llvm/unittests/Support/TrigramIndexTest.cpp

   1 //===- TrigramIndexTest.cpp - Unit tests for TrigramIndex -----------------===//
   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/TrigramIndex.h"
  10 #include "llvm/ADT/STLExtras.h"
  11 #include "gtest/gtest.h"
  12
  13 #include <string>
  14 #include <vector>
  15
  16 using namespace llvm;
  17
  18 namespace {
  19
  20 class TrigramIndexTest : public ::testing::Test {
  21 protected:
  22   std::unique_ptr<TrigramIndex> makeTrigramIndex(
  23       std::vector<std::string> Rules) {
  24     std::unique_ptr<TrigramIndex> TI =
  25         std::make_unique<TrigramIndex>();
  26     for (auto &Rule : Rules)
  27       TI->insert(Rule);
  28     return TI;
  29   }
  30 };
  31
  32 TEST_F(TrigramIndexTest, Empty) {
  33   std::unique_ptr<TrigramIndex> TI =
  34       makeTrigramIndex({});
  35   EXPECT_FALSE(TI->isDefeated());
  36   EXPECT_TRUE(TI->isDefinitelyOut("foo"));
  37 }
  38
  39 TEST_F(TrigramIndexTest, Basic) {
  40   std::unique_ptr<TrigramIndex> TI =
  41       makeTrigramIndex({"*hello*", "*wor.d*"});
  42   EXPECT_FALSE(TI->isDefeated());
  43   EXPECT_TRUE(TI->isDefinitelyOut("foo"));
  44 }
  45
  46 TEST_F(TrigramIndexTest, NoTrigramsInRules) {
  47   std::unique_ptr<TrigramIndex> TI =
  48       makeTrigramIndex({"b.r", "za*az"});
  49   EXPECT_TRUE(TI->isDefeated());
  50   EXPECT_FALSE(TI->isDefinitelyOut("foo"));
  51   EXPECT_FALSE(TI->isDefinitelyOut("bar"));
  52   EXPECT_FALSE(TI->isDefinitelyOut("zakaz"));
  53 }
  54
  55 TEST_F(TrigramIndexTest, NoTrigramsInARule) {
  56   std::unique_ptr<TrigramIndex> TI =
  57       makeTrigramIndex({"*hello*", "*wo.ld*"});
  58   EXPECT_TRUE(TI->isDefeated());
  59   EXPECT_FALSE(TI->isDefinitelyOut("foo"));
  60 }
  61
  62 TEST_F(TrigramIndexTest, RepetitiveRule) {
  63   std::unique_ptr<TrigramIndex> TI =
  64       makeTrigramIndex({"*bar*bar*bar*bar*bar", "bar*bar"});
  65   EXPECT_FALSE(TI->isDefeated());
  66   EXPECT_TRUE(TI->isDefinitelyOut("foo"));
  67   EXPECT_TRUE(TI->isDefinitelyOut("bar"));
  68   EXPECT_FALSE(TI->isDefinitelyOut("barbara"));
  69   EXPECT_FALSE(TI->isDefinitelyOut("bar+bar"));
  70 }
  71
  72 TEST_F(TrigramIndexTest, PopularTrigram) {
  73   std::unique_ptr<TrigramIndex> TI =
  74       makeTrigramIndex({"*aaa*", "*aaaa*", "*aaaaa*", "*aaaaa*", "*aaaaaa*"});
  75   EXPECT_TRUE(TI->isDefeated());
  76 }
  77
  78 TEST_F(TrigramIndexTest, PopularTrigram2) {
  79   std::unique_ptr<TrigramIndex> TI =
  80       makeTrigramIndex({"class1.h", "class2.h", "class3.h", "class4.h", "class.h"});
  81   EXPECT_TRUE(TI->isDefeated());
  82 }
  83
  84 TEST_F(TrigramIndexTest, TooComplicatedRegex) {
  85   std::unique_ptr<TrigramIndex> TI =
  86       makeTrigramIndex({"[0-9]+"});
  87   EXPECT_TRUE(TI->isDefeated());
  88 }
  89
  90 TEST_F(TrigramIndexTest, TooComplicatedRegex2) {
  91   std::unique_ptr<TrigramIndex> TI =
  92       makeTrigramIndex({"foo|bar"});
  93   EXPECT_TRUE(TI->isDefeated());
  94 }
  95
  96 TEST_F(TrigramIndexTest, EscapedSymbols) {
  97   std::unique_ptr<TrigramIndex> TI =
  98       makeTrigramIndex({"*c\\+\\+*", "*hello\\\\world*", "a\\tb", "a\\0b"});
  99   EXPECT_FALSE(TI->isDefeated());
 100   EXPECT_FALSE(TI->isDefinitelyOut("c++"));
 101   EXPECT_TRUE(TI->isDefinitelyOut("c\\+\\+"));
 102   EXPECT_FALSE(TI->isDefinitelyOut("hello\\world"));
 103   EXPECT_TRUE(TI->isDefinitelyOut("hello\\\\world"));
 104   EXPECT_FALSE(TI->isDefinitelyOut("atb"));
 105   EXPECT_TRUE(TI->isDefinitelyOut("a\\tb"));
 106   EXPECT_TRUE(TI->isDefinitelyOut("a\tb"));
 107   EXPECT_FALSE(TI->isDefinitelyOut("a0b"));
 108 }
 109
 110 TEST_F(TrigramIndexTest, Backreference1) {
 111   std::unique_ptr<TrigramIndex> TI =
 112       makeTrigramIndex({"*foo\\1*"});
 113   EXPECT_TRUE(TI->isDefeated());
 114 }
 115
 116 TEST_F(TrigramIndexTest, Backreference2) {
 117   std::unique_ptr<TrigramIndex> TI =
 118       makeTrigramIndex({"*foo\\2*"});
 119   EXPECT_TRUE(TI->isDefeated());
 120 }
 121
 122 TEST_F(TrigramIndexTest, Sequence) {
 123   std::unique_ptr<TrigramIndex> TI =
 124       makeTrigramIndex({"class1.h", "class2.h", "class3.h", "class4.h"});
 125   EXPECT_FALSE(TI->isDefeated());
 126   EXPECT_FALSE(TI->isDefinitelyOut("class1"));
 127   EXPECT_TRUE(TI->isDefinitelyOut("class.h"));
 128   EXPECT_TRUE(TI->isDefinitelyOut("class"));
 129 }
 130
 131 }  // namespace