python替换; 如果行以关键字开头

Question

I have a textfile containing thousands of entries like:

@INBOOK{Abu-Lughod1991,
  chapter = {Writing against culture},
  pages = {137-162},
  title = {Recapturing anthropology},
  publisher = {School of American Research Press},
  year = {1991},
  editor = {Richard Fox},
  author = {Abu-Lughod, Lila},
  address = {Santa Fe /NM},
  abstract = {Im Zusammenhang mit der Debatte um die writing culture fomuliert AL
        eine feministische Kritik und zeigt, wie von dort doch Anregungen
        für die Reflektion der Schreibweise und Repräsentation gekommen sind.*},
  crossref = {Rabinow1986},
  keywords = {Frauen; Feminismus; Erzählung als EG; Repräsentation; Roman; Schreibtechnik;
        James Clifford; writing culture; Dialog;},
  owner = {xko},
  systematik1 = {Anth\theor\Ethnographie},
  systematik2 = {Anth\theor\Text & Ges},
  timestamp = {1995-12-02}
}

I will replace all semicolons in the keywords - field to comma. But only in the keywords field - other fields should be untouched:

keywords = {Frauen, Feminismus, Erzählung als EG, Repräsentation, Roman, Schreibtechnik, James Clifford, writing culture, Dialog,},

I'm not a programmer and maybe the following code-snippet is a good starting point and i would really appreciate if someone could complete it.

outfile = open("literatur_comma.txt", "w") 
for line in open("literatur_semicolon.txt", "r"): 
    if line  # starts with "keywords" replace all semicolon with comma
        outfile.write(line) # write in new file
outfile.close() 

Thanks a lot!

EDIT: Thanks for all your answers and codes, that's great! I did a mistake in my thoughts and if i use my code-wrapper (with outfile), then it creates a new file with the keywords in it. How can i use the same file and replaces only the semicolons to comma in keywords line?

Answer 1

Something like this works for a single line.

if line.strip().startswith('keywords'):
    line = line.replace(';',',')
outfile.write(line) 

If keywords spans multiple lines in your actual text file though, this won't get the job done.

Answer 2

outfile = open("literatur_comma.txt", "w") 
for line in open("literatur_semicolon.txt", "r"): 
    if line.startswith('keywords'):  # starts with "keywords" replace all semicolon with comma
        outfile.write(line.replace(';',',')) # write in new file
outfile.close() 

Answer 3

using pyparsing

Note: this is one way to do it, but brain isn't in parsing mode - so this is an idea rather than a proper answer... It certainly needs some work, but might well be the right direction...

A somewhat messy example of using pyparsing ... (could be much nicer, with some @INBOOK and wotsit checking and parsing, but anyway...)

from pyparsing import *

keywords = originalTextFor(Keyword('keywords') + '=')
values = delimitedList(Regex('[^;}]+'), ';')
values.setParseAction(lambda L: ', '.join(L))

Where text is your example:

>>> print values.transformString(text)
@INBOOK{Abu-Lughod1991,
  chapter = {Writing against culture},
  pages = {137-162},
  title = {Recapturing anthropology},
  publisher = {School of American Research Press},
  year = {1991},
  editor = {Richard Fox},
  author = {Abu-Lughod, Lila},
  address = {Santa Fe /NM},
  abstract = {Im Zusammenhang mit der Debatte um die writing culture fomuliert AL
        eine feministische Kritik und zeigt, wie von dort doch Anregungen
        für die Reflektion der Schreibweise und Repräsentation gekommen sind.*},
  crossref = {Rabinow1986},
  keywords = {Frauen, Feminismus, Erzählung als EG, Repräsentation, Roman, Schreibtechnik, James Clifford, writing culture, Dialog;},
  owner = {xko},
  systematik1 = {Anth   heor\Ethnographie},
  systematik2 = {Anth   heor\Text & Ges},
  timestamp = {1995-12-02}