OLD | NEW |
---|---|
1 // Copyright 2015 The Chromium Authors. All rights reserved. | 1 // Copyright 2015 The Chromium Authors. All rights reserved. |
2 // Use of this source code is governed by a BSD-style license that can be | 2 // Use of this source code is governed by a BSD-style license that can be |
3 // found in the LICENSE file. | 3 // found in the LICENSE file. |
4 | 4 |
5 #include "courgette/third_party/qsufsort.h" | 5 #include "courgette/third_party/bsdiff/qsufsort.h" |
6 | 6 |
7 #include <stddef.h> | 7 #include <stddef.h> |
8 | 8 |
9 #include <algorithm> | 9 #include <algorithm> |
10 #include <cstring> | 10 #include <cstring> |
11 #include <string> | 11 #include <string> |
12 #include <vector> | 12 #include <vector> |
13 | 13 |
14 #include "base/macros.h" | 14 #include "base/macros.h" |
15 #include "testing/gtest/include/gtest/gtest.h" | 15 #include "testing/gtest/include/gtest/gtest.h" |
16 | 16 |
17 TEST(QSufSortTest, Sort) { | 17 TEST(QSufSortTest, Sort) { |
18 const char* test_cases[] = { | 18 const char* test_cases[] = { |
19 "", | 19 "", |
huangs
2016/05/10 18:17:03
Is this recommended by formatter? I'd think 2 spa
altimin
2016/05/11 17:48:47
I've just run "git cl format" :)
| |
20 "a", | 20 "a", |
21 "za", | 21 "za", |
22 "CACAO", | 22 "CACAO", |
23 "banana", | 23 "banana", |
24 "tobeornottobe", | 24 "tobeornottobe", |
25 "The quick brown fox jumps over the lazy dog.", | 25 "The quick brown fox jumps over the lazy dog.", |
26 "elephantelephantelephantelephantelephant", | 26 "elephantelephantelephantelephantelephant", |
27 "-------------------------", | 27 "-------------------------", |
28 "011010011001011010010110011010010", | 28 "011010011001011010010110011010010", |
29 "3141592653589793238462643383279502884197169399375105", | 29 "3141592653589793238462643383279502884197169399375105", |
30 "\xFF\xFE\xFF\xFE\xFD\x80\x30\x31\x32\x80\x30\xFF\x01\xAB\xCD", | 30 "\xFF\xFE\xFF\xFE\xFD\x80\x30\x31\x32\x80\x30\xFF\x01\xAB\xCD", |
31 }; | 31 }; |
32 | 32 |
33 for (size_t idx = 0; idx < arraysize(test_cases); ++idx) { | 33 for (size_t idx = 0; idx < arraysize(test_cases); ++idx) { |
34 int len = static_cast<int>(::strlen(test_cases[idx])); | 34 int len = static_cast<int>(::strlen(test_cases[idx])); |
35 const unsigned char* s = | 35 const unsigned char* s = |
36 reinterpret_cast<const unsigned char*>(test_cases[idx]); | 36 reinterpret_cast<const unsigned char*>(test_cases[idx]); |
37 | 37 |
38 // Generate the suffix array as I. | 38 // Generate the suffix array as I. |
39 std::vector<int> I(len + 1); | 39 std::vector<int> I(len + 1); |
40 std::vector<int> V(len + 1); | 40 std::vector<int> V(len + 1); |
(...skipping 20 matching lines...) Expand all Loading... | |
61 } | 61 } |
62 } | 62 } |
63 | 63 |
64 TEST(QSufSortTest, Search) { | 64 TEST(QSufSortTest, Search) { |
65 // Initialize main string and the suffix array. | 65 // Initialize main string and the suffix array. |
66 // Positions: 00000000001111111111122222222233333333334444 | 66 // Positions: 00000000001111111111122222222233333333334444 |
67 // 01234567890123456789012345678901234567890123 | 67 // 01234567890123456789012345678901234567890123 |
68 const char* old_str = "the quick brown fox jumps over the lazy dog."; | 68 const char* old_str = "the quick brown fox jumps over the lazy dog."; |
69 int old_size = static_cast<int>(::strlen(old_str)); | 69 int old_size = static_cast<int>(::strlen(old_str)); |
70 const unsigned char* old_buf = | 70 const unsigned char* old_buf = |
71 reinterpret_cast<const unsigned char*>(old_str); | 71 reinterpret_cast<const unsigned char*>(old_str); |
72 std::vector<int> I(old_size + 1); | 72 std::vector<int> I(old_size + 1); |
73 std::vector<int> V(old_size + 1); | 73 std::vector<int> V(old_size + 1); |
74 courgette::qsuf::qsufsort<int*>(&I[0], &V[0], old_buf, old_size); | 74 courgette::qsuf::qsufsort<int*>(&I[0], &V[0], old_buf, old_size); |
75 | 75 |
76 // Test queries. | 76 // Test queries. |
77 const struct { | 77 const struct { |
78 int exp_pos; // -1 means "don't care". | 78 int exp_pos; // -1 means "don't care". |
79 int exp_match_len; | 79 int exp_match_len; |
80 const char* query_str; | 80 const char* query_str; |
81 } test_cases[] = { | 81 } test_cases[] = { |
82 // Entire string. | 82 // Entire string. |
huangs
2016/05/10 18:17:03
Same here re. spacing.
altimin
2016/05/11 17:48:47
I'm not a big fan of deviating from what "git cl f
| |
83 {0, 44, "the quick brown fox jumps over the lazy dog."}, | 83 {0, 44, "the quick brown fox jumps over the lazy dog."}, |
84 // Empty string. | 84 // Empty string. |
85 {-1, 0, ""}, // Current algorithm does not enforce |pos| == 0. | 85 {-1, 0, ""}, // Current algorithm does not enforce |pos| == 0. |
86 // Exact and unique suffix match. | 86 // Exact and unique suffix match. |
87 {43, 1, "."}, | 87 {43, 1, "."}, |
88 {31, 13, "the lazy dog."}, | 88 {31, 13, "the lazy dog."}, |
89 // Exact and unique non-suffix match. | 89 // Exact and unique non-suffix match. |
90 {4, 5, "quick"}, | 90 {4, 5, "quick"}, |
91 {0, 9, "the quick"}, // Unique prefix. | 91 {0, 9, "the quick"}, // Unique prefix. |
92 // Entire word match with mutiple results: take lexicographical first. | 92 // Entire word match with mutiple results: take lexicographical first. |
93 {31, 3, "the"}, // Non-unique prefix: "the l"... < "the q"... | 93 {31, 3, "the"}, // Non-unique prefix: "the l"... < "the q"... |
94 {9, 1, " "}, // " brown"... wins. | 94 {9, 1, " "}, // " brown"... wins. |
95 // Partial and unique match of query prefix. | 95 // Partial and unique match of query prefix. |
96 {16, 10, "fox jumps with the hosps"}, | 96 {16, 10, "fox jumps with the hosps"}, |
97 // Partial and multiple match of query prefix: no guarantees on |pos|. | 97 // Partial and multiple match of query prefix: no guarantees on |pos|. |
98 // Take lexicographical first for matching portion *only*, so same results: | 98 // Take lexicographical first for matching portion *only*, so same |
99 {-1, 4, "the apple"}, // query < "the l"... < "the q"... | 99 // results: |
100 {-1, 4, "the opera"}, // "the l"... < query < "the q"... | 100 {-1, 4, "the apple"}, // query < "the l"... < "the q"... |
101 {-1, 4, "the zebra"}, // "the l"... < "the q"... < query | 101 {-1, 4, "the opera"}, // "the l"... < query < "the q"... |
102 // Prefix match dominates suffix match. | 102 {-1, 4, "the zebra"}, // "the l"... < "the q"... < query |
103 {26, 5, "over quick brown fox"}, | 103 // Prefix match dominates suffix match. |
104 // No match. | 104 {26, 5, "over quick brown fox"}, |
105 {-1, 0, ","}, | 105 // No match. |
106 {-1, 0, "1234"}, | 106 {-1, 0, ","}, |
107 {-1, 0, "THE QUICK BROWN FOX"}, | 107 {-1, 0, "1234"}, |
108 {-1, 0, "(the"}, | 108 {-1, 0, "THE QUICK BROWN FOX"}, |
109 {-1, 0, "(the"}, | |
109 }; | 110 }; |
110 | 111 |
111 for (size_t idx = 0; idx < arraysize(test_cases); ++idx) { | 112 for (size_t idx = 0; idx < arraysize(test_cases); ++idx) { |
112 const auto& test_case = test_cases[idx]; | 113 const auto& test_case = test_cases[idx]; |
113 int new_size = static_cast<int>(::strlen(test_case.query_str)); | 114 int new_size = static_cast<int>(::strlen(test_case.query_str)); |
114 const unsigned char* new_buf = | 115 const unsigned char* new_buf = |
115 reinterpret_cast<const unsigned char*>(test_case.query_str); | 116 reinterpret_cast<const unsigned char*>(test_case.query_str); |
116 | 117 |
117 // Perform the search. | 118 // Perform the search. |
118 int pos = 0; | 119 int pos = 0; |
119 int match_len = courgette::qsuf::search( | 120 int match_len = courgette::qsuf::search(&I[0], old_buf, old_size, new_buf, |
120 &I[0], old_buf, old_size, new_buf, new_size, &pos); | 121 new_size, &pos); |
121 | 122 |
122 // Check basic properties and match with expected values. | 123 // Check basic properties and match with expected values. |
123 EXPECT_GE(match_len, 0) << "test_case[" << idx << "]"; | 124 EXPECT_GE(match_len, 0) << "test_case[" << idx << "]"; |
124 EXPECT_LE(match_len, new_size) << "test_case[" << idx << "]"; | 125 EXPECT_LE(match_len, new_size) << "test_case[" << idx << "]"; |
125 if (match_len > 0) { | 126 if (match_len > 0) { |
126 EXPECT_GE(pos, 0) << "test_case[" << idx << "]"; | 127 EXPECT_GE(pos, 0) << "test_case[" << idx << "]"; |
127 EXPECT_LE(pos, old_size - match_len) << "test_case[" << idx << "]"; | 128 EXPECT_LE(pos, old_size - match_len) << "test_case[" << idx << "]"; |
128 EXPECT_EQ(0, ::memcmp(old_buf + pos, new_buf, match_len)) | 129 EXPECT_EQ(0, ::memcmp(old_buf + pos, new_buf, match_len)) << "test_case[" |
129 << "test_case[" << idx << "]"; | 130 << idx << "]"; |
130 } | 131 } |
131 if (test_case.exp_pos >= 0) { | 132 if (test_case.exp_pos >= 0) { |
132 EXPECT_EQ(test_case.exp_pos, pos) << "test_case[" << idx << "]"; | 133 EXPECT_EQ(test_case.exp_pos, pos) << "test_case[" << idx << "]"; |
133 } | 134 } |
134 EXPECT_EQ(test_case.exp_match_len, match_len) << "test_case[" << idx << "]"; | 135 EXPECT_EQ(test_case.exp_match_len, match_len) << "test_case[" << idx << "]"; |
135 } | 136 } |
136 } | 137 } |
OLD | NEW |