diff --git a/CHANGELOG.md b/CHANGELOG.md new file mode 100644 --- /dev/null +++ b/CHANGELOG.md @@ -0,0 +1,6 @@ +# Revision history for phladiprelio-general-shared + +## 0.1.0.0 -- 2023-03-10 + +* First version. Released on an unsuspecting world. + diff --git a/Phladiprelio/General/Parsing.hs b/Phladiprelio/General/Parsing.hs new file mode 100644 --- /dev/null +++ b/Phladiprelio/General/Parsing.hs @@ -0,0 +1,138 @@ +{-# OPTIONS_GHC -threaded -rtsopts #-} +{-# OPTIONS_HADDOCK show-extensions #-} +{-# LANGUAGE BangPatterns, NoImplicitPrelude #-} + +-- | +-- Module : Phladiprelio.General.Parsing +-- Copyright : (c) Oleksandr Zhabenko 2021-2023 +-- License : MIT +-- Stability : Experimental +-- Maintainer : oleksandr.zhabenko@yahoo.com +-- +-- The additional parsing library functions for the lineVariantsG3 executable. +-- Is taken from the Phonetic.Languages.Parsing module from the +-- @phonetic-languages-simplified-examples-array@ package to reduce dependencies in general case. +-- + +module Phladiprelio.General.Parsing ( + -- * Predicates + isClosingCurlyBracket + , isSlash + , isOpeningCurlyBracket + , variations + -- * Transformations + , breakGroupOfStrings + , breakInSlashes + , combineVariants + , combineHeadsWithNexts + , transformToVariations + -- * Files processment for specifications + , readLangSpecs + , innerProcessmentSimple + , argsProcessment +) where + +import GHC.Base +import GHC.List +import Phladiprelio.General.PrepareText +import System.Environment (getArgs) +import System.IO (FilePath, readFile) +import Data.List (sort,lines,unwords) +import GHC.Arr +import Phladiprelio.General.Base +import Phladiprelio.General.Syllables +import Text.Read (readMaybe,read) +import Data.Maybe (fromMaybe) +import Phladiprelio.General.SpecificationsRead + +isClosingCurlyBracket :: String -> Bool +isClosingCurlyBracket = (== "}") + +isSlash :: String -> Bool +isSlash (x:xs) + | x /= '/' = False + | null xs = True + | otherwise = False +isSlash _ = False + +isOpeningCurlyBracket :: String -> Bool +isOpeningCurlyBracket = (== "{") + +breakGroupOfStrings :: [String] -> (([String],[[String]]),[String]) +breakGroupOfStrings !xss = ((tss,breakInSlashes uss []), drop 1 zss) + where (!yss,!zss) = break isClosingCurlyBracket xss + (!tss,!uss) = (\(t1,t2) -> (t1,drop 1 t2)) . break isOpeningCurlyBracket $ yss + +breakInSlashes :: [String] -> [[String]] -> [[String]] +breakInSlashes !wss !usss + | null lss = kss : usss + | otherwise = breakInSlashes (drop 1 lss) (kss : usss) + where (!kss,!lss) = break isSlash wss + +combineVariants :: ([String],[[String]]) -> [[String]] +combineVariants (!xss, (!yss:ysss)) = (xss `mappend` yss) : combineVariants (xss, ysss) +combineVariants _ = [] + +combineHeadsWithNexts :: [[String]] -> [String] -> [[String]] +combineHeadsWithNexts !xsss !yss + | null yss = xsss + | otherwise = combineHeadsWithNexts [xss `mappend` zss | xss <- xsss, zss <- zsss] uss + where (!t,!uss) = breakGroupOfStrings yss + !zsss = combineVariants t + +transformToVariations :: [String] -> [[String]] +transformToVariations !yss + | null yss = [] + | otherwise = combineHeadsWithNexts xsss tss + where (!y,!tss) = breakGroupOfStrings yss + !xsss = combineVariants y + +variations :: [String] -> Bool +variations xss + | any isSlash xss = if any isOpeningCurlyBracket xss && any isClosingCurlyBracket xss then True else False + | otherwise = False + + +innerProcessmentSimple + :: String -- ^ Must be a valid 'GWritingSystemPRPLX' specifications 'String' representation only (see the gwrsysExample.txt file in the @phonetic-languages-phonetics-basics@ package as a schema); + -> String -- ^ Must be a 'String' with the 5 meaningful lines that are delimited with the \'~\' line one from another with the specifications for the possible allophones (if any), 'CharPhoneticClassification', white spaces information (two 'String's) and the 'String' of all the possible 'PLL' 'Char's; + -> String -- ^ Must be a 'String' with the 'SegmentRulesG' specifications only; + -> String -- ^ Must be a 'String' with the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be prepended to the next word. + -> String -- ^ Must be a 'String' with the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be appended to the previous word. + -> (GWritingSystemPRPLX, [(Char, Char)], CharPhoneticClassification, SegmentRulesG, String, String, Concatenations, Concatenations, String) +innerProcessmentSimple gwrsCnts controlConts segmentData concatenationsFileP concatenationsFileA = + let [allophonesGs, charClfs, jss, vss, wss] = groupBetweenChars '~' . lines $ controlConts + wrs = getGWritingSystem '~' gwrsCnts + ks = sort . fromMaybe [] $ (readMaybe (unwords allophonesGs)::Maybe [(Char, Char)]) + arr = read (unwords charClfs)::Array Int PRS -- The 'Array' must be previously sorted in the ascending order. + gs = read segmentData::SegmentRulesG + ysss = sort2Concat . fromMaybe [] $ (readMaybe concatenationsFileP::Maybe [[String]]) + zzzsss = sort2Concat . fromMaybe [] $ (readMaybe concatenationsFileA::Maybe [[String]]) + js = concat jss + vs = concat vss + ws = sort . concat $ wss + in (wrs, ks, arr, gs, js, vs, ysss, zzzsss, ws) + +{-| -} +argsProcessment + :: FilePath -- ^ With the 'GWritingSystemPRPLX' specifications only (see the gwrsysExample.txt file in the @phonetic-languages-phonetics-basics@ package as a schema); + -> FilePath -- ^ With the 5 meaningful lines that are delimited with the \'~\' line one from another with the specifications for the possible allophones (if any), 'CharPhoneticClassification', white spaces information (two 'String's) and the 'String' of all the possible 'PLL' 'Char's; + -> FilePath -- ^ With the 'SegmentRulesG' specifications only; + -> FilePath -- ^ With the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be prepended to the next word. + -> FilePath -- ^ With the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be appended to the previous word. + -> IO [String] +argsProcessment fileGWrSys controlFile segmentRulesFile concatenationsFileP concatenationsFileA = mapM readFile [controlFile, fileGWrSys, segmentRulesFile, concatenationsFileP, concatenationsFileA] + +-- | The function that is mostly intended to be used by the end user. Reads the specifications from +-- the5 given files and returns the data that can be used further for generalized PhLADiPreLiO. +readLangSpecs + :: FilePath -- ^ With the 'GWritingSystemPRPLX' specifications only (see the gwrsysExample.txt file in the @phonetic-languages-phonetics-basics@ package as a schema); + -> FilePath -- ^ With the 5 meaningful lines that are delimited with the \'~\' line one from another with the specifications for the possible allophones (if any), 'CharPhoneticClassification', white spaces information (two 'String's) and the 'String' of all the possible 'PLL' 'Char's; + -> FilePath -- ^ With the 'SegmentRulesG' specifications only; + -> FilePath -- ^ With the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be prepended to the next word. + -> FilePath -- ^ With the 'Concatenations' specifications only (see the data in the EnglishConcatenated.txt file in the @phonetic-languages-phonetics-basics@ package as a list of English equivalents of the needed 'String's). These are to be appended to the previous word. + -> IO (GWritingSystemPRPLX, [(Char, Char)], CharPhoneticClassification, SegmentRulesG, String, String, Concatenations, Concatenations, String) +readLangSpecs fileGWrSys controlFile segmentRulesFile concatenationsFileP concatenationsFileA = + argsProcessment fileGWrSys controlFile segmentRulesFile concatenationsFileP concatenationsFileA >>= \xss -> let [controlConts, gwrsCnts, segmentData, concatenationsFileP1, concatenationsFileA1] = xss in return $ innerProcessmentSimple gwrsCnts controlConts segmentData concatenationsFileP1 concatenationsFileA1 + + diff --git a/README.md b/README.md new file mode 100644 --- /dev/null +++ b/README.md @@ -0,0 +1,8 @@ +The module is intended to be used by both old and new projects related to PhLADiPreLiO. + +P.S.: the author would like to devote this project to support the Fund GASTROSTARS. +If you would like to share some financial support, please, contact the fund +using the URL: + +https://gastrostars.nl/hou-mij-op-de-hoogte + diff --git a/phladiprelio-general-shared.cabal b/phladiprelio-general-shared.cabal new file mode 100644 --- /dev/null +++ b/phladiprelio-general-shared.cabal @@ -0,0 +1,32 @@ +cabal-version: 2.4 +name: phladiprelio-general-shared +version: 0.1.0.0 + +-- A short (one-line) description of the package. +synopsis: A shared by different general implementations of the PhLADiPreLiO functionality. + +-- A longer description of the package. +description: Is intended to be used by both old phonetic-languages-simplified-generalized-examples-array and new phladiprelio-general-simple packages. + +-- A URL where users can report bugs. +-- bug-reports: + +-- The license under which the package is released. +license: MIT +author: Oleksandr Zhabenko +maintainer: oleksandr.zhabenko@yahoo.com + +-- A copyright notice. +copyright: Oleksandr Zhabenko +category: Language, Math, Data +extra-source-files: CHANGELOG.md, README.md + +library + exposed-modules: Phladiprelio.General.Parsing + + -- LANGUAGE extensions used by modules in this package. + other-extensions: BangPatterns, NoImplicitPrelude + build-depends: base >=4.13 && <5, phonetic-languages-phonetics-basics ==0.10.0.2 + hs-source-dirs: . + default-language: Haskell2010 +