使用 CSV 列在文本文件中搜索和替换

Question

Background

I have a two column CSV file like this:

Find	Replace
is	was
A	one
b	two

etc.

First column is text to find and second is text to replace.

I have second file with some text like this:

"This is A paragraph in a text file." (Please note the case sensitivity)

My requirement:

I want to use that csv file to search and replace in the text file with three conditions:-

1. whole word replacement.
2. case sensitive replacement.
3. Replace all instances of each entry in CSV

Script tried:

with open(CSV_file.csv', mode='r') as infile:
    reader = csv.reader(infile)
    mydict = {(r'\b' + rows[0] + r'\b'): (r'\b' + rows[1]+r'\b') for rows in reader}<--Requires Attention
    print(mydict)

with open('find.txt') as infile, open(r'resul_out.txt', 'w') as outfile:
    for line in infile:
        for src, target in mydict.items():
            line = re.sub(src, target, line)  <--Requires Attention
            # line = line.replace(src, target)
        outfile.write(line)

Description of script I have loaded my csv into a python dictionary and use regex to find whole words.

Problems

I used r'\\b' to make word boundry in order to make whole word replacement but output gives me "\\\\b" in the dictionary instead of '\\b' ??

using REPLACE function gives like:

"Thwas was one paragraph in a text file."

secondly I don't know how to make replacement case sensitive in regex pattern?

If anyone know better solution than this script or can improve the script?

Thanks for help if any..

Answer 1

I'd just put pure strings into mydict so it looks like

{'is': 'was', 'A': 'one', ...}

and replace this line:

# line = re.sub(src, target, line) # old
line = re.sub(r'\b' + src + r'\b', target, line) # new

Note that you don't need \\b in the replacement pattern. Regarding your other questions,

• regular expressions are case-sensitive by default,
• changing '\\b' to '\\\\b' is exactly what the r'' does. You can omit the r and write '\\\\b' , but that quickly gets ugly with more complex regexs.

Answer 2

Here's a more cumbersome approach (more code) but which is easier to read and does not rely on regular expressions. In fact, given the very simple nature of your CSV control file, I wouldn't normally bother using the csv module at all:-

import csv

with open('temp.csv', newline='') as c:
    reader = csv.DictReader(c, delimiter=' ')
    D = {}
    for row in reader:
        D[row['Find']] = row['Replace']
    with open('input.txt', newline='') as infile:
        with open('output.txt', 'w') as outfile:
            for line in infile:
                tokens = line.split()
                for i, t in enumerate(tokens):
                    if t in D:
                        tokens[i] = D[t]
                outfile.write(' '.join(tokens)+'\n')