internal/frontend: replace bluemonday with the simple sanitizer

The simple sanitizer has very few diffs compared to bluemonday and
they're all accounted for.

For #61399

Change-Id: Ib22b5828e8c1073a6fe3c2f87eee658eda3165a0
Reviewed-on: https://go-review.googlesource.com/c/pkgsite/+/547875
kokoro-CI: kokoro <[email protected]>
LUCI-TryBot-Result: Go LUCI <[email protected]>
Reviewed-by: Jonathan Amsterdam <[email protected]>

Author: matloob

go.mod

@@ -26,7 +26,6 @@ require (
 	github.com/jackc/pgx/v4 v4.14.1
 	github.com/jba/templatecheck v0.6.0
 	github.com/lib/pq v1.10.9
-	github.com/microcosm-cc/bluemonday v1.0.25
 	github.com/russross/blackfriday/v2 v2.1.0
 	github.com/yuin/goldmark v1.4.13
 	github.com/yuin/goldmark-emoji v1.0.1
@@ -55,7 +54,6 @@ require (
 	cloud.google.com/go/trace v1.9.0 // indirect
 	github.com/alicebob/gopher-json v0.0.0-20200520072559-a9ecdc9d1d3a // indirect
 	github.com/aws/aws-sdk-go v1.34.29 // indirect
-	github.com/aymerick/douceur v0.2.0 // indirect
 	github.com/beorn7/perks v1.0.1 // indirect
 	github.com/census-instrumentation/opencensus-proto v0.4.1 // indirect
 	github.com/cespare/xxhash/v2 v2.2.0 // indirect
@@ -68,7 +66,6 @@ require (
 	github.com/google/uuid v1.3.0 // indirect
 	github.com/googleapis/enterprise-certificate-proxy v0.2.3 // indirect
 	github.com/googleapis/gax-go/v2 v2.11.0 // indirect
-	github.com/gorilla/css v1.0.0 // indirect
 	github.com/hashicorp/errwrap v1.0.0 // indirect
 	github.com/hashicorp/go-multierror v1.1.0 // indirect
 	github.com/jackc/chunkreader/v2 v2.0.1 // indirect

go.sum

@@ -180,8 +180,6 @@ github.com/aws/aws-sdk-go-v2/service/sts v1.6.1/go.mod h1:hLZ/AnkIKHLuPGjEiyghNE
 github.com/aws/aws-sdk-go-v2/service/sts v1.7.2/go.mod h1:8EzeIqfWt2wWT4rJVu3f21TfrhJ8AEMzVybRNSb/b4g=
 github.com/aws/smithy-go v1.7.0/go.mod h1:SObp3lf9smib00L/v3U2eAKG8FyQ7iLrJnQiAmR5n+E=
 github.com/aws/smithy-go v1.8.0/go.mod h1:SObp3lf9smib00L/v3U2eAKG8FyQ7iLrJnQiAmR5n+E=
-github.com/aymerick/douceur v0.2.0 h1:Mv+mAeH1Q+n9Fr+oyamOlAkUNPWPlA8PPGR0QAaYuPk=
-github.com/aymerick/douceur v0.2.0/go.mod h1:wlT5vV2O3h55X9m7iVYN0TBM0NH/MmbLnd30/FjWUq4=
 github.com/beorn7/perks v0.0.0-20160804104726-4c0e84591b9a/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q=
 github.com/beorn7/perks v0.0.0-20180321164747-3a771d992973/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q=
 github.com/beorn7/perks v1.0.0/go.mod h1:KWe93zE9D1o94FZ5RNwFwVgaQK1VOXiVxmqh+CedLV8=
@@ -600,8 +598,6 @@ github.com/googleapis/gax-go/v2 v2.11.0 h1:9V9PWXEsWnPpQhu/PeQIkS4eGzMlTLGgt80cU
 github.com/googleapis/gax-go/v2 v2.11.0/go.mod h1:DxmR61SGKkGLa2xigwuZIQpkCI2S5iydzRfb3peWZJI=
 github.com/googleapis/gnostic v0.4.1/go.mod h1:LRhVm6pbyptWbWbuZ38d1eyptfvIytN3ir6b65WBswg=
 github.com/gopherjs/gopherjs v0.0.0-20181017120253-0766667cb4d1/go.mod h1:wJfORRmW1u3UXTncJ5qlYoELFm8eSnnEO6hX4iZ3EWY=
-github.com/gorilla/css v1.0.0 h1:BQqNyPTi50JCFMTw/b67hByjMVXZRwGha6wxVGkeihY=
-github.com/gorilla/css v1.0.0/go.mod h1:Dn721qIggHpt4+EFCcTLTU/vk5ySda2ReITrtgBl60c=
 github.com/gorilla/handlers v0.0.0-20150720190736-60c7bfde3e33/go.mod h1:Qkdc/uu4tH4g6mTK6auzZ766c4CA0Ng8+o/OAirnOIQ=
 github.com/gorilla/handlers v1.4.2/go.mod h1:Qkdc/uu4tH4g6mTK6auzZ766c4CA0Ng8+o/OAirnOIQ=
 github.com/gorilla/mux v1.7.2/go.mod h1:1lud6UwP+6orDFRuTfBEV8e9/aOM/c4fVVCaMa2zaAs=
@@ -801,8 +797,6 @@ github.com/mattn/go-sqlite3 v1.14.6/go.mod h1:NyWgC/yNuGj7Q9rpYnZvas74GogHl5/Z4A
 github.com/matttproud/golang_protobuf_extensions v1.0.1/go.mod h1:D8He9yQNgCq6Z5Ld7szi9bcBfOoFv/3dc6xSMkL2PC0=
 github.com/matttproud/golang_protobuf_extensions v1.0.2-0.20181231171920-c182affec369 h1:I0XW9+e1XWDxdcEniV4rQAIOPUGDq67JSCiRCgGCZLI=
 github.com/matttproud/golang_protobuf_extensions v1.0.2-0.20181231171920-c182affec369/go.mod h1:BSXmuO+STAnVfrANrmjBb36TMTDstsz7MSK+HVaYKv4=
-github.com/microcosm-cc/bluemonday v1.0.25 h1:4NEwSfiJ+Wva0VxN5B8OwMicaJvD8r9tlJWm9rtloEg=
-github.com/microcosm-cc/bluemonday v1.0.25/go.mod h1:ZIOjCQp1OrzBBPIJmfX4qDYFuhU02nx4bn030ixfHLE=
 github.com/miekg/pkcs11 v1.0.3/go.mod h1:XsNlhZGX73bx86s2hdc/FuaLm2CPZJemRLMA+WTFxgs=
 github.com/mistifyio/go-zfs v2.1.2-0.20190413222219-f784269be439+incompatible/go.mod h1:8AuVvqP/mXw1px98n46wfvcGfQ4ci2FwoAjKYxuo3Z4=
 github.com/mitchellh/go-homedir v1.1.0/go.mod h1:SfyaCUpYCn1Vlf4IUYiD9fPX4A5wJrkLzIz1N1q0pr0=

internal/frontend/readme.go

@@ -11,7 +11,6 @@ import (
 	"github.com/google/safehtml"
 	"github.com/google/safehtml/template"
 	"github.com/google/safehtml/uncheckedconversions"
-	"github.com/microcosm-cc/bluemonday"
 	"github.com/yuin/goldmark"
 	emoji "github.com/yuin/goldmark-emoji"
 	"github.com/yuin/goldmark/extension"
@@ -23,6 +22,7 @@ import (
 	"golang.org/x/pkgsite/internal"
 	"golang.org/x/pkgsite/internal/derrors"
 	"golang.org/x/pkgsite/internal/log"
+	"golang.org/x/pkgsite/internal/sanitizer"
 	"golang.org/x/pkgsite/internal/source"
 )
 
@@ -112,7 +112,7 @@ func processReadme(ctx context.Context, readme *internal.Readme, sourceInfo *sou
 			),
 		),
 		// These extensions lets users write HTML code in the README. This is
-		// fine since we process the contents using bluemonday after.
+		// fine since we process the contents using the sanitizer after.
 		goldmark.WithRendererOptions(goldmarkHtml.WithUnsafe(), goldmarkHtml.WithXHTML()),
 		goldmark.WithExtensions(
 			extension.GFM, // Support Github Flavored Markdown.
@@ -159,18 +159,6 @@ func processReadme(ctx context.Context, readme *internal.Readme, sourceInfo *sou
 
 // sanitizeHTML sanitizes HTML from a bytes.Buffer so that it is safe.
 func sanitizeHTML(b *bytes.Buffer) safehtml.HTML {
-	p := bluemonday.UGCPolicy()
-
-	p.AllowAttrs("width", "align").OnElements("img")
-	p.AllowAttrs("width", "align").OnElements("div")
-	p.AllowAttrs("width", "align").OnElements("p")
-	// Allow accessible headings (i.e <div role="heading" aria-level="7">).
-	p.AllowAttrs("width", "align", "role", "aria-level").OnElements("div")
-	for _, h := range []string{"h1", "h2", "h3", "h4", "h5", "h6"} {
-		// Needed to preserve github styles heading font-sizes
-		p.AllowAttrs("class").OnElements(h)
-	}
-
-	s := string(p.SanitizeBytes(b.Bytes()))
+	s := string(sanitizer.SanitizeBytes(b.Bytes()))
 	return uncheckedconversions.HTMLFromStringKnownToSatisfyTypeContract(s)
 }