1 // Copyright 2015 PDFium Authors. All rights reserved.
2 // Use of this source code is governed by a BSD-style license that can be
3 // found in the LICENSE file.
4
5 #include "core/fpdftext/cpdf_linkextract.h"
6
7 #include "testing/gtest/include/gtest/gtest.h"
8
9 // Class to help test functions in CPDF_LinkExtract class.
10 class CPDF_TestLinkExtract final : public CPDF_LinkExtract {
11 public:
CPDF_TestLinkExtract()12 CPDF_TestLinkExtract() : CPDF_LinkExtract(nullptr) {}
13
14 private:
15 // Add test cases as friends to access protected member functions.
16 // Access CheckMailLink and CheckWebLink.
17 FRIEND_TEST(CPDF_LinkExtractTest, CheckMailLink);
18 FRIEND_TEST(CPDF_LinkExtractTest, CheckWebLink);
19 };
20
TEST(CPDF_LinkExtractTest,CheckMailLink)21 TEST(CPDF_LinkExtractTest, CheckMailLink) {
22 CPDF_TestLinkExtract extractor;
23 // Check cases that fail to extract valid mail link.
24 const wchar_t* const invalid_strs[] = {
25 L"",
26 L"peter.pan", // '@' is required.
27 L"abc@server", // Domain name needs at least one '.'.
28 L"abc.@gmail.com", // '.' can not immediately precede '@'.
29 L"abc@xyz&q.org", // Domain name should not contain '&'.
30 L"abc@.xyz.org", // Domain name should not start with '.'.
31 L"fan@g..com" // Domain name should not have consecutive '.'
32 };
33 for (size_t i = 0; i < FX_ArraySize(invalid_strs); ++i) {
34 const wchar_t* const input = invalid_strs[i];
35 WideString text_str(input);
36 EXPECT_FALSE(extractor.CheckMailLink(&text_str)) << input;
37 }
38
39 // Check cases that can extract valid mail link.
40 // An array of {input_string, expected_extracted_email_address}.
41 const wchar_t* const valid_strs[][2] = {
42 {L"peter@abc.d", L"peter@abc.d"},
43 {L"red.teddy.b@abc.com", L"red.teddy.b@abc.com"},
44 {L"abc_@gmail.com", L"abc_@gmail.com"}, // '_' is ok before '@'.
45 {L"dummy-hi@gmail.com",
46 L"dummy-hi@gmail.com"}, // '-' is ok in user name.
47 {L"a..df@gmail.com", L"df@gmail.com"}, // Stop at consecutive '.'.
48 {L".john@yahoo.com", L"john@yahoo.com"}, // Remove heading '.'.
49 {L"abc@xyz.org?/", L"abc@xyz.org"}, // Trim ending invalid chars.
50 {L"fan{abc@xyz.org", L"abc@xyz.org"}, // Trim beginning invalid chars.
51 {L"fan@g.com..", L"fan@g.com"}, // Trim the ending periods.
52 {L"CAP.cap@Gmail.Com", L"CAP.cap@Gmail.Com"}, // Keep the original case.
53 };
54 for (size_t i = 0; i < FX_ArraySize(valid_strs); ++i) {
55 const wchar_t* const input = valid_strs[i][0];
56 WideString text_str(input);
57 WideString expected_str(L"mailto:");
58 expected_str += valid_strs[i][1];
59 EXPECT_TRUE(extractor.CheckMailLink(&text_str)) << input;
60 EXPECT_STREQ(expected_str.c_str(), text_str.c_str());
61 }
62 }
63
TEST(CPDF_LinkExtractTest,CheckWebLink)64 TEST(CPDF_LinkExtractTest, CheckWebLink) {
65 CPDF_TestLinkExtract extractor;
66 // Check cases that fail to extract valid web link.
67 // The last few are legit web addresses that we don't handle now.
68 const wchar_t* const invalid_cases[] = {
69 L"", L"http", L"www.", L"https-and-www",
70 L"http:/abc.com", // Missing slash.
71 L"http://((()),", // Only invalid chars in host name.
72 L"ftp://example.com", // Ftp scheme is not supported.
73 L"http:example.com", // Missing slashes.
74 L"http//[example.com", // Invalid IPv6 address.
75 L"http//[00:00:00:00:00:00", // Invalid IPv6 address.
76 L"http//[]", // Empty IPv6 address.
77 // Web addresses that in correct format that we don't handle.
78 L"abc.example.com", // URL without scheme.
79 };
80 const int32_t DEFAULT_VALUE = -42;
81 for (size_t i = 0; i < FX_ArraySize(invalid_cases); ++i) {
82 const wchar_t* const input = invalid_cases[i];
83 WideString text_str(input);
84 int32_t start_offset = DEFAULT_VALUE;
85 int32_t count = DEFAULT_VALUE;
86 EXPECT_FALSE(extractor.CheckWebLink(&text_str, &start_offset, &count))
87 << input;
88 EXPECT_EQ(DEFAULT_VALUE, start_offset) << input;
89 EXPECT_EQ(DEFAULT_VALUE, count) << input;
90 }
91
92 // Check cases that can extract valid web link.
93 // An array of {input_string, expected_extracted_web_link}.
94 struct ValidCase {
95 const wchar_t* const input_string;
96 const wchar_t* const url_extracted;
97 const int32_t start_offset;
98 const int32_t count;
99 };
100 const ValidCase valid_cases[] = {
101 {L"http://www.example.com", L"http://www.example.com", 0,
102 22}, // standard URL.
103 {L"http://www.example.com:88", L"http://www.example.com:88", 0,
104 25}, // URL with port number.
105 {L"http://test@www.example.com", L"http://test@www.example.com", 0,
106 27}, // URL with username.
107 {L"http://test:test@example.com", L"http://test:test@example.com", 0,
108 28}, // URL with username and password.
109 {L"http://example", L"http://example", 0,
110 14}, // URL with short domain name.
111 {L"http////www.server", L"http://www.server", 8,
112 10}, // URL starts with "www.".
113 {L"http:/www.abc.com", L"http://www.abc.com", 6,
114 11}, // URL starts with "www.".
115 {L"www.a.b.c", L"http://www.a.b.c", 0, 9}, // URL starts with "www.".
116 {L"https://a.us", L"https://a.us", 0, 12}, // Secure http URL.
117 {L"https://www.t.us", L"https://www.t.us", 0, 16}, // Secure http URL.
118 {L"www.example-test.com", L"http://www.example-test.com", 0,
119 20}, // '-' in host is ok.
120 {L"www.example.com,", L"http://www.example.com", 0,
121 15}, // Trim ending invalid chars.
122 {L"www.example.com;(", L"http://www.example.com", 0,
123 15}, // Trim ending invalid chars.
124 {L"test:www.abc.com", L"http://www.abc.com", 5,
125 11}, // Trim chars before URL.
126 {L"(http://www.abc.com)", L"http://www.abc.com", 1,
127 18}, // Trim external brackets.
128 {L"0(http://www.abc.com)0", L"http://www.abc.com", 2,
129 18}, // Trim chars outside brackets as well.
130 {L"0(www.abc.com)0", L"http://www.abc.com", 2,
131 11}, // Links without http should also have brackets trimmed.
132 {L"http://www.abc.com)0", L"http://www.abc.com)0", 0,
133 20}, // Do not trim brackets that were not opened.
134 {L"{(<http://www.abc.com>)}", L"http://www.abc.com", 3,
135 18}, // Trim chars with multiple levels of brackets.
136 {L"[http://www.abc.com/z(1)]", L"http://www.abc.com/z(1)", 1,
137 23}, // Brackets opened inside the URL should not be trimmed.
138 {L"(http://www.abc.com/z(1))", L"http://www.abc.com/z(1)", 1,
139 23}, // Brackets opened inside the URL should not be trimmed.
140 {L"\"http://www.abc.com\"", L"http://www.abc.com", 1,
141 18}, // External quotes can also be escaped
142 {L"www.g.com..", L"http://www.g.com..", 0, 11}, // Leave ending periods.
143
144 // Web links can contain IP addresses too.
145 {L"http://192.168.0.1", L"http://192.168.0.1", 0, 18}, // IPv4 address.
146 {L"http://192.168.0.1:80", L"http://192.168.0.1:80", 0,
147 21}, // IPv4 address with port.
148 {L"http://[aa::00:bb::00:cc:00]", L"http://[aa::00:bb::00:cc:00]", 0,
149 28}, // IPv6 reference.
150 {L"http://[aa::00:bb::00:cc:00]:12", L"http://[aa::00:bb::00:cc:00]:12",
151 0, 31}, // IPv6 reference with port.
152 {L"http://[aa]:12", L"http://[aa]:12", 0,
153 14}, // Not validate IP address.
154 {L"http://[aa]:12abc", L"http://[aa]:12", 0,
155 14}, // Trim for IPv6 address.
156 {L"http://[aa]:", L"http://[aa]", 0, 11}, // Trim for IPv6 address.
157
158 // Path and query parts can be anything.
159 {L"www.abc.com/#%%^&&*(", L"http://www.abc.com/#%%^&&*(", 0, 20},
160 {L"www.a.com/#a=@?q=rr&r=y", L"http://www.a.com/#a=@?q=rr&r=y", 0, 23},
161 {L"http://a.com/1/2/3/4\5\6", L"http://a.com/1/2/3/4\5\6", 0, 22},
162 {L"http://www.example.com/foo;bar", L"http://www.example.com/foo;bar", 0,
163 30},
164
165 // Invalid chars inside host name are ok as we don't validate them.
166 {L"http://ex[am]ple", L"http://ex[am]ple", 0, 16},
167 {L"http://:example.com", L"http://:example.com", 0, 19},
168 {L"http://((())/path?", L"http://((())/path?", 0, 18},
169 {L"http:////abc.server", L"http:////abc.server", 0, 19},
170
171 // Non-ASCII chars are not validated either.
172 {L"www.测试.net", L"http://www.测试.net", 0, 10},
173 {L"www.测试。net。", L"http://www.测试。net。", 0, 11},
174 {L"www.测试.net；", L"http://www.测试.net；", 0, 11},
175 };
176 for (size_t i = 0; i < FX_ArraySize(valid_cases); ++i) {
177 const wchar_t* const input = valid_cases[i].input_string;
178 WideString text_str(input);
179 int32_t start_offset = DEFAULT_VALUE;
180 int32_t count = DEFAULT_VALUE;
181 EXPECT_TRUE(extractor.CheckWebLink(&text_str, &start_offset, &count))
182 << input;
183 EXPECT_STREQ(valid_cases[i].url_extracted, text_str.c_str());
184 EXPECT_EQ(valid_cases[i].start_offset, start_offset) << input;
185 EXPECT_EQ(valid_cases[i].count, count) << input;
186 }
187 }
188