Compact, sorted-range representation of character sets enabling fast membership checks, set algebra (union/intersect/difference/invert), and deterministic partitioning via topology-based refinement for parsers, lexers, automata.
kotlin-parsing-charset is a Kotlin Multiplatform library for representing sets of Char values as
compact sorted ranges.
It is intended for parser, lexer, tokenizer, and automaton implementations that need fast character
membership checks, set operations, and deterministic partitioning of the Char space.
repositories {
mavenCentral()
}
dependencies {
implementation("one.wabbit:kotlin-parsing-charset:1.3.0")
}
import one.wabbit.parsing.charset.CharSet
val identifierStart = CharSet.letter union CharSet.one('_')
val identifierPart = identifierStart union CharSet.digit
check('a' in identifierStart)
check('_' in identifierStart)
check('9' !in identifierStart)
check('9' in identifierPart)
CharSet stores sorted, non-overlapping ranges, so adjacent characters are compacted into a single
range.
import one.wabbit.parsing.charset.CharSet
val lowercase = CharSet.range('a', 'z')
val vowels = CharSet.of("aeiou")
val consonants = lowercase difference vowels
check('b' in consonants)
check('a' !in consonants)
check((vowels union consonants) == lowercase)
The main operations are union, intersect, difference, invert, subset/superset predicates,
and overlap classification.
CharSetTop represents a partition of the whole Char.MIN_VALUE..Char.MAX_VALUE space. It is useful
when several character sets must be refined into disjoint basis ranges for parser or automaton
construction.
import one.wabbit.parsing.charset.CharSet
import one.wabbit.parsing.charset.CharSetTop
val top = CharSetTop.trivial
.refine(CharSet.digit)
.refine(CharSet.letter)
check(top.size >= 3)
The library is small and stable in scope. It focuses on character-set algebra and partitioning; it does not parse regular expressions or build automata by itself.
Generated API docs can be built locally with Dokka. See API reference notes for the command.
This project is licensed under the GNU Affero General Public License v3.0 (AGPL-3.0) for open source use.
For commercial use, contact Wabbit Consulting Corporation at wabbit@wabbit.one.
Before contributions can be merged, contributors need to agree to the repository CLA.
Surfaced from shared tags and platforms — no rankings paid for.