Swarm.Language.Parser.Lex

Token lexing and comment preservation for the Swarm language.

Parsing with source locations

parseLoc :: Parser Term -> Parser Syntax

Add SrcLoc to a Term parser

parseLocG :: Parser a -> Parser (SrcLoc, a)

Add SrcLoc to a parser

Whitespace + comments

getCommentSituation :: Parser CommentSituation

If we see a comment starting now, is it the first non-whitespace thing on the current line so far, or were there other non-whitespace tokens previously?

lineComment :: Text -> Parser ()

Parse a line comment, while appending it out-of-band to the list of comments saved in the custom state.

blockComment :: Text -> Text -> Parser ()

Parse a block comment, while appending it out-of-band to the list of comments saved in the custom state.

sc :: Parser ()

Skip spaces and comments.

Tokens

Lexemes

lexeme :: Parser a -> Parser a

In general, we follow the convention that every token parser assumes no leading whitespace and consumes all trailing whitespace. Concretely, we achieve this by wrapping every token parser using lexeme.

Also sets freshLine to False every time we see a non-whitespace token.

Specific token types

symbol :: Text -> Parser Text

A lexeme consisting of a literal string.

operator :: Text -> Parser Text

A lexeme consisting of a specific string, not followed by any other operator character.

reservedWords :: Set Text

List of reserved words that cannot be used as variable names.

reservedCS :: Text -> Parser ()

Parse a case-sensitive reserved word.

reserved :: Text -> Parser ()

Parse a case-insensitive reserved word.

data IdentifierType

What kind of identifier are we parsing?

Constructors

IDTyVar

IDTyName

IDTmVar

Instances

Show IdentifierType

Defined in Swarm.Language.Parser.Lex

Methods

showsPrec :: Int -> IdentifierType -> ShowS

show :: IdentifierType -> String

showList :: [IdentifierType] -> ShowS

Eq IdentifierType

Defined in Swarm.Language.Parser.Lex

Methods

(==) :: IdentifierType -> IdentifierType -> Bool

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

Ord IdentifierType

Defined in Swarm.Language.Parser.Lex

Methods

compare :: IdentifierType -> IdentifierType -> Ordering

(<) :: IdentifierType -> IdentifierType -> Bool

(<=) :: IdentifierType -> IdentifierType -> Bool

(>) :: IdentifierType -> IdentifierType -> Bool

(>=) :: IdentifierType -> IdentifierType -> Bool

max :: IdentifierType -> IdentifierType -> IdentifierType

min :: IdentifierType -> IdentifierType -> IdentifierType

locIdentifier :: IdentifierType -> Parser LocVar

Parse an identifier together with its source location info.

locTmVar :: Parser LocVar

Parse a term variable together with its source location info.

locTyName :: Parser LocVar

Parse a user-defined type name together with its source location info.

identifier :: IdentifierType -> Parser Var

Parse an identifier, i.e. any non-reserved string containing alphanumeric characters and underscores, not starting with a digit. The Bool indicates whether we are parsing a type variable.

tyVar :: Parser Var

Parse a type variable, which must start with an underscore or lowercase letter and cannot be the lowercase version of a type name.

tyName :: Parser Var

Parse a (user-defined) type constructor name, which must start with an uppercase letter.

tmVar :: Parser Var

Parse a term variable, which can start in any case and just cannot be the same (case-insensitively) as a lowercase reserved word.

textLiteral :: Parser Text

Parse a text literal (including escape sequences) in double quotes.

integer :: Parser Integer

Parse a positive integer literal token, in decimal, binary, octal, or hexadecimal notation. Note that negation is handled as a separate operator.

Combinators

braces :: Parser a -> Parser a

parens :: Parser a -> Parser a

brackets :: Parser a -> Parser a