use the new routine regexp.MustCompile to clean up some code that uses global regexps.

R=rsc, gri
CC=go-dev
http://go/go-review/1016025

3 files changed, 17 insertions, 45 deletions

diff --git a/src/pkg/go/doc/doc.go b/src/pkg/go/doc/doc.go
index a35ea8c68..0acf5cd58 100644
--- a/src/pkg/go/doc/doc.go
+++ b/src/pkg/go/doc/doc.go
@@ -245,28 +245,15 @@ func copyCommentList(list []*ast.Comment) []*ast.Comment {
 
 
 var (
-	// Regexp constructor needs threads - cannot use init expressions
-	bug_markers	*regexp.Regexp;
-	bug_content	*regexp.Regexp;
+	bug_markers = regexp.MustCompile("^/[/*][ \t]*BUG\\(.*\\):[ \t]*");	// BUG(uid):
+	bug_content = regexp.MustCompile("[^ \n\r\t]+");	// at least one non-whitespace char
 )
 
-func makeRex(s string) *regexp.Regexp {
-	re, err := regexp.Compile(s);
-	if err != nil {
-		panic("MakeRegexp ", s, " ", err.String());
-	}
-	return re;
-}
 
 // addFile adds the AST for a source file to the docReader.
 // Adding the same AST multiple times is a no-op.
 //
 func (doc *docReader) addFile(src *ast.File) {
-	if bug_markers == nil {
-		bug_markers = makeRex("^/[/*][ \t]*BUG\\(.*\\):[ \t]*");	// BUG(uid):
-		bug_content = makeRex("[^ \n\r\t]+");				// at least one non-whitespace char
-	}
-
 	// add package documentation
 	if src.Doc != nil {
 		// TODO(gri) This won't do the right thing if there is more
diff --git a/src/pkg/unicode/maketables.go b/src/pkg/unicode/maketables.go
index 73bfd2cbf..219140ea8 100644
--- a/src/pkg/unicode/maketables.go
+++ b/src/pkg/unicode/maketables.go
@@ -36,22 +36,22 @@ var url = flag.String("url",
 	"URL of Unicode database directory")
 var tablelist = flag.String("tables",
 	"all",
-	"comma-separated list of which tables to generate; can be letter");
+	"comma-separated list of which tables to generate; can be letter")
 var scriptlist = flag.String("scripts",
 	"all",
-	"comma-separated list of which script tables to generate");
+	"comma-separated list of which script tables to generate")
 var proplist = flag.String("props",
 	"all",
-	"comma-separated list of which property tables to generate");
+	"comma-separated list of which property tables to generate")
 var cases = flag.Bool("cases",
 	true,
-	"generate case tables");
+	"generate case tables")
 var test = flag.Bool("test",
 	false,
-	"test existing tables; can be used to compare web data with package data");
-var scriptRe *regexp.Regexp
+	"test existing tables; can be used to compare web data with package data")
 
-var die = log.New(os.Stderr, nil, "", log.Lexit|log.Lshortfile);
+var scriptRe = regexp.MustCompile(`([0-9A-F]+)(\.\.[0-9A-F]+)? *; ([A-Za-z_]+)`)
+var die = log.New(os.Stderr, nil, "", log.Lexit|log.Lshortfile)
 
 var category = map[string] bool{ "letter":true }	// Nd Lu etc. letter is a special case
 
@@ -125,8 +125,6 @@ var props = make(map[string] []Script)	// a property looks like a script; can sh
 
 var lastChar uint32 = 0
 
-const scriptParseExpression = `([0-9A-F]+)(\.\.[0-9A-F]+)? *; ([A-Za-z_]+)`
-
 // In UnicodeData.txt, some ranges are marked like this:
 //	3400;<CJK Ideograph Extension A, First>;Lo;0;L;;;;;N;;;;;
 //	4DB5;<CJK Ideograph Extension A, Last>;Lo;0;L;;;;;N;;;;;
@@ -166,7 +164,7 @@ func parseCategory(line string) (state State) {
 	switch char.category {
 	case "Nd":
 		// Decimal digit
-		v, err := strconv.Atoi(field[FNumericValue]);
+		_, err := strconv.Atoi(field[FNumericValue]);
 		if err != nil {
 			die.Log("U+%04x: bad numeric field: %s", point, err);
 		}
@@ -460,7 +458,7 @@ func fullCategoryTest(list []string) {
 }
 
 func verifyRange(name string, inCategory Op, table []unicode.Range) {
-	for i, c := range chars {
+	for i := range chars {
 		web := inCategory(i);
 		pkg := unicode.Is(table, i);
 		if web != pkg {
@@ -499,7 +497,7 @@ func parseScript(line string, scripts map[string] []Script) {
 	}
 	name := matches[3];
 	s, ok := scripts[name];
-	if len(s) == cap(s) {
+	if !ok || len(s) == cap(s) {
 		ns := make([]Script, len(s), len(s)+100);
 		for i, sc := range s {
 			ns[i] = sc
@@ -532,7 +530,7 @@ func fullScriptTest(list []string, installed map[string] []unicode.Range, script
 		if _, ok := scripts[name]; !ok {
 			die.Log("unknown script", name);
 		}
-		r, ok := installed[name];
+		_, ok := installed[name];
 		if !ok {
 			die.Log("unknown table", name);
 		}
@@ -564,10 +562,6 @@ func printScriptOrProperty(doProps bool) {
 		return
 	}
 	var err os.Error;
-	scriptRe, err = regexp.Compile(scriptParseExpression);
-	if err != nil {
-		die.Log("re error:", err)
-	}
 	resp, _, err := http.Get(*url + file);
 	if err != nil {
 		die.Log(err);
@@ -801,7 +795,7 @@ func printCases() {
 
 	var startState *caseState;	// the start of a run; nil for not active
 	var prevState = &caseState{};	// the state of the previous character
-	for i, c := range chars {
+	for i := range chars {
 		state := getCaseState(i);
 		if state.adjacent(prevState) {
 			prevState = state;
diff --git a/test/bench/regex-dna.go b/test/bench/regex-dna.go
index f42efc6d4..f04422c3f 100644
--- a/test/bench/regex-dna.go
+++ b/test/bench/regex-dna.go
@@ -43,15 +43,6 @@ import (
 	"strings";
 )
 
-func compile(s string) *regexp.Regexp {
-	r, err := regexp.Compile(s);
-	if err != nil {
-		fmt.Fprintf(os.Stderr, "can't compile pattern %q: %s\n", s, err);
-		os.Exit(2);
-	}
-	return r;
-}
-
 var variants = []string {
 	"agggtaaa|tttaccct",
 	"[cgt]gggtaaa|tttaccc[acg]",
@@ -83,7 +74,7 @@ var substs = [] Subst {
 }
 
 func countMatches(pat string, bytes []byte) int {
-	re := compile(pat);
+	re := regexp.MustCompile(pat);
 	n := 0;
 	for {
 		e := re.Execute(bytes);
@@ -104,13 +95,13 @@ func main() {
 	}
 	ilen := len(bytes);
 	// Delete the comment lines and newlines
-	bytes = compile("(>[^\n]+)?\n").ReplaceAll(bytes, []byte{});
+	bytes = regexp.MustCompile("(>[^\n]+)?\n").ReplaceAll(bytes, []byte{});
 	clen := len(bytes);
 	for _, s := range variants {
 		fmt.Printf("%s %d\n", s, countMatches(s, bytes));
 	}
 	for _, sub := range substs {
-		bytes = compile(sub.pat).ReplaceAll(bytes, strings.Bytes(sub.repl));
+		bytes = regexp.MustCompile(sub.pat).ReplaceAll(bytes, strings.Bytes(sub.repl));
 	}
 	fmt.Printf("\n%d\n%d\n%d\n", ilen, clen, len(bytes));
 }