Copyright (c) 2021 Tony Zorman BSD3 Tony Zorman experimental non-portable Safe-Inferred Haskell2010

Description

A small wrapper around the ReadP parser combinator in base, providing a more intuitive behaviour. While it's theoretically nice that ReadP is actually commutative, this makes a lot of parsing operations rather awkward—more often than not, one only wants the argument that's parsed "first".

Due to the left-biased nature of the chosen semigroup implementation, using functions like many or optional from Control.Applicative now yields more consistent behaviour with other parser combinator libraries.

Synopsis

# Usage

NOTE: This module is mostly intended for developing of other modules. If you are a users, you probably won't find much use here—you have been warned.

The high-level API tries to stay as close to ReadP as possible. If you are familiar with that then no functions here should surprise you.

One notable usability difference when forcing left-biasedness is when one wants to disambiguate a parse. For normal ReadP usage this happens after the actual parsing stage by going through the list of successful parses. For Parser it does when constructing the relevant combinators, leading to only one successful parse. As an example, consider the ReadP-based parser

pLangle = ReadP.string "<"
pLongerSequence = ReadP.char '<' *> ReadP.string "f" <* ReadP.char '>'
pCombination = pLangle ReadP.+++ pLongerSequence

Parsing the string "f" will return

>>> ReadP.readP_to_S pCombination "<f>"
[("<","f>"),("f","")]


One would now need to, for example, filter for the second (leftover) string being empty and take the head of the resulting list (which may still have more than one element).

With Parser, the same situation would look like the following

pLangle' = string "<"
pLongerSequence' = char '<' *> string "f" <* char '>'
pCombination' = pLongerSequence' <> pLangle'

Notice how pLangle' and pLongerSequence' have traded places—since we are not forcing pLangle' to consume the entire string and (<>) is left-biased, pLongerSequence' parses a superset of pLangle'! Running runParser pCombination' now yields the expected result:

>>> runParser pCombination' "<f>"
Just "f"


One might also define pLangle' as string "<" <* eof, which would enable a definition of pCombination' = pLangle' <> pLongerSequence'.

For example uses, see XMonad.Util.EZConfig or XMonad.Prompt.OrgMode.

# Running

data Parser a Source #

#### Instances

Instances details
 Alternative Parser Source # Instance detailsDefined in XMonad.Util.Parser Methodsempty :: Parser a(<|>) :: Parser a -> Parser a -> Parser asome :: Parser a -> Parser [a]many :: Parser a -> Parser [a] Source # Instance detailsDefined in XMonad.Util.Parser Methodspure :: a -> Parser a #(<*>) :: Parser (a -> b) -> Parser a -> Parser b #liftA2 :: (a -> b -> c) -> Parser a -> Parser b -> Parser c(*>) :: Parser a -> Parser b -> Parser b #(<*) :: Parser a -> Parser b -> Parser a # Source # Instance detailsDefined in XMonad.Util.Parser Methodsfmap :: (a -> b) -> Parser a -> Parser b #(<\$) :: a -> Parser b -> Parser a # Source # Instance detailsDefined in XMonad.Util.Parser Methods(>>=) :: Parser a -> (a -> Parser b) -> Parser b #(>>) :: Parser a -> Parser b -> Parser breturn :: a -> Parser a # Monoid (Parser a) Source # Instance detailsDefined in XMonad.Util.Parser Methodsmappend :: Parser a -> Parser a -> Parser a #mconcat :: [Parser a] -> Parser a # Source # Instance detailsDefined in XMonad.Util.Parser Methods(<>) :: Parser a -> Parser a -> Parser a #sconcat :: NonEmpty (Parser a) -> Parser astimes :: Integral b => b -> Parser a -> Parser a

runParser :: Parser a -> String -> Maybe a Source #

Run a parser on a given string.

# Primitive Parsers

eof :: Parser () Source #

Succeeds if and only if we are at the end of input.

num :: (Read a, Integral a) => Parser a Source #

Parse an integral number number.

Parse and return the specified character.

Parse and return the specified string.

Skip all whitespace.

Consume and return the next character. Fails if there is no input left.

Look-ahead: return the part of the input that is left, without consuming it.

# Combining Parsers

satisfy :: (Char -> Bool) -> Parser Char Source #

Consume and return the next character if it satisfies the specified predicate.

choice :: [Parser a] -> Parser a Source #

Combine all parsers in the given list in a left-biased way.

many1 :: Parser a -> Parser [a] Source #

Parse one or more occurrences of the given parser.

sepBy :: Parser a -> Parser sep -> Parser [a] Source #

sepBy p sep parses zero or more occurrences of p, separated by sep. Returns a list of values returned by p.

sepBy1 :: Parser a -> Parser sep -> Parser [a] Source #

sepBy1 p sep parses one or more occurrences of p, separated by sep. Returns a list of values returned by p.

endBy :: Parser a -> Parser sep -> Parser [a] Source #

endBy p sep parses zero or more occurrences of p, separated and ended by sep.

endBy1 :: Parser a -> Parser sep -> Parser [a] Source #

endBy p sep parses one or more occurrences of p, separated and ended by sep.

munch :: (Char -> Bool) -> Parser String Source #

Parse the first zero or more characters satisfying the predicate. Always succeeds; returns an empty string if the predicate returns False on the first character of input.

munch1 :: (Char -> Bool) -> Parser String Source #

Parse the first one or more characters satisfying the predicate. Fails if none, else succeeds exactly once having consumed all the characters.

Always fails