From 750a2639b0f46153c36f6c713f522db5bb2d8587 Mon Sep 17 00:00:00 2001
From: aarne <aarne@chalmers.se>
Date: Wed, 27 Jan 2010 17:51:15 +0000
Subject: command eb for example-based grammar conversion; see 'help eb' and
 the example in examples/animals/QuestionsI.gfe

---
 src/compiler/GF/Command/Commands.hs     | 31 +++++++++++++++++
 src/compiler/GF/Compile/ExampleBased.hs | 59 +++++++++++++++++++++++++++++++++
 2 files changed, 90 insertions(+)
 create mode 100644 src/compiler/GF/Compile/ExampleBased.hs

(limited to 'src/compiler')

diff --git a/src/compiler/GF/Command/Commands.hs b/src/compiler/GF/Command/Commands.hs
index 5d2b5421b..7f8722d00 100644
--- a/src/compiler/GF/Command/Commands.hs
+++ b/src/compiler/GF/Command/Commands.hs
@@ -23,6 +23,7 @@ import PGF.Printer
 import PGF.Probabilistic -- (getProbsFromFile,prProbabilities,defaultProbabilities)
 import PGF.Generate (genRandomProb) ----
 import GF.Compile.Export
+import GF.Compile.ExampleBased
 import GF.Infra.Option (noOptions, readOutputFormat, Encoding(..))
 import GF.Infra.UseIO
 import GF.Data.ErrM ----
@@ -231,6 +232,33 @@ allCommands cod env@(pgf, mos) = Map.fromList [
      longname = "empty",
      synopsis = "empty the environment"
      }),
+  ("eb", emptyCommandInfo {
+     longname = "example_based",
+     syntax = "eb (-probs=FILE | -lang=LANG)* -file=FILE.gfe",
+     synopsis = "converts .gfe files to .gf files by parsing examples to trees",
+     explanation = unlines [
+       "Reads FILE.gfe and writes FILE.gf. Each expression of form",
+       "'%ex CAT QUOTEDSTRING' in FILE.gfe is replaced by a syntax tree.",
+       "This tree is the first one returned by the parser; a biased ranking",
+       "can be used to regulate the order. If there are more than one parses",
+       "the rest are shown in comments, with probabilities if the order is biased.",  
+       "The probabilities flag and configuration file is similar to the commands",
+       "gr and rt. Notice that the command doesn't change the environment,",
+       "but the resulting .gf file must be imported separately."
+       ],
+     flags = [
+       ("file","the file to be converted (suffix .gfe must be given)"),
+       ("lang","the language in which to parse"),
+       ("probs","file with probabilities to rank the parses")
+       ],
+     exec = \opts _ -> do
+       let file = optFile opts
+       mprobs <- optProbs opts pgf
+       let conf = configureExBased pgf mprobs (optLang opts)
+       file' <- parseExamplesInGrammar conf file
+       return (fromString ("wrote " ++ file')),
+     needsTypeCheck = False
+     }),
   ("gr", emptyCommandInfo {
      longname = "generate_random",
      synopsis = "generate random trees in the current abstract syntax",
@@ -871,6 +899,8 @@ allCommands cod env@(pgf, mos) = Map.fromList [
 --       putStrLn $ prProbabilities ps 
        return $ Just ps
 
+   optFile opts = valStrOpts "file" "_gftmp" opts
+
    optType opts = 
      let str = valStrOpts "cat" (showCId $ lookStartCat pgf) opts
      in case readType str of
@@ -988,3 +1018,4 @@ prMorphoAnalysis (w,lps) =
 
 morphoMissing :: Morpho -> [String] -> [String]
 morphoMissing mo ws = [w | w <- ws, null (lookupMorpho mo w)]
+
diff --git a/src/compiler/GF/Compile/ExampleBased.hs b/src/compiler/GF/Compile/ExampleBased.hs
new file mode 100644
index 000000000..10d7cdc88
--- /dev/null
+++ b/src/compiler/GF/Compile/ExampleBased.hs
@@ -0,0 +1,59 @@
+module GF.Compile.ExampleBased (parseExamplesInGrammar,configureExBased) where
+
+import PGF
+import PGF.Probabilistic
+
+parseExamplesInGrammar :: ExConfiguration -> FilePath -> IO FilePath
+parseExamplesInGrammar conf file = do
+  src <- readFile file                             -- .gfe
+  let file' = take (length file - 3) file ++ "gf"  -- .gf
+  convertFile conf src file'
+  return file'
+
+convertFile :: ExConfiguration -> String -> FilePath -> IO ()
+convertFile conf src file = do
+  writeFile file "" -- "-- created by example-based grammar writing in GF\n"
+  conv src
+ where
+  conv s = do
+    (cex,end) <- findExample s
+    if null end then return () else do
+      convEx cex
+      conv end
+  findExample s = case s of
+    '%':'e':'x':cs -> return $ getExample cs
+    c:cs -> appf [c] >> findExample cs
+    _ -> return (undefined,s)
+  getExample s = 
+    let 
+      (cat,exend) = break (=='"') s
+      (ex,   end) = break (=='"') (tail exend)
+    in ((unwords (words cat),ex), tail end)  -- quotes ignored
+  pgf = resource_pgf conf
+  lang = language conf 
+  convEx (cat,ex) = do
+    appn "("
+    let typ = maybe (error "no valid cat") id $ readType cat
+    let ts = rank $ parse pgf lang typ ex
+    case ts of
+      []   -> appv ("WARNING: cannot parse example " ++ ex)
+      t:tt -> appn t >> mapM_ (appn . ("  --- " ++)) tt 
+    appn ")"
+  rank ts = case probs conf of
+    Just probs -> [showExpr [] t ++ "  -- " ++ show p | (t,p) <- rankTreesByProbs probs ts]
+    _ -> map (showExpr []) ts
+  appf = appendFile file
+  appn s = appf s >> appf "\n"
+  appv s = appn s >> putStrLn s
+
+data ExConfiguration = ExConf {
+  resource_file :: FilePath,
+  resource_pgf  :: PGF,
+  probs    :: Maybe Probabilities,
+  verbose  :: Bool,
+  language :: Language
+  }
+
+configureExBased :: PGF -> Maybe Probabilities -> Language -> ExConfiguration
+configureExBased pgf mprobs lang = ExConf [] pgf mprobs False lang
+
-- 
cgit v1.2.3