forks/miniflux
1
0
Fork 0
Code Releases Activity

Improve HTML sanitizer

Browse source
This commit is contained in:
Frédéric Guillot 2017-11-25 18:08:59 -08:00
parent 1f015d5dfe
commit bd663b43a0
4 changed files with 28 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
reader/json/parser_test.go
View file

@ -148,7 +148,7 @@ func TestParsePodcast(t *testing.T) {
t.Errorf(`Incorrect entry title, got: "%s"`, feed.Entries[0].Title) t.Errorf(`Incorrect entry title, got: "%s"`, feed.Entries[0].Title)
} }
if feed.Entries[0].Content != `Chris has worked at <a href="http://adobe.com/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Adobe</a> and as a founder of Rogue Sheep, which won an Apple Design Award for Postage. Chris’s new company is Aged & Distilled with Guy English — which shipped <a href="http://aged-and-distilled.com/napkin/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Napkin</a>, a Mac app for visual collaboration. Chris is also the co-host of The Record. He lives on <a href="http://www.ci.bainbridge-isl.wa.us/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Bainbridge Island</a>, a quick ferry ride from Seattle.` { if feed.Entries[0].Content != `Chris has worked at <a href="http://adobe.com/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Adobe</a> and as a founder of Rogue Sheep, which won an Apple Design Award for Postage. Chris’s new company is Aged &amp; Distilled with Guy English — which shipped <a href="http://aged-and-distilled.com/napkin/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Napkin</a>, a Mac app for visual collaboration. Chris is also the co-host of The Record. He lives on <a href="http://www.ci.bainbridge-isl.wa.us/" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Bainbridge Island</a>, a quick ferry ride from Seattle.` {
t.Errorf(`Incorrect entry content, got: "%s"`, feed.Entries[0].Content) t.Errorf(`Incorrect entry content, got: "%s"`, feed.Entries[0].Content)
} }

2
reader/rss/parser_test.go
View file

@ -94,7 +94,7 @@ func TestParseRss2Sample(t *testing.T) {
t.Errorf("Incorrect entry title, got: %s", feed.Entries[0].Title) t.Errorf("Incorrect entry title, got: %s", feed.Entries[0].Title)
} }
if feed.Entries[0].Content != `How do Americans get ready to work with Russians aboard the International Space Station? They take a crash course in culture, language and protocol at Russia's <a href="http://howe.iki.rssi.ru/GCTC/gctc_e.htm" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Star City</a>.` { if feed.Entries[0].Content != `How do Americans get ready to work with Russians aboard the International Space Station? They take a crash course in culture, language and protocol at Russia&#39;s <a href="http://howe.iki.rssi.ru/GCTC/gctc_e.htm" rel="noopener noreferrer" target="_blank" referrerpolicy="no-referrer">Star City</a>.` {
t.Errorf("Incorrect entry content, got: %s", feed.Entries[0].Content) t.Errorf("Incorrect entry content, got: %s", feed.Entries[0].Content)
} }
} }

11
reader/sanitizer/sanitizer.go
View file

@ -7,10 +7,11 @@ package sanitizer
import ( import (
"bytes" "bytes"
"fmt" "fmt"
"github.com/miniflux/miniflux2/reader/url"
"io" "io"
"strings" "strings"
"github.com/miniflux/miniflux2/reader/url"
"golang.org/x/net/html" "golang.org/x/net/html"
) )
@ -33,7 +34,7 @@ func Sanitize(baseURL, input string) string {
token := tokenizer.Token() token := tokenizer.Token()
switch token.Type { switch token.Type {
case html.TextToken: case html.TextToken:
buffer.WriteString(token.Data) buffer.WriteString(html.EscapeString(token.Data))
case html.StartTagToken: case html.StartTagToken:
tagName := token.DataAtom.String() tagName := token.DataAtom.String()
@ -72,8 +73,8 @@ func Sanitize(baseURL, input string) string {
} }
} }
func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute) (attrNames []string, html string) { func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute) ([]string, string) {
var htmlAttrs []string var htmlAttrs, attrNames []string
var err error var err error
for _, attribute := range attributes { for _, attribute := range attributes {
@ -99,7 +100,7 @@ func sanitizeAttributes(baseURL, tagName string, attributes []html.Attribute) (a
} }
attrNames = append(attrNames, attribute.Key) attrNames = append(attrNames, attribute.Key)
htmlAttrs = append(htmlAttrs, fmt.Sprintf(`%s="%s"`, attribute.Key, value)) htmlAttrs = append(htmlAttrs, fmt.Sprintf(`%s="%s"`, attribute.Key, html.EscapeString(value)))
} }
extraAttrNames, extraHTMLAttributes := getExtraAttributes(tagName) extraAttrNames, extraHTMLAttributes := getExtraAttributes(tagName)

20
reader/sanitizer/sanitizer_test.go
View file

@ -142,3 +142,23 @@ func TestPixelTracker(t *testing.T) {
t.Errorf(`Wrong output: "%s" != "%s"`, expected, output) t.Errorf(`Wrong output: "%s" != "%s"`, expected, output)
} }
} }
func TestXmlEntities(t *testing.T) {
input := `<pre>echo "test" &gt; /etc/hosts</pre>`
expected := `<pre>echo &#34;test&#34; &gt; /etc/hosts</pre>`
output := Sanitize("http://example.org/", input)
if expected != output {
t.Errorf(`Wrong output: "%s" != "%s"`, expected, output)
}
}
func TestEspaceAttributes(t *testing.T) {
input := `<td rowspan="<b>test</b>">test</td>`
expected := `<td rowspan="&lt;b&gt;test&lt;/b&gt;">test</td>`
output := Sanitize("http://example.org/", input)
if expected != output {
t.Errorf(`Wrong output: "%s" != "%s"`, expected, output)
}
}