megaparsec/tests/Test/Hspec/Megaparsec/AdHoc.hs
2016-09-25 18:42:21 +03:00

186 lines
5.9 KiB
Haskell
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

--
-- Tests for Megaparsec's expression parsers.
--
-- Copyright © 20152016 Megaparsec contributors
--
-- Redistribution and use in source and binary forms, with or without
-- modification, are permitted provided that the following conditions are
-- met:
--
-- * Redistributions of source code must retain the above copyright notice,
-- this list of conditions and the following disclaimer.
--
-- * Redistributions in binary form must reproduce the above copyright
-- notice, this list of conditions and the following disclaimer in the
-- documentation and/or other materials provided with the distribution.
--
-- THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS “AS IS” AND ANY
-- EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
-- WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
-- DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDERS BE LIABLE FOR ANY
-- DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
-- DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
-- OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
-- HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
-- STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
-- ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
-- POSSIBILITY OF SUCH DAMAGE.
{-# LANGUAGE CPP #-}
{-# LANGUAGE FlexibleContexts #-}
{-# LANGUAGE RankNTypes #-}
module Test.Hspec.Megaparsec.AdHoc
( -- * Helpers to run parsers
prs
, prs'
, prs_
, grs
, grs'
-- * Working with source position
, updatePosString
, pos1
, nes
-- * Other
, abcRow
, toFirstMismatch )
where
import Control.Monad
import Control.Monad.Reader
import Control.Monad.Trans.Identity
import Data.Foldable (foldl')
import Data.List.NonEmpty (NonEmpty (..))
import Test.Hspec
import Test.Hspec.Megaparsec
import Text.Megaparsec.Error
import Text.Megaparsec.Pos
import Text.Megaparsec.Prim
import qualified Control.Monad.State.Lazy as L
import qualified Control.Monad.State.Strict as S
import qualified Control.Monad.Writer.Lazy as L
import qualified Control.Monad.Writer.Strict as S
#if !MIN_VERSION_base(4,8,0)
import Control.Applicative
#endif
----------------------------------------------------------------------------
-- Helpers to run parsers
-- | Apply parser to given input. This is a specialized version of 'parse'
-- that assumes empty file name.
prs
:: Parsec Dec String a -- ^ Parser to run
-> String -- ^ Input for the parser
-> Either (ParseError Char Dec) a -- ^ Result of parsing
prs p = parse p ""
{-# INLINE prs #-}
-- | Just like 'prs', but allows to inspect final state of the parser.
prs'
:: Parsec Dec String a -- ^ Parser to run
-> String -- ^ Input for the parser
-> (State String, Either (ParseError Char Dec) a) -- ^ Result of parsing
prs' p s = runParser' p (initialState s)
{-# INLINE prs' #-}
-- | Just like 'prs', but forces the parser to consume all input by adding
-- 'eof':
--
-- > prs_ p = parse (p <* eof) ""
prs_
:: Parsec Dec String a -- ^ Parser to run
-> String -- ^ Input for the parser
-> Either (ParseError Char Dec) a -- ^ Result of parsing
prs_ p = parse (p <* eof) ""
{-# INLINE prs_ #-}
-- | Just like 'prs', but interprets given parser as various monads (tries
-- all supported monads transformers in turn).
grs
:: (forall m. MonadParsec Dec String m => m a) -- ^ Parser to run
-> String -- ^ Input for the parser
-> (Either (ParseError Char Dec) a -> Expectation)
-- ^ How to check result of parsing
-> Expectation
grs p s r = do
r (prs p s)
r (prs (runIdentityT p) s)
r (prs (runReaderT p ()) s)
r (prs (L.evalStateT p ()) s)
r (prs (S.evalStateT p ()) s)
r (prs (evalWriterTL p) s)
r (prs (evalWriterTS p) s)
-- | 'grs'' to 'grs' as 'prs'' to 'prs'.
grs'
:: (forall m. MonadParsec Dec String m => m a) -- ^ Parser to run
-> String -- ^ Input for the parser
-> ((State String, Either (ParseError Char Dec) a) -> Expectation)
-- ^ How to check result of parsing
-> Expectation
grs' p s r = do
r (prs' p s)
r (prs' (runIdentityT p) s)
r (prs' (runReaderT p ()) s)
r (prs' (L.evalStateT p ()) s)
r (prs' (S.evalStateT p ()) s)
r (prs' (evalWriterTL p) s)
r (prs' (evalWriterTS p) s)
evalWriterTL :: Monad m => L.WriterT [Int] m a -> m a
evalWriterTL = liftM fst . L.runWriterT
evalWriterTS :: Monad m => S.WriterT [Int] m a -> m a
evalWriterTS = liftM fst . S.runWriterT
----------------------------------------------------------------------------
-- Working with source position
-- | A helper function that is used to advance 'SourcePos' given a 'String'.
updatePosString
:: Pos -- ^ Tab width
-> SourcePos -- ^ Initial position
-> String -- ^ 'String' — collection of tokens to process
-> SourcePos -- ^ Final position
updatePosString w = foldl' f
where f p t = snd (defaultUpdatePos w p t)
-- | Position with minimal value.
pos1 :: Pos
pos1 = unsafePos 1
-- | Make a singleton non-empty list from a value.
nes :: a -> NonEmpty a
nes x = x :| []
{-# INLINE nes #-}
----------------------------------------------------------------------------
-- Other
-- | @abcRow a b c@ generates string consisting of character “a” repeated
-- @a@ times, character “b” repeated @b@ times, and character “c” repeated
-- @c@ times.
abcRow :: Int -> Int -> Int -> String
abcRow a b c = replicate a 'a' ++ replicate b 'b' ++ replicate c 'c'
-- | Given a comparing function, get prefix of one string till first
-- mismatch with another string (including first mismatching character).
toFirstMismatch
:: (Char -> Char -> Bool) -- ^ Comparing function
-> String -- ^ First string
-> String -- ^ Second string
-> String -- ^ Resulting prefix
toFirstMismatch f str s = take (n + 1) s
where n = length (takeWhile (uncurry f) (zip str s))