1 //===- llvm/unittest/Support/RegexTest.cpp - Regex tests --===//
2 //
3 // The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9
10 #include "llvm/Support/Regex.h"
11 #include "llvm/ADT/SmallVector.h"
12 #include "gtest/gtest.h"
13 #include <cstring>
14
15 using namespace llvm;
16 namespace {
17
18 class RegexTest : public ::testing::Test {
19 };
20
TEST_F(RegexTest,Basics)21 TEST_F(RegexTest, Basics) {
22 Regex r1("^[0-9]+$");
23 EXPECT_TRUE(r1.match("916"));
24 EXPECT_TRUE(r1.match("9"));
25 EXPECT_FALSE(r1.match("9a"));
26
27 SmallVector<StringRef, 1> Matches;
28 Regex r2("[0-9]+");
29 EXPECT_TRUE(r2.match("aa216b", &Matches));
30 EXPECT_EQ(1u, Matches.size());
31 EXPECT_EQ("216", Matches[0].str());
32
33 Regex r3("[0-9]+([a-f])?:([0-9]+)");
34 EXPECT_TRUE(r3.match("9a:513b", &Matches));
35 EXPECT_EQ(3u, Matches.size());
36 EXPECT_EQ("9a:513", Matches[0].str());
37 EXPECT_EQ("a", Matches[1].str());
38 EXPECT_EQ("513", Matches[2].str());
39
40 EXPECT_TRUE(r3.match("9:513b", &Matches));
41 EXPECT_EQ(3u, Matches.size());
42 EXPECT_EQ("9:513", Matches[0].str());
43 EXPECT_EQ("", Matches[1].str());
44 EXPECT_EQ("513", Matches[2].str());
45
46 Regex r4("a[^b]+b");
47 std::string String="axxb";
48 String[2] = '\0';
49 EXPECT_FALSE(r4.match("abb"));
50 EXPECT_TRUE(r4.match(String, &Matches));
51 EXPECT_EQ(1u, Matches.size());
52 EXPECT_EQ(String, Matches[0].str());
53
54 std::string NulPattern="X[0-9]+X([a-f])?:([0-9]+)";
55 String="YX99a:513b";
56 NulPattern[7] = '\0';
57 Regex r5(NulPattern);
58 EXPECT_FALSE(r5.match(String));
59 EXPECT_FALSE(r5.match("X9"));
60 String[3]='\0';
61 EXPECT_TRUE(r5.match(String));
62 }
63
TEST_F(RegexTest,Backreferences)64 TEST_F(RegexTest, Backreferences) {
65 Regex r1("([a-z]+)_\\1");
66 SmallVector<StringRef, 4> Matches;
67 EXPECT_TRUE(r1.match("abc_abc", &Matches));
68 EXPECT_EQ(2u, Matches.size());
69 EXPECT_FALSE(r1.match("abc_ab", &Matches));
70
71 Regex r2("a([0-9])b\\1c\\1");
72 EXPECT_TRUE(r2.match("a4b4c4", &Matches));
73 EXPECT_EQ(2u, Matches.size());
74 EXPECT_EQ("4", Matches[1].str());
75 EXPECT_FALSE(r2.match("a2b2c3"));
76
77 Regex r3("a([0-9])([a-z])b\\1\\2");
78 EXPECT_TRUE(r3.match("a6zb6z", &Matches));
79 EXPECT_EQ(3u, Matches.size());
80 EXPECT_EQ("6", Matches[1].str());
81 EXPECT_EQ("z", Matches[2].str());
82 EXPECT_FALSE(r3.match("a6zb6y"));
83 EXPECT_FALSE(r3.match("a6zb7z"));
84 }
85
TEST_F(RegexTest,Substitution)86 TEST_F(RegexTest, Substitution) {
87 std::string Error;
88
89 EXPECT_EQ("aNUMber", Regex("[0-9]+").sub("NUM", "a1234ber"));
90
91 // Standard Escapes
92 EXPECT_EQ("a\\ber", Regex("[0-9]+").sub("\\\\", "a1234ber", &Error));
93 EXPECT_EQ("", Error);
94 EXPECT_EQ("a\nber", Regex("[0-9]+").sub("\\n", "a1234ber", &Error));
95 EXPECT_EQ("", Error);
96 EXPECT_EQ("a\tber", Regex("[0-9]+").sub("\\t", "a1234ber", &Error));
97 EXPECT_EQ("", Error);
98 EXPECT_EQ("ajber", Regex("[0-9]+").sub("\\j", "a1234ber", &Error));
99 EXPECT_EQ("", Error);
100
101 EXPECT_EQ("aber", Regex("[0-9]+").sub("\\", "a1234ber", &Error));
102 EXPECT_EQ(Error, "replacement string contained trailing backslash");
103
104 // Backreferences
105 EXPECT_EQ("aa1234bber", Regex("a[0-9]+b").sub("a\\0b", "a1234ber", &Error));
106 EXPECT_EQ("", Error);
107
108 EXPECT_EQ("a1234ber", Regex("a([0-9]+)b").sub("a\\1b", "a1234ber", &Error));
109 EXPECT_EQ("", Error);
110
111 EXPECT_EQ("aber", Regex("a[0-9]+b").sub("a\\100b", "a1234ber", &Error));
112 EXPECT_EQ(Error, "invalid backreference string '100'");
113 }
114
TEST_F(RegexTest,IsLiteralERE)115 TEST_F(RegexTest, IsLiteralERE) {
116 EXPECT_TRUE(Regex::isLiteralERE("abc"));
117 EXPECT_FALSE(Regex::isLiteralERE("a(bc)"));
118 EXPECT_FALSE(Regex::isLiteralERE("^abc"));
119 EXPECT_FALSE(Regex::isLiteralERE("abc$"));
120 EXPECT_FALSE(Regex::isLiteralERE("a|bc"));
121 EXPECT_FALSE(Regex::isLiteralERE("abc*"));
122 EXPECT_FALSE(Regex::isLiteralERE("abc+"));
123 EXPECT_FALSE(Regex::isLiteralERE("abc?"));
124 EXPECT_FALSE(Regex::isLiteralERE("abc."));
125 EXPECT_FALSE(Regex::isLiteralERE("a[bc]"));
126 EXPECT_FALSE(Regex::isLiteralERE("abc\\1"));
127 EXPECT_FALSE(Regex::isLiteralERE("abc{1,2}"));
128 }
129
TEST_F(RegexTest,Escape)130 TEST_F(RegexTest, Escape) {
131 EXPECT_EQ("a\\[bc\\]", Regex::escape("a[bc]"));
132 EXPECT_EQ("abc\\{1\\\\,2\\}", Regex::escape("abc{1\\,2}"));
133 }
134
TEST_F(RegexTest,IsValid)135 TEST_F(RegexTest, IsValid) {
136 std::string Error;
137 EXPECT_FALSE(Regex("(foo").isValid(Error));
138 EXPECT_EQ("parentheses not balanced", Error);
139 EXPECT_FALSE(Regex("a[b-").isValid(Error));
140 EXPECT_EQ("invalid character range", Error);
141 }
142
TEST_F(RegexTest,MoveConstruct)143 TEST_F(RegexTest, MoveConstruct) {
144 Regex r1("^[0-9]+$");
145 Regex r2(std::move(r1));
146 EXPECT_TRUE(r2.match("916"));
147 }
148
TEST_F(RegexTest,MoveAssign)149 TEST_F(RegexTest, MoveAssign) {
150 Regex r1("^[0-9]+$");
151 Regex r2("abc");
152 r2 = std::move(r1);
153 EXPECT_TRUE(r2.match("916"));
154 }
155
156 }
157