147
|
1 //===---------- llvm/unittest/Support/DJBTest.cpp -------------------------===//
|
|
2 //
|
|
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
4 // See https://llvm.org/LICENSE.txt for license information.
|
|
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
|
6 //
|
|
7 //===----------------------------------------------------------------------===//
|
|
8
|
|
9 #include "llvm/Support/DJB.h"
|
|
10 #include "llvm/ADT/Twine.h"
|
|
11 #include "gtest/gtest.h"
|
|
12
|
|
13 using namespace llvm;
|
|
14
|
|
15 TEST(DJBTest, caseFolding) {
|
|
16 struct TestCase {
|
|
17 StringLiteral One;
|
|
18 StringLiteral Two;
|
|
19 };
|
|
20
|
|
21 static constexpr TestCase Tests[] = {
|
|
22 {{"ASDF"}, {"asdf"}},
|
|
23 {{"qWeR"}, {"QwEr"}},
|
|
24 {{"qqqqqqqqqqqqqqqqqqqq"}, {"QQQQQQQQQQQQQQQQQQQQ"}},
|
|
25
|
|
26 {{"I"}, {"i"}},
|
|
27 // Latin Small Letter Dotless I
|
|
28 {{u8"\u0130"}, {"i"}},
|
|
29 // Latin Capital Letter I With Dot Above
|
|
30 {{u8"\u0131"}, {"i"}},
|
|
31
|
|
32 // Latin Capital Letter A With Grave
|
|
33 {{u8"\u00c0"}, {u8"\u00e0"}},
|
|
34 // Latin Capital Letter A With Macron
|
|
35 {{u8"\u0100"}, {u8"\u0101"}},
|
|
36 // Latin Capital Letter L With Acute
|
|
37 {{u8"\u0139"}, {u8"\u013a"}},
|
|
38 // Cyrillic Capital Letter Ie
|
|
39 {{u8"\u0415"}, {u8"\u0435"}},
|
|
40 // Latin Capital Letter A With Circumflex And Grave
|
|
41 {{u8"\u1ea6"}, {u8"\u1ea7"}},
|
|
42 // Kelvin Sign
|
|
43 {{u8"\u212a"}, {u8"\u006b"}},
|
|
44 // Glagolitic Capital Letter Chrivi
|
|
45 {{u8"\u2c1d"}, {u8"\u2c4d"}},
|
|
46 // Fullwidth Latin Capital Letter M
|
|
47 {{u8"\uff2d"}, {u8"\uff4d"}},
|
|
48 // Old Hungarian Capital Letter Ej
|
|
49 {{u8"\U00010c92"}, {u8"\U00010cd2"}},
|
|
50 };
|
|
51
|
|
52 for (const TestCase &T : Tests) {
|
|
53 SCOPED_TRACE("Comparing '" + T.One + "' and '" + T.Two + "'");
|
|
54 EXPECT_EQ(caseFoldingDjbHash(T.One), caseFoldingDjbHash(T.Two));
|
|
55 }
|
|
56 }
|
|
57
|
|
58 TEST(DJBTest, knownValuesLowerCase) {
|
|
59 struct TestCase {
|
|
60 StringLiteral Text;
|
|
61 uint32_t Hash;
|
|
62 };
|
|
63 static constexpr TestCase Tests[] = {
|
|
64 {{""}, 5381u},
|
|
65 {{"f"}, 177675u},
|
|
66 {{"fo"}, 5863386u},
|
|
67 {{"foo"}, 193491849u},
|
|
68 {{"foob"}, 2090263819u},
|
|
69 {{"fooba"}, 259229388u},
|
|
70 {{"foobar"}, 4259602622u},
|
|
71 {{"pneumonoultramicroscopicsilicovolcanoconiosis"}, 3999417781u},
|
|
72 };
|
|
73
|
|
74 for (const TestCase &T : Tests) {
|
|
75 SCOPED_TRACE("Text: '" + T.Text + "'");
|
|
76 EXPECT_EQ(T.Hash, djbHash(T.Text));
|
|
77 EXPECT_EQ(T.Hash, caseFoldingDjbHash(T.Text));
|
|
78 EXPECT_EQ(T.Hash, caseFoldingDjbHash(T.Text.upper()));
|
|
79 }
|
|
80 }
|
|
81
|
|
82 TEST(DJBTest, knownValuesUnicode) {
|
|
83 EXPECT_EQ(5866553u, djbHash(u8"\u0130"));
|
|
84 EXPECT_EQ(177678u, caseFoldingDjbHash(u8"\u0130"));
|
|
85 EXPECT_EQ(
|
|
86 1302161417u,
|
|
87 djbHash(
|
|
88 u8"\u0130\u0131\u00c0\u00e0\u0100\u0101\u0139\u013a\u0415\u0435\u1ea6"
|
|
89 u8"\u1ea7\u212a\u006b\u2c1d\u2c4d\uff2d\uff4d\U00010c92\U00010cd2"));
|
|
90 EXPECT_EQ(
|
|
91 1145571043u,
|
|
92 caseFoldingDjbHash(
|
|
93 u8"\u0130\u0131\u00c0\u00e0\u0100\u0101\u0139\u013a\u0415\u0435\u1ea6"
|
|
94 u8"\u1ea7\u212a\u006b\u2c1d\u2c4d\uff2d\uff4d\U00010c92\U00010cd2"));
|
|
95 }
|