text/template: efficient reporting of line numbers

Instead of scanning the text to count newlines, which is n², keep track as we go
and store the line number in the token.

benchmark                 old ns/op      new ns/op     delta
BenchmarkParseLarge-4     1589721293     38783310      -97.56%

Fixes #17851

Change-Id: Ieaf89a35e371b405ad92e38baa1e3fa98d18cfb4
Reviewed-on: https://go-review.googlesource.com/32923Reviewed-by: Robert Griesemer <gri@golang.org>

src/text/template/exec_test.go

@@ -1152,7 +1152,7 @@ func TestUnterminatedStringError(t *testing.T) {
 		t.Fatal("expected error")
 	}
 	str := err.Error()
-	if !strings.Contains(str, "X:3: unexpected unterminated raw quoted strin") {
+	if !strings.Contains(str, "X:3: unexpected unterminated raw quoted string") {
 		t.Fatalf("unexpected error: %s", str)
 	}
 }

src/text/template/parse/lex.go

@@ -13,9 +13,10 @@ import (
 
 // item represents a token or text string returned from the scanner.
 type item struct {
-	typ itemType // The type of this item.
-	pos Pos      // The starting position, in bytes, of this item in the input string.
-	val string   // The value of this item.
+	typ  itemType // The type of this item.
+	pos  Pos      // The starting position, in bytes, of this item in the input string.
+	val  string   // The value of this item.
+	line int      // The line number at the start of this item.
 }
 
 func (i item) String() string {
@@ -116,6 +117,7 @@ type lexer struct {
 	lastPos    Pos       // position of most recent item returned by nextItem
 	items      chan item // channel of scanned items
 	parenDepth int       // nesting depth of ( ) exprs
+	line       int       // 1+number of newlines seen
 }
 
 // next returns the next rune in the input.
@@ -127,6 +129,9 @@ func (l *lexer) next() rune {
 	r, w := utf8.DecodeRuneInString(l.input[l.pos:])
 	l.width = Pos(w)
 	l.pos += l.width
+	if r == '\n' {
+		l.line++
+	}
 	return r
 }
 
@@ -140,11 +145,20 @@ func (l *lexer) peek() rune {
 // backup steps back one rune. Can only be called once per call of next.
 func (l *lexer) backup() {
 	l.pos -= l.width
+	// Correct newline count.
+	if l.width == 1 && l.input[l.pos] == '\n' {
+		l.line--
+	}
 }
 
 // emit passes an item back to the client.
 func (l *lexer) emit(t itemType) {
-	l.items <- item{t, l.start, l.input[l.start:l.pos]}
+	l.items <- item{t, l.start, l.input[l.start:l.pos], l.line}
+	// Some items contain text internally. If so, count their newlines.
+	switch t {
+	case itemText, itemRawString, itemLeftDelim, itemRightDelim:
+		l.line += strings.Count(l.input[l.start:l.pos], "\n")
+	}
 	l.start = l.pos
 }
 
@@ -169,17 +183,10 @@ func (l *lexer) acceptRun(valid string) {
 	l.backup()
 }
 
-// lineNumber reports which line we're on, based on the position of
-// the previous item returned by nextItem. Doing it this way
-// means we don't have to worry about peek double counting.
-func (l *lexer) lineNumber() int {
-	return 1 + strings.Count(l.input[:l.lastPos], "\n")
-}
-
 // errorf returns an error token and terminates the scan by passing
 // back a nil pointer that will be the next state, terminating l.nextItem.
 func (l *lexer) errorf(format string, args ...interface{}) stateFn {
-	l.items <- item{itemError, l.start, fmt.Sprintf(format, args...)}
+	l.items <- item{itemError, l.start, fmt.Sprintf(format, args...), l.line}
 	return nil
 }
 
@@ -212,6 +219,7 @@ func lex(name, input, left, right string) *lexer {
 		leftDelim:  left,
 		rightDelim: right,
 		items:      make(chan item),
+		line:       1,
 	}
 	go l.run()
 	return l
@@ -602,10 +610,14 @@ Loop:
 
 // lexRawQuote scans a raw quoted string.
 func lexRawQuote(l *lexer) stateFn {
+	startLine := l.line
 Loop:
 	for {
 		switch l.next() {
 		case eof:
+			// Restore line number to location of opening quote.
+			// We will error out so it's ok just to overwrite the field.
+			l.line = startLine
 			return l.errorf("unterminated raw quoted string")
 		case '`':
 			break Loop

src/text/template/parse/parse.go

@@ -157,7 +157,7 @@ func (t *Tree) ErrorContext(n Node) (location, context string) {
 // errorf formats the error and terminates processing.
 func (t *Tree) errorf(format string, args ...interface{}) {
 	t.Root = nil
-	format = fmt.Sprintf("template: %s:%d: %s", t.ParseName, t.lex.lineNumber(), format)
+	format = fmt.Sprintf("template: %s:%d: %s", t.ParseName, t.lex.line, format)
 	panic(fmt.Errorf(format, args...))
 }
 
@@ -376,15 +376,17 @@ func (t *Tree) action() (n Node) {
 		return t.withControl()
 	}
 	t.backup()
+	token := t.peek()
 	// Do not pop variables; they persist until "end".
-	return t.newAction(t.peek().pos, t.lex.lineNumber(), t.pipeline("command"))
+	return t.newAction(token.pos, token.line, t.pipeline("command"))
 }
 
 // Pipeline:
 //	declarations? command ('|' command)*
 func (t *Tree) pipeline(context string) (pipe *PipeNode) {
 	var decl []*VariableNode
-	pos := t.peekNonSpace().pos
+	token := t.peekNonSpace()
+	pos := token.pos
 	// Are there declarations?
 	for {
 		if v := t.peekNonSpace(); v.typ == itemVariable {
@@ -413,7 +415,7 @@ func (t *Tree) pipeline(context string) (pipe *PipeNode) {
 		}
 		break
 	}
-	pipe = t.newPipeline(pos, t.lex.lineNumber(), decl)
+	pipe = t.newPipeline(pos, token.line, decl)
 	for {
 		switch token := t.nextNonSpace(); token.typ {
 		case itemRightDelim, itemRightParen:
@@ -450,7 +452,6 @@ func (t *Tree) checkPipeline(pipe *PipeNode, context string) {
 
 func (t *Tree) parseControl(allowElseIf bool, context string) (pos Pos, line int, pipe *PipeNode, list, elseList *ListNode) {
 	defer t.popVars(len(t.vars))
-	line = t.lex.lineNumber()
 	pipe = t.pipeline(context)
 	var next Node
 	list, next = t.itemList()
@@ -479,7 +480,7 @@ func (t *Tree) parseControl(allowElseIf bool, context string) (pos Pos, line int
 			t.errorf("expected end; found %s", next)
 		}
 	}
-	return pipe.Position(), line, pipe, list, elseList
+	return pipe.Position(), pipe.Line, pipe, list, elseList
 }
 
 // If:
@@ -521,9 +522,10 @@ func (t *Tree) elseControl() Node {
 	peek := t.peekNonSpace()
 	if peek.typ == itemIf {
 		// We see "{{else if ... " but in effect rewrite it to {{else}}{{if ... ".
-		return t.newElse(peek.pos, t.lex.lineNumber())
+		return t.newElse(peek.pos, peek.line)
 	}
-	return t.newElse(t.expect(itemRightDelim, "else").pos, t.lex.lineNumber())
+	token := t.expect(itemRightDelim, "else")
+	return t.newElse(token.pos, token.line)
 }
 
 // Block:
@@ -550,7 +552,7 @@ func (t *Tree) blockControl() Node {
 	block.add()
 	block.stopParse()
 
-	return t.newTemplate(token.pos, t.lex.lineNumber(), name, pipe)
+	return t.newTemplate(token.pos, token.line, name, pipe)
 }
 
 // Template:
@@ -567,7 +569,7 @@ func (t *Tree) templateControl() Node {
 		// Do not pop variables; they persist until "end".
 		pipe = t.pipeline(context)
 	}
-	return t.newTemplate(token.pos, t.lex.lineNumber(), name, pipe)
+	return t.newTemplate(token.pos, token.line, name, pipe)
 }
 
 func (t *Tree) parseTemplateName(token item, context string) (name string) {

src/text/template/parse/parse_test.go

@@ -484,3 +484,37 @@ func TestBlock(t *testing.T) {
 		t.Errorf("inner template = %q, want %q", g, w)
 	}
 }
+
+func TestLineNum(t *testing.T) {
+	const count = 100
+	text := strings.Repeat("{{printf 1234}}\n", count)
+	tree, err := New("bench").Parse(text, "", "", make(map[string]*Tree), builtins)
+	if err != nil {
+		t.Fatal(err)
+	}
+	// Check the line numbers. Each line is an action containing a template, followed by text.
+	// That's two nodes per line.
+	nodes := tree.Root.Nodes
+	for i := 0; i < len(nodes); i += 2 {
+		line := 1 + i/2
+		// Action first.
+		action := nodes[i].(*ActionNode)
+		if action.Line != line {
+			t.Fatalf("line %d: action is line %d", line, action.Line)
+		}
+		pipe := action.Pipe
+		if pipe.Line != line {
+			t.Fatalf("line %d: pipe is line %d", line, pipe.Line)
+		}
+	}
+}
+
+func BenchmarkParseLarge(b *testing.B) {
+	text := strings.Repeat("{{1234}}\n", 10000)
+	for i := 0; i < b.N; i++ {
+		_, err := New("bench").Parse(text, "", "", make(map[string]*Tree), builtins)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}