[bugfix] html escape special characters in text instead of totally removing them (#719)

* remove minify dependency

* tidy up some tests

* remove pre + postformat funcs

* rework sanitization + formatting

* update tests

* add some more markdown tests
This commit is contained in:
tobi 2022-07-19 15:21:17 +02:00 committed by GitHub
commit c84384e660
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
51 changed files with 129 additions and 7419 deletions

View file

@ -178,7 +178,7 @@ func (suite *StatusCreateTestSuite) TestPostAnotherNewStatus() {
err = json.Unmarshal(b, statusReply)
suite.NoError(err)
suite.Equal("<p><a href=\"http://localhost:8080/tags/test\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>test</span></a> alright, should be able to post <a href=\"http://localhost:8080/tags/links\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>links</span></a> with fragments in them now, let's see........<br><br><a href=\"https://docs.gotosocial.org/en/latest/user_guide/posts/#links\" rel=\"noopener nofollow noreferrer\" target=\"_blank\">docs.gotosocial.org/en/latest/user_guide/posts/#links</a><br><br><a href=\"http://localhost:8080/tags/gotosocial\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>gotosocial</span></a><br><br>(tobi remember to pull the docker image challenge)</p>", statusReply.Content)
suite.Equal("<p><a href=\"http://localhost:8080/tags/test\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>test</span></a> alright, should be able to post <a href=\"http://localhost:8080/tags/links\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>links</span></a> with fragments in them now, let&#39;s see........<br/><br/><a href=\"https://docs.gotosocial.org/en/latest/user_guide/posts/#links\" rel=\"noopener nofollow noreferrer\" target=\"_blank\">docs.gotosocial.org/en/latest/user_guide/posts/#links</a><br/><br/><a href=\"http://localhost:8080/tags/gotosocial\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>gotosocial</span></a><br/><br/>(tobi remember to pull the docker image challenge)</p>", statusReply.Content)
}
func (suite *StatusCreateTestSuite) TestPostNewStatusWithEmoji() {
@ -211,7 +211,7 @@ func (suite *StatusCreateTestSuite) TestPostNewStatusWithEmoji() {
suite.NoError(err)
suite.Equal("", statusReply.SpoilerText)
suite.Equal("<p>here is a rainbow emoji a few times! :rainbow: :rainbow: :rainbow:<br>here's an emoji that isn't in the db: :test_emoji:</p>", statusReply.Content)
suite.Equal("<p>here is a rainbow emoji a few times! :rainbow: :rainbow: :rainbow: <br/> here&#39;s an emoji that isn&#39;t in the db: :test_emoji:</p>", statusReply.Content)
suite.Len(statusReply.Emojis, 1)
apiEmoji := statusReply.Emojis[0]
@ -330,7 +330,7 @@ func (suite *StatusCreateTestSuite) TestAttachNewMediaSuccess() {
suite.NoError(err)
suite.Equal("", statusResponse.SpoilerText)
suite.Equal("<p>here's an image attachment</p>", statusResponse.Content)
suite.Equal("<p>here&#39;s an image attachment</p>", statusResponse.Content)
suite.False(statusResponse.Sensitive)
suite.Equal(model.VisibilityPublic, statusResponse.Visibility)

View file

@ -73,13 +73,12 @@ func (suite *AccountUpdateTestSuite) TestAccountUpdateSimple() {
func (suite *AccountUpdateTestSuite) TestAccountUpdateWithMention() {
testAccount := suite.testAccounts["local_account_1"]
locked := true
displayName := "new display name"
note := `#hello here i am!
go check out @1happyturtle, they have a cool account!
`
noteExpected := `<p><a href="http://localhost:8080/tags/hello" class="mention hashtag" rel="tag nofollow noreferrer noopener" target="_blank">#<span>hello</span></a> here i am!<br><br>go check out <span class="h-card"><a href="http://localhost:8080/@1happyturtle" class="u-url mention" rel="nofollow noreferrer noopener" target="_blank">@<span>1happyturtle</span></a></span>, they have a cool account!</p>`
var (
locked = true
displayName = "new display name"
note = "#hello here i am!\n\ngo check out @1happyturtle, they have a cool account!\n"
noteExpected = "<p><a href=\"http://localhost:8080/tags/hello\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>hello</span></a> here i am!<br/><br/>go check out <span class=\"h-card\"><a href=\"http://localhost:8080/@1happyturtle\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>1happyturtle</span></a></span>, they have a cool account!</p>"
)
form := &apimodel.UpdateCredentialsRequest{
DisplayName: &displayName,

View file

@ -23,31 +23,20 @@ import (
"fmt"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/suite"
"github.com/superseriousbusiness/gotosocial/internal/api/model"
"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
)
const statusText1 = `Another test @foss_satan@fossbros-anonymous.io
#Hashtag
Text`
const (
statusText1ExpectedFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br><br><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br><br>Text</p>"
statusText1ExpectedPartial = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br><br>#Hashtag<br><br>Text</p>"
statusText1 = "Another test @foss_satan@fossbros-anonymous.io\n\n#Hashtag\n\nText"
statusText1ExpectedFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br/><br/><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br/><br/>Text</p>"
statusText1ExpectedPartial = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br/><br/>#Hashtag<br/><br/>Text</p>"
statusText2 = "Another test @foss_satan@fossbros-anonymous.io\n\n#Hashtag\n\n#hashTAG"
status2TextExpectedFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br/><br/><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br/><br/><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>hashTAG</span></a></p>"
status2TextExpectedPartial = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br/><br/>#Hashtag<br/><br/>#hashTAG</p>"
)
const statusText2 = `Another test @foss_satan@fossbros-anonymous.io
#Hashtag
#hashTAG`
const status2TextExpectedFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br><br><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br><br><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>hashTAG</span></a></p>"
type UtilTestSuite struct {
StatusStandardTestSuite
}
@ -82,21 +71,21 @@ func (suite *UtilTestSuite) TestProcessMentions1() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
suite.NoError(err)
assert.Len(suite.T(), status.Mentions, 1)
suite.Len(status.Mentions, 1)
newMention := status.Mentions[0]
assert.Equal(suite.T(), mentionedAccount.ID, newMention.TargetAccountID)
assert.Equal(suite.T(), creatingAccount.ID, newMention.OriginAccountID)
assert.Equal(suite.T(), creatingAccount.URI, newMention.OriginAccountURI)
assert.Equal(suite.T(), status.ID, newMention.StatusID)
assert.Equal(suite.T(), fmt.Sprintf("@%s@%s", mentionedAccount.Username, mentionedAccount.Domain), newMention.NameString)
assert.Equal(suite.T(), mentionedAccount.URI, newMention.TargetAccountURI)
assert.Equal(suite.T(), mentionedAccount.URL, newMention.TargetAccountURL)
assert.NotNil(suite.T(), newMention.OriginAccount)
suite.Equal(mentionedAccount.ID, newMention.TargetAccountID)
suite.Equal(creatingAccount.ID, newMention.OriginAccountID)
suite.Equal(creatingAccount.URI, newMention.OriginAccountURI)
suite.Equal(status.ID, newMention.StatusID)
suite.Equal(fmt.Sprintf("@%s@%s", mentionedAccount.Username, mentionedAccount.Domain), newMention.NameString)
suite.Equal(mentionedAccount.URI, newMention.TargetAccountURI)
suite.Equal(mentionedAccount.URL, newMention.TargetAccountURL)
suite.NotNil(newMention.OriginAccount)
assert.Len(suite.T(), status.MentionIDs, 1)
assert.Equal(suite.T(), newMention.ID, status.MentionIDs[0])
suite.Len(status.MentionIDs, 1)
suite.Equal(newMention.ID, status.MentionIDs[0])
}
func (suite *UtilTestSuite) TestProcessContentFull1() {
@ -131,20 +120,20 @@ func (suite *UtilTestSuite) TestProcessContentFull1() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
suite.NoError(err)
suite.Empty(status.Content) // shouldn't be set yet
err = suite.status.ProcessTags(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
suite.NoError(err)
suite.Empty(status.Content) // shouldn't be set yet
/*
ACTUAL TEST
*/
err = suite.status.ProcessContent(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Equal(suite.T(), statusText1ExpectedFull, status.Content)
suite.NoError(err)
suite.Equal(statusText1ExpectedFull, status.Content)
}
func (suite *UtilTestSuite) TestProcessContentPartial1() {
@ -179,16 +168,16 @@ func (suite *UtilTestSuite) TestProcessContentPartial1() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
suite.NoError(err)
suite.Empty(status.Content) // shouldn't be set yet
/*
ACTUAL TEST
*/
err = suite.status.ProcessContent(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Equal(suite.T(), statusText1ExpectedPartial, status.Content)
suite.NoError(err)
suite.Equal(statusText1ExpectedPartial, status.Content)
}
func (suite *UtilTestSuite) TestProcessMentions2() {
@ -221,21 +210,21 @@ func (suite *UtilTestSuite) TestProcessMentions2() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
suite.NoError(err)
assert.Len(suite.T(), status.Mentions, 1)
suite.Len(status.Mentions, 1)
newMention := status.Mentions[0]
assert.Equal(suite.T(), mentionedAccount.ID, newMention.TargetAccountID)
assert.Equal(suite.T(), creatingAccount.ID, newMention.OriginAccountID)
assert.Equal(suite.T(), creatingAccount.URI, newMention.OriginAccountURI)
assert.Equal(suite.T(), status.ID, newMention.StatusID)
assert.Equal(suite.T(), fmt.Sprintf("@%s@%s", mentionedAccount.Username, mentionedAccount.Domain), newMention.NameString)
assert.Equal(suite.T(), mentionedAccount.URI, newMention.TargetAccountURI)
assert.Equal(suite.T(), mentionedAccount.URL, newMention.TargetAccountURL)
assert.NotNil(suite.T(), newMention.OriginAccount)
suite.Equal(mentionedAccount.ID, newMention.TargetAccountID)
suite.Equal(creatingAccount.ID, newMention.OriginAccountID)
suite.Equal(creatingAccount.URI, newMention.OriginAccountURI)
suite.Equal(status.ID, newMention.StatusID)
suite.Equal(fmt.Sprintf("@%s@%s", mentionedAccount.Username, mentionedAccount.Domain), newMention.NameString)
suite.Equal(mentionedAccount.URI, newMention.TargetAccountURI)
suite.Equal(mentionedAccount.URL, newMention.TargetAccountURL)
suite.NotNil(newMention.OriginAccount)
assert.Len(suite.T(), status.MentionIDs, 1)
assert.Equal(suite.T(), newMention.ID, status.MentionIDs[0])
suite.Len(status.MentionIDs, 1)
suite.Equal(newMention.ID, status.MentionIDs[0])
}
func (suite *UtilTestSuite) TestProcessContentFull2() {
@ -270,21 +259,21 @@ func (suite *UtilTestSuite) TestProcessContentFull2() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
suite.NoError(err)
suite.Empty(status.Content) // shouldn't be set yet
err = suite.status.ProcessTags(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
suite.NoError(err)
suite.Empty(status.Content) // shouldn't be set yet
/*
ACTUAL TEST
*/
err = suite.status.ProcessContent(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
suite.NoError(err)
assert.Equal(suite.T(), status2TextExpectedFull, status.Content)
suite.Equal(status2TextExpectedFull, status.Content)
}
func (suite *UtilTestSuite) TestProcessContentPartial2() {
@ -319,18 +308,13 @@ func (suite *UtilTestSuite) TestProcessContentPartial2() {
}
err := suite.status.ProcessMentions(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
assert.Empty(suite.T(), status.Content) // shouldn't be set yet
/*
ACTUAL TEST
*/
suite.NoError(err)
suite.Empty(status.Content)
err = suite.status.ProcessContent(context.Background(), form, creatingAccount.ID, status)
assert.NoError(suite.T(), err)
suite.NoError(err)
fmt.Println(status.Content)
// assert.Equal(suite.T(), statusText2ExpectedPartial, status.Content)
suite.Equal(status2TextExpectedPartial, status.Content)
}
func TestUtilTestSuite(t *testing.T) {

View file

@ -21,7 +21,6 @@ package text
import (
"bytes"
"context"
"html"
"strings"
"unicode"
@ -30,38 +29,6 @@ import (
"github.com/superseriousbusiness/gotosocial/internal/regexes"
)
// preformat contains some common logic for making a string ready for formatting, which should be used for all user-input text.
func preformat(in string) string {
// do some preformatting of the text
// 1. unescape everything that might be html escaped
s := html.UnescapeString(in)
// 2. trim leading or trailing whitespace
s = strings.TrimSpace(s)
return s
}
// postformat contains some common logic for html sanitization of text, wrapping elements, and trimming newlines and whitespace
func postformat(in string) string {
// do some postformatting of the text
// 1. sanitize html to remove potentially dangerous elements
s := SanitizeHTML(in)
// 2. the sanitize step tends to escape characters inside codeblocks, which is behavior we don't want, so unescape everything again
s = html.UnescapeString(s)
// 3. minify html to remove any trailing newlines, spaces, unnecessary elements, etc etc
mini, err := MinifyHTML(s)
if err != nil {
// if the minify failed, just return what we have
return s
}
// return minified version of the html
return mini
}
func (f *formatter) ReplaceTags(ctx context.Context, in string, tags []*gtsmodel.Tag) string {
return regexes.ReplaceAllStringFunc(regexes.HashtagFinder, in, func(match string, buf *bytes.Buffer) string {
// we have a match

View file

@ -28,44 +28,14 @@ import (
)
const (
replaceMentionsString = `Another test @foss_satan@fossbros-anonymous.io
#Hashtag
Text`
replaceMentionsExpected = `Another test <span class="h-card"><a href="http://fossbros-anonymous.io/@foss_satan" class="u-url mention">@<span>foss_satan</span></a></span>
#Hashtag
Text`
replaceHashtagsExpected = `Another test @foss_satan@fossbros-anonymous.io
<a href="http://localhost:8080/tags/Hashtag" class="mention hashtag" rel="tag">#<span>Hashtag</span></a>
Text`
replaceHashtagsAfterMentionsExpected = `Another test <span class="h-card"><a href="http://fossbros-anonymous.io/@foss_satan" class="u-url mention">@<span>foss_satan</span></a></span>
<a href="http://localhost:8080/tags/Hashtag" class="mention hashtag" rel="tag">#<span>Hashtag</span></a>
Text`
replaceMentionsWithLinkString = `Another test @foss_satan@fossbros-anonymous.io
http://fossbros-anonymous.io/@foss_satan/statuses/6675ee73-fccc-4562-a46a-3e8cd9798060`
replaceMentionsWithLinkStringExpected = `Another test <span class="h-card"><a href="http://fossbros-anonymous.io/@foss_satan" class="u-url mention">@<span>foss_satan</span></a></span>
http://fossbros-anonymous.io/@foss_satan/statuses/6675ee73-fccc-4562-a46a-3e8cd9798060`
replaceMentionsWithLinkSelfString = `Mentioning myself: @the_mighty_zork
and linking to my own status: https://localhost:8080/@the_mighty_zork/statuses/01FGXKJRX2PMERJQ9EQF8Y6HCR`
replaceMemtionsWithLinkSelfExpected = `Mentioning myself: <span class="h-card"><a href="http://localhost:8080/@the_mighty_zork" class="u-url mention">@<span>the_mighty_zork</span></a></span>
and linking to my own status: https://localhost:8080/@the_mighty_zork/statuses/01FGXKJRX2PMERJQ9EQF8Y6HCR`
replaceMentionsString = "Another test @foss_satan@fossbros-anonymous.io\n\n#Hashtag\n\nText"
replaceMentionsExpected = "Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\">@<span>foss_satan</span></a></span>\n\n#Hashtag\n\nText"
replaceHashtagsExpected = "Another test @foss_satan@fossbros-anonymous.io\n\n<a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag\">#<span>Hashtag</span></a>\n\nText"
replaceHashtagsAfterMentionsExpected = "Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\">@<span>foss_satan</span></a></span>\n\n<a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag\">#<span>Hashtag</span></a>\n\nText"
replaceMentionsWithLinkString = "Another test @foss_satan@fossbros-anonymous.io\n\nhttp://fossbros-anonymous.io/@foss_satan/statuses/6675ee73-fccc-4562-a46a-3e8cd9798060"
replaceMentionsWithLinkStringExpected = "Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\">@<span>foss_satan</span></a></span>\n\nhttp://fossbros-anonymous.io/@foss_satan/statuses/6675ee73-fccc-4562-a46a-3e8cd9798060"
replaceMentionsWithLinkSelfString = "Mentioning myself: @the_mighty_zork\n\nand linking to my own status: https://localhost:8080/@the_mighty_zork/statuses/01FGXKJRX2PMERJQ9EQF8Y6HCR"
replaceMemtionsWithLinkSelfExpected = "Mentioning myself: <span class=\"h-card\"><a href=\"http://localhost:8080/@the_mighty_zork\" class=\"u-url mention\">@<span>the_mighty_zork</span></a></span>\n\nand linking to my own status: https://localhost:8080/@the_mighty_zork/statuses/01FGXKJRX2PMERJQ9EQF8Y6HCR"
)
type CommonTestSuite struct {

View file

@ -71,16 +71,16 @@ type LinkTestSuite struct {
func (suite *LinkTestSuite) TestParseSimple() {
f := suite.formatter.FromPlain(context.Background(), simple, nil, nil)
assert.Equal(suite.T(), simpleExpected, f)
suite.Equal(simpleExpected, f)
}
func (suite *LinkTestSuite) TestParseURLsFromText1() {
urls := text.FindLinks(text1)
assert.Equal(suite.T(), "https://example.org/link/to/something#fragment", urls[0].String())
assert.Equal(suite.T(), "http://test.example.org?q=bahhhhhhhhhhhh", urls[1].String())
assert.Equal(suite.T(), "https://another.link.example.org/with/a/pretty/long/path/at/the/end/of/it", urls[2].String())
assert.Equal(suite.T(), "https://example.orghttps://google.com", urls[3].String())
suite.Equal("https://example.org/link/to/something#fragment", urls[0].String())
suite.Equal("http://test.example.org?q=bahhhhhhhhhhhh", urls[1].String())
suite.Equal("https://another.link.example.org/with/a/pretty/long/path/at/the/end/of/it", urls[2].String())
suite.Equal("https://example.orghttps://google.com", urls[3].String())
}
func (suite *LinkTestSuite) TestParseURLsFromText2() {
@ -99,7 +99,7 @@ func (suite *LinkTestSuite) TestParseURLsFromText3() {
func (suite *LinkTestSuite) TestReplaceLinksFromText1() {
replaced := suite.formatter.ReplaceLinks(context.Background(), text1)
assert.Equal(suite.T(), `
suite.Equal(`
This is a text with some links in it. Here's link number one: <a href="https://example.org/link/to/something#fragment" rel="noopener">example.org/link/to/something#fragment</a>
Here's link number two: <a href="http://test.example.org?q=bahhhhhhhhhhhh" rel="noopener">test.example.org?q=bahhhhhhhhhhhh</a>
@ -114,7 +114,7 @@ really.cool.website <-- this one shouldn't be parsed as a link because it doesn'
func (suite *LinkTestSuite) TestReplaceLinksFromText2() {
replaced := suite.formatter.ReplaceLinks(context.Background(), text2)
assert.Equal(suite.T(), `
suite.Equal(`
this is one link: <a href="https://example.org" rel="noopener">example.org</a>
this is the same link again: <a href="https://example.org" rel="noopener">example.org</a>
@ -126,14 +126,14 @@ these should be deduplicated
func (suite *LinkTestSuite) TestReplaceLinksFromText3() {
// we know mailto links won't be replaced with hrefs -- we only accept https and http
replaced := suite.formatter.ReplaceLinks(context.Background(), text3)
assert.Equal(suite.T(), `
suite.Equal(`
here's a mailto link: mailto:whatever@test.org
`, replaced)
}
func (suite *LinkTestSuite) TestReplaceLinksFromText4() {
replaced := suite.formatter.ReplaceLinks(context.Background(), text4)
assert.Equal(suite.T(), `
suite.Equal(`
two similar links:
<a href="https://example.org" rel="noopener">example.org</a>
@ -145,7 +145,7 @@ two similar links:
func (suite *LinkTestSuite) TestReplaceLinksFromText5() {
// we know this one doesn't work properly, which is why html should always be sanitized before being passed into the ReplaceLinks function
replaced := suite.formatter.ReplaceLinks(context.Background(), text5)
assert.Equal(suite.T(), `
suite.Equal(`
what happens when we already have a link within an href?
<a href="<a href="https://example.org" rel="noopener">example.org</a>"><a href="https://example.org" rel="noopener">example.org</a></a>

View file

@ -26,13 +26,11 @@ import (
)
func (f *formatter) FromMarkdown(ctx context.Context, md string, mentions []*gtsmodel.Mention, tags []*gtsmodel.Tag) string {
content := preformat(md)
// do the markdown parsing *first*
contentBytes := blackfriday.Run([]byte(content))
contentBytes := blackfriday.Run([]byte(md))
// format tags nicely
content = f.ReplaceTags(ctx, string(contentBytes), tags)
content := f.ReplaceTags(ctx, string(contentBytes), tags)
// format mentions nicely
content = f.ReplaceMentions(ctx, content, mentions)

View file

@ -44,15 +44,19 @@ that was some JSON :)
`
const (
simpleMarkdown = "# Title\n\nHere's a simple text in markdown.\n\nHere's a [link](https://example.org)."
simpleMarkdownExpected = "<h1>Title</h1>\n\n<p>Heres a simple text in markdown.</p>\n\n<p>Heres a <a href=\"https://example.org\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">link</a>.</p>\n"
withCodeBlockExpected = "<h1>Title</h1>\n\n<p>Below is some JSON.</p>\n\n<pre><code class=\"language-json\">{\n &#34;key&#34;: &#34;value&#34;,\n &#34;another_key&#34;: [\n &#34;value1&#34;,\n &#34;value2&#34;\n ]\n}\n</code></pre>\n\n<p>that was some JSON :)</p>\n"
withInlineCode = "`Nobody tells you about the <code><del>SECRET CODE</del></code>, do they?`"
withInlineCodeExpected = "<p><code>Nobody tells you about the &lt;code&gt;&lt;del&gt;SECRET CODE&lt;/del&gt;&lt;/code&gt;, do they?</code></p>\n"
withInlineCode2 = "`Nobody tells you about the </code><del>SECRET CODE</del><code>, do they?`"
withInlineCode2Expected = "<p><code>Nobody tells you about the &lt;/code&gt;&lt;del&gt;SECRET CODE&lt;/del&gt;&lt;code&gt;, do they?</code></p>\n"
withHashtag = "# Title\n\nhere's a simple status that uses hashtag #Hashtag!"
withHashtagExpected = "<h1>Title</h1>\n\n<p>heres a simple status that uses hashtag <a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a>!</p>\n"
simpleMarkdown = "# Title\n\nHere's a simple text in markdown.\n\nHere's a [link](https://example.org)."
simpleMarkdownExpected = "<h1>Title</h1>\n\n<p>Heres a simple text in markdown.</p>\n\n<p>Heres a <a href=\"https://example.org\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">link</a>.</p>\n"
withCodeBlockExpected = "<h1>Title</h1>\n\n<p>Below is some JSON.</p>\n\n<pre><code class=\"language-json\">{\n &#34;key&#34;: &#34;value&#34;,\n &#34;another_key&#34;: [\n &#34;value1&#34;,\n &#34;value2&#34;\n ]\n}\n</code></pre>\n\n<p>that was some JSON :)</p>\n"
withInlineCode = "`Nobody tells you about the <code><del>SECRET CODE</del></code>, do they?`"
withInlineCodeExpected = "<p><code>Nobody tells you about the &lt;code&gt;&lt;del&gt;SECRET CODE&lt;/del&gt;&lt;/code&gt;, do they?</code></p>\n"
withInlineCode2 = "`Nobody tells you about the </code><del>SECRET CODE</del><code>, do they?`"
withInlineCode2Expected = "<p><code>Nobody tells you about the &lt;/code&gt;&lt;del&gt;SECRET CODE&lt;/del&gt;&lt;code&gt;, do they?</code></p>\n"
withHashtag = "# Title\n\nhere's a simple status that uses hashtag #Hashtag!"
withHashtagExpected = "<h1>Title</h1>\n\n<p>heres a simple status that uses hashtag <a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a>!</p>\n"
mdWithHTML = "# Title\n\nHere's a simple text in markdown.\n\nHere's a <a href=\"https://example.org\">link</a>.\n\nHere's an image: <img src=\"https://gts.superseriousbusiness.org/assets/logo.png\" alt=\"The GoToSocial sloth logo.\" width=\"500\" height=\"600\">"
mdWithHTMLExpected = "<h1>Title</h1>\n\n<p>Heres a simple text in markdown.</p>\n\n<p>Heres a <a href=\"https://example.org\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">link</a>.</p>\n\n<p>Heres an image: <img src=\"https://gts.superseriousbusiness.org/assets/logo.png\" alt=\"The GoToSocial sloth logo.\" width=\"500\" height=\"600\" crossorigin=\"anonymous\"></p>\n"
mdWithCheekyHTML = "# Title\n\nHere's a simple text in markdown.\n\nHere's a cheeky little script: <script>alert(ahhhh)</script>"
mdWithCheekyHTMLExpected = "<h1>Title</h1>\n\n<p>Heres a simple text in markdown.</p>\n\n<p>Heres a cheeky little script: </p>\n"
)
type MarkdownTestSuite struct {
@ -88,6 +92,16 @@ func (suite *MarkdownTestSuite) TestParseWithHashtag() {
suite.Equal(withHashtagExpected, s)
}
func (suite *MarkdownTestSuite) TestParseWithHTML() {
s := suite.formatter.FromMarkdown(context.Background(), mdWithHTML, nil, nil)
suite.Equal(mdWithHTMLExpected, s)
}
func (suite *MarkdownTestSuite) TestParseWithCheekyHTML() {
s := suite.formatter.FromMarkdown(context.Background(), mdWithCheekyHTML, nil, nil)
suite.Equal(mdWithCheekyHTMLExpected, s)
}
func TestMarkdownTestSuite(t *testing.T) {
suite.Run(t, new(MarkdownTestSuite))
}

View file

@ -1,39 +0,0 @@
/*
GoToSocial
Copyright (C) 2021-2022 GoToSocial Authors admin@gotosocial.org
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>.
*/
package text
import (
"github.com/tdewolff/minify/v2"
"github.com/tdewolff/minify/v2/html"
)
var m *minify.M
// MinifyHTML runs html through a minifier, reducing it in size.
func MinifyHTML(in string) (string, error) {
if m == nil {
m = minify.New()
m.Add("text/html", &html.Minifier{
KeepQuotes: true,
KeepEndTags: true,
KeepDocumentTags: true,
})
}
return m.String("text/html", in)
}

View file

@ -20,6 +20,7 @@ package text
import (
"context"
"html"
"strings"
"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
@ -32,10 +33,11 @@ var breakReplacer = strings.NewReplacer(
)
func (f *formatter) FromPlain(ctx context.Context, plain string, mentions []*gtsmodel.Mention, tags []*gtsmodel.Tag) string {
content := preformat(plain)
// trim any crap
content := strings.TrimSpace(plain)
// sanitize any html elements
content = removeHTML(content)
// clean 'er up
content = html.EscapeString(content)
// format links nicely
content = f.ReplaceLinks(ctx, content)
@ -52,5 +54,5 @@ func (f *formatter) FromPlain(ctx context.Context, plain string, mentions []*gts
// wrap the whole thing in a pee
content = `<p>` + content + `</p>`
return postformat(content)
return SanitizeHTML(content)
}

View file

@ -20,27 +20,21 @@ package text_test
import (
"context"
"fmt"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/suite"
"github.com/superseriousbusiness/gotosocial/internal/gtsmodel"
)
const (
simple = "this is a plain and simple status"
simpleExpected = "<p>this is a plain and simple status</p>"
withTag = "here's a simple status that uses hashtag #welcome!"
withTagExpected = "<p>here's a simple status that uses hashtag <a href=\"http://localhost:8080/tags/welcome\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>welcome</span></a>!</p>"
moreComplex = `Another test @foss_satan@fossbros-anonymous.io
#Hashtag
Text`
moreComplexFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br><br><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br><br>Text</p>"
simple = "this is a plain and simple status"
simpleExpected = "<p>this is a plain and simple status</p>"
withTag = "here's a simple status that uses hashtag #welcome!"
withTagExpected = "<p>here&#39;s a simple status that uses hashtag <a href=\"http://localhost:8080/tags/welcome\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>welcome</span></a>!</p>"
withHTML = "<div>blah this should just be html escaped blah</div>"
withHTMLExpected = "<p>&lt;div&gt;blah this should just be html escaped blah&lt;/div&gt;</p>"
moreComplex = "Another test @foss_satan@fossbros-anonymous.io\n\n#Hashtag\n\nText"
moreComplexFull = "<p>Another test <span class=\"h-card\"><a href=\"http://fossbros-anonymous.io/@foss_satan\" class=\"u-url mention\" rel=\"nofollow noreferrer noopener\" target=\"_blank\">@<span>foss_satan</span></a></span><br/><br/><a href=\"http://localhost:8080/tags/Hashtag\" class=\"mention hashtag\" rel=\"tag nofollow noreferrer noopener\" target=\"_blank\">#<span>Hashtag</span></a><br/><br/>Text</p>"
)
type PlainTestSuite struct {
@ -49,7 +43,7 @@ type PlainTestSuite struct {
func (suite *PlainTestSuite) TestParseSimple() {
f := suite.formatter.FromPlain(context.Background(), simple, nil, nil)
assert.Equal(suite.T(), simpleExpected, f)
suite.Equal(simpleExpected, f)
}
func (suite *PlainTestSuite) TestParseWithTag() {
@ -58,7 +52,12 @@ func (suite *PlainTestSuite) TestParseWithTag() {
}
f := suite.formatter.FromPlain(context.Background(), withTag, nil, foundTags)
assert.Equal(suite.T(), withTagExpected, f)
suite.Equal(withTagExpected, f)
}
func (suite *PlainTestSuite) TestParseWithHTML() {
f := suite.formatter.FromPlain(context.Background(), withHTML, nil, nil)
suite.Equal(withHTMLExpected, f)
}
func (suite *PlainTestSuite) TestParseMoreComplex() {
@ -71,10 +70,7 @@ func (suite *PlainTestSuite) TestParseMoreComplex() {
}
f := suite.formatter.FromPlain(context.Background(), moreComplex, foundMentions, foundTags)
fmt.Println(f)
assert.Equal(suite.T(), moreComplexFull, f)
suite.Equal(moreComplexFull, f)
}
func TestPlainTestSuite(t *testing.T) {

View file

@ -19,7 +19,9 @@
package text
import (
"html"
"regexp"
"strings"
"github.com/microcosm-cc/bluemonday"
)
@ -59,7 +61,8 @@ func SanitizeHTML(in string) string {
// SanitizePlaintext runs text through basic sanitization. This removes
// any html elements that were in the string, and returns clean plaintext.
func SanitizePlaintext(in string) string {
content := preformat(in)
content := html.UnescapeString(in)
content = removeHTML(content)
return postformat(content)
content = html.UnescapeString(content)
return strings.TrimSpace(content)
}