golang/0050-Backport-html-template-support-HTML-like-comments-in.patch
2023-12-05 11:44:14 +08:00

268 lines
10 KiB
Diff

From fc0e2bd566a8e8777acc824211f6027184840a49 Mon Sep 17 00:00:00 2001
From: Roland Shoemaker <bracewell@google.com>
Date: Fri, 4 Aug 2023 03:24:13 +0800
Subject: [PATCH 1/2] [Backport] html/template: support HTML-like comments in
script contexts
Reference: https://open.codehub.huawei.com/OpenSourceCenter/golang/go/merge_requests/89
Conflict: NA
Per Appendix B.1.1 of the ECMAScript specification, support HTML-like
comments in script contexts. Also per section 12.5, support hashbang
comments. This brings our parsing in-line with how browsers treat these
comment types.
Thanks to Takeshi Kaneko (GMO Cybersecurity by Ierae, Inc.) for
reporting this issue.
Note: The upstream does not submit this change to go1.17 according to the rules of MinorReleases.
Corego3.x are based on go1.17.8. Therefore, it need to submit the change to corego3.x.
Edited-by: machangwang m00509938
Fixes #62196
Fixes #62395
Fixes CVE-2023-39318
Change-Id: Id512702c5de3ae46cf648e268cb10e1eb392a181
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1976593
Run-TryBot: Roland Shoemaker <bracewell@google.com>
Reviewed-by: Tatiana Bradley <tatianabradley@google.com>
Reviewed-by: Damien Neil <dneil@google.com>
Reviewed-by: Dmitri Shuralyov <dmitshur@google.com>
Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/2014620
Reviewed-on: https://go-review.googlesource.com/c/go/+/526098
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Gopher Robot <gobot@golang.org>
Signed-off-by: Ma Chang Wang machangwang@huawei.com
---
src/html/template/context.go | 6 ++-
src/html/template/escape.go | 5 +-
src/html/template/escape_test.go | 10 ++++
src/html/template/state_string.go | 26 +++++-----
src/html/template/transition.go | 80 ++++++++++++++++++++-----------
5 files changed, 84 insertions(+), 43 deletions(-)
diff --git a/src/html/template/context.go b/src/html/template/context.go
index a67b5a7871..472b5a9772 100644
--- a/src/html/template/context.go
+++ b/src/html/template/context.go
@@ -124,6 +124,10 @@ const (
stateJSBlockCmt
// stateJSLineCmt occurs inside a JavaScript // line comment.
stateJSLineCmt
+ // stateJSHTMLOpenCmt occurs inside a JavaScript <!-- HTML-like comment.
+ stateJSHTMLOpenCmt
+ // stateJSHTMLCloseCmt occurs inside a JavaScript --> HTML-like comment.
+ stateJSHTMLCloseCmt
// stateCSS occurs inside a <style> element or style attribute.
stateCSS
// stateCSSDqStr occurs inside a CSS double quoted string.
@@ -151,7 +155,7 @@ const (
// authors & maintainers, not for end-users or machines.
func isComment(s state) bool {
switch s {
- case stateHTMLCmt, stateJSBlockCmt, stateJSLineCmt, stateCSSBlockCmt, stateCSSLineCmt:
+ case stateHTMLCmt, stateJSBlockCmt, stateJSLineCmt, stateJSHTMLOpenCmt, stateJSHTMLCloseCmt, stateCSSBlockCmt, stateCSSLineCmt:
return true
}
return false
diff --git a/src/html/template/escape.go b/src/html/template/escape.go
index bdccc65a57..3249c67008 100644
--- a/src/html/template/escape.go
+++ b/src/html/template/escape.go
@@ -700,9 +700,12 @@ func (e *escaper) escapeText(c context, n *parse.TextNode) context {
if c.state != c1.state && isComment(c1.state) && c1.delim == delimNone {
// Preserve the portion between written and the comment start.
cs := i1 - 2
- if c1.state == stateHTMLCmt {
+ if c1.state == stateHTMLCmt || c1.state == stateJSHTMLOpenCmt {
// "<!--" instead of "/*" or "//"
cs -= 2
+ } else if c1.state == stateJSHTMLCloseCmt {
+ // "-->" instead of "/*" or "//"
+ cs -= 1
}
b.Write(s[written:cs])
written = i1
diff --git a/src/html/template/escape_test.go b/src/html/template/escape_test.go
index 3e17aee8f2..e38888a13f 100644
--- a/src/html/template/escape_test.go
+++ b/src/html/template/escape_test.go
@@ -503,6 +503,16 @@ func TestEscape(t *testing.T) {
"<script>var a/*b*///c\nd</script>",
"<script>var a \nd</script>",
},
+ {
+ "JS HTML-like comments",
+ "<script>before <!-- beep\nbetween\nbefore-->boop\n</script>",
+ "<script>before \nbetween\nbefore\n</script>",
+ },
+ {
+ "JS hashbang comment",
+ "<script>#! beep\n</script>",
+ "<script>\n</script>",
+ },
{
"CSS comments",
"<style>p// paragraph\n" +
diff --git a/src/html/template/state_string.go b/src/html/template/state_string.go
index 6fb1a6eeb0..be7a920511 100644
--- a/src/html/template/state_string.go
+++ b/src/html/template/state_string.go
@@ -25,21 +25,23 @@ func _() {
_ = x[stateJSRegexp-14]
_ = x[stateJSBlockCmt-15]
_ = x[stateJSLineCmt-16]
- _ = x[stateCSS-17]
- _ = x[stateCSSDqStr-18]
- _ = x[stateCSSSqStr-19]
- _ = x[stateCSSDqURL-20]
- _ = x[stateCSSSqURL-21]
- _ = x[stateCSSURL-22]
- _ = x[stateCSSBlockCmt-23]
- _ = x[stateCSSLineCmt-24]
- _ = x[stateError-25]
- _ = x[stateDead-26]
+ _ = x[stateJSHTMLOpenCmt-17]
+ _ = x[stateJSHTMLCloseCmt-18]
+ _ = x[stateCSS-19]
+ _ = x[stateCSSDqStr-20]
+ _ = x[stateCSSSqStr-21]
+ _ = x[stateCSSDqURL-22]
+ _ = x[stateCSSSqURL-23]
+ _ = x[stateCSSURL-24]
+ _ = x[stateCSSBlockCmt-25]
+ _ = x[stateCSSLineCmt-26]
+ _ = x[stateError-27]
+ _ = x[stateDead-28]
}
-const _state_name = "stateTextstateTagstateAttrNamestateAfterNamestateBeforeValuestateHTMLCmtstateRCDATAstateAttrstateURLstateSrcsetstateJSstateJSDqStrstateJSSqStrstateJSBqStrstateJSRegexpstateJSBlockCmtstateJSLineCmtstateCSSstateCSSDqStrstateCSSSqStrstateCSSDqURLstateCSSSqURLstateCSSURLstateCSSBlockCmtstateCSSLineCmtstateErrorstateDead"
+const _state_name = "stateTextstateTagstateAttrNamestateAfterNamestateBeforeValuestateHTMLCmtstateRCDATAstateAttrstateURLstateSrcsetstateJSstateJSDqStrstateJSSqStrstateJSBqStrstateJSRegexpstateJSBlockCmtstateJSLineCmtstateJSHTMLOpenCmtstateJSHTMLCloseCmtstateCSSstateCSSDqStrstateCSSSqStrstateCSSDqURLstateCSSSqURLstateCSSURLstateCSSBlockCmtstateCSSLineCmtstateErrorstateDead"
-var _state_index = [...]uint16{0, 9, 17, 30, 44, 60, 72, 83, 92, 100, 111, 118, 130, 142, 154, 167, 182, 196, 204, 217, 230, 243, 256, 267, 283, 298, 308, 317}
+var _state_index = [...]uint16{0, 9, 17, 30, 44, 60, 72, 83, 92, 100, 111, 118, 130, 142, 154, 167, 182, 196, 214, 233, 241, 254, 267, 280, 293, 304, 320, 335, 345, 354}
func (i state) String() string {
if i >= state(len(_state_index)-1) {
diff --git a/src/html/template/transition.go b/src/html/template/transition.go
index 92eb351906..12aa4c41fe 100644
--- a/src/html/template/transition.go
+++ b/src/html/template/transition.go
@@ -14,32 +14,34 @@ import (
// the updated context and the number of bytes consumed from the front of the
// input.
var transitionFunc = [...]func(context, []byte) (context, int){
- stateText: tText,
- stateTag: tTag,
- stateAttrName: tAttrName,
- stateAfterName: tAfterName,
- stateBeforeValue: tBeforeValue,
- stateHTMLCmt: tHTMLCmt,
- stateRCDATA: tSpecialTagEnd,
- stateAttr: tAttr,
- stateURL: tURL,
- stateSrcset: tURL,
- stateJS: tJS,
- stateJSDqStr: tJSDelimited,
- stateJSSqStr: tJSDelimited,
- stateJSBqStr: tJSDelimited,
- stateJSRegexp: tJSDelimited,
- stateJSBlockCmt: tBlockCmt,
- stateJSLineCmt: tLineCmt,
- stateCSS: tCSS,
- stateCSSDqStr: tCSSStr,
- stateCSSSqStr: tCSSStr,
- stateCSSDqURL: tCSSStr,
- stateCSSSqURL: tCSSStr,
- stateCSSURL: tCSSStr,
- stateCSSBlockCmt: tBlockCmt,
- stateCSSLineCmt: tLineCmt,
- stateError: tError,
+ stateText: tText,
+ stateTag: tTag,
+ stateAttrName: tAttrName,
+ stateAfterName: tAfterName,
+ stateBeforeValue: tBeforeValue,
+ stateHTMLCmt: tHTMLCmt,
+ stateRCDATA: tSpecialTagEnd,
+ stateAttr: tAttr,
+ stateURL: tURL,
+ stateSrcset: tURL,
+ stateJS: tJS,
+ stateJSDqStr: tJSDelimited,
+ stateJSSqStr: tJSDelimited,
+ stateJSBqStr: tJSDelimited,
+ stateJSRegexp: tJSDelimited,
+ stateJSBlockCmt: tBlockCmt,
+ stateJSLineCmt: tLineCmt,
+ stateJSHTMLOpenCmt: tLineCmt,
+ stateJSHTMLCloseCmt: tLineCmt,
+ stateCSS: tCSS,
+ stateCSSDqStr: tCSSStr,
+ stateCSSSqStr: tCSSStr,
+ stateCSSDqURL: tCSSStr,
+ stateCSSSqURL: tCSSStr,
+ stateCSSURL: tCSSStr,
+ stateCSSBlockCmt: tBlockCmt,
+ stateCSSLineCmt: tLineCmt,
+ stateError: tError,
}
var commentStart = []byte("<!--")
@@ -263,7 +265,7 @@ func tURL(c context, s []byte) (context, int) {
// tJS is the context transition function for the JS state.
func tJS(c context, s []byte) (context, int) {
- i := bytes.IndexAny(s, "\"`'/")
+ i := bytes.IndexAny(s, "\"`'/<-#")
if i == -1 {
// Entire input is non string, comment, regexp tokens.
c.jsCtx = nextJSCtx(s, c.jsCtx)
@@ -293,6 +295,26 @@ func tJS(c context, s []byte) (context, int) {
err: errorf(ErrSlashAmbig, nil, 0, "'/' could start a division or regexp: %.32q", s[i:]),
}, len(s)
}
+ // ECMAScript supports HTML style comments for legacy reasons, see Appendix
+ // B.1.1 "HTML-like Comments". The handling of these comments is somewhat
+ // confusing. Multi-line comments are not supported, i.e. anything on lines
+ // between the opening and closing tokens is not considered a comment, but
+ // anything following the opening or closing token, on the same line, is
+ // ignored. As such we simply treat any line prefixed with "<!--" or "-->"
+ // as if it were actually prefixed with "//" and move on.
+ case '<':
+ if i+3 < len(s) && bytes.Equal(commentStart, s[i:i+4]) {
+ c.state, i = stateJSHTMLOpenCmt, i+3
+ }
+ case '-':
+ if i+2 < len(s) && bytes.Equal(commentEnd, s[i:i+3]) {
+ c.state, i = stateJSHTMLCloseCmt, i+2
+ }
+ // ECMAScript also supports "hashbang" comment lines, see Section 12.5.
+ case '#':
+ if i+1 < len(s) && s[i+1] == '!' {
+ c.state, i = stateJSLineCmt, i+1
+ }
default:
panic("unreachable")
}
@@ -372,12 +394,12 @@ func tBlockCmt(c context, s []byte) (context, int) {
return c, i + 2
}
-// tLineCmt is the context transition function for //comment states.
+// tLineCmt is the context transition function for //comment states, and the JS HTML-like comment state.
func tLineCmt(c context, s []byte) (context, int) {
var lineTerminators string
var endState state
switch c.state {
- case stateJSLineCmt:
+ case stateJSLineCmt, stateJSHTMLOpenCmt, stateJSHTMLCloseCmt:
lineTerminators, endState = "\n\r\u2028\u2029", stateJS
case stateCSSLineCmt:
lineTerminators, endState = "\n\f\r", stateCSS
--
2.33.0