python regex - 如何将txt文件中的组替换为另一个txt文件中的另一个组？

Question

So, I have the following txt files:

test1.txt (It's all in the same line.)

(hello)(bye)

text2.txt (It's in two different lines.)

(This actually works)
(Amazing!)

And I have the following regex pattern

\((.*?)\)

Which obviously selects all the words that are inside the parenthesis.

What I want to do is to replace the words inside the () in test1.txt with the words inside the () in test2.txt, leaving test1.txt like:

(This actually works)(Amazing!)

I tried the following code, but it doesn't seem to work. What did I do wrong?

import re

pattern = re.compile("\((.*?)\)")

for line in enumerate(open("test1.txt")):
    match = re.finditer(pattern, line)

for line in enumerate(open("test2.txt")):
    pattern.sub(match, line)

I think I made a very big error, it's one of my first programs in python.

Answer 1

Okay, there are several problems:

1. finditer method returns a match object, not a string. findall returns a list of matched string groups
2. you do the contrary you said. Do you want to replace data in test1 by data from test2 don't you?
3. enumerate returns a tuple so your var line was not a line but a list of [line_number, line_string_content] . I use it in last code block.

So you can try to first catch the content:

pattern = re.compile("\((.*?)\)")
for line in open("test2.txt"):
    match = pattern.findall(line)
#match contains the list ['Amazing!'] from the last line of test2, your variable match is overwritten on each line of the file...

note: If you compile your pattern, you can use it as object to call the re methods.

If you want to do it line by line (big file?).
An other option whould be to load the entire file and create a multiline regex.

matches = []
for line in open("test2.txt"):
    matches.extend(pattern.findall(line))
#matches contains the list ['This actually works','Amazing!']

Then replace the content of the parenthesis by you matches items:

for line in open("test1.txt"):
    for i, match in enumerate(pattern.findall(line)):
        re.sub(match, matches[i], line)

note: doing this will raise exception if there is more (string in parenthesis) in test1.txt than in test2.txt...

If you want to write an output file you should do

with open('fileout.txt', 'w') as outfile:
    for line in enumerate(open("test1.txt")):
        #another writing for the same task (in one line!)
        newline = [re.sub(match, matches[i], line) for i, match in enumerate(pattern.findall(line))][0]
        outfile.write(newline)

Answer 2

You can use the feature of re.sub() to allow a callable as a replacement pattern and create on-the-spot lambda function to go through your matches from test2.txt to achieve your result, eg

import re

# slightly changed to use lookahead and lookbehind groups for a proper match/substitution
pattern = re.compile(r"(?<=\()(.*?)(?=\))")
# you can also use r"(\(.*?\))" if you're preserving the brackets

with open("test2.txt", "r") as f:  # open test2.txt for reading
    words = pattern.findall(f.read())  # grabs all the found words in test2.txt

with open("test1.txt", "r+") as f:  # open test1.txt for reading and writing
    # read the content of test1.txt and replace each match with the next `words` list value
    content = pattern.sub(lambda x: words.pop(0) if words else x.group(), f.read())
    f.seek(0)  # rewind the file to the beginning
    f.write(content)  # write the new, 'updated' content
    f.truncate()  # truncate the rest of the file (if any)

For test1.txt containing:

(hello)(bye)

and test2.txt containing:

(This actually works)
(Amazing!)

executing the above script will change test1.txt to:

(This actually works)(Amazing!)

It will also account for mismatches in the files by iterative replacing only up to how many matches were found in test2.txt (eg if your test1.txt contained (hello)(bye)(pie) it will be changed to (This actually works)(Amazing!)(pie) ).