Fix #451: Java/ANTLR: unescape Char and String literals in parser

BNFC · Jul 26, 2023 · d5ba1ac · d5ba1ac
1 parent 9cb7658
commit d5ba1ac
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 14 deletions.
diff --git a/source/CHANGELOG.md b/source/CHANGELOG.md
@@ -2,8 +2,8 @@
 
 Unreleased
 
-* C/C++/Java: escape newline etc. when printing `String` and `Char` literals
-  [[#449](https://github.com/BNFC/bnfc/issues/449)]
+* C/C++/Java: escape newline etc. when printing `String` and `Char` literals [[#449](https://github.com/BNFC/bnfc/issues/449)]
+* Java/ANTLR: unescape `String` and `Char` literals in parser [[#451](https://github.com/BNFC/bnfc/issues/451)]
 * Haskell/GADT: generated `ComposOp.hs` no longer needs `mtl` library [[#438](https://github.com/BNFC/bnfc/pull/438)]
 * Ocaml: fixed a crash in printer with unbalanced `}` [[#439](https://github.com/BNFC/bnfc/issues/439)]
 * Ocaml: lex escape sequences in `Char` [[#452](https://github.com/BNFC/bnfc/issues/452)]

diff --git a/source/src/BNFC/Backend/Java/CFtoAntlr4Parser.hs b/source/src/BNFC/Backend/Java/CFtoAntlr4Parser.hs
@@ -138,19 +138,16 @@ generateAction packageAbsyn nt f ms rev
      p_1               = resultvalue $ ms!!0
      p_2               = resultvalue $ ms!!1
      add               = if rev then "addLast" else "addFirst"
-     gettext           = "getText()"
-     removeQuotes x    = "substring(1, "++ x +.+ gettext +.+ "length()-1)"
-     parseint x        = "Integer.parseInt("++x++")"
-     parsedouble x     = "Double.parseDouble("++x++")"
-     charat            = "charAt(1)"
+     unescape x        = concat [ "org.antlr.v4.misc.CharSupport.getStringFromGrammarStringLiteral(", x, ")" ]
      resultvalue (n,c) = case c of
-                          TokenCat "Ident"   -> n'+.+gettext
-                          TokenCat "Integer" -> parseint $ n'+.+gettext
-                          TokenCat "Char"    -> n'+.+gettext+.+charat
-                          TokenCat "Double"  -> parsedouble $ n'+.+gettext
-                          TokenCat "String"  -> n'+.+gettext+.+removeQuotes n'
-                          _         -> (+.+) n' (if isTokenCat c then gettext else "result")
-                          where n' = '$':n
+                          TokenCat "Double"  -> concat [ "Double.parseDouble(", txt, ")" ]
+                          TokenCat "Integer" -> concat [ "Integer.parseInt("  , txt, ")" ]
+                          TokenCat "Char"    -> unescape txt +.+ "charAt(0)"
+                          TokenCat "String"  -> unescape txt
+                          TokenCat "Ident"   -> txt
+                          c | isTokenCat c   -> txt
+                            | otherwise      -> concat [ "$", n, ".result" ]
+                          where txt = '$':n +.+ "getText()"
 
 -- | Generate patterns and a set of metavariables indicating
 -- where in the pattern the non-terminal