Copyright	(c) Nadia Yvette Chambers 2026
License	Apache-2.0
Safe Haskell	None
Language	GHC2021

Smirk.Core.Bytecode

Description

Defines the low-level virtual machine bytecode instruction set (Opcode) used by the Pike-style non-backtracking / thread-stepped VM in VM. Includes relative jumping, nondeterministic thread branching (OpSplit), capture slot registers (OpSaveStart, OpSaveEnd), and lookaround subprograms.

Synopsis

data Opcode
- = OpChar !Char
- | OpClass !CharClass
- | OpAnchor !Anchor
- | OpSplit !Int !Int
- | OpJump !Int
- | OpSaveStart !Int
- | OpSaveEnd !Int
- | OpAssertLookahead !(Vector Opcode) !Bool
- | OpAssertLookbehind !(Vector Opcode) !Bool
- | OpMatchBackRef !Int
- | OpMatch
type Program = Vector Opcode
data Regex = Regex {
- originalPattern :: !Text
- program :: !Program
- groupNames :: !(Map Text Int)
- maxGroupIndex :: !Int
}

Documentation

data Opcode Source #

Low-level bytecode instruction for the Smirk regex virtual machine.

Constructors

OpChar !Char	Match a single literal character
OpClass !CharClass	Match a character class or set
OpAnchor !Anchor	Assert a zero-width anchor condition
OpSplit !Int !Int	Fork execution: try relative offset 1, alternative at offset 2
OpJump !Int	Unconditional relative jump by specified offset
OpSaveStart !Int	Save current input position as start of capture group $n$
OpSaveEnd !Int	Save current input position as end of capture group $n$
OpAssertLookahead !(Vector Opcode) !Bool	Zero-width lookahead assertion: subprogram and polarity
OpAssertLookbehind !(Vector Opcode) !Bool	Zero-width lookbehind assertion: subprogram and polarity
OpMatchBackRef !Int	Match dynamic slice previously captured by group $n$
OpMatch	Successful match reached; terminate thread

Instances

Instances details

Eq Opcode Source #

Instance details

Defined in Smirk.Core.Bytecode

Methods

(==) :: Opcode -> Opcode -> Bool

(/=) :: Opcode -> Opcode -> Bool

Generic Opcode Source #

Instance details

Defined in Smirk.Core.Bytecode

Associated Types

type Rep Opcode

Instance details

Defined in Smirk.Core.Bytecode

type Rep Opcode = D1 ('MetaData "Opcode" "Smirk.Core.Bytecode" "smirk-0.1.0.0-inplace" 'False) (((C1 ('MetaCons "OpChar" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Char)) :+: C1 ('MetaCons "OpClass" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 CharClass))) :+: (C1 ('MetaCons "OpAnchor" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Anchor)) :+: (C1 ('MetaCons "OpSplit" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpJump" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int))))) :+: ((C1 ('MetaCons "OpSaveStart" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: (C1 ('MetaCons "OpSaveEnd" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpAssertLookahead" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Vector Opcode)) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Bool)))) :+: (C1 ('MetaCons "OpAssertLookbehind" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Vector Opcode)) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Bool)) :+: (C1 ('MetaCons "OpMatchBackRef" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpMatch" 'PrefixI 'False) (U1 :: Type -> Type)))))

Methods

from :: Opcode -> Rep Opcode x

to :: Rep Opcode x -> Opcode

Show Opcode Source #

Instance details

Defined in Smirk.Core.Bytecode

Methods

showsPrec :: Int -> Opcode -> ShowS

show :: Opcode -> String

showList :: [Opcode] -> ShowS

type Rep Opcode Source #

Instance details

Defined in Smirk.Core.Bytecode

type Rep Opcode = D1 ('MetaData "Opcode" "Smirk.Core.Bytecode" "smirk-0.1.0.0-inplace" 'False) (((C1 ('MetaCons "OpChar" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Char)) :+: C1 ('MetaCons "OpClass" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 CharClass))) :+: (C1 ('MetaCons "OpAnchor" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Anchor)) :+: (C1 ('MetaCons "OpSplit" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpJump" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int))))) :+: ((C1 ('MetaCons "OpSaveStart" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: (C1 ('MetaCons "OpSaveEnd" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpAssertLookahead" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Vector Opcode)) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Bool)))) :+: (C1 ('MetaCons "OpAssertLookbehind" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Vector Opcode)) :*: S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Bool)) :+: (C1 ('MetaCons "OpMatchBackRef" 'PrefixI 'False) (S1 ('MetaSel ('Nothing :: Maybe Symbol) 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int)) :+: C1 ('MetaCons "OpMatch" 'PrefixI 'False) (U1 :: Type -> Type)))))

type Program = Vector Opcode Source #

Compiled bytecode sequence indexed as a contiguous vector.

data Regex Source #

Complete compiled regular expression ready for VM execution.

Constructors

Regex
Fields originalPattern :: !Text Original PCRE source pattern string program :: !Program Compiled bytecode instruction vector groupNames :: !(Map Text Int) Mapping from named capture groups to 1-based indices maxGroupIndex :: !Int Highest capture group index allocated in the pattern

Instances

Instances details

Eq Regex Source #

Instance details

Defined in Smirk.Core.Bytecode

Methods

(==) :: Regex -> Regex -> Bool

(/=) :: Regex -> Regex -> Bool

Generic Regex Source #

Instance details

Defined in Smirk.Core.Bytecode

Associated Types

type Rep Regex

Instance details

Defined in Smirk.Core.Bytecode

type Rep Regex = D1 ('MetaData "Regex" "Smirk.Core.Bytecode" "smirk-0.1.0.0-inplace" 'False) (C1 ('MetaCons "Regex" 'PrefixI 'True) ((S1 ('MetaSel ('Just "originalPattern") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Text) :*: S1 ('MetaSel ('Just "program") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Program)) :*: (S1 ('MetaSel ('Just "groupNames") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Map Text Int)) :*: S1 ('MetaSel ('Just "maxGroupIndex") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int))))

Methods

from :: Regex -> Rep Regex x

to :: Rep Regex x -> Regex

Show Regex Source #

Instance details

Defined in Smirk.Core.Bytecode

Methods

showsPrec :: Int -> Regex -> ShowS

show :: Regex -> String

showList :: [Regex] -> ShowS

type Rep Regex Source #

Instance details

Defined in Smirk.Core.Bytecode

type Rep Regex = D1 ('MetaData "Regex" "Smirk.Core.Bytecode" "smirk-0.1.0.0-inplace" 'False) (C1 ('MetaCons "Regex" 'PrefixI 'True) ((S1 ('MetaSel ('Just "originalPattern") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Text) :*: S1 ('MetaSel ('Just "program") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Program)) :*: (S1 ('MetaSel ('Just "groupNames") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 (Map Text Int)) :*: S1 ('MetaSel ('Just "maxGroupIndex") 'NoSourceUnpackedness 'SourceStrict 'DecidedStrict) (Rec0 Int))))