Extract table key-values from LUA code

Question

I have multiple strings from LUA code, each one with a LUA table item, something like:

atable['akeyofthetable'] = { 'name' = 'a name', 'thevalue' = 34, 'anotherkey' = 'something' } 

The string might be spanned in multiple lines, meaning it might be:

atable['akeyofthetable'] = { 'name' = 'a name', 
      'thevalue' = 34, 
      "anotherkey" = 'something' }

How to get some (ex: only name and anotherkey in the above example) of the keys with their values as "re.match" objects in python3 from that string? Because this is taken from code, the existence of keys is not guarantied, the "quoting" of keys and values (double or single quotes) may vary, even from key to key, and there may be empty values ('name' = '') or non quoted strings as values ('thevalue' = anonquotedstringasvalue). Even the order of the keys is not guarantied. Split using commas (,) is not working because some string values have commas (ex: 'anotherkey' = 'my beloved, strange, value' or even 'anotherkey' = "my beloved, 'strange' = 34, value"). Also keys may or may not be quoted (depends, if names are in ASCII probably will not be quoted).

Is it possible to do this using one regex or I must do multiple searches for every key needed?

Answer 1

Code

If there is a possibility of escaped quotes \' or \" within the string, you can substitute the respective capture groups for '((?:[^'\\]|\\.)*)' as seen here.

See regex in use here

['\"](?:name|anotherkey)['\"]\s*=\s*(?:'([^']*)'|\"([^\"]*)\")

Usage

See code in use here

import re

keys = [
    "name",
    "anotherkey"
]
r = r"['\"](" + "|".join([re.escape(key) for key in keys]) + r")['\"]\s*=\s*(?:'([^']*)'|\"([^\"]*)\")"
s = "atable['akeyofthetable'] = { 'name' = 'a name',\n\t  'thevalue' = 34, \n\t  \"anotherkey\" = 'something' }"

print(re.findall(r, s))

---

Explanation

The second point below is replaced by a join of the keys array.

• ['\"] Match any character in the set '"
• (name|anotherkey) Capture the key into capture group 1
• ['\"] Match any character in the set '"
• \s* Match any number of whitespace characters
• = Match this literally
• \s* Match any number of whitespace characters
• (?:'([^']*)'|\"([^\"]*)\") Match either of the following
  • '([^']*)' Match ', followed by any character except ' any number of times, followed by '
  • \"([^\"]*)\" Match ", followed by any character except " any number of times, followed by "