From ce7bd33345416e6d8cac901792060591cafc2797 Mon Sep 17 00:00:00 2001 From: Roland Shoemaker Date: Tue, 11 Apr 2023 16:27:43 +0100 Subject: [PATCH] [release-branch.go1.19] html/template: handle all JS whitespace characters Rather than just a small set. Character class as defined by \s [0]. Thanks to Juho Nurminen of Mattermost for reporting this. For #59721 Fixes #59813 Fixes CVE-2023-24540 [0] https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Regular_Expressions/Character_Classes Change-Id: I56d4fa1ef08125b417106ee7dbfb5b0923b901ba Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1821459 Reviewed-by: Julie Qiu Run-TryBot: Roland Shoemaker Reviewed-by: Damien Neil Reviewed-on: https://team-review.git.corp.google.com/c/golang/go-private/+/1851497 Run-TryBot: Damien Neil Reviewed-by: Roland Shoemaker Reviewed-on: https://go-review.googlesource.com/c/go/+/491355 Reviewed-by: Dmitri Shuralyov Reviewed-by: Carlos Amedee TryBot-Bypass: Carlos Amedee Run-TryBot: Carlos Amedee Upstream-Status: Backport [https://github.com/golang/go/commit/ce7bd33345416e6d8cac901792060591cafc2797] CVE: CVE-2023-24540 Signed-off-by: Vijay Anusuri --- src/html/template/js.go | 8 +++++++- src/html/template/js_test.go | 11 +++++++---- 2 files changed, 14 insertions(+), 5 deletions(-) diff --git a/src/html/template/js.go b/src/html/template/js.go index fe7054efe5cd8..4e05c1455723f 100644 --- a/src/html/template/js.go +++ b/src/html/template/js.go @@ -13,6 +13,11 @@ import ( "unicode/utf8" ) +// jsWhitespace contains all of the JS whitespace characters, as defined +// by the \s character class. +// See https://developer.mozilla.org/en-US/docs/Web/JavaScript/Guide/Regular_expressions/Character_classes. +const jsWhitespace = "\f\n\r\t\v\u0020\u00a0\u1680\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a\u2028\u2029\u202f\u205f\u3000\ufeff" + // nextJSCtx returns the context that determines whether a slash after the // given run of tokens starts a regular expression instead of a division // operator: / or /=. @@ -26,7 +31,8 @@ import ( // JavaScript 2.0 lexical grammar and requires one token of lookbehind: // https://www.mozilla.org/js/language/js20-2000-07/rationale/syntax.html func nextJSCtx(s []byte, preceding jsCtx) jsCtx { - s = bytes.TrimRight(s, "\t\n\f\r \u2028\u2029") + // Trim all JS whitespace characters + s = bytes.TrimRight(s, jsWhitespace) if len(s) == 0 { return preceding } diff --git a/src/html/template/js_test.go b/src/html/template/js_test.go index e07c695f7a77d..e52180cc113b5 100644 --- a/src/html/template/js_test.go +++ b/src/html/template/js_test.go @@ -81,14 +81,17 @@ func TestNextJsCtx(t *testing.T) { {jsCtxDivOp, "0"}, // Dots that are part of a number are div preceders. {jsCtxDivOp, "0."}, + // Some JS interpreters treat NBSP as a normal space, so + // we must too in order to properly escape things. + {jsCtxRegexp, "=\u00A0"}, } for _, test := range tests { - if nextJSCtx([]byte(test.s), jsCtxRegexp) != test.jsCtx { - t.Errorf("want %s got %q", test.jsCtx, test.s) + if ctx := nextJSCtx([]byte(test.s), jsCtxRegexp); ctx != test.jsCtx { + t.Errorf("%q: want %s got %s", test.s, test.jsCtx, ctx) } - if nextJSCtx([]byte(test.s), jsCtxDivOp) != test.jsCtx { - t.Errorf("want %s got %q", test.jsCtx, test.s) + if ctx := nextJSCtx([]byte(test.s), jsCtxDivOp); ctx != test.jsCtx { + t.Errorf("%q: want %s got %s", test.s, test.jsCtx, ctx) } }