----- Original Message ----- From: "Bob" Hi Rob, Thanks very much for your help. I'm going to have to work though it at a later date. I'm just managing to check my email daily, and that's it at the moment.
I transfer replies to my colour coded editor, then print them out for reference. I have noticed on your past posts that you have some interesting techniques also, which I have saved. Thanks again. Regards, Bob E. ----------------------------------- Hi Bob, Most of what I post is hand written code, often bugs and all! But that last script for php 4.x.x was straight off the php.net web site. I really don't spend any time with php 4.x.x any more. I looked to find one of my servers that still had it available but no good. And I didn't want to install an older version on a WAMP. Someone mentioned the use of $d = implode($d); and said it would use extra memory. I don't see why it would although I agree something like $new_var = implode($d); would. Perhaps they can enlighten me? In any case I only write strings to files. For objects I use serialize(); and then write the resulting string. The code I sent is a bit of a dog with the itineration over an array. Php 5.x.x has far superior array manipulating functions. You may be able to achieve the same if you push the new value and old array into an empty stack in php 4.x.x If I get time I will see what I can do in php 4.x.x but in reality I think it is time for you to look for new hosting! Someone mentioned using fseek. The basic principle is to feed straight from one stream to another. This vastly improves speed and the memory usage in absolutely minimal compared to other methods. Because it is file steam to file stream, the large data file never has to sit in memory. The script I sent was only intended for about 100 or less data items. The problem with fseek is that you need to determine where to seek to. To use this efficiently you need to pad your data items out to a fixed record length then you can calculate the seek position easily without loading the stream into memory. Remember that NULL is stored as a NULL value and uses the same storage as a normal character. If you don't want to used fixed length records then you need a delimiter that only exists between records. Then you read character by character and count delimiters. Alternatively you can focus on a file size rather than a number of records. This way you can seek to a position say 20KB from the end of the file and then just go character by character until immediately past the first found delimiter. Basic principle for fixed record- Get new data item Pad it to a fixed record length fopen input file flock input file get size of input file f seek to size of input file - (desired number of records -1) * record length + any needed common adjustment fopen output file pass through remaining data from input file to output file write new record to output file fclose file streams unlink input file rename output file to input file release flocks In the past I have done this and there can be server issues from time to time so it is best to have a backup file. new file becomes > log file log file becomes > backup file backup file is unlinked Hope this helps ------------------------------------ Please remember to write your response BELOW the previous text. Community email addresses: Post message: php-list@yahoogroups.com Subscribe: [EMAIL PROTECTED] Unsubscribe: [EMAIL PROTECTED] List owner: [EMAIL PROTECTED] Shortcut URL to this page: http://groups.yahoo.com/group/php-listYahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/php-list/ <*> Your email settings: Individual Email | Traditional <*> To change settings online go to: http://groups.yahoo.com/group/php-list/join (Yahoo! ID required) <*> To change settings via email: mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
