Text file manipulation.

[Joe1986]

Hi there, im trying to create a python program that can read a text file line by line and search for specified words/text/strings and remove them from the text file. Then finally save the modified text file to an output file. The only problem is, the text file contains code for a BACKSPACE typed in the text. e.g. "<BACKSPACE& gt;" this needs to be removed which sounds quite simple, but often there are numbers involved. e.g. "<BACKSP ACE: 4>" which would represent 4 backspaces, and so the string needs to be removed and 4 backspaces take place on the text before the code. e.g. "repatre<BAC KSPACE: 4>resent" would become "represent" I know a search and delete script that can search for a word at the beggining of the line and then delete the whole line, but im not sure about this. Any ideas would be great.
Cheers,
Joe

[kudos]

hi,
start by searching for <BACKSPAC E (for instance by using the find method). Note the index of it, then check what the value of the char after BACKSPACE (you handle it differently if it is & or :). substring the string and clue it together without the backspace, and go n characters back, given what the value is. I would recommend to string replace every <BACKSPACE&g t; with <BACKSPAC E: 1> (so you only need to handle one case)

Did this help, or was my post confusing?
-kudos

Hi there, im trying to create a python program that can read a text file line by line and search for specified words/text/strings and remove them from the text file. Then finally save the modified text file to an output file. The only problem is, the text file contains code for a BACKSPACE typed in the text. e.g. "<BACKSPACE& gt;" this needs to be removed which sounds quite simple, but often there are numbers involved. e.g. "<BACKSP ACE: 4>" which would represent 4 backspaces, and so the string needs to be removed and 4 backspaces take place on the text before the code. e.g. "repatre<BAC KSPACE: 4>resent" would become "represent" I know a search and delete script that can search for a word at the beggining of the line and then delete the whole line, but im not sure about this. Any ideas would be great.
Cheers,
Joe

[Joe1986]

hi,
start by searching for <BACKSPAC E (for instance by using the find method). Note the index of it, then check what the value of the char after BACKSPACE (you handle it differently if it is & or :). substring the string and clue it together without the backspace, and go n characters back, given what the value is. I would recommend to string replace every <BACKSPACE&g t; with <BACKSPAC E: 1> (so you only need to handle one case)

Did this help, or was my post confusing?
-kudos

Hi there,

Im a little confused. Im not sure I understand the string replace method you mentioned at the end of your post.
cheers,
Joe

[Joe1986]

Hi,
Here is a solution. I can write results to a txt file which is great. All I need to do now is read from a text which im guessing will just go in place of the 'test' section in the code and to be able to iterate this code so that I can run through a whole paragraph containing multiple <BACKSPACE> sections removing all (rather than getting to the first one and terminating.) Would you have any ideas looking at this to do such a thing? Any help would be great. Cheers. Joe

#! /usr/bin/python

import re
# Global variable


bs = re.compile('<BACKSPACE(:[ ]*[0-9]+)?>')


#theInFile = open("test2.txt", "r")

theOutFile = open("backspace_out.txt", "w")



tests = ['re <BACKSPACE>present', 'This is bound to repatreaaaabbbbcccc <BACKSPACE: 17>resent']

def bs_remove(s):

    global bs

    for m in bs.finditer(s):

        if m.groups()[0] is None:

            return s[:m.start() - 1] + s[m.end():]

        else:

            return s[:m.start() - int(m.groups()[0][1:])] + s[m.end():]



for s in tests:

      theOutFile.write (bs_remove(s)+ " ")

[bvdet]

Hi,
Here is a solution. I can write results to a txt file which is great. All I need to do now is read from a text which im guessing will just go in place of the 'test' section in the code and to be able to iterate this code so that I can run through a whole paragraph containing multiple <BACKSPACE> sections removing all (rather than getting to the first one and terminating.) Would you have any ideas looking at this to do such a thing? Any help would be great. Cheers. Joe

#! /usr/bin/python

import re
# Global variable


bs = re.compile('<BACKSPACE(:[ ]*[0-9]+)?>')


#theInFile = open("test2.txt", "r")

theOutFile = open("backspace_out.txt", "w")



tests = ['re <BACKSPACE>present', 'This is bound to repatreaaaabbbbcccc <BACKSPACE: 17>resent']

def bs_remove(s):

    global bs

    for m in bs.finditer(s):

        if m.groups()[0] is None:

            return s[:m.start() - 1] + s[m.end():]

        else:

            return s[:m.start() - int(m.groups()[0][1:])] + s[m.end():]



for s in tests:

      theOutFile.write (bs_remove(s)+ " ")

To account for single or multiple occurrences of backspaces in each line:[code=Python]# &lt;BACKSPACE&g t; 1 backspace
# &lt;BACKSPAC E: 4&gt; 4 backspaces

import re

def remove_bs(s):
patt = re.compile(r'&l t;BACKSPACE:? ?(\d+)?&gt;')
while True:
m = patt.search(s)
if m:
if m.group(1):
start = m.start()-int(m.group(1))
else:
start = m.start()-1
s = ''.join([s[:start], s[m.end():]])
else:
break
return s

def parse_file(fn):
return ''.join([remove_bs(line) for line in open(fn).readli nes()])

fn = 'input.txt'
fnOut = 'output.txt
f = open(fnOut, 'w')
f.write(parse_f ile(fn))
f.close()[/code]

[thisissuma]

Hi there, im trying to create a python program that can read a text file line by line and search for specified words/text/strings and place them in another text file.
where line length should be 65 characters and number of lines per page are 70.