Octavian Rasnita wrote: > Hi, > > I need to create a program which reads a file from the LAN, and that file is > continuously updated (kind of log). > The file increases continuously, and it can become very big. > > On every read, I need to read just the lines for the current day (which is > specified in each line) and jump over the lines from the previous days. > > The problem is that if I just open() the file and verify each line to see if > it is a good one, takes very much time, and I need to find a faster > solution. > > Checking each line might take a long time, but I think that verifying each > line means downloading the whole content of the file on my computer, and > this will also make the program lose much time. > > Do you have any idea what can I do to make the program run faster? > > I am thinking to make the program in such a way that after the first run, it > gets the line number of the first line from the current day, or the number > of the first bit of the start of that line, then on next times it runs, it > jumps directly to that line somehow (?), or directly to that bit (using > seek), but I don't know if in that case the content of the file won't be > also downloaded to my computer. > > Do you have any idea how I can download just a part of a file from the > network? >
If my understanding is correct that would have to be something built into the protocol of whatever network protocol you are using. Your seek idea should be correct, and I know for instance modern FTP servers have the ability to seek to a different part of a file, in which case the FTP server only provides that remaining portion of the file on a get (this is how pause/resume downloaders are implemented). And some HTTP servers have this capability. You should probably check with the protocol docs to see if it has the capability. Assuming the protocol is advanced enough then it should respect a seek and not provide the whole file. The best thing to do in this kind of case is to setup some sort of network monitoring and then just try it. See if you open a large file, seek to a location near the end and then see if the network gets slammed or if only the small sample of file is sent. perldoc -f seek For more info. > Help! > > Thank you. > > Teddy > Good luck, http://danconia.org -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>
