summaryrefslogtreecommitdiff
path: root/internal/text/link_test.go
blob: dfb4656b8982654a30776a05a295627218dd4b2e (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
/*
   GoToSocial
   Copyright (C) 2021-2023 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_test

import (
	"context"
	"testing"

	"github.com/stretchr/testify/assert"
	"github.com/stretchr/testify/suite"
	"github.com/superseriousbusiness/gotosocial/internal/text"
)

const text1 = `
This is a text with some links in it. Here's link number one: https://example.org/link/to/something#fragment

Here's link number two: http://test.example.org?q=bahhhhhhhhhhhh

https://another.link.example.org/with/a/pretty/long/path/at/the/end/of/it

really.cool.website <-- this one shouldn't be parsed as a link because it doesn't contain the scheme

https://example.orghttps://google.com <-- this shouldn't work either, but it does?! OK
`

const text2 = `
this is one link: https://example.org

this is the same link again: https://example.org

these should be deduplicated
`

const text3 = `
here's a mailto link: mailto:whatever@test.org
`

const text4 = `
two similar links:

https://example.org

https://example.org/test
`

const text5 = `
what happens when we already have a link within an href?

<a href="https://example.org">https://example.org</a>
`

type LinkTestSuite struct {
	TextStandardTestSuite
}

func (suite *LinkTestSuite) TestParseSimple() {
	f := suite.formatter.FromPlain(context.Background(), simple, nil, nil)
	suite.Equal(simpleExpected, f)
}

func (suite *LinkTestSuite) TestParseURLsFromText1() {
	urls := text.FindLinks(text1)

	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() {
	urls := text.FindLinks(text2)

	// assert length 1 because the found links will be deduplicated
	assert.Len(suite.T(), urls, 1)
}

func (suite *LinkTestSuite) TestParseURLsFromText3() {
	urls := text.FindLinks(text3)

	// assert length 0 because `mailto:` isn't accepted
	assert.Len(suite.T(), urls, 0)
}

func (suite *LinkTestSuite) TestReplaceLinksFromText1() {
	replaced := suite.formatter.ReplaceLinks(context.Background(), text1)
	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>

<a href="https://another.link.example.org/with/a/pretty/long/path/at/the/end/of/it" rel="noopener">another.link.example.org/with/a/pretty/long/path/at/the/end/of/it</a>

really.cool.website <-- this one shouldn't be parsed as a link because it doesn't contain the scheme

<a href="https://example.orghttps://google.com" rel="noopener">example.orghttps://google.com</a> <-- this shouldn't work either, but it does?! OK
`, replaced)
}

func (suite *LinkTestSuite) TestReplaceLinksFromText2() {
	replaced := suite.formatter.ReplaceLinks(context.Background(), text2)
	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>

these should be deduplicated
`, replaced)
}

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)
	suite.Equal(`
here's a mailto link: mailto:whatever@test.org
`, replaced)
}

func (suite *LinkTestSuite) TestReplaceLinksFromText4() {
	replaced := suite.formatter.ReplaceLinks(context.Background(), text4)
	suite.Equal(`
two similar links:

<a href="https://example.org" rel="noopener">example.org</a>

<a href="https://example.org/test" rel="noopener">example.org/test</a>
`, replaced)
}

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)
	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>
`, replaced)
}

func TestLinkTestSuite(t *testing.T) {
	suite.Run(t, new(LinkTestSuite))
}