as: handle .ascii and .asciz directives

erlang/apps/as/priv/test6.s

@@ -0,0 +1,23 @@
+/*
+ * test6.s
+ * Copyright (C) 2019  Mikael Pettersson
+ *
+ * This file is part of pdp10-tools.
+ *
+ * pdp10-tools is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * pdp10-tools is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with pdp10-tools.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+	.text
+	.ascii "bar"
+	.asciz "foo", "fie", "fum"

erlang/apps/as/src/assemble.erl

@@ -100,12 +100,21 @@ stmts_image([Stmt | Stmts], Tunit, Acc) ->
 
 stmt_image(Stmt, Tunit) ->
   case Stmt of
+    #s_dot_ascii{} -> dot_ascii_image(Stmt, Tunit);
     #s_dot_byte{} -> dot_byte_image(Stmt, Tunit);
     #s_dot_long{} -> dot_long_image(Stmt, Tunit);
     #s_dot_short{} -> dot_short_image(Stmt, Tunit);
     #s_insn{} -> insn_image(Stmt, Tunit)
   end.
 
+dot_ascii_image(#s_dot_ascii{z = Z, strings = Strings}, _Tunit) ->
+  Image =
+    case Z of
+      true -> [String ++ [0] || String <- Strings];
+      false -> Strings
+    end,
+  {ok, Image}.
+
 dot_byte_image(#s_dot_byte{exprs = Exprs}, Tunit) ->
   integer_data_directive(Exprs, Tunit, fun(Value) -> Value band ?PDP10_UINT9_MAX end).
 

erlang/apps/as/src/input.erl

@@ -313,6 +313,7 @@ pass2_stmts([{Location, Stmt} | Stmts], Tunit) ->
 
 pass2_stmt(Location, Tunit, Stmt) ->
   case Stmt of
+    #s_dot_ascii{} -> dot_ascii(Location, Tunit, Stmt);
     #s_dot_byte{} -> dot_byte(Location, Tunit, Stmt);
     #s_dot_file{} -> dot_file(Location, Tunit, Stmt);
     #s_dot_globl{} -> dot_globl(Location, Tunit, Stmt);
@@ -326,6 +327,21 @@ pass2_stmt(Location, Tunit, Stmt) ->
     #s_insn{} -> insn(Location, Tunit, Stmt)
   end.
 
+dot_ascii(_Location, Tunit, #s_dot_ascii{z = Z, strings = Strings} = Stmt) ->
+  #tunit{cursect = Cursect} = Tunit,
+  #section{data = {stmts, Stmts}, dot = Dot} = Section = tunit:get_section(Tunit, Cursect),
+  Size0 = lists:sum(lists:map(fun erlang:length/1, Strings)),
+  Size =
+    case Z of
+      true -> Size0 + length(Strings);
+      false -> Size0
+    end,
+  NewSection =
+    Section#section{ data = {stmts, [Stmt | Stmts]}
+                   , dot = Dot + Size
+                   },
+  {ok, tunit:put_section(Tunit, NewSection)}.
+
 dot_byte(Location, Tunit, #s_dot_byte{} = Stmt0) ->
   integer_data_directive(Location, Tunit, Stmt0, 1, ".byte",
                          fun(Stmt) -> Stmt#s_dot_byte.exprs end,

erlang/apps/as/src/parse.erl

@@ -34,6 +34,8 @@
       -> {ok, {location(), stmt()}} | eof | {error, {module(), term()}}.
 stmt(ScanState) ->
   case scan:token(ScanState) of
+    {ok, {Location, ?T_DOT_ASCII}} -> dot_ascii(ScanState, Location);
+    {ok, {Location, ?T_DOT_ASCIZ}} -> dot_asciz(ScanState, Location);
     {ok, {Location, ?T_DOT_BYTE}} -> dot_byte(ScanState, Location);
     {ok, {Location, ?T_DOT_DATA}} -> dot_data(ScanState, Location);
     {ok, {Location, ?T_DOT_FILE}} -> dot_file(ScanState, Location);
@@ -274,6 +276,23 @@ badinsn(Location, Fmt, Mnemonic) ->
 
 %% Directives ------------------------------------------------------------------
 
+dot_ascii(ScanState, Location) ->
+  dot_ascii(ScanState, Location, ".ascii", _Z = false).
+
+dot_asciz(ScanState, Location) ->
+  dot_ascii(ScanState, Location, ".asciz", _Z = true).
+
+dot_ascii(ScanState, Location, Lexeme, Z) ->
+  case string_list(ScanState) of
+    {ok, {Strings, Follow}} ->
+      case Follow of
+        {_Location, ?T_NEWLINE} ->
+          {ok, {Location, #s_dot_ascii{z = Z, strings = Strings}}};
+        _ -> badtok("junk after " ++ Lexeme ++ " <strings>", {ok, Follow})
+      end;
+    {error, _Reason} = Error -> Error
+  end.
+
 dot_byte(ScanState, Location) ->
   integer_data_directive(ScanState, Location, ".byte",
                          fun(Exprs) -> #s_dot_byte{exprs = Exprs} end).
@@ -500,6 +519,29 @@ expr_opt(ScanState) ->
       Error
   end.
 
+%% String Lists ----------------------------------------------------------------
+
+%% <string_list> ::= (<string> ("," <string>)*)?
+string_list(ScanState) ->
+  case scan:token(ScanState) of
+    {ok, {_Location, {?T_STRING, String}}} -> string_list(ScanState, [String]);
+    {ok, Follow} -> {ok, {[], Follow}};
+    {error, _Reason} = Error -> Error
+  end.
+
+string_list(ScanState, Strings) ->
+  case scan:token(ScanState) of
+    {ok, {_Location1, ?T_COMMA}} ->
+      case scan:token(ScanState) of
+        {ok, {_Location2, {?T_STRING, String}}} ->
+          string_list(ScanState, [String | Strings]);
+        {ok, First} -> badtok("expected string literal", {ok, First});
+        {error, _Reason} = Error -> Error
+      end;
+    {ok, Follow} -> {ok, {lists:reverse(Strings), Follow}};
+    {error, _Reason} = Error -> Error
+  end.
+
 %% Error reporting -------------------------------------------------------------
 
 badtok(_ErrMsg, {error, _Reason} = Error) -> Error;

erlang/apps/as/src/token.erl

@@ -29,6 +29,8 @@
 -spec from_symbol(string()) -> token().
 from_symbol(Name) ->
   case Name of
+    ".ascii"            -> ?T_DOT_ASCII;
+    ".asciz"            -> ?T_DOT_ASCIZ;
     ".byte"             -> ?T_DOT_BYTE;
     ".data"             -> ?T_DOT_DATA;
     ".file"             -> ?T_DOT_FILE;
@@ -51,6 +53,8 @@ from_symbol(Name) ->
 -spec format(token()) -> io_lib:chars().
 format(Token) ->
   case Token of
+    ?T_DOT_ASCII        -> ".ascii";
+    ?T_DOT_ASCIZ        -> ".asciz";
     ?T_DOT_BYTE         -> ".byte";
     ?T_DOT_DATA         -> ".data";
     ?T_DOT_FILE         -> ".file";

erlang/apps/as/src/token.hrl

@@ -26,6 +26,8 @@
 %% with attributes (ordinary symbols, literals, errors).
 
 %% reserved symbols including directives
+-define(T_DOT_ASCII,            'T_DOT_ASCII').         % .ascii
+-define(T_DOT_ASCIZ,            'T_DOT_ASCIZ').         % .asciz
 -define(T_DOT_BYTE,             'T_DOT_BYTE').          % .byte
 -define(T_DOT_DATA,             'T_DOT_DATA').          % .data
 -define(T_DOT_FILE,             'T_DOT_FILE').          % .file
@@ -64,7 +66,9 @@
 -define(T_NEWLINE,      'T_NEWLINE').   % <end-of-line>
 -define(T_EOF,          'T_EOF').       % <end-of-file>
 
--type token() :: ?T_DOT_BYTE
+-type token() :: ?T_DOT_ASCII
+               | ?T_DOT_ASCIZ
+               | ?T_DOT_BYTE
                | ?T_DOT_DATA
                | ?T_DOT_FILE
                | ?T_DOT_GLOBL

erlang/apps/as/src/tunit.hrl

@@ -36,6 +36,9 @@
 %% either interpreted immediately or appended to the representation of the
 %% current section.
 
+%% .asci{i,z} [string (, string)*]
+-record(s_dot_ascii, {z :: boolean(), strings :: [string()]}).
+
 %% .byte [expr (, expr)*]
 -record(s_dot_byte, {exprs :: [expr()]}).
 
@@ -90,7 +93,8 @@
         , index         :: 0..((1 bsl 4) - 1)
         }).
 
--type stmt() :: #s_dot_data{}
+-type stmt() :: #s_dot_ascii{}
+              | #s_dot_data{}
               | #s_dot_file{}
               | #s_dot_globl{}
               | #s_dot_ident{}