First commit ; includes first TP and minijazz compiler

3 files changed, 383 insertions, 0 deletions

diff --git a/minijazz/src/parser/_tags b/minijazz/src/parser/_tags
new file mode 100644
index 0000000..d4a6ba1
--- /dev/null
+++ b/minijazz/src/parser/_tags
@@ -0,0 +1,2 @@
+<parser.ml>: use_menhirLib
+true: use_menhir
diff --git a/minijazz/src/parser/lexer.mll b/minijazz/src/parser/lexer.mll
new file mode 100644
index 0000000..f7b0e82
--- /dev/null
+++ b/minijazz/src/parser/lexer.mll
@@ -0,0 +1,196 @@
+(* lexer.mll *)
+
+
+{
+open Location
+open Lexing
+open Parser
+open Errors
+
+exception Lexical_error of lexical_error * location;;
+
+let comment_depth = ref 0
+
+let keyword_table = ((Hashtbl.create 149) : (string, token) Hashtbl.t);;
+
+List.iter (fun (str,tok) -> Hashtbl.add keyword_table str tok) [
+ "ram", RAM;
+ "rom", ROM;
+ "where", WHERE;
+ "end", END;
+ "true", BOOL(true);
+ "false", BOOL(false);
+ "reg", REG;
+ "not", NOT;
+ "const", CONST;
+ "and", AND;
+ "nand", NAND;
+ "or", OR;
+ "xor", XOR;
+ "if", IF;
+ "then", THEN;
+ "else", ELSE;
+ "inlined", INLINED;
+ "probing", PROBING
+]
+
+
+(* To buffer string literals *)
+
+let initial_string_buffer = String.create 256
+let string_buff = ref initial_string_buffer
+let string_index = ref 0
+
+let reset_string_buffer () =
+  string_buff := initial_string_buffer;
+  string_index := 0;
+  ()
+
+(*
+let incr_linenum lexbuf =
+  let pos = lexbuf.Lexing.lex_curr_p in
+    lexbuf.Lexing.lex_curr_p <- { pos with
+      Lexing.pos_lnum = pos.Lexing.pos_lnum + 1;
+      Lexing.pos_bol = pos.Lexing.pos_cnum;
+    }
+*)
+
+let store_string_char c =
+  if !string_index >= String.length (!string_buff) then begin
+    let new_buff = String.create (String.length (!string_buff) * 2) in
+      String.blit (!string_buff) 0 new_buff 0 (String.length (!string_buff));
+      string_buff := new_buff
+  end;
+  String.set (!string_buff) (!string_index) c;
+  incr string_index
+
+
+let get_stored_string () =
+  let s = String.sub (!string_buff) 0 (!string_index) in
+    string_buff := initial_string_buffer;
+    s
+
+let char_for_backslash = function
+    'n' -> '\010'
+  | 'r' -> '\013'
+  | 'b' -> '\008'
+  | 't' -> '\009'
+  | c   -> c
+
+let char_for_decimal_code lexbuf i =
+  let c =
+    100 * (int_of_char(Lexing.lexeme_char lexbuf i) - 48) +
+     10 * (int_of_char(Lexing.lexeme_char lexbuf (i+1)) - 48) +
+          (int_of_char(Lexing.lexeme_char lexbuf (i+2)) - 48) in
+  char_of_int(c land 0xFF)
+
+}
+
+let newline = '\n' | '\r' '\n'
+
+rule token = parse
+  | newline         { new_line lexbuf; token lexbuf }
+  | [' ' '\t'] +    { token lexbuf }
+  | "("             { LPAREN }
+  | ")"             { RPAREN }
+  | "*"             { STAR }
+  | "+"             { PLUS }
+  | "&"             { AND }
+  | "/"             { SLASH }
+  | "<"             { LESS }
+  | ">"             { GREATER }
+  | "["             { LBRACKET }
+  | "]"             { RBRACKET }
+  | ":"             { COLON }
+  | ";"             { SEMICOL }
+  | "="             { EQUAL }
+  | ","             { COMMA }
+  | "-"             { MINUS }
+  | "^"             { POWER }
+  | "<="            { LEQ }
+  | "."             { DOT }
+  | ".."            { DOTDOT }
+  | (['A'-'Z']('_' ? ['A'-'Z' 'a'-'z' ''' '0'-'9']) * as id)
+      {NAME id}
+  | (['A'-'Z' 'a'-'z']('_' ? ['A'-'Z' 'a'-'z' ''' '0'-'9']) * as id)
+      { let s = Lexing.lexeme lexbuf in
+        try Hashtbl.find keyword_table s
+        with Not_found -> NAME id }
+  | '0' ['b' 'B'] (['0'-'1']+ as lit)
+      { BOOL_INT lit }
+  | ['0'-'9']+
+  | '0' ['x' 'X'] ['0'-'9' 'A'-'F' 'a'-'f']+
+  | '0' ['o' 'O'] ['0'-'7']+
+      { INT (int_of_string(Lexing.lexeme lexbuf)) }
+  | "\""
+      { reset_string_buffer();
+        let string_start = lexbuf.lex_start_p in
+         (* string_start_loc := Location.curr lexbuf; *)
+          string lexbuf;
+          lexbuf.lex_start_p <- string_start;
+          STRING (get_stored_string()) }
+  | "(*"
+      { let comment_start = lexbuf.lex_curr_p in
+        comment_depth := 1;
+        begin try
+          comment lexbuf
+        with Lexical_error(Unterminated_comment, (Loc (_, comment_end))) ->
+          raise(Lexical_error(Unterminated_comment,
+                              Loc (comment_start, comment_end)))
+        end;
+        token lexbuf }
+  | eof            {EOF}
+  | _              {raise (Lexical_error (Illegal_character,
+                      Loc (Lexing.lexeme_start_p lexbuf,
+                      Lexing.lexeme_end_p lexbuf)))}
+
+and comment = parse
+    "(*"
+      { comment_depth := succ !comment_depth; comment lexbuf }
+  | "*)"
+      { comment_depth := pred !comment_depth;
+        if !comment_depth > 0 then comment lexbuf }
+   | "\""
+      { reset_string_buffer();
+        let string_start = lexbuf.lex_curr_p in
+        begin try
+          string lexbuf
+        with Lexical_error(Unterminated_string, Loc (_, string_end)) ->
+          raise(Lexical_error
+            (Unterminated_string, Loc (string_start, string_end)))
+        end;
+        comment lexbuf }
+  | "''"
+      { comment lexbuf }
+  | "'" [^ '\\' '\''] "'"
+      { comment lexbuf }
+  | "'" '\\' ['\\' '\'' 'n' 't' 'b' 'r'] "'"
+      { comment lexbuf }
+  | "'" '\\' ['0'-'9'] ['0'-'9'] ['0'-'9'] "'"
+      { comment lexbuf }
+  | eof
+      { raise(Lexical_error(Unterminated_comment, Loc(dummy_pos,
+                            Lexing.lexeme_start_p lexbuf))) }
+  | _
+      { comment lexbuf }
+
+and string = parse
+    '"'
+      { () }
+  | '\\' ("\010" | "\013" | "\013\010") [' ' '\009'] *
+      { string lexbuf }
+  | '\\' ['\\' '"'  'n' 't' 'b' 'r']
+      { store_string_char(char_for_backslash(Lexing.lexeme_char lexbuf 1));
+        string lexbuf }
+  | '\\' ['0'-'9'] ['0'-'9'] ['0'-'9']
+      { store_string_char(char_for_decimal_code lexbuf 1);
+         string lexbuf }
+  | eof
+      { raise (Lexical_error(Unterminated_string, Loc (dummy_pos,
+                              Lexing.lexeme_start_p lexbuf))) }
+  | _
+      { store_string_char(Lexing.lexeme_char lexbuf 0);
+        string lexbuf }
+
+(* eof *)
+
diff --git a/minijazz/src/parser/parser.mly b/minijazz/src/parser/parser.mly
new file mode 100644
index 0000000..126ab36
--- /dev/null
+++ b/minijazz/src/parser/parser.mly
@@ -0,0 +1,185 @@
+%{
+
+open Ident
+open Static
+open Ast
+open Location
+open Misc
+
+let fresh_param () =
+  mk_static_exp (SVar ("_n"^(Misc.gen_symbol ())))
+
+%}
+
+%token INLINED ROM RAM WHERE END CONST PROBING
+%token LPAREN RPAREN COLON COMMA EQUAL REG OR XOR NAND AND POWER SLASH
+%token EOF RBRACKET LBRACKET GREATER LESS NOT SEMICOL PLUS MINUS STAR
+%token IF THEN ELSE LEQ DOT DOTDOT
+%token <string> NAME
+%token <string> STRING
+%token <int> INT
+%token <string> BOOL_INT
+%token <bool> BOOL
+
+%left DOT
+%left OR PLUS
+%left LEQ EQUAL
+%right MINUS
+%left NAND XOR AND
+%left STAR SLASH
+%right NOT REG
+%right POWER
+
+%start program
+%type <Ast.program> program
+
+%%
+
+/** Tools **/
+%inline slist(S, x)        : l=separated_list(S, x)                    {l}
+%inline snlist(S, x)       : l=separated_nonempty_list(S, x)           {l}
+%inline tuple(x)           : LPAREN h=x COMMA t=snlist(COMMA,x) RPAREN { h::t }
+%inline tag_option(P,x):
+  |/* empty */    { None }
+  | P v=x         { Some(v) }
+
+localize(x): y=x { y, (Loc($startpos(y),$endpos(y))) }
+
+program:
+  | c=const_decs n=node_decs EOF
+      { mk_program c n }
+
+const_decs: c=list(const_dec) {c}
+const_dec:
+  | CONST n=name EQUAL se=static_exp option(SEMICOL)
+      { mk_const_dec ~loc:(Loc($startpos,$endpos)) n se }
+
+name: n=NAME { n }
+
+ident:
+  | n=name { ident_of_string n }
+
+type_ident: LBRACKET se=static_exp RBRACKET { TBitArray se }
+
+node_name:
+  | n=name { reset_symbol_table (); n }
+
+node_decs: ns=list(node_dec) { ns }
+node_dec:
+  inlined=inlined_status n=node_name p=params LPAREN args=args RPAREN
+  EQUAL out=node_out WHERE b=block probes=probe_decls END WHERE option(SEMICOL)
+      { mk_node n (Loc ($startpos,$endpos)) inlined args out p b probes }
+
+node_out:
+  | a=arg { [a] }
+  | LPAREN out=args RPAREN { out }
+
+inlined_status:
+  | INLINED { Inlined }
+  | /*empty*/ { NotInlined }
+
+params:
+  | /*empty*/ { [] }
+  | LESS pl=snlist(COMMA,param) GREATER { pl }
+
+param:
+  n=NAME { mk_param n }
+
+args: vl=slist(COMMA, arg) { vl }
+
+arg:
+  | n=ident COLON t=type_ident { mk_var_dec n t }
+  | n=ident { mk_var_dec n TBit }
+
+block:
+  | eqs=equs { BEqs (eqs, []) }
+  | IF se=static_exp THEN thenb=block ELSE elseb=block END IF { BIf(se, thenb, elseb) }
+
+equs: eq=equ tl=equ_tail { eq::tl }
+equ_tail:
+  | /*empty*/ { [] }
+  | SEMICOL { [] }
+  | SEMICOL eq=equ tl=equ_tail { eq::tl }
+equ: p=pat EQUAL e=exp { mk_equation p e }
+
+pat:
+  | n=ident                              { Evarpat n }
+  | LPAREN p=snlist(COMMA, ident) RPAREN { Etuplepat p }
+
+static_exp: se=_static_exp { mk_static_exp ~loc:(Loc ($startpos,$endpos)) se }
+_static_exp :
+  | i=INT { SInt i }
+  | n=NAME { SVar n }
+  | LPAREN se=_static_exp RPAREN { se }
+  /*integer ops*/
+  | se1=static_exp POWER se2=static_exp { SBinOp(SPower, se1, se2) }
+  | se1=static_exp PLUS se2=static_exp { SBinOp(SAdd, se1, se2) }
+  | se1=static_exp MINUS se2=static_exp { SBinOp(SMinus, se1, se2) }
+  | se1=static_exp STAR se2=static_exp { SBinOp(SMult, se1, se2) }
+  | se1=static_exp SLASH se2=static_exp { SBinOp(SDiv, se1, se2) }
+  /*bool ops*/
+  | se1=static_exp EQUAL se2=static_exp { SBinOp(SEqual, se1, se2) }
+  | se1=static_exp LEQ se2=static_exp { SBinOp(SLeq, se1, se2) }
+
+exps: LPAREN e=slist(COMMA, exp) RPAREN {e}
+
+exp: e=_exp { mk_exp ~loc:(Loc ($startpos,$endpos)) e }
+_exp:
+  | e=_simple_exp  { e }
+  | c=const { Econst c }
+  | REG e=exp { Ereg e }
+  | n=NAME p=call_params a=exps { Ecall (n, p, a) }
+  | e1=exp PLUS e2=exp { Ecall ("or", [], [e1; e2]) }
+  | e1=exp OR e2=exp { Ecall ("or", [], [e1; e2]) }
+  | e1=exp AND e2=exp { Ecall ("and", [], [e1; e2]) }
+  | e1=exp POWER e2=exp { Ecall("xor", [], [e1; e2]) }
+  | e1=exp XOR e2=exp { Ecall ("xor", [], [e1; e2]) }
+  | e1=exp NAND e2=exp { Ecall ("nand", [], [e1; e2]) }
+  | NOT a=exp     { Ecall ("not", [], [a])}
+  | e1=exp DOT e2=exp
+    { Ecall("concat", [fresh_param(); fresh_param(); fresh_param ()], [e1; e2]) }
+  | e1=simple_exp LBRACKET idx=static_exp RBRACKET
+    { Ecall ("select", [idx; fresh_param()], [e1]) }
+  | e1=simple_exp LBRACKET low=static_exp DOTDOT high=static_exp RBRACKET
+    { Ecall("slice", [low; high; fresh_param()], [e1]) }
+  | e1=simple_exp LBRACKET low=static_exp DOTDOT RBRACKET
+    { let n = fresh_param () in
+      let high = mk_static_exp (SBinOp(SMinus, n, mk_static_exp (SInt 1))) in
+      Ecall("slice", [low; high; n], [e1]) }
+  | e1=simple_exp LBRACKET DOTDOT high=static_exp RBRACKET
+    {
+      let params = [mk_static_exp (SInt 0); high; fresh_param ()] in
+      Ecall("slice", params, [e1])
+    }
+  | ro=rom_or_ram LESS addr_size=static_exp
+    COMMA word_size=static_exp input_file=tag_option(COMMA, STRING) GREATER a=exps
+    { Emem(ro, addr_size, word_size, input_file, a) }
+
+simple_exp: e=_simple_exp { mk_exp ~loc:(Loc ($startpos,$endpos)) e }
+_simple_exp:
+  | n=ident                   { Evar n }
+  | LPAREN e=_exp RPAREN      { e }
+
+const:
+  | b=BOOL { VBit b }
+  | b=BOOL_INT { VBitArray (bool_array_of_string b) }
+  | i=INT
+    { match i with
+      | 0 -> VBit false
+      | 1 -> VBit true
+      | _ -> raise Parsing.Parse_error
+    }
+  | LBRACKET RBRACKET { VBitArray (Array.make 0 false) }
+
+rom_or_ram :
+  | ROM { MRom }
+  | RAM { MRam }
+
+call_params:
+  | /*empty*/ { [] }
+  | LESS pl=snlist(COMMA,static_exp) GREATER { pl }
+
+probe_decls:
+  | /*empty*/ { [] }
+  | PROBING l=separated_nonempty_list(COMMA, ident) { l }
+%%