Python regular expression search doesn't works on a string contains colon(:)

Question

I have a string which can be like "some string {{key:value}} again some string {{key:value}}{{key:value}} some string" I want to find all the {{key:value}} matching sub strings from the given string. for this what pattern I am trying is like :

string ="fdfd{{dsfdss:dssssasa}} fdsfdsfds"
pattern = re.compile("\\\\[a-z]\w+\:\[a-z]\w+\}}")
abc = re.search( pattern, string)

I tried it using this also

pattern = re.compile("\\\\[a-z]\w+:[a-zA-Z]\w+\}}")
abc = re.search( pattern, string)

But each time it is returning None

Suggested me the best way to accomplish it.

You are trying to match a ``\`` with your regex but it is missing in the input. — Wiktor Stribiżew, Nov 23 '15 at 12:37
What is with the leading \\\\? And you never grab the open curly brace — OneCricketeer, Nov 23 '15 at 12:38
Read https://docs.python.org/3/library/re.html about which escapes you need: You need to escape `{` and `}`, but not `[`. `\w` matches too much (also `:`). — jofel, Nov 23 '15 at 12:42
Yes @Mayur Kostri, i want to grab the only key value pair from the string. it will be like "{{key:value}}" or [(key, value), (key, value)]. — Sajid Ahmad, Nov 23 '15 at 12:51
@SajidAhmad: Cyrbil's answer is correct and yields what you requested above (`[(key, value), ...]`). See [this demo](https://ideone.com/aXIT5T). Just remove the outer parentheses. — Wiktor Stribiżew, Nov 23 '15 at 13:01
@stribizhev , the link you given works but fails when string is changed to string = u'fdfd{{dsfdss:dssssasa}} fdsfdsfds {{aaaa:bbbb}}' in this case it returns an empty strng — Sajid Ahmad, Nov 23 '15 at 13:16
@SajidAhmad: Are you sure? [I have not noticed any bug like that.](https://ideone.com/Lf3cym) — Wiktor Stribiżew, Nov 23 '15 at 13:19
yes, it works this time, don't know why i got the null value at that time. Thanks — Sajid Ahmad, Nov 23 '15 at 13:46

Cyrbil · Accepted Answer · 2015-11-23T12:57:47.000

6

Try this simple regex:

({{(\w+):(\w+)}})

It returns the full match and elements:

MATCH 1
1. [4-23] {{dsfdss:dssssasa}}
2. [6-12] dsfdss
3. [13-21] dssssasa

Try it live here

Python code:

import re
string = u'fdfd{{dsfdss:dssssasa}} fdsfdsfds'
pattern = re.compile(ur'({{(\w+):(\w+)}})')
print(re.findall(pattern, string))

Edit: For only the inner part, the regex is {{(\w+:\w+)}}

edited Nov 23 '15 at 12:57

answered Nov 23 '15 at 12:38

Cyrbil

6,341
1
24
40

1

Fun fact: you can remove extra backslashes and use `({{(\w+):(\w+)}})`. Python re module is still a bit smart and unlike Java recognizes the `{` not followed by a digit and closing brace is a literal `{`. And `}` do not have to be escaped. – Wiktor Stribiżew Nov 23 '15 at 12:49
Thank you, but why is that possible ? Is it because the regex engine understand that it is not a quantifier ? How ? – Cyrbil Nov 23 '15 at 12:52
1

Ok, you edit your answer and answered my interrogation :) – Cyrbil Nov 23 '15 at 12:52
Hi @Cybril, thanks for your answer. It works also but what i need is a bit different. I need only the text enclosed in {{}}. – Sajid Ahmad Nov 23 '15 at 12:53
Not sure what you mean @SajidAhmad ? Just `dsfdss:dssssasa` ? – Cyrbil Nov 23 '15 at 12:54
@SajidAhmad: It is even easier then, and your question is a practically a dupe of [*regex in python dealing with curly braces*](http://stackoverflow.com/questions/10643553/regex-in-python-dealing-with-curly-braces): use `{{(.*?)}}` with `re.findall`. – Wiktor Stribiżew Nov 23 '15 at 12:56
yes, I need only this key value pair, not other strings. – Sajid Ahmad Nov 23 '15 at 12:56
Move the parenthesis a bit ... `{{(\w+:\w+)}}` you could have figure that by yourself... – Cyrbil Nov 23 '15 at 12:58

Mayur Koshti · Answer 2 · 2015-11-23T13:03:24.350

0

You can use the following regex:

(\{\{([^\:]+)\:([^\}]+)\}\}

edited Nov 23 '15 at 13:03

answered Nov 23 '15 at 12:39

Mayur Koshti

1,794
15
20

Python regular expression search doesn't works on a string contains colon(:)

2 Answers2