@lang:str.split

Split a string by a delimiter into an array of strings, or select one segment when an index is given.

On-chain (@lang:str.split!): The segment index is required; the delimiter may be a live call, which costs three reads of it per segment.

Returns: array | string

Syntax

@lang:str.split(s delim index?)

Arguments

Name	Type	Description
s	string	Source string
delim	string	Exact, non-empty delimiter byte sequence
[index]	number	Segment to select instead of the whole array: zero-based from the start, or negative from the end (-1 = last, -2 = second-last, and so on)

See Also

• @str.join — join array into a string

On-chain face (@str.split!)

Split the string return of a call on a delimiter and select one segment, on-chain. Segment indexes are 0, 1, 2, … from the start, or -1, -2, … from the end (-1 is the last segment).

Examples

load lang

set $pool 0x44fA8E6f47987339850636F88629646662444217

# "Uniswap LP Token" -> segment 1 is "LP"
assert @str.split!($pool::{name()(string)} " " 1) == "LP"

# The name ends with "Token": negative index counts from the end, on-chain
assert @str.split!($pool::{name()(string)} " " -1) == "Token"

Notes

• A live delimiter costs three reads of it per segment: the two indexOf bounds and its own length. Index 0 is the exception, needing neither the length nor a second indexOf.
• Splits on the exact byte sequence; adjacent delimiters produce empty segments and an out-of-range index (in either direction) reverts with SegmentIndexOutOfBounds.
• Negative indices resolve against the segment count at assertion time, so -1 is the last segment however many the live value has.
• String-valued: compare with ==/!=. At the top level the core judges the string directly; nested inside @bool! the comparison compiles to an on-chain keccak comparison of the two sides.

See Also

• assert, @hash!