaaat atttcaagaa
61 cttgaatgta ctaaaaattg agacaaacag tagcaaatca taaaaaaaaa ttgaagtgaa
121 ttttacaact ggattcatgt gcctaatatt ttcattggga agtggattca tgtttaacat
181 ttccattggg <snippet>

i wrote a program

<?php
session_start() ;

if (isset ($_POST['seq']) )
$seq = $_POST['seq'];
else
$seq= $_GET['seq'];

$seq = preg_replace("/[\s\r\n0-9]/", "", $seq);
echo $seq;

?>

but it generates an output of fragmented sequences (i.e. partially processed
result), what is the problem?

[Benjamin Esham]

Re: regular expression inquiry

vito wrote:

i wrote a program
>
$seq = preg_replace("/[\s\r\n0-9]/", "", $seq);
>
but it generates an output of fragmented sequences (i.e. partially
processed result), what is the problem?

Your regular expression may be giving unexpected results: since the RE is
double-quoted, the \r and \n are converted to their respective special
characters *before* being sent to preg_replace(), while the \s is left to be
processed by preg_replace(). You might try

$seq = preg_replace('/[[:space:]0-9]/', '', $seq);

which sidesteps the issue entirely by using [:space:]. Personally, I would
just use

$seq = preg_replace('/[^acgt]/', '', $seq);

which removes everything except for the characters [acgt]. This will work
no matter what other stuff happens to be present in the input file.

HTH,
--
Benjamin D. Esham
bdesham@gmail.c om | AIM: bdesham128 | Jabber: same as e-mail
Más sabe el diablo por viejo que por diablo. (Spanish proverb)

[yawnmoth]

Re: regular expression inquiry

vito wrote:

I'm processing the following sequence with length more than 100k
>
1 cagatgctga taaaaaagtg tgttcctcat agcatttatt taattgaaat atttcaagaa
61 cttgaatgta ctaaaaattg agacaaacag tagcaaatca taaaaaaaaa ttgaagtgaa
121 ttttacaact ggattcatgt gcctaatatt ttcattggga agtggattca tgtttaacat
181 ttccattggg <snippet>
>
i wrote a program
>
<?php
session_start() ;
>
if (isset ($_POST['seq']) )
$seq = $_POST['seq'];
else
$seq= $_GET['seq'];
>
$seq = preg_replace("/[\s\r\n0-9]/", "", $seq);
echo $seq;
>
?>
>
but it generates an output of fragmented sequences (i.e. partially processed
result), what is the problem?

To quote from php.net's "Pattern Syntax" article:

"By default, a whitespace character (eg. \s) is any character that the
C library function isspace() recognizes, though it is possible to
compile PCRE with alternative character type tables. Normally isspace()
matches space, formfeed, newline, carriage return, horizontal tab, and
vertical tab."

Per that, including \r and \n in the class definition is redundant.

Anyway, what you should get with the code you wrote is an uninterupted
sequence of a, g, t, and c's. What would you get had you provided the
above sequence? And what's the output that you want?

[Colin McKinnon]

Re: regular expression inquiry

vito wrote:

I'm processing the following sequence with length more than 100k
>
1 cagatgctga taaaaaagtg tgttcctcat agcatttatt taattgaaat atttcaagaa
61 cttgaatgta ctaaaaattg agacaaacag tagcaaatca taaaaaaaaa
ttgaagtgaa
121 ttttacaact ggattcatgt gcctaatatt ttcattggga agtggattca
tgtttaacat 181 ttccattggg <snippet>
>

Nice genes.

i wrote a program
>
<?php
session_start() ;
>
if (isset ($_POST['seq']) )
$seq = $_POST['seq'];
else
$seq= $_GET['seq'];
>
$seq = preg_replace("/[\s\r\n0-9]/", "", $seq);
echo $seq;
>
?>

You provided an example of the input but not the output, so I'm not quite
sure. But...

Since you're using PCRE, why not:
/[\s\r\n\d]/

or even

/[^ctag]/

It improbable, but you do know that \s doesn't match ascii chr(11) but
[:space:] does.

HTH
C.