Re: Keeping file pointers open after script end

On Sat, 2007-06-09 at 22:23 +0100, Stut wrote: > Tijnema wrote: > > On 6/9/07, Stut <stuttle@gmail.com> wrote: > >> Tijnema wrote: > >> > On 6/9/07, Stut <stuttle@gmail.com> wrote: > >> >> Tijnema wrote: > >> >> > On 6/9/07, Stut <stuttle@gmail.com> wrote: > >> >> >> Tijnema wrote: > >> >> >> > Hmm, fseek seems cool, but what about FTP resources? If I open > >> them > >> >> >> > with ftp_connect, do I need to fetch all data from FTP again, > >> >> >> > and > >> >> then > >> >> >> > just trash all data I don't need? > >> >> >> > >> >> >> Yes, but depending on what you're actually doing you may be > >> >> >> able to > >> >> >> cache enough to skip large chunks in subsequent requests. > >> >> >> > >> >> >> -Stut > >> >> > > >> >> > Well, I'm working on a script that can transfer large files over > >> HTTP. > >> >> > The point is that I have a HTTP proxy at school, and so I can use > >> HTTP > >> >> > only. Moreover, they blocked downloading files bigger than 1MB, and > >> >> > so, I wanted to transfer files in packets of lets say 990KB, so that > >> >> > they can be downloaded fine through the HTTP proxy. This means i'm > >> >> > calling repeatly the script for a next chunk, and that's why I > >> wanted > >> >> > to keep the FTP connection open. > >> >> > >> >> Write it to a temporary file and store that temporary filename in the > >> >> session. When the page is called, get the filesize of the temporary > >> file > >> >> and use that as the start of the next chunk to download. > >> >> > >> >> -Stut > >> > > >> > Do you think this is the best way to do? > >> > Even for large files , let's say 1 DVD (4,7GB)? > >> > >> I can't think of a better way. This way you're building up the file on > >> disk chunk by chunk, not storing much in the session and it should scale > >> to any size of file quite well. The only thing you might have to watch > >> is cleaning up the files created by aborted downloads. > >> > >> -Stut > > > > What about creating a new file for each chunk? > > Would that be faster and less resource intensive then using fseek all > > the time? > > No, it would be more resource intensive and probably slower. Seeking > through a file is a very cheap thing for most OS's to do. As long as it's not on a tape drive >:) Cheers, Rob. -- .------------------------------------------------------------. | InterJinn Application Framework - http://www.interjinn.com | :------------------------------------------------------------: | An application and templating framework for PHP. Boasting | | a powerful, scalable system for accessing system services | | such as forms, properties, sessions, and caches. InterJinn | | also provides an extremely flexible architecture for | | creating re-usable components quickly and easily. | `------------------------------------------------------------'