How to extract key value pairs using regular expression?

Question

Let's be honest, I'm struggling with the regular expression I need to extract parts of the character sequence. The sequence contains pairs of key and value pairs enclosed with / character. So the pair could be /KEY/VALUE/, but also /KEY/VAL/UE/. The pairs sits next to each other in the sequence.

Let's look at the example sequence:

/ABCD/value1//ECFG/value2//HIJK/value3a/value3b/

What I'd like to be able to do is to get the list of the key value pairs like this:

ABCD -> value1

ECFG -> value2

HIJK -> value3a/value3b

Answer 1

This should work:

/(.+?)/(.+?)/(?=/|$)

The first paren will capture the key, the second the value. The lookahead matches either a 2nd /, indicating a new key/value pair or the string end for the last key/value pair.

Edit: Here some python code:

s = "/ABCD/value1//ECFG/value2//HIJK/value3a/value3b/"
re.findall('/(.+?)/(.+?)/(?=/|$)', s)
# [('ABCD', 'value1'), ('ECFG', 'value2'), ('HIJK', 'value3a/value3b')]

Answer 2

Try this: /(.*?)/(.*?)/

Here's how you would use it with sed:

sed -e 's,/(.*?)/(.*?)/,$1 --> $2\n,g' inputfile.txt

The key is non-greedy matches .*? (instead of greedy .*)