Add block-comment support, various additional operators. Rustc can lex itself now.

2 files changed, 164 insertions, 60 deletions

diff --git a/src/comp/fe/lexer.rs b/src/comp/fe/lexer.rs
index 0deb5401..95aac8c5 100644
--- a/src/comp/fe/lexer.rs
+++ b/src/comp/fe/lexer.rs
@@ -238,16 +238,48 @@ fn consume_any_whitespace(reader rdr) {
 
 fn consume_any_line_comment(reader rdr) {
     if (rdr.curr() == '/') {
-        if (rdr.next() == '/') {
-            while (rdr.curr() != '\n') {
+        alt (rdr.next()) {
+            case ('/') {
+                while (rdr.curr() != '\n') {
+                    rdr.bump();
+                }
+                // Restart whitespace munch.
+                be consume_any_whitespace(rdr);
+            }
+            case ('*') {
                 rdr.bump();
+                rdr.bump();
+                be consume_block_comment(rdr);
+            }
+            case (_) {
+                ret;
             }
-            // Restart whitespace munch.
-            be consume_any_whitespace(rdr);
         }
     }
 }
 
+
+fn consume_block_comment(reader rdr) {
+    let int level = 1;
+    while (level > 0) {
+        if (rdr.curr() == '/' && rdr.next() == '*') {
+            rdr.bump();
+            rdr.bump();
+            level += 1;
+        } else {
+            if (rdr.curr() == '*' && rdr.next() == '/') {
+                rdr.bump();
+                rdr.bump();
+                level -= 1;
+            } else {
+                rdr.bump();
+            }
+        }
+    }
+    // restart whitespace munch.
+    be consume_any_whitespace(rdr);
+}
+
 fn next_token(reader rdr) -> token.token {
     auto accum_str = "";
     auto accum_int = 0;
@@ -310,18 +342,19 @@ fn next_token(reader rdr) -> token.token {
     }
 
 
-    fn op_or_opeq(reader rdr, token.op op) -> token.token {
+    fn binop(reader rdr, token.binop op) -> token.token {
         rdr.bump();
         if (rdr.next() == '=') {
             rdr.bump();
-            ret token.OPEQ(op);
+            ret token.BINOPEQ(op);
         } else {
-            ret token.OP(op);
+            ret token.BINOP(op);
         }
     }
 
     alt (c) {
         // One-byte tokens.
+        case (':') { rdr.bump(); ret token.COLON(); }
         case (';') { rdr.bump(); ret token.SEMI(); }
         case (',') { rdr.bump(); ret token.COMMA(); }
         case ('.') { rdr.bump(); ret token.DOT(); }
@@ -334,16 +367,74 @@ fn next_token(reader rdr) -> token.token {
         case ('@') { rdr.bump(); ret token.AT(); }
         case ('#') { rdr.bump(); ret token.POUND(); }
         case ('_') { rdr.bump(); ret token.UNDERSCORE(); }
+        case ('~') { rdr.bump(); ret token.TILDE(); }
+
 
         // Multi-byte tokens.
         case ('=') {
-            if (rdr.next() == '=') {
-                rdr.bump();
+            rdr.bump();
+            if (rdr.curr() == '=') {
                 rdr.bump();
-                ret token.OP(token.EQEQ());
+                ret token.EQEQ();
             } else {
+                ret token.EQ();
+            }
+        }
+
+        case ('!') {
+            rdr.bump();
+            if (rdr.curr() == '=') {
                 rdr.bump();
-                ret token.OP(token.EQ());
+                ret token.NE();
+            } else {
+                ret token.NOT();
+            }
+        }
+
+        case ('<') {
+            rdr.bump();
+            alt (rdr.curr()) {
+                case ('=') {
+                    rdr.bump();
+                    ret token.LE();
+                }
+                case ('<') {
+                    ret binop(rdr, token.LSL());
+                }
+                case ('-') {
+                    rdr.bump();
+                    ret token.LARROW();
+                }
+                case ('|') {
+                    rdr.bump();
+                    ret token.SEND();
+                }
+                case (_) {
+                    ret token.LT();
+                }
+            }
+        }
+
+        case ('>') {
+            rdr.bump();
+            alt (rdr.curr()) {
+                case ('=') {
+                    rdr.bump();
+                    ret token.GE();
+                }
+
+                case ('>') {
+                    if (rdr.next() == '>') {
+                        rdr.bump();
+                        ret binop(rdr, token.ASR());
+                    } else {
+                        ret binop(rdr, token.LSR());
+                    }
+                }
+
+                case (_) {
+                    ret token.GT();
+                }
             }
         }
 
@@ -426,7 +517,7 @@ fn next_token(reader rdr) -> token.token {
                 rdr.bump();
                 ret token.RARROW();
             } else {
-                ret op_or_opeq(rdr, token.MINUS());
+                ret binop(rdr, token.MINUS());
             }
         }
 
@@ -434,34 +525,40 @@ fn next_token(reader rdr) -> token.token {
             if (rdr.next() == '&') {
                 rdr.bump();
                 rdr.bump();
-                ret token.OP(token.ANDAND());
+                ret token.ANDAND();
             } else {
-                ret op_or_opeq(rdr, token.AND());
+                ret binop(rdr, token.AND());
+            }
+        }
+
+        case ('|') {
+            if (rdr.next() == '|') {
+                rdr.bump();
+                rdr.bump();
+                ret token.OROR();
+            } else {
+                ret binop(rdr, token.OR());
             }
         }
 
         case ('+') {
-            ret op_or_opeq(rdr, token.PLUS());
+            ret binop(rdr, token.PLUS());
         }
 
         case ('*') {
-            ret op_or_opeq(rdr, token.STAR());
+            ret binop(rdr, token.STAR());
         }
 
         case ('/') {
-            ret op_or_opeq(rdr, token.STAR());
-        }
-
-        case ('!') {
-            ret op_or_opeq(rdr, token.NOT());
+            ret binop(rdr, token.STAR());
         }
 
         case ('^') {
-            ret op_or_opeq(rdr, token.CARET());
+            ret binop(rdr, token.CARET());
         }
 
         case ('%') {
-            ret op_or_opeq(rdr, token.PERCENT());
+            ret binop(rdr, token.PERCENT());
         }
 
     }
diff --git a/src/comp/fe/token.rs b/src/comp/fe/token.rs
index 2ec492fa..e9e6f222 100644
--- a/src/comp/fe/token.rs
+++ b/src/comp/fe/token.rs
@@ -3,12 +3,21 @@ import util.common.ty_mach_to_str;
 import std._int;
 import std._uint;
 
-type op = tag
+type binop = tag
     (PLUS(),
      MINUS(),
      STAR(),
      SLASH(),
      PERCENT(),
+     CARET(),
+     AND(),
+     OR(),
+     LSL(),
+     LSR(),
+     ASR());
+
+type token = tag
+    (/* Expression-operator symbols. */
      EQ(),
      LT(),
      LE(),
@@ -16,20 +25,14 @@ type op = tag
      NE(),
      GE(),
      GT(),
-     NOT(),
-     TILDE(),
-     CARET(),
-     AND(),
      ANDAND(),
-     OR(),
      OROR(),
-     LSL(),
-     LSR(),
-     ASR());
+     NOT(),
+     TILDE(),
+
+     BINOP(binop),
+     BINOPEQ(binop),
 
-type token = tag
-    (OP(op),
-     OPEQ(op),
      AS(),
      WITH(),
 
@@ -152,40 +155,44 @@ type token = tag
      BRACEQUOTE(str),
      EOF());
 
-fn op_to_str(op o) -> str {
+fn binop_to_str(binop o) -> str {
     alt (o) {
-         case (PLUS()) { ret "+"; }
-         case (MINUS()) { ret "-"; }
-         case (STAR()) { ret "*"; }
-         case (SLASH()) { ret "/"; }
-         case (PERCENT()) { ret "%"; }
-         case (EQ()) { ret "="; }
-         case (LT()) { ret "<"; }
-         case (LE()) { ret "<="; }
-         case (EQEQ()) { ret "=="; }
-         case (NE()) { ret "!="; }
-         case (GE()) { ret ">="; }
-         case (GT()) { ret ">"; }
-         case (NOT()) { ret "!"; }
-         case (TILDE()) { ret "~"; }
-         case (CARET()) { ret "^"; }
-         case (AND()) { ret "&"; }
-         case (ANDAND()) { ret "&&"; }
-         case (OR()) { ret "|"; }
-         case (OROR()) { ret "||"; }
-         case (LSL()) { ret "<<"; }
-         case (LSR()) { ret ">>"; }
-         case (ASR()) { ret ">>>"; }
+        case (PLUS()) { ret "+"; }
+        case (MINUS()) { ret "-"; }
+        case (STAR()) { ret "*"; }
+        case (SLASH()) { ret "/"; }
+        case (PERCENT()) { ret "%"; }
+        case (CARET()) { ret "^"; }
+        case (AND()) { ret "&"; }
+        case (OR()) { ret "|"; }
+        case (LSL()) { ret "<<"; }
+        case (LSR()) { ret ">>"; }
+        case (ASR()) { ret ">>>"; }
     }
 }
 
 fn to_str(token t) -> str {
     alt (t) {
-        case (OP(op)) { ret op_to_str(op); }
-        case (OPEQ(op)) { ret op_to_str(op) + "="; }
+
+        case (EQ()) { ret "="; }
+        case (LT()) { ret "<"; }
+        case (LE()) { ret "<="; }
+        case (EQEQ()) { ret "=="; }
+        case (NE()) { ret "!="; }
+        case (GE()) { ret ">="; }
+        case (GT()) { ret ">"; }
+        case (NOT()) { ret "!"; }
+        case (TILDE()) { ret "~"; }
+        case (OROR()) { ret "||"; }
+        case (ANDAND()) { ret "&&"; }
+
+        case (BINOP(op)) { ret binop_to_str(op); }
+        case (BINOPEQ(op)) { ret binop_to_str(op) + "="; }
+
         case (AS()) { ret "as"; }
         case (WITH()) { ret "with"; }
 
+
         /* Structural symbols */
         case (AT()) { ret "@"; }
         case (DOT()) { ret "."; }