* pretty printing snapshot: towards printing comments nicely

- implemented elastic tabstops algorithm, now correct and documented
- first cut at printing comments (use -comments flag, disabled for now)
- struct field types are now aligned (using elastic tab stops)
- needs more fine-tuning

* fixed a bug in test script
* added quick smoke test to makefile and invoke it in run.bash
  instead of the full test

R=r
OCL=19220
CL=19220

src/run.bash

@@ -32,7 +32,7 @@ time make
 (xcd ../usr/gri/pretty
 make clean
 time make
-make test
+make smoketest
 ) || exit $?
 
 (xcd ../test

usr/gri/pretty/Makefile

@@ -11,6 +11,9 @@ pretty: pretty.6
 test: pretty
 	./test.sh
 
+smoketest: pretty
+	./test.sh parser.go
+
 install: pretty
 	cp pretty $(HOME)/bin/pretty
 

usr/gri/pretty/ast.go

@@ -43,6 +43,11 @@ func (p *List) at(i int) Any {
 }
 
 
+func (p *List) last() Any {
+	return p.a[len(p.a) - 1];
+}
+
+
 func (p *List) set(i int, x Any) {
 	p.a[i] = x;
 }

usr/gri/pretty/parser.go

@@ -77,13 +77,8 @@ func (P *Parser) Next0() {
 
 
 func (P *Parser) Next() {
-	P.Next0();
-	if P.tok == Scanner.COMMENT {
-		pos, s := P.pos, P.val;
-		for P.Next0(); P.tok == Scanner.COMMENT; P.Next0() {
-			s += P.val;
-		}
-		P.comments.Add(AST.NewComment(pos, s));
+	for P.Next0(); P.tok == Scanner.COMMENT; P.Next0() {
+		P.comments.Add(AST.NewComment(P.pos, P.val));
 	}
 }
 

usr/gri/pretty/scanner.go

@@ -518,22 +518,29 @@ func (S *Scanner) Expect(ch int) {
 }
 
 
-func (S *Scanner) SkipWhitespace() {
-	for S.ch == ' ' || S.ch == '\r' {
+func (S *Scanner) SkipWhitespace() int {
+	pos := -1;  // no new line position yet
+	
+	if S.chpos == 0 {
+		// file beginning is always start of a new line
+		pos = 0;
+	}
+	
+	for {
+		switch S.ch {
+		case '\t', '\r', ' ':  // nothing to do
+		case '\n': pos = S.pos;  // remember start of new line
+		default: goto exit;
+		}
 		S.Next();
 	}
-}
-
 
-func (S *Scanner) ScanWhitespace() string {
-	// first char ('\n' or '\t', 1 byte) already consumed
-	pos := S.chpos - 1;
-	S.SkipWhitespace();
-	return S.src[pos : S.chpos];
+exit:
+	return pos;
 }
 
 
-func (S *Scanner) ScanComment() string {
+func (S *Scanner) ScanComment(nlpos int) string {
 	// first '/' already consumed
 	pos := S.chpos - 1;
 	
@@ -543,6 +550,9 @@ func (S *Scanner) ScanComment() string {
 		for S.ch >= 0 {
 			S.Next();
 			if S.ch == '\n' {
+				// '\n' terminates comment but we do not include
+				// it in the comment (otherwise we cannot see the
+				// start of a newline in SkipWhitespace()).
 				goto exit;
 			}
 		}
@@ -554,6 +564,7 @@ func (S *Scanner) ScanComment() string {
 			ch := S.ch;
 			S.Next();
 			if ch == '*' && S.ch == '/' {
+				S.Next();
 				goto exit;
 			}
 		}
@@ -562,7 +573,6 @@ func (S *Scanner) ScanComment() string {
 	S.Error(pos, "comment not terminated");
 
 exit:
-	S.Next();
 	comment := S.src[pos : S.chpos];
 
 	if S.testmode {
@@ -586,6 +596,16 @@ exit:
 		}
 	}
 	
+	if nlpos < 0 {
+		// not only whitespace before comment on this line
+		comment = " " + comment;
+	} else if nlpos == pos {
+		// comment starts at the beginning of the line
+		comment = "\n" + comment;
+	} else {
+		// only whitespace before comment on this line
+		comment = "\t" + comment;
+	}
 	return comment;
 }
 
@@ -815,20 +835,17 @@ func (S *Scanner) Select4(tok0, tok1, ch2, tok2, tok3 int) int {
 
 
 func (S *Scanner) Scan() (pos, tok int, val string) {
-	S.SkipWhitespace();
+	nlpos := S.SkipWhitespace();
 	
-	ch := S.ch;
-	pos = S.chpos;
-	tok = ILLEGAL;
+	pos, tok = S.chpos, ILLEGAL;
 	
-	switch {
+	switch ch := S.ch; {
 	case is_letter(ch): tok, val = S.ScanIdentifier();
 	case digit_val(ch) < 10: tok, val = S.ScanNumber(false);
 	default:
 		S.Next();  // always make progress
 		switch ch {
 		case -1: tok = EOF;
-		case '\n', '\t': tok, val = COMMENT, S.ScanWhitespace();
 		case '"': tok, val = STRING, S.ScanString();
 		case '\'': tok, val = INT, S.ScanChar();
 		case '`': tok, val = STRING, S.ScanRawString();
@@ -858,7 +875,7 @@ func (S *Scanner) Scan() (pos, tok int, val string) {
 		case '*': tok = S.Select2(MUL, MUL_ASSIGN);
 		case '/':
 			if S.ch == '/' || S.ch == '*' {
-				tok, val = COMMENT, S.ScanComment();
+				tok, val = COMMENT, S.ScanComment(nlpos);
 			} else {
 				tok = S.Select2(QUO, QUO_ASSIGN);
 			}

usr/gri/pretty/selftest2.go

+// Copyright 2009 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.
+
+package main
+
+import Fmt "fmt"
+
+
+type T struct {
+	x, y int;
+	s string;
+	next_t *T
+}
+
+
+var (
+	A = 5;
+	a, b, c int = 0, 0, 0;
+	foo = "foo";
+)
+
+
+func main() {
+// the prolog
+	for i := 0; i <= 10 /* limit */; i++ {
+		println(i);  // the index
+		println(i + 1);  // the index + 1
+		println(i + 1000);  // the index + 1000
+		println();
+	}
+// the epilog
+	println("foo");  // foo
+	println("foobar");  // foobar
+var x int;
+}

usr/gri/pretty/test.sh

@@ -96,11 +96,11 @@ valid() {
 runtest() {
 	#echo "Testing silent mode"
 	cleanup
-	$1 silent
+	$1 silent $2
 
 	#echo "Testing idempotency"
 	cleanup
-	$1 idempotent
+	$1 idempotent $2
 }