Scripting HelpersARCHIVE
A little piece of Roblox history. Preserved for everyone, one question at a time.Read about the archive →
All questions/Question #85065
✓ ANSWER ACCEPTED

String Patterns - Finding the Complement of a Character Sequence?

Asked 683 views1 answers

I understand that you can find the complement of a character set. For example, the pattern [^abc]+ will find all characters that are neither "a" nor "b" nor "c".

I'm wondering if or how it's possible to find the complement of a sequence, such that unlike above, I could match all characters that are not the literal sequence "abc" instead of matching everything that's not the character "a" or "b" or "c".

If this is still confusing, this is essentially what I'm trying to do and what kind of result I expect:

local text = "Hello, world! Wonderful day, isn't it?"

for a in string.gmatch(text, "Complement of 'day'") do
	print(a)
end

-- Output ->
-- Hello, world! Wonderful 
-- Isn't it?

I hope this makes my question clear. Please let me know if you need more clarification.

AnonymousEdited

4 comments

1

i think string.split may be what you want https://developer.roblox.com/en-us/api-reference/lua-docs/string/ you may be aware of it but im leaving this here just in case

Anonymous#
0

You kind of start to not make sense here: "... such that unlike above, I could match all characters that are not the literal sequence "abc" instead of matching everything that's not the character "a" or "b" or "c"." Can you clarify what you meant?.

Anonymous#
0

The closest solution for your preferred method is by using Frontier pattern matching. I will post an example after I come back home.

Anonymous#
0

@incap I'm not really sure how else to phrase it. I'm essentially just trying to match all characters that are not in a given sequence. For example, if I wanted to ignore "Wo" in "Hello, World!", the expected result should be "Hello, " \n "rld!"

Anonymous#

1 answer

✓ Accepted answer

To simply remove it from the string:

s="Hello abc, howabc are you?"
--Remove 'abc':
print(s:gsub("(.-)(abc)", "%1"))

Roblox has added string.split to Lua [thanks to GoldAngelInDisguise for mentioning this], meaning you can also do:

print(unpack(s:split("abc")))
-- or
local list = s:split("abc")
for i = 1, #list do
	print(list[i])
end

In normal Lua you can do what's shown here: https://stackoverflow.com/a/1647577/3377142 (Essentially, you can use string.find to get the index of the next instance abc (or whatever pattern you choose) and then return the strings in between that pattern.)

AnonymousEdited Permalink

6 comments

0

split seems more useful, it was more a matter of parsing a string by using a sequence of characters as a delimiter, instead of just removing the sequence. Either way, super useful answer, thank you.

Anonymous#
0

the annoying part seemed to be finding a method that included the delimiters in the parsing. I think the best route at this time is just to loop through all the characters.

Anonymous#
0

If the delimiter is a few characters (but not a pattern), you can use split and know that for every loop iteration except the last, the delimiter is present after the string given by split. If it's a pattern, you can use repeated calls to 'find' to find the next instance of the delimiter pattern and use sub to extract the characters in between (instead of looking at one char at a time)

Anonymous#
0

yeah, the only problem is my delimiter is "[...]" so what's between the two brackets is variable. Essentially, I want to be able to split "hello [world] there" into "hello", "[world]", "there"

Anonymous#
0

I developed the pattern: ".-[^%[%]]+]?" to solve exactly that, but it doesn't account for nested cases like "[[ ]]", so I'm pretty sure I have to stick with looping through the string and keeping track when the delimiter starts/ends

Anonymous#
0

Ah, I see - I believe you're right. Of interest, some implementations of regex outside of Lua can handle that with "recursion" or "balanced groups": https://stackoverflow.com/questions/546433/regular-expression-to-match-balanced-parentheses?noredirect=1&lq=1

Anonymous#

This discussion is preserved as it was. New answers, comments, and votes are closed.