gc: initial BOM is legal.

Fixes #4040.

R=rsc
CC=golang-dev
https://golang.org/cl/6497098

diff --git a/src/cmd/gc/lex.c b/src/cmd/gc/lex.c
index f07a23c150f294dbd83fe4227290ed8576bab9f9..0788a615146b9303b58badf052fc61f4390dd88b 100644 (file)
--- a/src/cmd/gc/lex.c
+++ b/src/cmd/gc/lex.c
@@ -30,6 +30,8 @@ static void   addidir(char*);
 static int     getlinepragma(void);
 static char *goos, *goarch, *goroot;
 
+#define        BOM     0xFEFF
+
 // Compiler experiments.
 // These are controlled by the GOEXPERIMENT environment
 // variable recorded when the compiler is built.
@@ -319,6 +321,10 @@ main(int argc, char *argv[])
                curio.peekc1 = 0;
                curio.nlsemi = 0;
 
+               // Skip initial BOM if present.
+               if(Bgetrune(curio.bin) != BOM)
+                       Bungetrune(curio.bin);
+
                block = 1;
                iota = -1000000;
 
@@ -1200,7 +1206,7 @@ talph:
                        rune = getr();
                        // 0xb7 · is used for internal names
                        if(!isalpharune(rune) && !isdigitrune(rune) && (importpkg == nil || rune != 0xb7))
-                               yyerror("invalid identifier character 0x%ux", rune);
+                               yyerror("invalid identifier character U+%04x", rune);
                        cp += runetochar(cp, &rune);
                } else if(!yy_isalnum(c) && c != '_')
                        break;
@@ -1583,6 +1589,10 @@ loop:
        if(!fullrune(str, i))
                goto loop;
        c = chartorune(&rune, str);
+       if(rune == BOM) {
+               lineno = lexlineno;
+               yyerror("Unicode (UTF-8) BOM in middle of file");
+       }
        if(rune == Runeerror && c == 1) {
                lineno = lexlineno;
                yyerror("illegal UTF-8 sequence");

diff --git a/test/bom.go b/test/bom.go
new file mode 100644 (file)
index 0000000..37f73bc
--- /dev/null
+++ b/test/bom.go
@@ -0,0 +1,26 @@
+// runoutput
+
+// Copyright 2011 The Go Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Test source file beginning with a byte order mark.
+
+package main
+
+import (
+       "fmt"
+       "strings"
+)
+
+func main() {
+       prog = strings.Replace(prog, "BOM", "\uFEFF", -1)
+       fmt.Print(prog)
+}
+
+var prog = `BOM
+package main
+
+func main() {
+}
+`

diff --git a/test/bombad.go b/test/bombad.go
new file mode 100644 (file)
index 0000000..b894d9b
--- /dev/null
+++ b/test/bombad.go
@@ -0,0 +1,18 @@
+// errorcheck
+
+// Copyright 2012 The Go Authors.  All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Here for reference, but hard to test automatically
+// because the BOM muddles the
+// processing done by ../run.
+
+package main
+
+func main() {
+       // There's a bom here.       // ERROR "BOM"
+       // And here. // ERROR "BOM"
+       /* And here.*/       // ERROR "BOM"
+       println("hi there") // and here      // ERROR "BOM"
+}