php-general Digest 1 Jun 2009 02:54:13 -0000 Issue 6152
php-general Digest 1 Jun 2009 02:54:13 - Issue 6152 Topics (messages 293399 through 293403): spawning a process that uses pipes - doesn't terminate when webpage download is canceled 293399 by: flint 293402 by: bruce Re: mysql_query takes long time... 293400 by: flint 293401 by: Phpster Directing form to different handlers? 293403 by: Angus Mann Administrivia: To subscribe to the digest, e-mail: php-general-digest-subscr...@lists.php.net To unsubscribe from the digest, e-mail: php-general-digest-unsubscr...@lists.php.net To post to the list, e-mail: php-gene...@lists.php.net -- ---BeginMessage--- sent this before, don't know if it went through... someone please reply if it went, even if they don't know answer?... so here's the scenario.. I have a site that uses php with a database to offer sound files to users using streaming methods. the request page has options, allowing the user to modify the sound file in various ways, before having it sent to them Here's the problem: The method i'm using to feed the data to the user is to run the source file through various piped commands, with the resulting audio being dumped to stdout, and then using passthru in php to get that data to the enduser. here's an example, for serving an MP3 with its pitch/speed changed by sox: passthru(lame --quiet --decode \ . $in_file . \ - | . sox -V -S -t wav - -t wav - speed . $speed_factor . | . lame --quiet . $lame_params . - -); This works just fine, except the problem is if the end user aborts the transfer (e.g. stops playback in the media player, cancels download of the mp3, whatever) then it leaves behind both the sox process and the decoder LAMe process along with the sh that's running them. the only process that exits is the final encoding lame process. If the sound file runs to completion, everythign exits properly. But this obviously means enough cancelling of downloads means the server ends up with a huge batch of stuck processes! And I even tried simply killing the 'host' sh process, and the lame and sox processes remain anyway. The only way I've been able to deal with this is manually killing the lame and sox processes directly. is there any way I can make this work, such so that if the user cancels the transfer, all relavent processes are killed rather than just the single process that's feeding output into php? -FM ---End Message--- ---BeginMessage--- we answered this a number of times... was there something in the replies that didn't satisfy you? -Original Message- From: flint [mailto:fmill...@gmail.com] Sent: Sunday, May 31, 2009 6:53 AM To: PHP-General List Subject: [PHP] spawning a process that uses pipes - doesn't terminate when webpage download is canceled sent this before, don't know if it went through... someone please reply if it went, even if they don't know answer?... so here's the scenario.. I have a site that uses php with a database to offer sound files to users using streaming methods. the request page has options, allowing the user to modify the sound file in various ways, before having it sent to them Here's the problem: The method i'm using to feed the data to the user is to run the source file through various piped commands, with the resulting audio being dumped to stdout, and then using passthru in php to get that data to the enduser. here's an example, for serving an MP3 with its pitch/speed changed by sox: passthru(lame --quiet --decode \ . $in_file . \ - | . sox -V -S -t wav - -t wav - speed . $speed_factor . | . lame --quiet . $lame_params . - -); This works just fine, except the problem is if the end user aborts the transfer (e.g. stops playback in the media player, cancels download of the mp3, whatever) then it leaves behind both the sox process and the decoder LAMe process along with the sh that's running them. the only process that exits is the final encoding lame process. If the sound file runs to completion, everythign exits properly. But this obviously means enough cancelling of downloads means the server ends up with a huge batch of stuck processes! And I even tried simply killing the 'host' sh process, and the lame and sox processes remain anyway. The only way I've been able to deal with this is manually killing the lame and sox processes directly. is there any way I can make this work, such so that if the user cancels the transfer, all relavent processes are killed rather than just the single process that's feeding output into php? -FM -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php ---End Message--- ---BeginMessage--- - Original Message - From: flint fmill...@gmail.com To: דניאל דנון danondan...@gmail.com Sent: Sunday, May 31, 2009 9:21 AM Subject: Re: [PHP]
Re: [PHP] mysql_query takes long time...
You can also stack the queries to run multiple rows in one insert Insert into table values (row1col1, row1col2,'row1col3'), (row2col1,row2col2,'row2col3'),...(rowNcol1,rowNcol2,'rowNcol3') Bastien Sent from my iPod On May 31, 2009, at 8:18, דניאל דנון danondan...@gmail.com wrote: I've a file of about 500,000 lines, each line contains a string in variety of lengths, but no less then 3 characters and usually no more then 120. average of about 80, and maximum of about 250. I made a PHP script to fetch the data (using fgets), process it and insert it to a MySQL database. The problem is inserting to MySQL takes about 0.02 seconds, which looks like nothing - but when you have 500,000 lines to insert... The while goes like that: fgets from file x1 = some function about the string x2 = some other function about the string x3 = the string insert into table (field1, field2, field3) VALUES (x1, x2, x3) (pseudo-code) I was wondering, is there any faster way to perform it, assuming I have to do it with PHP? also, if it matters - the MySQL table got id in auto increment. Yours, Daniel. -- Use ROT26 for best security -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] spawning a process that uses pipes - doesn't terminate when webpage download is canceled
we answered this a number of times... was there something in the replies that didn't satisfy you? -Original Message- From: flint [mailto:fmill...@gmail.com] Sent: Sunday, May 31, 2009 6:53 AM To: PHP-General List Subject: [PHP] spawning a process that uses pipes - doesn't terminate when webpage download is canceled sent this before, don't know if it went through... someone please reply if it went, even if they don't know answer?... so here's the scenario.. I have a site that uses php with a database to offer sound files to users using streaming methods. the request page has options, allowing the user to modify the sound file in various ways, before having it sent to them Here's the problem: The method i'm using to feed the data to the user is to run the source file through various piped commands, with the resulting audio being dumped to stdout, and then using passthru in php to get that data to the enduser. here's an example, for serving an MP3 with its pitch/speed changed by sox: passthru(lame --quiet --decode \ . $in_file . \ - | . sox -V -S -t wav - -t wav - speed . $speed_factor . | . lame --quiet . $lame_params . - -); This works just fine, except the problem is if the end user aborts the transfer (e.g. stops playback in the media player, cancels download of the mp3, whatever) then it leaves behind both the sox process and the decoder LAMe process along with the sh that's running them. the only process that exits is the final encoding lame process. If the sound file runs to completion, everythign exits properly. But this obviously means enough cancelling of downloads means the server ends up with a huge batch of stuck processes! And I even tried simply killing the 'host' sh process, and the lame and sox processes remain anyway. The only way I've been able to deal with this is manually killing the lame and sox processes directly. is there any way I can make this work, such so that if the user cancels the transfer, all relavent processes are killed rather than just the single process that's feeding output into php? -FM -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Directing form to different handlers?
Hi all. I realize this is more an HTML question than PHP but I'm sure someone here can help. I have several forms with lots (dozens) of text inputs. If the user presses the Update button I want the form handled by update.php but if they press Delete it needs to be handled by delete.php or add.php and so-on depending on the button they press. But when establishing the form I can only have form method=POST action=delete.php or add.php or whatever. Is there a way to direct the content of the form do a different handler depending on the button? I know I can use javascript to direct to a constructed URL and append ?name=smithaddress=hishousetelephone=28376.and so on but this is not practical when there are dozens of entriesthe URL becomes massive. I prefer to use POST and then use PHP to extract the POST array. Any ideas? Much appreciated. Angus
Re: [PHP] Directing form to different handlers?
On Mon, 2009-06-01 at 12:53 +1000, Angus Mann wrote: Hi all. I realize this is more an HTML question than PHP but I'm sure someone here can help. I have several forms with lots (dozens) of text inputs. If the user presses the Update button I want the form handled by update.php but if they press Delete it needs to be handled by delete.php or add.php and so-on depending on the button they press. But when establishing the form I can only have form method=POST action=delete.php or add.php or whatever. Is there a way to direct the content of the form do a different handler depending on the button? I know I can use javascript to direct to a constructed URL and append ?name=smithaddress=hishousetelephone=28376.and so on but this is not practical when there are dozens of entriesthe URL becomes massive. I prefer to use POST and then use PHP to extract the POST array. Any ideas? Yes, have the form submit to the URL that presented the form. Have THAT PHP script determine which button was clicked. Then delegate to the appropriate handler function. Cheers, Rob. -- http://www.interjinn.com Application and Templating Framework for PHP -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
Re: [PHP] spawning a process that uses pipes - doesn't terminate when webpage download is canceled
On Sun, 2009-05-31 at 08:52 -0500, flint wrote: sent this before, don't know if it went through... someone please reply if it went, even if they don't know answer?... so here's the scenario.. I have a site that uses php with a database to offer sound files to users using streaming methods. the request page has options, allowing the user to modify the sound file in various ways, before having it sent to them Here's the problem: The method i'm using to feed the data to the user is to run the source file through various piped commands, with the resulting audio being dumped to stdout, and then using passthru in php to get that data to the enduser. here's an example, for serving an MP3 with its pitch/speed changed by sox: passthru(lame --quiet --decode \ . $in_file . \ - | . sox -V -S -t wav - -t wav - speed . $speed_factor . | . lame --quiet . $lame_params . - -); This works just fine, except the problem is if the end user aborts the transfer (e.g. stops playback in the media player, cancels download of the mp3, whatever) then it leaves behind both the sox process and the decoder LAMe process along with the sh that's running them. the only process that exits is the final encoding lame process. If the sound file runs to completion, everythign exits properly. But this obviously means enough cancelling of downloads means the server ends up with a huge batch of stuck processes! And I even tried simply killing the 'host' sh process, and the lame and sox processes remain anyway. The only way I've been able to deal with this is manually killing the lame and sox processes directly. is there any way I can make this work, such so that if the user cancels the transfer, all relavent processes are killed rather than just the single process that's feeding output into php? Use something else to pass the data back to the user... popen() comes to mind or proc_open(). Then disable auto abort on user disconnect via ignore_user_abort(). Then after sending periodic data chunks, check the user connection status via connection_aborted(). If your script finds that the user has aborted, then kill all the processes in the pipeline from the PHP script before finally aborting the PHP script itself. Cheers, Rob. -- http://www.interjinn.com Application and Templating Framework for PHP -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
RE: [PHP] spawning a process that uses pipes - doesn't terminatewhen webpage download is canceled
hi robert.,, now you've got me curious.. you state... -Use something else to pass the data back to the user... popen() comes to -mind or proc_open(). Then disable auto abort on user disconnect via -ignore_user_abort(). Then after sending periodic data chunks, check the -user connection status via connection_aborted(). If your script finds -that the user has aborted, then kill all the processes in the pipeline -from the PHP script before finally aborting the PHP script itself. but if the user is using a browser session... which has a web server connection... are you siggesting that the server app spawns off a child process via the popen, and that the child somehow connects to the existing browser session?? walk through the psuedo logic/flow of this if you don't mind.. i must be missing something.. thanks -Original Message- From: Robert Cummings [mailto:rob...@interjinn.com] Sent: Sunday, May 31, 2009 8:16 PM To: flint Cc: PHP-General List Subject: Re: [PHP] spawning a process that uses pipes - doesn't terminatewhen webpage download is canceled On Sun, 2009-05-31 at 08:52 -0500, flint wrote: sent this before, don't know if it went through... someone please reply if it went, even if they don't know answer?... so here's the scenario.. I have a site that uses php with a database to offer sound files to users using streaming methods. the request page has options, allowing the user to modify the sound file in various ways, before having it sent to them Here's the problem: The method i'm using to feed the data to the user is to run the source file through various piped commands, with the resulting audio being dumped to stdout, and then using passthru in php to get that data to the enduser. here's an example, for serving an MP3 with its pitch/speed changed by sox: passthru(lame --quiet --decode \ . $in_file . \ - | . sox -V -S -t wav - -t wav - speed . $speed_factor . | . lame --quiet . $lame_params . - -); This works just fine, except the problem is if the end user aborts the transfer (e.g. stops playback in the media player, cancels download of the mp3, whatever) then it leaves behind both the sox process and the decoder LAMe process along with the sh that's running them. the only process that exits is the final encoding lame process. If the sound file runs to completion, everythign exits properly. But this obviously means enough cancelling of downloads means the server ends up with a huge batch of stuck processes! And I even tried simply killing the 'host' sh process, and the lame and sox processes remain anyway. The only way I've been able to deal with this is manually killing the lame and sox processes directly. is there any way I can make this work, such so that if the user cancels the transfer, all relavent processes are killed rather than just the single process that's feeding output into php? Use something else to pass the data back to the user... popen() comes to mind or proc_open(). Then disable auto abort on user disconnect via ignore_user_abort(). Then after sending periodic data chunks, check the user connection status via connection_aborted(). If your script finds that the user has aborted, then kill all the processes in the pipeline from the PHP script before finally aborting the PHP script itself. Cheers, Rob. -- http://www.interjinn.com Application and Templating Framework for PHP -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Re: PHP vs ASP.NET
Hello, on 05/28/2009 10:20 AM Olexandr Heneralov said the following: Hi! Guys, you of course, know that ASP.NET becomes more and more popular in the world. I have a question for everyone: Can it happen so that PHP will be replaced with ASP.NET? ASP.NET is not a language. It is more like a framework that can run multiple languages. It can run VB.NET, C# and even PHP (although it is not usual). I am not sure what are the current numbers, but the latest statistics that I have seen Apache was running on 72% of the Internet Web servers against only 17% of Microsoft IIS. Although you can run ASP.NET on Apache via mono, that is unusual. Also PHP is the most popular Apache extension present in between 40% and 50% of Apache installations. This means that PHP is present in 1/3 of the Internet Web servers, which represents about half of the PHP market share. This article provides more details about PHP market share. http://www.phpclasses.org/blog/post/95-How-large-is-the-PHP-market.html As of ASP.NET becoming more popular than PHP, I don't think that even Microsoft believes that is possible. Actually Microsoft is very concerned that PHP runs well on Windows and is sponsoring a lot of PHP activity. This other article talks about what is Microsoft so interested in PHP now. http://www.phpclasses.org/blog/post/85-What-is-Microsoft-up-to-with-PHP.html -- Regards, Manuel Lemos Find and post PHP jobs http://www.phpclasses.org/jobs/ PHP Classes - Free ready to use OOP components written in PHP http://www.phpclasses.org/ -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
[PHP] Jacob's Calendar
Hi I am creating a birthday calendar of all my friends and family. Can you please click on the link below to enter your birthday for me? http://www.birthdayalarm.com/bd2/85206071a420999425b1469532889c603775600d905 Thanks, Jacob -- PHP General Mailing List (http://www.php.net/) To unsubscribe, visit: http://www.php.net/unsub.php
