Wordlist format

doc/gf-history.html

@@ -14,6 +14,19 @@ Changes in functionality since May 17, 2005, release of GF Version 2.2
 
 <p>
 
+18/5 (AR) Introduced a wordlist format <tt>gfwl</tt> for
+quick creation of language exercises and (in future) multilingual lexica.
+The format is now very simple:
+<pre>
+  # Svenska - Franska - Finska
+  berg      - montagne            - vuori
+  klättra   - grimper / escalader - kiivetä / kiipeillä
+</pre>
+but can be extended to cover paradigm functions in addition to just
+words.
+
+<p>
+
 3/4 (AR) The predefined abstract syntax type <tt>Int</tt> now has two
 inherent parameters indicating its last digit and its size. The (hard-coded)
 linearization type is

lib/resource-1.0/Makefile

@@ -19,7 +19,7 @@ langs:
 
 present:
 	chmod u+x mkPresent
-	$(GF) -make -src -preproc=./mkPresent */Lang??*.gf
+	$(GF) -make -src -preproc=./mkPresent */Lang??*.gf +RTS -M800M -K100M
 	mv */*.gfc */*.gfr ../present
 
 mathematical:

src/GF.hs

@@ -22,6 +22,7 @@ import GF.API.IOGrammar
 import GF.Compile.ShellState
 import GF.Compile.Compile
 import GF.Compile.MkConcrete
+import GF.Compile.Wordlist
 import GF.Shell
 import GF.Shell.SubShell
 import GF.Shell.ShellCommands
@@ -91,6 +92,11 @@ main = do
         es <- liftM (nub . concat) $ mapM (getGFEFiles os) fs
         mkConcretes os es
         doGF (removeOption fromExamples os) fs
+  -- preprocessing gfwl
+    else if (length fs == 1 && fileSuffix (head fs) == "gfwl")
+    then do
+      fs' <- mkWordlist (head fs)
+      doGF os fs'
     else doGF os fs
 
 helpMsg = unlines [

src/GF/Compile/Wordlist.hs

@@ -0,0 +1,107 @@
+----------------------------------------------------------------------
+-- |
+-- Module      : Wordlist
+-- Maintainer  : AR
+-- Stability   : (stable)
+-- Portability : (portable)
+--
+-- > CVS $Date: 
+-- > CVS $Author: 
+-- > CVS $Revision: 
+--
+-- Compile a gfwl file (multilingual word list) to an abstract + concretes
+-----------------------------------------------------------------------------
+
+module GF.Compile.Wordlist (mkWordlist) where
+
+import GF.Data.Operations
+import GF.Infra.UseIO
+import Data.List
+import Data.Char
+
+-- read File.gfwl, write File.gf (abstract) and a set of concretes
+-- return the names of the concretes
+
+mkWordlist :: FilePath -> IO [FilePath]
+mkWordlist file = do
+  s <- readFileIf file
+  let abs = fileBody file
+  let (cnchs,wlist) = pWordlist abs $ filter notComment $ lines s
+  let (gr,grs) = mkGrammars abs cnchs wlist
+  let cncfs = [cnc ++ ".gf" | (cnc,_) <- cnchs]
+  mapM_ (uncurry writeFile) $ (abs ++ ".gf",gr) : zip cncfs grs
+  putStrLn $ "wrote " ++ unwords ((abs ++ ".gf") : cncfs)
+  return cncfs
+
+{-
+-- syntax of files, e.g.
+
+  # Svenska - Franska - Finska  -- names of concretes
+ 
+  berg  - montagne - vuori      -- word entry
+
+-- this creates:
+
+  cat S ; 
+  fun berg_S : S ; 
+  lin berg_S = {s = ["berg"]} ;
+  lin berg_S = {s = ["montagne"]} ;
+  lin berg_S = {s = ["vuori"]} ;
+
+-- support for different categories to be elaborated. The syntax it
+
+  Verb . klättra - grimper / escalader - kiivetä / kiipeillä
+
+-- notice that a word can have several alternative (separator /)
+-- and that an alternative can consist of several words
+-}
+
+type CncHeader = (String,String) -- module name, module header
+
+type Wordlist = [(String, [[String]])] -- cat, variants for each cnc
+
+
+pWordlist :: String -> [String] -> ([CncHeader],Wordlist)
+pWordlist abs ls = (headers,rules) where
+  (hs,rs) = span ((=="#") . take 1) ls
+  headers = map mkHeader $ chunks "-" $ filter (/="#") $ words $ concat hs
+  rules   = map (mkRule   . words) rs
+
+  mkHeader ws = case ws of
+    w:ws2 -> (w, unwords ("concrete":w:"of":abs:"=":ws2))
+  mkRule ws = case ws of
+    cat:".":vs -> (cat, mkWords vs)
+    _          -> ("S", mkWords ws)
+  mkWords = map (map unwords . chunks "/") . chunks "-"
+
+
+mkGrammars :: String -> [CncHeader] -> Wordlist -> (String,[String])
+mkGrammars ab hs wl = (abs,cncs) where
+  abs = unlines $ map unwords $
+    ["abstract",ab,"=","{"]:
+    cats ++
+    funs ++
+    [["}"]]
+
+  cncs = [unlines $ (h ++ " {") : map lin rs ++ ["}"] | ((_,h),rs) <- zip hs rss]
+
+  cats = [["cat",c,";"] | c <- nub $ map fst wl]
+  funs = [["fun", f , ":", c,";"] | (f,c,_) <- wlf]
+
+  wlf = [(ident f c, c, ws) | (c,ws@(f:_)) <- wl]
+
+  rss = [[(f, wss !! i) | (f,_,wss) <- wlf] | i <- [0..length hs - 1]]
+
+  lin (f,ss) = unwords ["lin", f, "=", "{s", "=", val ss, "}", ";"]
+
+  val ss = case ss of
+    [w] -> quote w
+    _   -> "variants {" ++ unwords (intersperse ";" (map quote ss)) ++ "}"
+
+  quote w = "[" ++ prQuotedString w ++ "]" 
+
+  ident f c = concat $ intersperse "_" $ words (head f) ++ [c]
+
+
+notComment s = not (all isSpace s) && take 2 s /= "--"
+

src/GF/Shell.hs

@@ -50,6 +50,7 @@ import GF.Infra.Option
 import GF.UseGrammar.Information
 import GF.Shell.HelpFile
 import GF.Compile.PrOld
+import GF.Compile.Wordlist
 import GF.Grammar.PrGrammar
 
 import Control.Monad (foldM,liftM)
@@ -189,6 +190,10 @@ execLine put (c@(co, os), arg, cs) (outps,st) = do
 execC :: CommandOpt -> ShellIO
 execC co@(comm, opts0) sa@(sh@(st,(h,_,_,_)),a) = checkOptions st co >> case comm of
 
+  CImport file | fileSuffix file == "gfwl" -> do
+    fs <- mkWordlist file
+    foldM (\x y -> execC (CImport y, opts) x) sa fs
+
   CImport file | oElem fromExamples opts -> do
     es <- liftM nub $ getGFEFiles opts file
     system $ "gf -examples" +++ unlines es

src/GF/Shell/HelpFile.hs

@@ -49,6 +49,7 @@ txtHelpFile =
   "\n        .gfr   precompiled GF resource  " ++
   "\n        .gfcm  multilingual canonical GF" ++
   "\n        .gfe   example-based grammar files (only with the -ex option)" ++
+  "\n        .gfwl  multilingual word list (preprocessed to abs + cncs)" ++
   "\n        .ebnf  Extended BNF format" ++
   "\n        .cf    Context-free (BNF) format" ++
   "\n        .trc   TransferCore format" ++

src/HelpFile

@@ -20,6 +20,7 @@ i,  import: i File
         .gfr   precompiled GF resource  
         .gfcm  multilingual canonical GF
         .gfe   example-based grammar files (only with the -ex option)
+        .gfwl  multilingual word list (preprocessed to abs + cncs)
         .ebnf  Extended BNF format
         .cf    Context-free (BNF) format
         .trc   TransferCore format