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
|
#include "tokenizer.h"
#include "Parameters.h"
#include <memory>
#include <vector>
#include <cctype>
#include <cstring>
#ifdef TOKENIZER_NAMESPACE
using namespace TOKENIZER_NAMESPACE ;
#endif
void
usage(const char *path)
{
std::cerr << "Usage: " << path << "[-{v|x|p|a|e|s|u|n|N]* [LL] [-{c|o} PATH]* INFILE*" << std::endl;
std::cerr << " -a -- aggressive hyphenization" << std::endl;
std::cerr << " -b -- drop bad bytes" << std::endl;
std::cerr << " -c DIR -- config (pattern) file directory" << std::endl;
std::cerr << " -d -- downcase" << std::endl;
std::cerr << " -D -- detokenize" << std::endl;
std::cerr << " -e -- do not escape entities during tokenization" << std::endl;
std::cerr << " -k -- narrow kana" << std::endl;
std::cerr << " -n -- narrow latin" << std::endl;
std::cerr << " -N -- normalize" << std::endl;
std::cerr << " -o OUT -- output file path" << std::endl;
std::cerr << " -p -- penn treebank style" << std::endl;
std::cerr << " -r -- refined contraction and quantity conjoining" << std::endl;
std::cerr << " -s -- super- and sub-script conjoining" << std::endl;
std::cerr << " -u -- disable url handling" << std::endl;
std::cerr << " -U -- unescape entities before tokenization, after detokenization" << std::endl;
std::cerr << " -v -- verbose" << std::endl;
std::cerr << " -w -- word filter" << std::endl;
std::cerr << " -x -- skip xml tag lines" << std::endl;
std::cerr << " -y -- skip all xml tags" << std::endl;
std::cerr << "Default is -c ., stdin, stdout." << std::endl;
std::cerr << "LL in en,fr,it affect contraction. LL selects nonbreaking prefix file" << std::endl;
std::cerr << "nonbreaking_prefix.LL is sought in getenv('TOKENIZER_SHARED_DIR')." << std::endl;
return;
}
std::string token_word(const std::string& in) {
int pos = -1;
int digits_prefixed = 0;
int nalpha = 0;
int len = in.size();
std::vector<char> cv;
int last_quirk = -1;
while (++pos < len) {
char ch = in.at(pos);
if (std::isdigit(ch)) {
if (digits_prefixed > 0) {
last_quirk = pos;
break;
}
digits_prefixed--;
cv.push_back(std::tolower(ch));
} else if (std::isalpha(ch)) {
if (digits_prefixed < 0)
digits_prefixed = -digits_prefixed;
cv.push_back(std::tolower(ch));
nalpha++;
} else {
if (digits_prefixed < 0)
digits_prefixed = -digits_prefixed;
last_quirk = pos;
if ((ch == '-' || ch == '\'') && pos != 0) {
cv.push_back(ch);
} else {
break;
}
}
}
if (last_quirk == pos || (digits_prefixed > 0 && nalpha == 0))
cv.clear(); // invalid word
return std::string(cv.begin(),cv.end());
}
int
copy_words(Tokenizer& tize, std::istream& ifs, std::ostream& ofs) {
int nlines = 0;
std::string line;
while (ifs.good() && std::getline(ifs,line)) {
if (line.empty()) continue;
std::vector<std::string> tokens(tize.tokens(line));
int count = 0;
for (auto& token: tokens) {
std::string word(token_word(token));
if (word.empty()) continue;
ofs << word << ' ';
count++;
}
if (count) {
ofs << std::endl;
nlines++;
}
}
return nlines;
}
int main(int ac, char **av)
{
int rc = 0;
Parameters params;
const char *prog = av[0];
bool next_cfg_p = false;
bool next_output_p = false;
bool detokenize_p = std::strstr(av[0],"detokenize") != 0;
while (++av,--ac) {
if (**av == '-') {
switch (av[0][1]) {
case 'a':
params.aggro_p = true;
break;
case 'b':
params.drop_bad_p = true;
break;
case 'c':
next_cfg_p = true;
break;
case 'd':
params.downcase_p = true;
break;
case 'D':
detokenize_p = true;
break;
case 'e':
params.escape_p = false;
break;
case 'h':
usage(prog);
exit(0);
case 'k':
params.narrow_kana_p = true;
break;
case 'n':
params.narrow_latin_p = true;
break;
case 'N':
params.normalize_p = true;
break;
case 'o':
next_output_p = true;
break;
case 'p':
params.penn_p = true;
break;
case 'r':
params.refined_p = true;
break;
case 's':
params.supersub_p = true;
break;
case 'U':
params.unescape_p = true;
break;
case 'u':
params.url_p = false;
break;
case 'v':
params.verbose_p = true;
break;
case 'w':
params.words_p = true;
break;
case 'x':
params.detag_p = true;
break;
case 'y':
params.alltag_p = true;
break;
case 'l':
// ignored
break;
default:
std::cerr << "Unknown option: " << *av << std::endl;
::exit(1);
}
} else if (params.lang_iso.empty() && strlen(*av) == 2) {
params.lang_iso = *av;
} else if (next_output_p) {
next_output_p = false;
params.out_path = *av;
} else if (next_cfg_p) {
next_cfg_p = false;
params.cfg_path = *av;
} else {
params.args.push_back(std::string(*av));
}
}
if (!params.cfg_path) {
params.cfg_path = getenv("TOKENIZER_SHARED_DIR");
}
if (!params.cfg_path) {
if (!::access("../share/.",X_OK)) {
if (!::access("../share/moses/.",X_OK)) {
params.cfg_path = "../share/moses";
} else {
params.cfg_path = "../share";
}
} else if (!::access("./scripts/share/.",X_OK)) {
params.cfg_path = "./scripts/share";
} else if (!::access("./nonbreaking_prefix.en",R_OK)) {
params.cfg_path = ".";
} else {
const char *slash = std::strrchr(prog,'/');
if (slash) {
std::string cfg_dir_str(prog,slash-prog);
std::string cfg_shr_str(cfg_dir_str);
cfg_shr_str.append("/shared");
std::string cfg_mos_str(cfg_shr_str);
cfg_mos_str.append("/moses");
if (!::access(cfg_mos_str.c_str(),X_OK)) {
params.cfg_path = strdup(cfg_mos_str.c_str());
} else if (!::access(cfg_shr_str.c_str(),X_OK)) {
params.cfg_path = strdup(cfg_shr_str.c_str());
} else if (!::access(cfg_dir_str.c_str(),X_OK)) {
params.cfg_path = strdup(cfg_dir_str.c_str());
}
}
}
}
if (params.cfg_path) {
if (params.verbose_p) {
std::cerr << "config path: " << params.cfg_path << std::endl;
}
Tokenizer::set_config_dir(std::string(params.cfg_path));
}
std::unique_ptr<std::ofstream> pofs = 0;
if (!params.out_path.empty()) {
pofs.reset(new std::ofstream(params.out_path.c_str()));
}
std::ostream& ofs(pofs ? *pofs : std::cout);
if (params.lang_iso.empty())
params.lang_iso = "en";
Tokenizer tize(params);
tize.init();
size_t nlines = 0;
if (params.words_p) {
if (params.args.empty()) {
nlines += copy_words(tize,std::cin,ofs);
} else {
for (std::string& arg : params.args) {
try {
std::ifstream ifs(arg.c_str());
nlines += copy_words(tize,ifs,ofs);
} catch (...) {
std::cerr << "Exception extracting words from path " << arg << std::endl;
}
}
}
} else if (params.args.empty()) {
if (detokenize_p) {
nlines = tize.detokenize(std::cin,ofs);
} else {
nlines = tize.tokenize(std::cin,ofs);
}
} else {
for (std::string& arg : params.args) {
try {
std::ifstream ifs(arg.c_str());
if (detokenize_p) {
nlines = tize.detokenize(ifs,ofs);
} else {
nlines = tize.tokenize(ifs,ofs);
}
} catch (...) {
std::cerr << "Exception tokenizing from path " << arg << std::endl;
}
}
}
if (params.verbose_p)
std::cerr << "%%% " << nlines << " lines." << std::endl;
return rc;
}
|