Python: find a string between 2 strings in text - python

I have a text like this
s = """
...
(1) Literature
1. a.
2. b.
3. c.
...
"""
I want to cut Literature section but I have some problem with detection.
I use here
re.search("(1) Literature\n\n(.*).\n\n", s).group(1)
but search return None.
Desire output is
(1) Literature
1. a.
2. b.
3. c.
What did I do wrong?

You could match (1) Literature and 2 newlines, and then capture all lines that start with digits followed by a dot.
\(1\) Literature\n\n((?:\d+\..*(?:\n|$))+)
The pattern matches:
\(1\) Literature\n\n Match (1) Literature and 2 newlines
( Capture group 1
(?: Non capture group
\d+\..*(?:\n|$) Match 1+ digits and a dot followed by either a newline or end of string
)+ Close non capture group and repeat it 1 or more times to match all the lines
) Close group 1
Regex demo
Another option is to capture all following lines that do not start with ( digits ) using a negative lookahead, and then trim the leading and trailing whitespaces.
\(1\) Literature((?:\n(?!\(\d+\)).*)*)
Regex demo

Parentheses have a special meaning in regex. They are used to group matches.
(1) - Capture 1 as the first capturing group.
Since the string has parentheses in it, the match is not successful. And .* capturing end with line end.
Check Demo
Based on your regex, I assumed you wanted to capture the line with the word Literature, 5 lines below it. Here is a regex to do so.
\(1\) Literature(.*\n){5}
Regex Demo
Note the scape characters used on parentheses around 1.
EDIT
Based on zr0gravity7's comment, I came up with this regex to capture the middle section on the string.
\(1\)\sLiterature\n+((.*\n){3})
This regex will capture the below string in capturing group 1.
1. a.
2. b.
3. c.
Regex Demo

You may use this regex with a capture group:
r'\(1\)\s+Literature\s+((?:.+\n)+)'
RegEx Demo
Explanation:
\(1\): Match (1) text
\s+: Match 1+ whitespaces
Literature:
\s+:
(: Start capture group #1
(?:.+\n)+: Match a line with 1+ character followed by newline. Repeat this 1 or more times to allow it to match multiple such lines
): End capture group #1

Regex for capturing the generic question with that structure:
\(\d+\)\s+(\w+)\s+((?:\d+\.\s.+\n)+)
It will capture the title "Literature", then the choices in another group (for a total of 2 groups).
It is not possible to capture repeating groups, so in order to get each of your "1. a." in a separate group you would have to match the second group from above again, with this pattern:
((\d+\.\s+.+)\n)+) then globally match to get all groups.

Related

Regex pattern matching comma delimited values with spaces allowed around comma

I am trying to write a Regex validator (Python 3.8) to accept strings like these:
foo
foo,bar
foo, bar
foo , bar
foo , bar
foo, bar,foobar
This is what I have so far (but it matches only the first two cases):
^[a-zA-Z][0-9a-zA-Z]+(,[a-zA-Z][0-9a-zA-Z]+)*$|^[a-zA-Z][0-9a-zA-Z]+
However, when I add the whitespace match \w, it stops matching altogether:
^[a-zA-Z][0-9a-zA-Z]+(\w+,\w+[a-zA-Z][0-9a-zA-Z]+)*$|^[a-zA-Z][0-9a-zA-Z]+
What is the pattern to use (with explanation as to why my second pattern above is not matching).
\w matches [0-9a-zA-Z_] and it doesn't include whitespaces.
What you need is this regex:
^[a-zA-Z][0-9a-zA-Z]*(?:\s*,\s*[a-zA-Z][0-9a-zA-Z]*)*$
RegEx Demo
RegEx Details:
^: Start
[a-zA-Z][0-9a-zA-Z]*: Match a text starting with a letter followed by 0 or more alphanumeric characters
(?:: Start non-capture group
\s*,\s*: Match a comma optionally surrounded with 0 or more whitespaces on both sides
[a-zA-Z][0-9a-zA-Z]*: Match a text starting with a letter followed by 0 or more alphanumeric characters
)*: End non-capture group. Repeat this group 0 or more times
$: End

How to capture a group only if occurs twice in a line

import re
text = """
Tumble Trouble Twwixt Two Towns!
Was the Moon soon in the Sea
Or soon in the sky?
Nobody really knows YET.
"""
How should I make the match happen only when the occurence is found twice in a line?
Regular expression that highlights two 'o's that appear beside each other only if there is another occurence of two 'o's appearing beside each other subsequently in the same line
You can match a single word char with a backreference, and group that again.
The word character will become group 2 as the groups are nested, then the outer group will be group 1.
Then you can assert group 1 using a positive lookahead again in the line.
((\w+)\2)(?=.*?\1)
The pattern matches:
( Capture group 1
(\w+)\2 Match 1+ word chars in capture group 2 followed by a backreference to group 2 to match the same again
) Close group 1
(?=.*?\1) Positive lookahead to assert the captured value of group 1 in the line
See a regex demo and a Python demo.
Example
print(re.compile(r"((\w+)\2)(?=.*?\1)").sub('{\g<1>}', text.rstrip()))
Output
Tumble Trouble Twwixt Two Towns!
Was the M{oo}n soon in the Sea
Or soon in the sky?
Nobody really knows YET.

Regex to find sentences of a minimum length

I am trying to create a regular expression that finds sentences with a minimum length.
Really my conditions are:
there must at least be 5 words in a sequence
words in sequence must be distinct
sequence must be followed by some punctuation character.
So far I have tried
^(\b\w*\b\s?){5,}\s?[.?!]$
If my sample text is:
This is a sentence I would like to parse.
This is too short.
Single word
Not not not distinct distinct words words.
Another sentence that I would be interested in.
I would like to match on strings 1 and 5.
I am using the python re library. I am using regex101 to test and it appears the regex I have above is doing quite a bit of work regards to backtracking so I imagine those knowledgable in regex may be a bit appalled (my apologies).
You can use the following regex to identify the strings that meet all three conditions:
^(?!.*\b(\w+)\b.+\b\1\b)(?:.*\b\w+\b){5}.*[.?!]\s*$
with the case-indifferent flag set.
Demo
Python's regex engine performs the following operations.
^ # match beginning of line
(?! # begin negative lookahead
.+ # match 1+ chars
\b(\w+)\b # match a word in cap grp 1
.+ # match 1+ chars
\b\1\b # match the contents of cap grp 1 with word breaks
) # end negative lookahead
(?: # begin non-cap grp
.+ # match 1+ chars
\b\w+\b # match a word
) # end non-cap grp
{5} # execute non-cap grp 5 times
.* # match 0+ chars
[.?!] # match a punctuation char
\s* # match 0+ whitespaces
$ # match end of line
Items 1. and 3. are easily done by regex, but
2. words in sequence must be distinct
I don't see how you could do it with a regex pattern. Remember that regex is a string-matching operation; it doesn't do heavy logic. This problem doesn't sound like a regex problem to me.
I recommend splitting the string in the character " " and checking word by word. Quickier, no sweat.
Edit
can be done with a lookahead as Cary said.

Capturing repeated pattern in Python

I'm trying to implement some kind of markdown like behavior for a Python log formatter.
Let's take this string as example:
**This is a warning**: Virus manager __failed__
A few regexes later the string has lost the markdown like syntax and been turned into bash code:
\033[33m\033[1mThis is a warning\033[0m: Virus manager \033[4mfailed\033[0m\033[0m
But that should be compressed to
\033[33;1mThis is a warning\033[0m: Virus manager \033[4mfailed\033[0m
I tried these, beside many other non working solutions:
(\\033\[([\d]+)m){2,} => Capture: \033[33m\033[1m with g1 '\033[1m' and g2 '1' and \033[0m\033[0mwith g1 '\033[0m' and g2 '0'
(\\033\[([\d]+)m)+ many results, not ok
(?:(\\033\[([\d]+)m)+) many results, although this is the recommended way for repeated patterns if I understood correctly, not ok
and others..
My goal is to have as results:
Input
\033[33m\033[1mThis is a warning\033[0m: Virus manager \033[4mfailed\033[0m\033[0m
Output
Match 1
033[33m\033[1m
Group1: 33
Group2: 1
Match 2
033[0m\033[0m
Group1: 0
Group2: 0
In other words, capture the ones that are "duplicated" and not the ones alone, so I can fuse them with a regex sub.
You want to match consectuively repeating \033[\d+m chunks of text and join the numbers after [ with a semi-colon.
You may use
re.sub(r'(?:\\033\[\d+m){2,}', lambda m: r'\033['+";".join(set(re.findall(r"\[(\d+)", m.group())))+'m', text)
See the Python demo online
The (?:\\033\[\d+m){2,} pattern will match two or more sequences of \033[ + one or more digits + m chunks of texts and then, the match will be passed to the lambda expression, where the output will be: 1) \033[, 2) all the numbers after [ extracted with re.findall(r"\[(\d+)", m.group()) and deduplicated with the set, and then 3) m.
The patterns in the string to be modified have not been made clear from the question. For example, is 033 fixed or might it be 025 or even 25? I've made certain assumptions in using the regex
r" ^(\\0(\d+)\[\2)[a-z]\\0\2\[(\d[a-z].+)
to obtain two capture groups that are to be combined, separated by a semi-colon. I've attempted to make clear my assumptions below, in part to help the OP modify this regex to satisfy alternative requirements.
Demo
The regex performs the following operations:
^ # match beginning of line
( # begin cap grp 1
\\0 # match '\0'
(\d+) # match 1+ digits in cap grp 2
\[ # match '['
\2 # match contents of cap grp 2
) # end cap grp 1
[a-z] # match a lc letter
\\0 # match '\0'
\2 # match contents of cap grp 2
\[ # match '['
(\d[a-z].+) # match a digit, then lc letter then 1+ chars to the
# end of the line in cap grp 3
As you see, the portion of the string captured in group 1 is
\033[33
I've assumed that the part of this string that is now 033 must be two or more digits beginning with a zero, and the second appearance of a string of digits consists of the same digits after the zero. This is done by capturing the digits following '0' (33) in capture group 2 and then using a back-reference \2.
The next part of the string is to be replaced and therefore is not captured:
m\\033[
I've assumed that m must be one lower case letter (or should it be a literal m?), the backslash and zero and required and the following digits must again match the content of capture group 2.
The remainder of the string,
1mThis is a warning\033[0m: Virus manager \033[4mfailed\033[0m\033[0m
is captured in capture group 3. Here I've assumed it begins with one digit (perhaps it should be \d+) followed by one lower case letter that needn't be the same as the lower case letter matched earlier (though that could be enforced with another capture group). At that point I match the remainder of the line with .+, having given up matching patterns in that part of the string.
One may alternatively have just two capture groups, the capture group that is now #2, becoming #1, and #2 being the part of the string that is to be replaced with a semicolon.
This is pretty straightforward for the cases you desribe here; simply write out from left to right what you want to match and capture. Repeating capturing blocks won't help you here, because only the most recently captured values would be returned as a result.
\\033\[(\d+)m\\033\[(\d+)m

Regex - How do i find this specific slice of string inside a bigger whole string

following my previous question (How do i find multiple occurences of this specific string and split them into a list?), I'm now going to ask something more since the rule has been changed.
Here's the string, and the bold words are the ones that I want to extract.
text|p1_1_1120170AS074192161A0Z20|C M E -
Rectifier|#|text|p1_2_1120170AS074192161A0Z20|Huawei|#|text|p1_3_1120170AS074192161A0Z20|Rectifier
Module 3KW|#|text|p1_4_1120170AS074192161A0Z20|Shuangdeng
6-FMX-170|#|text|p1_5_1120170AS074192161A0Z20|24021665|#|text|p1_6_1120170AS074192161A0Z20|1120170AS074192161A0Z20|#|text|p1_7_1120170AS074192161A0Z20|OK|#|text|p1_8_1120170AS074192161A0Z20||#|text|p1_9_1120170AS074192161A0Z20|ACTIVE|#|text|p1_10_1120170AS074192161A0Z20|-OK|#|text|site_id|20MJK110|#|text|barcode_flag|auto|#|text|movement_flag||#|text|unit_of_measurement||#|text|flag_waste|no|#|text|req_qty_db|2|#|text|req_qty|2
Here's my current regex:
(?<=p1\_1\_.*)[^|]+(?=\|\#\|.*|$)
After trying it out in https://regexr.com/, I found the result instead :
text|p1_1_1120170AS074192161A0Z20|C M E -
Rectifier|#|text|p1_2_1120170AS074192161A0Z20|Huawei|#|text|p1_3_1120170AS074192161A0Z20|Rectifier
Module 3KW|#|text|p1_4_1120170AS074192161A0Z20|Shuangdeng
6-FMX-170|#|text|p1_5_1120170AS074192161A0Z20|24021665|#|text|p1_6_1120170AS074192161A0Z20|1120170AS074192161A0Z20|#|text|p1_7_1120170AS074192161A0Z20|OK|#|text|p1_8_1120170AS074192161A0Z20||#|text|p1_9_1120170AS074192161A0Z20|ACTIVE|#|text|p1_10_1120170AS074192161A0Z20|-OK|#|text|site_id|20MJK110|#|text|barcode_flag|auto|#|text|movement_flag||#|text|unit_of_measurement||#|text|flag_waste|no|#|text|req_qty_db|2|#|text|req_qty|2
The question remains: "Why don't just return the first matched occurrence ?".
Let's consider that if the value between the first "bar section" is empty, then it'll return the value of the next bar section.
Example :
text|p1_1_1120170AS074192161A0Z20||#|text|p1_2_1120170AS074192161A0Z20|Huawei|#|text . . .
And I don't want that. Let it be just return nothing instead (nothing match).
What's the correct regex to acquire such a match?
Thank you :).
This data looks more structured than you are giving it credit for. A regular expression is great for e.g. extracting email addresses from unstructured text, but this data seems delimited in a straightforward manner.
If there is structure it will be simpler, faster, and more reliable to just split on | and perhaps #:
text = 'text|p1_1_1120170AS074192161A0Z20|C M E - Rectifier|#|text|p1_2_1120170AS074192161A0Z20|Huawei|#|text|p1_3_1120170AS074192161A0Z20|Rectifier Module 3KW|#|text|p1_4_11201...'
lines = text.split('|#|')
words = [line.split('|')[-1] for line in lines]
doc='text|p1_1_1120170AS074192161A0Z20|C M E - Rectifier|#|text|p1_2_1120170AS074192161A0Z20|Huawei|#|text|...'
re.findall('[^|]+(?=\|\#\|)', doc)
In the re expression:
[^|]+finds chunks of text not containing the separator
(?=...) is a "lookahead assertion" (match the text but do not include in result)
About the pattern you tried
This part of the pattern [^|]+ states to match any char other than |
Then (?=\|\#\|.*|$) asserts using a positive lookahead what is on the right is |#|.* or the end of the string.
The positive lookbehind (?<=p1\_1\_.*) asserts what is on the left is p1_1_ followed by any char except a newline using a quantifier in the lookbehind.
As the pattern is not anchored, you will get all the matches for this logic because the p1_1_ assertion is true as it precedes all the|#| parts
Note that using the quantifier in the lookbehind will require the pypi regex module.
If you want the first match using a quantifier in the positive lookbehind you could for example use an anchor in combination with a negative lookahead to not cross the |#| or match || in case it is empty:
(?<=^.*?p1_1_(?:(?!\|#\|).|\|{2})*\|)[^|]+(?=\|\#\||$)
Python demo
You could use your original pattern using re.search getting the first match.
(?<=p1_1_.*)[^|]+(?=\|\#\||$)
Note that you don't have to escape the underscore in your original pattern and you can omit .* from the positive lookahead
Python demo
But to get the first match you don't have to use a positive lookbehind. You could also use an anchor, match and capturing group.
^.*?p1_1_(?:(?!\|#\|).|\|{2})*\|([^|]+)(?:\|#\||$)
^ Start of string
.*? Match any char except a newline
p1_1_ Match literally
(?: Non capturing group
(?!\|#\|).|\|{2} If what is on the right is not |#| match any char, or match 2 times ||
)* Close non capturing group and repeat 0+ times
\| Match |
( Capture group 1 (This will contain your value
[^|]+ Match 1+ times any char except |
) Close group
(?:\|#\||$) Match either |#|
Regex demo

Categories