forked from HIT-SCIR/ltp
-
Notifications
You must be signed in to change notification settings - Fork 0
/
template.hpp
318 lines (278 loc) · 7.81 KB
/
template.hpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
//! A light weight template engine which support few template functions. It's
//! designed for feature extraction in various NLP tasks. Speed is mainly
//! concerned. Since template is treated as an internal program-specified, we
//! didn't use the safe functions like strnlen, strncmp, strncpy just for
//! consideration of speed.
#ifndef __TEMPLATE_HPP__
#define __TEMPLATE_HPP__
#include <iostream>
#include <string.h>
#include <stdlib.h>
#include <map>
#include "unordered_map.hpp"
#include "hasher.hpp"
namespace ltp {
namespace utility {
// a singleton used to story (token, id) mapping.
// A Template::Data should make a copy of the cache class.
template <typename T>
class __Template_Token_Cache {
private:
typedef std::unordered_map<const char *, int,
__Default_CharArray_HashFunction,
__Default_CharArray_EqualFunction> indexing_t;
public:
/**
* The entry function for the singleton. It's a typical get_instance
* method in design pattern.
*
* @return __Template_Token_Cache* The cache singleton.
*/
static __Template_Token_Cache * get_cache() {
if (0 == _instance) {
_instance = new __Template_Token_Cache;
}
return _instance;
}
/**
* Push a key into the cache.
*
*
*
*/
int push_back(const char * key) {
// allocate new data
int new_num;
int len = strlen(key) + 1;
for (int i = 0; i < _num_tokens; ++ i) {
if (!strcmp(key, _tokens[i])) {
return i;
}
}
// The self-implemented allocator.
if (_cap_tokens <= (new_num = (_num_tokens + 1))) {
_cap_tokens = (new_num << 1);
char ** new_tokens = new char *[_cap_tokens];
if (_tokens) {
memcpy(new_tokens, _tokens, sizeof(char *) * _num_tokens);
delete [](_tokens);
}
_tokens = new_tokens;
// [TRICK] Should re-hash all the index. However, the cost for
// forcely rehashing is not too high because number of entries
// in templates are relatively low.
for (int i = 0; i< _num_tokens; ++ i) {
_indexing[_tokens[i]] = i;
}
}
_tokens[_num_tokens] = new char[len];
memcpy(_tokens[_num_tokens], key, len);
_indexing[_tokens[_num_tokens]] = _num_tokens;
++ _num_tokens;
return _num_tokens - 1;
}
int index(const char* key) {
indexing_t::const_iterator itx = _indexing.find(key);
if (itx != _indexing.end()) {
return itx->second;
}
return -1;
}
const char* at(int idx) {
if (idx < 0 || idx >= _num_tokens) {
return 0;
}
return _tokens[idx];
}
int num_tokens() {
return _num_tokens;
}
// this is still a shit that so many stars in the arguments.
void copy(char ** & tokens) {
if (0 == tokens) {
tokens = new char *[_num_tokens];
}
for (int i = 0; i < _num_tokens; ++ i) {
int len = strlen(_tokens[i]) + 1;
tokens[i] = new char[len];
memcpy(tokens[i], _tokens[i], len);
}
}
protected:
__Template_Token_Cache() :
_tokens(0),
_num_tokens(0),
_cap_tokens(0) {}
private:
static __Template_Token_Cache * _instance;
char ** _tokens;
int _num_tokens;
int _cap_tokens;
indexing_t _indexing;
};
template<typename T> __Template_Token_Cache<T>* __Template_Token_Cache<T>::_instance = NULL;
// The template class
class Template {
public:
// The template data class
class Data {
public:
/**
* Constructor for Template::Data
*/
// make a copy from the Token_Cache and linke all value to the key.
Data() : _keys(0), _values(0) {
_num_tokens = __Template_Token_Cache<void>::get_cache()->num_tokens();
__Template_Token_Cache<void>::get_cache()->copy( _keys );
_values = new char*[_num_tokens];
for (int i = 0; i < _num_tokens; ++ i) {
_values[i] = _keys[i];
}
}
~Data() {
for (int i = 0; i < _num_tokens; ++ i) {
if (_values[i] != _keys[i]) {
delete [](_values[i]);
}
delete [](_keys[i]);
}
delete [](_keys);
delete [](_values);
}
/**
* set (key, value) pair to the Template::Data
*
* @param[in] key the key
* @param[in] val the value
* @return bool true on success, otherwise false
*/
bool set(const char* key, const char* val) {
if (!val) {
return false;
}
int i = __Template_Token_Cache<void>::get_cache()->index(key);
if (i >= 0) {
int len = strlen(val) + 1;
char * new_key = new char[len];
memcpy(new_key, val, len);
if (_values[i] != _keys[i]) {
delete [](_values[i]);
}
_values[i] = new_key;
return true;
}
// [TRICK] Performance benchmark on linear search vs. hashmap indexing.
// The linear search still achieve better performance. Analysis shows
// more time is consumed by the allocation and deallocation of the map
// structure.
/*for (int i = 0; i < _num_tokens; ++ i) {
// i didnt check case like "pid={pid}", for speed concerns.
// users should guarantee no such template is used.
if (!strcmp(_keys[i], key)) {
int len = strlen(val) + 1;
char * new_key = new char[len];
memcpy(new_key, val, len);
if (_values[i] != _keys[i]) {
delete [](_values[i]);
}
_values[i] = new_key;
return true;
}
}*/
return false;
}
/**
* A string wrapper for bool set(const char * key, const char * val)
*
* @param[in] key the key
* @param[in] val the value
* @return bool true on success, otherwise false
*/
bool set(const char * key, const std::string & val) {
return set( key, val.c_str() );
}
const char* at(int i) const {
if (i < 0 || i >= _num_tokens) {
return 0;
}
return _values[i];
}
private:
char ** _keys;
char ** _values;
int _num_tokens;
};
public:
Template(const char * tempstr) : items(0), buffer(0) {
int len = strlen(tempstr);
buffer = new char[len + 1];
memcpy(buffer, tempstr, len + 1);
num_items = 0;
const char * s = NULL;
int right_bracket = -1;
int left_bracket = -1;
// get number of tokens in the template
for (s = buffer; *s; ++ s) {
if ((*s) == '{') {
left_bracket = s - buffer;
if (right_bracket + 1 < left_bracket) {
++ num_items;
}
}
if ((*s) == '}') {
right_bracket = s - buffer;
++ num_items;
}
}
items = new int[num_items];
right_bracket = -1;
int i = 0;
// loop over all the tokens and push them into the cache.
for (s = buffer; *s; ++ s) {
if ((*s) == '{') {
left_bracket = s - buffer;
if (right_bracket + 1 < left_bracket) {
buffer[left_bracket] = 0;
items[i] = __Template_Token_Cache<void>::get_cache()->push_back(
buffer + right_bracket + 1);
++ i;
}
}
if ((*s) == '}') {
right_bracket = s - buffer;
buffer[right_bracket] = 0;
items[i] = __Template_Token_Cache<void>::get_cache()->push_back(
buffer + left_bracket + 1);
++ i;
}
}
}
~Template() {
if (items) {
delete [](items);
}
if (buffer) {
delete [](buffer);
}
}
/**
* Generate the template from templates and save it into a string
*
* @param[in] data the template data
* @param[out] ret the output string
*/
inline bool render(const Data & data, std::string & ret) {
ret.clear();
for (int i = 0; i < num_items; ++ i) {
ret.append( data.at(items[i]) );
}
return true;
}
private:
int num_items;
int * items;
char * buffer;
};
} // end for namespace utility
} // end for namespace ltp
#endif // end for __TEMPLATE_HPP__