Re: [Dlang] Internationalisation WIP

From 9c0d524d055525a4574cfce211f2688608110020 Mon Sep 17 00:00:00 2001
From: Adela Vais <[email protected]>
Date: Sat, 28 Nov 2020 21:09:39 +0200
Subject: [PATCH for Dlang support 1/3] d: add internationalisation support

The D parser implements this feature similarly to the C parser,
by using Gettext. Functions gettext() and dgettext() are
imported using extern(C). The internationalisation uses yysymbol_name
to report the name of the SymbolKinds.

* data/skeletons/d.m4 (SymbolKind.toString.yytranslatable,
SymbolKind.toString.yysymbol_name: New), data/skeletons/lalr1.d: Here.
* doc/bison.texi: Document it.
* tests/calc.at: Test it.
---
 data/skeletons/d.m4    | 22 ++++++++++++----
 data/skeletons/lalr1.d | 59 ++++++++++++++++++++++++++++++++++-------
 doc/bison.texi         | 60 ++++++++++++++++++++++++++++++++++++++++++
 tests/calc.at          | 19 ++++++++++++-
 4 files changed, 145 insertions(+), 15 deletions(-)

diff --git a/data/skeletons/d.m4 b/data/skeletons/d.m4
index 331c06a8..5320c483 100644
--- a/data/skeletons/d.m4
+++ b/data/skeletons/d.m4
@@ -192,7 +192,12 @@ b4_symbol_foreach([b4_token_enum])dnl
 }
 ])
 
-
+# b4_symbol_translate(STRING)
+# ---------------------------
+# Used by "bison" in the array of symbol names to mark those that
+# require translation.
+m4_define([b4_symbol_translate],
+[[_($1)]])
 
 ## -------------- ##
 ## Symbol kinds.  ##
@@ -252,8 +257,17 @@ m4_define([b4_declare_symbol_enum],
     final void toString(W)(W sink) const
     if (isOutputRange!(W, char))
     {
+      immutable string[] yy_sname = @{
+  ]b4_symbol_names[
+      @};]b4_has_translations_if([[
+      /* YYTRANSLATABLE[SYMBOL-NUM] -- Whether YY_SNAME[SYMBOL-NUM] is
+        internationalizable.  */
+      immutable ]b4_int_type_for([b4_translatable])[[] yytranslatable = @{
+  ]b4_translatable[
+      @};
+
+        put(sink, yy_sname[yycode_]);]], [[
       string yystr = yytname_[yycode_];
-
       if (yystr[0] == '"')
         {
         strip_quotes:
@@ -280,9 +294,7 @@ m4_define([b4_declare_symbol_enum],
       {
         put(sink, "end of input");
         return;
-      }
-
-      put(sink, yystr);
+      }]])[
     }
   }
 ]])
diff --git a/data/skeletons/lalr1.d b/data/skeletons/lalr1.d
index fd0038d6..ede7d8c7 100644
--- a/data/skeletons/lalr1.d
+++ b/data/skeletons/lalr1.d
@@ -40,6 +40,29 @@ version(D_Version2) {
 ]b4_user_post_prologue[
 ]b4_percent_code_get([[imports]])[
 import std.format;
+import std.conv;
+
+/**
+ * Handle error message internationalisation.
+ */
+static if (!is(typeof(YY_))) {
+  version(YYENABLE_NLS)
+  {
+    version(ENABLE_NLS)
+    {
+      extern(C) char* dgettext(const char*, const char*);
+      string YY_(const char* s)
+      {
+        return to!string(dgettext("bison-runtime", s));
+      }
+    }
+  }
+  static if (!is(typeof(YY_)))
+  {
+    pragma(inline, true)
+    string YY_(string msg) { return msg; }
+  }
+}
 
 /**
  * A Bison parser, automatically generated from <tt>]m4_bpatsubst(b4_file_name, [^"\(.*\)"$], [\1])[</tt>.
@@ -680,20 +703,38 @@ m4_popdef([b4_at_dollar])])dnl
       immutable int argmax = 5;
       SymbolKind[] yyarg = new SymbolKind[argmax];
       int yycount = yysyntaxErrorArguments(yyctx, yyarg, argmax);
-      string res = "syntax error, unexpected ";
-      res ~= format!"%s"(yyarg[0]);
-      if (yycount < argmax + 1)
+      string res, yyformat;
+      import std.string;
+      switch (yycount)
       {
-        for (int yyi = 1; yyi < yycount; yyi++)
-        {
-          res ~= yyi == 1 ? ", expecting " : " or ";
-          res ~= format!"%s"(SymbolKind(yyarg[yyi]));
-        }
+        case  1:
+          yyformat = YY_("syntax error, unexpected %s");
+          res = format(yyformat, yyarg[0]);
+         break;
+        case  2:
+          yyformat = YY_("syntax error, unexpected %s, expecting %s");
+          res = format(yyformat, yyarg[0], yyarg[1]);
+          break;
+        case  3:
+          yyformat = YY_("syntax error, unexpected %s, expecting %s or %s");
+          res = format(yyformat, yyarg[0], yyarg[1], yyarg[2]);
+          break;
+        case  4:
+          yyformat = YY_("syntax error, unexpected %s, expecting %s or %s or %s");
+          res = format(yyformat, yyarg[0], yyarg[1], yyarg[2], yyarg[3]);
+          break;
+        case  5:
+          yyformat = YY_("syntax error, unexpected %s, expecting %s or %s or %s or %s");
+          res = format(yyformat, yyarg[0], yyarg[1], yyarg[2], yyarg[3], yyarg[4]);
+          break;
+        default:
+          res = YY_("syntax error");
+          break;
       }
       yyerror(]b4_locations_if([yyctx.getLocation(), ])[res);
     }]],
 [[simple]], [[
-    yyerror(]b4_locations_if([yyctx.getLocation(), ])["syntax error");]])[
+    yyerror(]b4_locations_if([yyctx.getLocation(), ])[YY_("syntax error"));]])[
   }
 
 ]b4_parse_error_bmatch(
diff --git a/doc/bison.texi b/doc/bison.texi
index 0a55f343..c749079f 100644
--- a/doc/bison.texi
+++ b/doc/bison.texi
@@ -13953,6 +13953,66 @@ or nonzero, full tracing.
 Identify the Bison version and skeleton used to generate this parser.
 @end deftypecv
 
+The internationalization in D is very simmilar to the one in C. The D
+parser uses @code{dgettext} for translating Bison messages.
+
+To enable internationalisation, compile using
+@code{-version ENABLE_NLS -version YYENABLE_NLS} and import
+@code{bindtextdomain} and @code{textdomain} from C:
+
+@example
+extern(C) char* bindtextdomain(const char* domainname, const char* dirname);
+extern(C) char* textdomain(const char* domainname);
+@end example
+
+The main function should load the translation catalogues, similarly to the
+@file{c/bistromathic} example:
+
+@example
+int main()
+@{
+  import core.stdc.locale;
+
+  // Set up internationalization.
+  setlocale(LC_ALL, "");
+  // Use Bison's standard translation catalogue for error messages
+  // (the generated messages).
+  bindtextdomain("bison-runtime", BISON_LOCALEDIR);
+  // For the translation catalogue of your own project, use the
+  // name of your project.
+  bindtextdomain("bison", LOCALEDIR);
+  textdomain("bison");
+
+  // usual main content
+  ...
+@}
+@end example
+
+For user messages translations, the user must implement the
+@code{string} _(@code{const char*} @var{msg}) function and it is recommended
+to use @code{gettext}:
+
+@example
+%code imports @{
+  static if (!is(typeof(_)))
+  @{
+    version(ENABLE_NLS)
+    @{
+      extern(C) char* gettext(const char*);
+      string _(const char* s)
+      @{
+        return to!string(gettext(s));
+      @}
+    @}
+  @}
+  static if (!is(typeof(_)))
+  @{
+    pragma(inline, true)
+    string _(string msg) @{ return msg; @}
+  @}
+@}
+@end example
+
 @node D Parser Context Interface
 @subsection D Parser Context Interface
 The parser context provides information to build error reports when you
diff --git a/tests/calc.at b/tests/calc.at
index b1428316..3d319645 100644
--- a/tests/calc.at
+++ b/tests/calc.at
@@ -654,6 +654,23 @@ m4_define([_AT_DATA_CALC_Y(d)],
 };
 %printer { fprintf (yyo, "%d", $$); } <ival>;
 
+%code {
+]AT_TOKEN_TRANSLATE_IF([[
+  static string _(string s)
+  {
+    switch (s)
+    {
+      case "end of input":
+        return "end of file";
+      case "number":
+        return "nombre";
+      default:
+        return s;
+    }
+  }
+]])[
+}
+
 /* Bison Declarations */
 %token EOF 0 ]AT_TOKEN_TRANSLATE_IF([_("end of file")], ["end of input"])[
 %token <ival> NUM   "number"
@@ -665,7 +682,7 @@ m4_define([_AT_DATA_CALC_Y(d)],
        STAR   "*"
        SLASH  "/"
        POW    "^"
-       EOL    "\n"
+       EOL    "'\\n'"
        LPAR   "("
        RPAR   ")"
        NOT    "!"
-- 
2.17.1

From 05b159624c4f82874d99b84d4af82e093c825ee6 Mon Sep 17 00:00:00 2001
From: Adela Vais <[email protected]>
Date: Tue, 5 Jan 2021 16:47:29 +0200
Subject: [PATCH for Dlang support 2/3] d: remove yytnamerr usage

It is a backwards-compatible feature for the other parsers.
D should not support this option.

* data/skeletons/d.m4: Here.
---
 data/skeletons/d.m4 | 40 +++-------------------------------------
 1 file changed, 3 insertions(+), 37 deletions(-)

diff --git a/data/skeletons/d.m4 b/data/skeletons/d.m4
index 5320c483..cfb87944 100644
--- a/data/skeletons/d.m4
+++ b/data/skeletons/d.m4
@@ -243,12 +243,6 @@ m4_define([b4_declare_symbol_enum],
       yycode_ = code;
     }
 
-    /* YYTNAME[SYMBOL-NUM] -- String name of the symbol SYMBOL-NUM.
-       First, the terminals, then, starting at \a YYNTOKENS_, nonterminals.  */
-    static immutable string[] yytname_ = @{
-  ]b4_tname[
-    @};
-
     /* Return YYSTR after stripping away unnecessary quotes and
      backslashes, so that it's suitable for yyerror.  The heuristic is
      that double-quoting is unnecessary unless the string contains an
@@ -264,37 +258,9 @@ m4_define([b4_declare_symbol_enum],
         internationalizable.  */
       immutable ]b4_int_type_for([b4_translatable])[[] yytranslatable = @{
   ]b4_translatable[
-      @};
-
-        put(sink, yy_sname[yycode_]);]], [[
-      string yystr = yytname_[yycode_];
-      if (yystr[0] == '"')
-        {
-        strip_quotes:
-          for (int i = 1; i < yystr.length; i++)
-            switch (yystr[i])
-              {
-              case '\'':
-              case ',':
-                break strip_quotes;
-
-              case '\\':
-                if (yystr[++i] != '\\')
-                  break strip_quotes;
-                goto default;
-              default:
-                put(sink, yystr[i]);
-                break;
-
-              case '"':
-                return;
-              }
-        }
-      else if (yystr == "$end")
-      {
-        put(sink, "end of input");
-        return;
-      }]])[
+      @};]])[
+
+      put(sink, yy_sname[yycode_]);
     }
   }
 ]])
-- 
2.17.1

From 6ef304b24eb52531348e51a0cfee2755c69f8660 Mon Sep 17 00:00:00 2001
From: Adela Vais <[email protected]>
Date: Tue, 5 Jan 2021 17:20:11 +0200
Subject: [PATCH for Dlang support 3/3] d: remove unnecessary imports

* data/skeletons/lalr1.d: Here.
---
 data/skeletons/lalr1.d | 5 -----
 1 file changed, 5 deletions(-)

diff --git a/data/skeletons/lalr1.d b/data/skeletons/lalr1.d
index ede7d8c7..e97613b6 100644
--- a/data/skeletons/lalr1.d
+++ b/data/skeletons/lalr1.d
@@ -382,7 +382,6 @@ b4_user_union_members
     }
 
 ]b4_parse_trace_if([[
-    import std.conv : to;
     yy_symbol_print ("-> $$ =", to!SymbolKind (yyr1_[yyn]), yyval]b4_locations_if([, yyloc])[);]])[
 
     yystack.pop (yylen);
@@ -664,7 +663,6 @@ m4_popdef([b4_at_dollar])])dnl
 
         /* Shift the error token.  */]b4_lac_if([[
         yylacDiscard("error recovery");]])[]b4_parse_trace_if([[
-        import std.conv : to;
         yy_symbol_print ("Shifting", to!SymbolKind (yystos_[yyn]), yylval]b4_locations_if([, yyloc])[);]])[
         yystate = yyn;
         yystack.push (yyn, yylval]b4_locations_if([, yyloc])[);
@@ -704,7 +702,6 @@ m4_popdef([b4_at_dollar])])dnl
       SymbolKind[] yyarg = new SymbolKind[argmax];
       int yycount = yysyntaxErrorArguments(yyctx, yyarg, argmax);
       string res, yyformat;
-      import std.string;
       switch (yycount)
       {
         case  1:
@@ -882,7 +879,6 @@ m4_popdef([b4_at_dollar])])dnl
     destroy(yylacStack);
     // Reduce until we encounter a shift and thereby accept the token.
 ]b4_parse_trace_if([[
-    import std.conv;
     yycdebug("LAC: checking lookahead " ~ format("%s", yytoken) ~ ":");]])[
     int lacTop = 0;
     while (true)
@@ -1073,7 +1069,6 @@ m4_popdef([b4_at_dollar])])dnl
                 yyrule - 1, yylno));
 
     /* The symbols being reduced.  */
-    import std.conv : to;
     for (int yyi = 0; yyi < yynrhs; yyi++)
       yy_symbol_print (format("   $%d =", yyi + 1),
                        to!SymbolKind (yystos_[yystack.stateAt(yynrhs - (yyi + 1))]),
-- 
2.17.1