| Left: | ||
| Right: |
| LEFT | RIGHT |
|---|---|
| 1 #pragma once | 1 #pragma once |
| 2 | 2 |
| 3 #include <cstddef> | 3 #include <cstddef> |
| 4 #include <cmath> | 4 #include <cmath> |
| 5 #include <initializer_list> | 5 #include <initializer_list> |
| 6 #include <memory> | 6 #include <memory> |
| 7 | 7 |
| 8 #include "String.h" | 8 #include "String.h" |
| 9 #include "debug.h" | 9 #include "debug.h" |
| 10 | 10 |
| (...skipping 45 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 56 { | 56 { |
| 57 return mPos != it.mPos; | 57 return mPos != it.mPos; |
| 58 } | 58 } |
| 59 }; | 59 }; |
| 60 | 60 |
| 61 template<typename Entry> | 61 template<typename Entry> |
| 62 struct HashContainerReference | 62 struct HashContainerReference |
| 63 { | 63 { |
| 64 typedef Entry entry_type; | 64 typedef Entry entry_type; |
| 65 | 65 |
| 66 const entry_type* mEntry; | 66 entry_type* mEntry; |
| 67 | 67 |
| 68 explicit HashContainerReference(const entry_type* entry) | 68 explicit HashContainerReference(entry_type* entry) |
| 69 : mEntry(entry) | 69 : mEntry(entry) |
| 70 { | 70 { |
| 71 } | 71 } |
| 72 | 72 |
| 73 const entry_type* operator->() const | 73 const entry_type* operator->() const |
| 74 { | 74 { |
| 75 return mEntry; | 75 return mEntry; |
| 76 } | 76 } |
| 77 | 77 |
| 78 operator bool() const | 78 operator bool() const |
| 79 { | 79 { |
| 80 return !mEntry->first.is_invalid(); | 80 return !mEntry->first.is_invalid(); |
| 81 } | 81 } |
| 82 }; | 82 }; |
| 83 | 83 |
| 84 template<typename Entry> | 84 template<typename Entry> |
| 85 class HashContainer | 85 class HashContainer |
| 86 { | 86 { |
| 87 public: | 87 public: |
| 88 typedef Entry entry_type; | 88 typedef Entry entry_type; |
| 89 typedef size_t size_type; | 89 typedef size_t size_type; |
| 90 typedef HashContainerIterator<Entry> const_iterator; | 90 typedef HashContainerIterator<Entry> const_iterator; |
| 91 typedef HashContainerReference<Entry> const_reference; | 91 typedef HashContainerReference<const Entry> const_reference; |
| 92 | 92 |
| 93 private: | 93 private: |
| 94 HashContainer(const HashContainer& other); | 94 explicit HashContainer(const HashContainer& other); |
| 95 void operator=(const HashContainer& other); | 95 void operator=(const HashContainer& other); |
| 96 | 96 |
| 97 protected: | 97 protected: |
| 98 static constexpr size_type MIN_BUCKETS = 1; | 98 static constexpr size_type MIN_BUCKETS = 1; |
| 99 static constexpr double LOAD_FACTOR = 0.8; | 99 static constexpr double LOAD_FACTOR = 0.8; |
| 100 std::unique_ptr<entry_type[]> mBuckets; | 100 std::unique_ptr<entry_type[]> mBuckets; |
| 101 size_type mBucketCount; | 101 size_type mBucketCount; |
| 102 size_type mEntryCount; | 102 size_type mEntryCount; |
| 103 | 103 |
| 104 #if defined(DEBUG) | |
| 105 size_type mInsertCounter; | |
| 106 #endif | |
| 107 | |
| 104 explicit HashContainer(size_type expectedEntries = 0) | 108 explicit HashContainer(size_type expectedEntries = 0) |
| 105 : mEntryCount(0) | 109 : mEntryCount(0) |
| 106 { | 110 { |
| 107 expectedEntries = ceil(expectedEntries / LOAD_FACTOR); | 111 expectedEntries = ceil(expectedEntries / LOAD_FACTOR); |
| 108 mBucketCount = MIN_BUCKETS; | 112 mBucketCount = MIN_BUCKETS; |
| 109 while (mBucketCount < expectedEntries) | 113 while (mBucketCount < expectedEntries) |
| 110 mBucketCount <<= 1; | 114 mBucketCount <<= 1; |
| 111 | 115 |
| 112 mBuckets.reset(new entry_type[mBucketCount]); | 116 mBuckets.reset(new entry_type[mBucketCount]); |
| 113 // Working around https://github.com/waywardmonkeys/emscripten-trace-colle ctor/issues/2 here | 117 // Working around https://github.com/waywardmonkeys/emscripten-trace-colle ctor/issues/2 here |
| (...skipping 12 matching lines...) Expand all Loading... | |
| 126 entry_type* find_bucket(const String& key) const | 130 entry_type* find_bucket(const String& key) const |
| 127 { | 131 { |
| 128 size_type h = hash(key); | 132 size_type h = hash(key); |
| 129 | 133 |
| 130 // This does quadratic probing, effectively the following formula is used: | 134 // This does quadratic probing, effectively the following formula is used: |
| 131 // pos = (hash + 1/2 i + 1/2 i ^ 2) mod bucketCount | 135 // pos = (hash + 1/2 i + 1/2 i ^ 2) mod bucketCount |
| 132 for (size_type i = 0; ; ++i) | 136 for (size_type i = 0; ; ++i) |
| 133 { | 137 { |
| 134 // mBucketCount is 2^n so (h & mBucketCount - 1) is equivalent to | 138 // mBucketCount is 2^n so (h & mBucketCount - 1) is equivalent to |
| 135 // h % mBucketCount but significantly faster. | 139 // h % mBucketCount but significantly faster. |
| 136 entry_type* entry = &mBuckets[h & mBucketCount - 1]; | 140 entry_type* entry = &mBuckets[h & (mBucketCount - 1)]; |
| 137 if (entry->first.is_invalid() || entry->first.equals(key)) | 141 if (entry->first.is_invalid() || entry->first.equals(key)) |
| 138 return entry; | 142 return entry; |
| 139 h += i; | 143 h += i; |
| 140 } | 144 } |
| 141 } | 145 } |
| 142 | 146 |
| 143 void resize(size_type bucketCount) | 147 void resize(size_type bucketCount) |
| 144 { | 148 { |
| 145 std::unique_ptr<entry_type[]> oldBuckets(std::move(mBuckets)); | 149 std::unique_ptr<entry_type[]> oldBuckets(std::move(mBuckets)); |
| 146 size_type oldCount = mBucketCount; | 150 size_type oldCount = mBucketCount; |
| 147 | 151 |
| 148 mEntryCount = 0; | 152 mEntryCount = 0; |
| 149 mBucketCount = bucketCount; | 153 mBucketCount = bucketCount; |
| 150 mBuckets.reset(new entry_type[mBucketCount]); | 154 mBuckets.reset(new entry_type[mBucketCount]); |
| 151 // Working around https://github.com/waywardmonkeys/emscripten-trace-colle ctor/issues/2 here | 155 // Working around https://github.com/waywardmonkeys/emscripten-trace-colle ctor/issues/2 here |
| 152 annotate_address(reinterpret_cast<size_type*>(mBuckets.get()) - 1, "Hash t able buffer"); | 156 annotate_address(reinterpret_cast<size_type*>(mBuckets.get()) - 1, "Hash t able buffer"); |
| 153 | 157 |
| 154 // Copy old entries into the new buffer | 158 // Copy old entries into the new buffer |
| 155 for (size_type i = 0; i < oldCount; i++) | 159 for (size_type i = 0; i < oldCount; i++) |
| 156 { | 160 { |
| 157 entry_type& entry = oldBuckets[i]; | 161 entry_type& entry = oldBuckets[i]; |
| 158 if (!entry.first.is_invalid()) | 162 if (!entry.first.is_invalid() && !entry.first.is_deleted()) |
| 159 { | 163 { |
| 160 *find_bucket(entry.first) = entry; | 164 *find_bucket(entry.first) = entry; |
| 161 mEntryCount++; | 165 mEntryCount++; |
| 162 } | 166 } |
| 163 } | 167 } |
| 164 } | 168 } |
| 165 | 169 |
| 166 entry_type* assign(entry_type* existing, const entry_type& entry) | 170 entry_type* assign(entry_type* existing, const entry_type& entry) |
| 167 { | 171 { |
| 168 if (existing->first.is_invalid()) | 172 if (existing->first.is_invalid()) |
| 169 { | 173 { |
| 170 if (mEntryCount + 1 >= mBucketCount * LOAD_FACTOR) | 174 if (mEntryCount + 1 >= mBucketCount * LOAD_FACTOR) |
| 171 { | 175 { |
| 172 resize(mBucketCount << 1); | 176 resize(mBucketCount << 1); |
| 173 existing = find_bucket(entry.first); | 177 existing = find_bucket(entry.first); |
| 174 } | 178 } |
| 175 mEntryCount++; | 179 mEntryCount++; |
| 180 #if defined(DEBUG) | |
| 181 mInsertCounter++; | |
| 182 #endif | |
| 176 } | 183 } |
| 177 *existing = entry; | 184 *existing = entry; |
| 178 return existing; | 185 return existing; |
| 179 } | 186 } |
| 180 | 187 |
| 181 public: | 188 public: |
| 182 void insert(const entry_type& entry) | 189 void insert(const entry_type& entry) |
| 183 { | 190 { |
| 184 assign(find_bucket(entry.first), entry); | 191 assign(find_bucket(entry.first), entry); |
| 185 } | 192 } |
| 186 | 193 |
| 194 bool erase(const String& key) | |
| 195 { | |
| 196 entry_type* entry = find_bucket(key); | |
| 197 if (entry->first.is_invalid()) | |
| 198 return false; | |
| 199 | |
| 200 entry->first.erase(); | |
| 201 return true; | |
| 202 } | |
| 203 | |
| 187 const_reference find(const String& key) const | 204 const_reference find(const String& key) const |
| 188 { | 205 { |
| 189 return const_reference(find_bucket(key)); | 206 return const_reference(find_bucket(key)); |
| 190 } | 207 } |
| 191 | 208 |
| 192 const_iterator begin() const | 209 const_iterator begin() const |
| 193 { | 210 { |
| 194 return const_iterator(&mBuckets[0], &mBuckets[mBucketCount]); | 211 return const_iterator(&mBuckets[0], &mBuckets[mBucketCount]); |
| 195 } | 212 } |
| 196 | 213 |
| (...skipping 12 matching lines...) Expand all Loading... | |
| 209 } | 226 } |
| 210 | 227 |
| 211 DependentString first; | 228 DependentString first; |
| 212 }; | 229 }; |
| 213 | 230 |
| 214 template<typename T> | 231 template<typename T> |
| 215 struct StringMapEntry | 232 struct StringMapEntry |
| 216 { | 233 { |
| 217 StringMapEntry() {} | 234 StringMapEntry() {} |
| 218 StringMapEntry(const String& key) | 235 StringMapEntry(const String& key) |
| 219 : first(key) | 236 : first(key), second() |
| 220 { | 237 { |
| 221 } | 238 } |
| 222 StringMapEntry(const String& key, T value) | 239 StringMapEntry(const String& key, T value) |
| 223 : first(key), second(value) | 240 : first(key), second(value) |
| 224 { | 241 { |
| 225 } | 242 } |
| 226 | 243 |
| 227 DependentString first; | 244 DependentString first; |
| 228 T second; | 245 T second; |
| 229 }; | 246 }; |
| 230 | 247 |
| 231 template<typename T> | 248 template<typename T> |
| 232 struct StringMapEntryReference : public HashContainerReference<StringMapEntry< T>> | 249 struct StringMapEntryReference : public HashContainerReference<StringMapEntry< T>> |
| 233 { | 250 { |
| 234 typedef HashContainerReference<StringMapEntry<T>> super; | 251 typedef HashContainerReference<StringMapEntry<T>> super; |
| 235 typedef typename super::entry_type entry_type; | 252 typedef typename super::entry_type entry_type; |
| 236 typedef StringMap<T> map_type; | 253 typedef StringMap<T> map_type; |
| 237 | 254 |
| 238 map_type* mMap; | 255 map_type* mMap; |
| 239 | 256 |
| 240 StringMapEntryReference(map_type* map, const entry_type* entry) | 257 #if defined(DEBUG) |
| 258 typename map_type::size_type mInsertCounter; | |
| 259 typename map_type::size_type mHash; | |
| 260 #endif | |
| 261 | |
| 262 StringMapEntryReference(map_type* map, const String& key, entry_type* entry) | |
| 241 : super(entry), mMap(map) | 263 : super(entry), mMap(map) |
| 242 { | 264 { |
| 265 #if defined(DEBUG) | |
| 266 mInsertCounter = mMap->mInsertCounter; | |
| 267 mHash = mMap->hash(key); | |
| 268 #endif | |
| 243 } | 269 } |
| 244 | 270 |
| 245 void assign(const String& key, const T& value) | 271 void assign(const String& key, const T& value) |
|
sergei
2016/02/22 12:46:03
I have noticed a couple of asserts here in the rep
Wladimir Palant
2016/02/23 12:37:36
As explained under https://codereview.adblockplus.
| |
| 246 { | 272 { |
| 247 mMap->assign(const_cast<entry_type*>(super::mEntry), entry_type(key, value )); | 273 #if defined(DEBUG) |
|
sergei
2016/02/22 12:46:02
`super::mEntry` is also accessible by `this->mEntr
sergei
2016/02/22 12:46:05
It's not necessary to use `const_cast` here. We ca
Wladimir Palant
2016/02/23 12:37:34
Done.
Wladimir Palant
2016/02/23 12:37:37
Done.
| |
| 274 assert(mInsertCounter == mMap->mInsertCounter, | |
| 275 u"There should be no insert operations performed between map.find() an d assign()"_str); | |
| 276 assert(mHash == mMap->hash(key), | |
| 277 u"The keys used in map.find() and assign() should be identical"_str); | |
| 278 #endif | |
| 279 | |
| 280 mMap->assign(this->mEntry, entry_type(key, value)); | |
| 248 } | 281 } |
| 249 }; | 282 }; |
| 250 } | 283 } |
| 251 | 284 |
| 252 class StringSet | 285 class StringSet |
| 253 : public StringMap_internal::HashContainer<StringMap_internal::StringSetEntry> | 286 : public StringMap_internal::HashContainer<StringMap_internal::StringSetEntry> |
| 254 { | 287 { |
| 255 }; | 288 }; |
| 256 | 289 |
| 257 template<typename T> | 290 template<typename T> |
| (...skipping 32 matching lines...) Expand 10 before | Expand all | Expand 10 after Loading... | |
| 290 return entry->second; | 323 return entry->second; |
| 291 } | 324 } |
| 292 | 325 |
| 293 const_reference find(const String& key) const | 326 const_reference find(const String& key) const |
| 294 { | 327 { |
| 295 return super::find(key); | 328 return super::find(key); |
| 296 } | 329 } |
| 297 | 330 |
| 298 reference find(const String& key) | 331 reference find(const String& key) |
| 299 { | 332 { |
| 300 return reference(this, super::find_bucket(key)); | 333 return reference(this, key, super::find_bucket(key)); |
| 301 } | 334 } |
| 302 }; | 335 }; |
| LEFT | RIGHT |