Faster read of filenames

[Sonnich]

Hi all

I use the opendir and readdir to search for files... but they read in
all files, so sometimes I get several hundreds.
If there is a way to give a wildcard, it might be a lot faster.
Since I do this search 100+ times at certain times, this is the thing
which slows me down.
Disableing this part speeds my code up by 90%.

Any ideas?

Sonnich


if($handle = opendir($sSearc hPath))
{
while (false !== ($file = readdir($handle )))
{
$file2 = strtoupper($fil e);
if( (strpos($file2, $code)!==False) && (strpos($file2,
".PDF")!==False ))
{
$files[] = $file;
$paths[] = $sSearchPath;
}
}
closedir($handl e);
}

[Jeff North]

Re: Faster read of filenames

On 31 Oct 2006 03:22:03 -0800, in comp.lang.php "Sonnich"
<sonnich.jensen @elektrobit.com >
<1162293723.742 405.257780@e3g2 000cwe.googlegr oups.comwrote:

>| Hi all
>|
>| I use the opendir and readdir to search for files... but they read in
>| all files, so sometimes I get several hundreds.
>| If there is a way to give a wildcard, it might be a lot faster.
>| Since I do this search 100+ times at certain times, this is the thing
>| which slows me down.
>| Disableing this part speeds my code up by 90%.
>|
>| Any ideas?
>|
>| Sonnich
>|
>|
>| if($handle = opendir($sSearc hPath))
>| {
>| while (false !== ($file = readdir($handle )))
>| {
>| $file2 = strtoupper($fil e);
>| if( (strpos($file2, $code)!==False) && (strpos($file2,
>| ".PDF")!==False ))
>| {
>| $files[] = $file;
>| $paths[] = $sSearchPath;
>| }
>| }
>| closedir($handl e);
>| }

glob() maybe?
foreach (glob("*.txt") as $filename)
{
echo "$filename size " . filesize($filen ame) . "\n";
}
---------------------------------------------------------------
jnorthau@yourpa ntsyahoo.com.au : Remove your pants to reply
---------------------------------------------------------------

[Pedro Graca]

Re: Faster read of filenames

Sonnich wrote:

I use the opendir and readdir to search for files... but they read in
all files, so sometimes I get several hundreds.
If there is a way to give a wildcard, it might be a lot faster.
Since I do this search 100+ times at certain times, this is the thing
which slows me down.
Disableing this part speeds my code up by 90%.
>
Any ideas?

glob()? http://php.net/glob

If your files do not change very often, it may be better to have a
file (or database) with the current files and use that instead of
searching in the file system.

<?php
$filelist = file($path . '/filelist.txt');
foreach ($filelist as $fname) {
$files[] = trim($fname);
$paths[] = $path;
}
?>

Of course you'd have to keep filelist.txt updated.

--
I (almost) never check the dodgeit address.
If you *really* need to mail me, use the address in the Reply-To
header with a message in *plain* *text* *without* *attachments*.

[Sonnich]

Re: Faster read of filenames


Jeff North wrote:

On 31 Oct 2006 03:22:03 -0800, in comp.lang.php "Sonnich"
<sonnich.jensen @elektrobit.com >
<1162293723.742 405.257780@e3g2 000cwe.googlegr oups.comwrote:
>

| Hi all
|
| I use the opendir and readdir to search for files... but they read in
| all files, so sometimes I get several hundreds.
| If there is a way to give a wildcard, it might be a lot faster.
| Since I do this search 100+ times at certain times, this is the thing
| which slows me down.
| Disableing this part speeds my code up by 90%.
|
| Any ideas?
|
| Sonnich
|
|
| if($handle = opendir($sSearc hPath))
| {
| while (false !== ($file = readdir($handle )))
| {
| $file2 = strtoupper($fil e);
| if( (strpos($file2, $code)!==False) && (strpos($file2,
| ".PDF")!==False ))
| {
| $files[] = $file;
| $paths[] = $sSearchPath;
| }
| }
| closedir($handl e);
| }

>
glob() maybe?
foreach (glob("*.txt") as $filename)
{
echo "$filename size " . filesize($filen ame) . "\n";
}

That might do it, but it is case sensitive. And when trying something
like:

foreach (glob($search, GLOB_BRACE ) as $filename)
{
echo "$filename <br>";
}

I get absolutlely nothing. My search paths are something like:
\\server\somewh ere\AB123*.pdf or \\server\somewh ere\AB123*.PDF
Both returns a number of files when runned alone, but adding them
togther
$sx = "{".$s1.",".$s2 ."}";
does not do the trick.

BR
Sonnich

[NC]

Re: Faster read of filenames

Sonnich wrote:

Jeff North wrote:

glob() maybe?

>
That might do it, but it is case sensitive.

But of course it is. On Unix, file names are case-sensitive...

And when trying something like:
>
foreach (glob($search, GLOB_BRACE ) as $filename)
{
echo "$filename <br>";
}
>
I get absolutlely nothing. My search paths are something like:
\\server\somewh ere\AB123*.pdf or \\server\somewh ere\AB123*.PDF
Both returns a number of files when runned alone, but adding them
togther
$sx = "{".$s1.",".$s2 ."}";
does not do the trick.

Try one or more of these:

\\server\somewh ere\*123*.pdf
\\server\somewh ere\??123*.pdf
\\server\somewh ere\{a,A}{b,B}1 23*.pdf

Cheers,
NC

[Sonnich]

Re: Faster read of filenames

After some heavy testing I found that opendir/readdir is app twice as
fast as glob, just as mentioned on www.php.net.
So my advice would be to use opendir/readdir in stead of glob.

BR
Sonnich