delete data from file

baxy77bax has asked for the wisdom of the Perl Monks concerning the following question:

Hi, my question is how to delete parts of a large txt file. Example: let say I have a large txt file (a whole novel), and every new paragraph starts with a tab:

  Julie and Jolie were playing in the
shadow when a large spider jomped on 
jolie and started dancing on her head
  Than something else happened , i 
don't know why or how...
[download]

now i want to delete the whole paragraphs that contain the word Julie in the first line. plus i do not want to create a new file but i want to save all changes in the same file. thank you! robert

Comment on delete data from file Download Code

Replies are listed 'Best First'.
Re: delete data from file by moritz (Cardinal) on Aug 31, 2007 at 10:09 UTC
You could try something like Tie::File and set the record separator to `\n\t`.	[reply] [d/l]
Re: delete data from file by andreas1234567 (Vicar) on Aug 31, 2007 at 11:19 UTC
`use strict; use warnings; use English; $INPUT_RECORD_SEPARATOR = "\n\t"; while (<DATA>) { next if (m/Julie/); print; } __DATA__ Julie and Jolie were playing in the shadow when a large spider jomped on jolie and started dancing on her head Than something else happened , i don't know why or how... To my surprise Perl saved the day. Thank you Julie! Yet another reason to use Perl.` [download] `$ perl 636280.pl Than something else happened , i don't know why or how... Yet another reason to use Perl.` [download] Perl inline replace is described in perlrun. -- Andreas	[reply] [d/l] [select]
Re: delete data from file by Taulmarill (Deacon) on Aug 31, 2007 at 09:58 UTC
Most of the time, there is no such thing as changing a file. Whenever you save changes in a file, the program you use will most likely create a new file and overwrite the old one with the new version. That's what i would recommend.	[reply]
Re: delete data from file by cdarke (Prior) on Aug 31, 2007 at 11:30 UTC
You can't really. I will qualify that. If the data you wish to delete is at the end of the file then you could truncate the file, in other words you just move the position of EOF. I guess that is not the case. You could replace the characters that should be deleted with something like 0x00, but that is not guaranteed to be ignored by the code that will read it, unless you have control over that. The only practical way is to create a new file containing only those lines of interest, then rename it to the same name as the original.	[reply]
Re: delete data from file by perl_fan (Novice) on Aug 31, 2007 at 11:32 UTC
hi there, you can use the perl-inline functionality. from the command line just type perl -pi -e 's/Julie and Jolie were playing in//ig' Just complete the paragraph in the above line. cheers, perl_fan	[reply]
Re: delete data from file by MidLifeXis (Monsignor) on Aug 31, 2007 at 18:04 UTC
School is starting again. So in general terms... If you keep two file pointers (one read and one write), open the file read/write, and scan until your first match. Move the read file pointer to the end of the paragraph and the writer to the beginning of the paragraph. Now read from one and write to the other, skipping the write when the read runs into your key string. At the end, make sure you truncate to the end of the file. Update: Forgot to truncate --MidLifeXis	[reply]


"be consistent"
	PerlMonks