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
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
|
// -*- C++ -*-
// Boost general library 'format' ---------------------------
// See http://www.boost.org for updates, documentation, and revision history.
// (C) Samuel Krempp 2001
// krempp@crans.ens-cachan.fr
// Permission to copy, use, modify, sell and
// distribute this software is granted provided this copyright notice appears
// in all copies. This software is provided "as is" without express or implied
// warranty, and with no claim as to its suitability for any purpose.
// ideas taken from Rudiger Loos's format class
// and Karl Nelson's ofstream (also took its parsing code as basis for printf parsing)
// ------------------------------------------------------------------------------
// parsing.hpp : implementation of the parsing member functions
// ( parse, parse_printf_directive)
// ------------------------------------------------------------------------------
#ifndef BOOST_FORMAT_PARSING_HPP
#define BOOST_FORMAT_PARSING_HPP
#include <boost/format/format_class.hpp>
//#include <boost/throw_exception.hpp>
//#include <boost/assert.hpp>
namespace boost {
namespace io {
namespace detail {
template<class Ch, class Stream> inline
bool wrap_isdigit(Ch c, Stream &os)
{
#ifndef BOOST_NO_LOCALE_ISIDIGIT
return std::isdigit(c, os.rdbuf()->getloc() );
# else
using namespace std;
return isdigit(c);
#endif
} //end- wrap_isdigit(..)
template<class Res, class Ch, class Tr> inline
Res str2int(const std::basic_string<Ch, Tr>& s,
typename std::basic_string<Ch, Tr>::size_type start,
BOOST_IO_STD basic_ios<Ch,Tr> &os,
const Res = Res(0) )
// Input : char string, with starting index
// a basic_ios& merely to call its widen/narrow member function in the desired locale.
// Effects : reads s[start:] and converts digits into an integral n, of type Res
// Returns : n
{
Res n = 0;
while(start<s.size() && wrap_isdigit(s[start], os) ) {
char cur_ch = os.narrow( s[start], 0);
BOOST_ASSERT(cur_ch != 0 ); // since we called isdigit, this should not happen.
n *= 10;
n += cur_ch - '0'; // 22.2.1.1.2 of the C++ standard
++start;
}
return n;
}
template<class Ch, class Tr>
void skip_asterisk(const std::basic_string<Ch,Tr> & buf,
typename std::basic_string<Ch,Tr>::size_type * pos_p,
BOOST_IO_STD basic_ios<Ch, Tr> &os)
// skip printf's "asterisk-fields" directives in the format-string buf
// Input : char string, with starting index *pos_p
// a basic_ios& merely to call its widen/narrow member function in the desired locale.
// Effects : advance *pos_p by skipping printf's asterisk fields.
// Returns : nothing
{
using namespace std;
BOOST_ASSERT( pos_p != 0);
if(*pos_p >= buf.size() ) return;
if(buf[ *pos_p]==os.widen('*')) {
++ (*pos_p);
while (*pos_p < buf.size() && wrap_isdigit(buf[*pos_p],os)) ++(*pos_p);
if(buf[*pos_p]==os.widen('$')) ++(*pos_p);
}
}
inline void maybe_throw_exception( unsigned char exceptions)
// auxiliary func called by parse_printf_directive
// for centralising error handling
// it either throws if user sets the corresponding flag, or does nothing.
{
if(exceptions & io::bad_format_string_bit)
boost::throw_exception(io::bad_format_string());
}
template<class Ch, class Tr>
bool parse_printf_directive(const std::basic_string<Ch, Tr> & buf,
typename std::basic_string<Ch, Tr>::size_type * pos_p,
detail::format_item<Ch, Tr> * fpar,
BOOST_IO_STD basic_ios<Ch,Tr> &os,
unsigned char exceptions)
// Input : a 'printf-directive' in the format-string, starting at buf[ *pos_p ]
// a basic_ios& merely to call its widen/narrow member function in the desired locale.
// a bitset'excpetions' telling whether to throw exceptions on errors.
// Returns : true if parse somehow succeeded (possibly ignoring errors if exceptions disabled)
// false if it failed so bad that the directive should be printed verbatim
// Effects : - *pos_p is incremented so that buf[*pos_p] is the first char after the directive
// - *fpar is set with the parameters read in the directive
{
typedef format_item<Ch, Tr> format_item_t;
BOOST_ASSERT( pos_p != 0);
typename std::basic_string<Ch, Tr>::size_type &i1 = *pos_p,
i0;
fpar->argN_ = format_item_t::argN_no_posit; // if no positional-directive
bool in_brackets=false;
if(buf[i1]==os.widen('|'))
{
in_brackets=true;
if( ++i1 >= buf.size() ) {
maybe_throw_exception(exceptions);
return false;
}
}
// the flag '0' would be picked as a digit for argument order, but here it's a flag :
if(buf[i1]==os.widen('0'))
goto parse_flags;
// handle argument order (%2$d) or possibly width specification: %2d
i0 = i1; // save position before digits
while (i1 < buf.size() && wrap_isdigit(buf[i1], os))
++i1;
if (i1!=i0)
{
if( i1 >= buf.size() ) {
maybe_throw_exception(exceptions);
return false;
}
int n=str2int(buf,i0, os, int(0) );
// %N% case : this is already the end of the directive
if( buf[i1] == os.widen('%') )
{
fpar->argN_ = n-1;
++i1;
if( in_brackets)
maybe_throw_exception(exceptions);
// but don't return. maybe "%" was used in lieu of '$', so we go on.
else return true;
}
if ( buf[i1]==os.widen('$') )
{
fpar->argN_ = n-1;
++i1;
}
else
{
// non-positionnal directive
fpar->ref_state_.width_ = n;
fpar->argN_ = format_item_t::argN_no_posit;
goto parse_precision;
}
}
parse_flags:
// handle flags
while ( i1 <buf.size()) // as long as char is one of + - = # 0 l h or ' '
{
// misc switches
switch (os.narrow(buf[i1], 0))
{
case '\'' : break; // no effect yet. (painful to implement)
case 'l':
case 'h': // short/long modifier : for printf-comaptibility (no action needed)
break;
case '-':
fpar->ref_state_.flags_ |= std::ios_base::left;
break;
case '=':
fpar->pad_scheme_ |= format_item_t::centered;
break;
case ' ':
fpar->pad_scheme_ |= format_item_t::spacepad;
break;
case '+':
fpar->ref_state_.flags_ |= std::ios_base::showpos;
break;
case '0':
fpar->pad_scheme_ |= format_item_t::zeropad;
// need to know alignment before really setting flags,
// so just add 'zeropad' flag for now, it will be processed later.
break;
case '#':
fpar->ref_state_.flags_ |= std::ios_base::showpoint | std::ios_base::showbase;
break;
default:
goto parse_width;
}
++i1;
} // loop on flag.
if( i1>=buf.size()) {
maybe_throw_exception(exceptions);
return true;
}
parse_width:
// handle width spec
skip_asterisk(buf, &i1, os); // skips 'asterisk fields' : *, or *N$
i0 = i1; // save position before digits
while (i1<buf.size() && wrap_isdigit(buf[i1], os))
i1++;
if (i1!=i0)
{ fpar->ref_state_.width_ = str2int( buf,i0, os, std::streamsize(0) ); }
parse_precision:
if( i1>=buf.size()) {
maybe_throw_exception(exceptions);
return true;
}
// handle precision spec
if (buf[i1]==os.widen('.'))
{
++i1;
skip_asterisk(buf, &i1, os);
i0 = i1; // save position before digits
while (i1<buf.size() && wrap_isdigit(buf[i1], os))
++i1;
if(i1==i0)
fpar->ref_state_.precision_ = 0;
else
fpar->ref_state_.precision_ = str2int(buf,i0, os, std::streamsize(0) );
}
// handle formatting-type flags :
while( i1<buf.size() &&
( buf[i1]==os.widen('l') || buf[i1]==os.widen('L') || buf[i1]==os.widen('h')) )
++i1;
if( i1>=buf.size()) {
maybe_throw_exception(exceptions);
return true;
}
if( in_brackets && buf[i1]==os.widen('|') )
{
++i1;
return true;
}
switch (os.narrow(buf[i1], 0) )
{
case 'X':
fpar->ref_state_.flags_ |= std::ios_base::uppercase;
case 'p': // pointer => set hex.
case 'x':
fpar->ref_state_.flags_ &= ~std::ios_base::basefield;
fpar->ref_state_.flags_ |= std::ios_base::hex;
break;
case 'o':
fpar->ref_state_.flags_ &= ~std::ios_base::basefield;
fpar->ref_state_.flags_ |= std::ios_base::oct;
break;
case 'E':
fpar->ref_state_.flags_ |= std::ios_base::uppercase;
case 'e':
fpar->ref_state_.flags_ &= ~std::ios_base::floatfield;
fpar->ref_state_.flags_ |= std::ios_base::scientific;
fpar->ref_state_.flags_ &= ~std::ios_base::basefield;
fpar->ref_state_.flags_ |= std::ios_base::dec;
break;
case 'f':
fpar->ref_state_.flags_ &= ~std::ios_base::floatfield;
fpar->ref_state_.flags_ |= std::ios_base::fixed;
case 'u':
case 'd':
case 'i':
fpar->ref_state_.flags_ &= ~std::ios_base::basefield;
fpar->ref_state_.flags_ |= std::ios_base::dec;
break;
case 'T':
++i1;
if( i1 >= buf.size())
maybe_throw_exception(exceptions);
else
fpar->ref_state_.fill_ = buf[i1];
fpar->pad_scheme_ |= format_item_t::tabulation;
fpar->argN_ = format_item_t::argN_tabulation;
break;
case 't':
fpar->ref_state_.fill_ = os.widen(' ');
fpar->pad_scheme_ |= format_item_t::tabulation;
fpar->argN_ = format_item_t::argN_tabulation;
break;
case 'G':
fpar->ref_state_.flags_ |= std::ios_base::uppercase;
break;
case 'g': // 'g' conversion is default for floats.
fpar->ref_state_.flags_ &= ~std::ios_base::basefield;
fpar->ref_state_.flags_ |= std::ios_base::dec;
// CLEAR all floatield flags, so stream will CHOOSE
fpar->ref_state_.flags_ &= ~std::ios_base::floatfield;
break;
case 'C':
case 'c':
fpar->truncate_ = 1;
break;
case 'S':
case 's':
fpar->truncate_ = fpar->ref_state_.precision_;
fpar->ref_state_.precision_ = -1;
break;
case 'n' :
fpar->argN_ = format_item_t::argN_ignored;
break;
default:
maybe_throw_exception(exceptions);
}
++i1;
if( in_brackets )
{
if( i1<buf.size() && buf[i1]==os.widen('|') )
{
++i1;
return true;
}
else maybe_throw_exception(exceptions);
}
return true;
}
} // detail namespace
} // io namespace
// -----------------------------------------------
// format :: parse(..)
template<class Ch, class Traits>
void basic_format<Ch, Traits> ::parse(const string_t & buf)
// parse the format-string
{
using namespace std;
const Ch arg_mark = oss_.widen('%');
bool ordered_args=true;
int max_argN=-1;
typename string_t::size_type i1=0;
int num_items=0;
// A: find upper_bound on num_items and allocates arrays
i1=0;
while( (i1=buf.find(arg_mark,i1)) != string::npos )
{
if( i1+1 >= buf.size() ) {
if(exceptions() & io::bad_format_string_bit)
boost::throw_exception(io::bad_format_string()); // must not end in "bla bla %"
else break; // stop there, ignore last '%'
}
if(buf[i1+1] == buf[i1] ) { i1+=2; continue; } // escaped "%%" / "##"
++i1;
// in case of %N% directives, dont count it double (wastes allocations..) :
while(i1 < buf.size() && io::detail::wrap_isdigit(buf[i1],oss_)) ++i1;
if( i1 < buf.size() && buf[i1] == arg_mark ) ++ i1;
++num_items;
}
items_.assign( num_items, format_item_t() );
// B: Now the real parsing of the format string :
num_items=0;
i1 = 0;
typename string_t::size_type i0 = i1;
bool special_things=false;
int cur_it=0;
while( (i1=buf.find(arg_mark,i1)) != string::npos )
{
string_t & piece = (cur_it==0) ? prefix_ : items_[cur_it-1].appendix_;
if( buf[i1+1] == buf[i1] ) // escaped mark, '%%'
{
piece += buf.substr(i0, i1-i0) + buf[i1];
i1+=2; i0=i1;
continue;
}
BOOST_ASSERT( static_cast<unsigned int>(cur_it) < items_.size() || cur_it==0);
if(i1!=i0) piece += buf.substr(i0, i1-i0);
++i1;
bool parse_ok;
parse_ok = io::detail::parse_printf_directive(buf, &i1, &items_[cur_it], oss_, exceptions());
if( ! parse_ok ) continue; // the directive will be printed verbatim
i0=i1;
items_[cur_it].compute_states(); // process complex options, like zeropad, into stream params.
int argN=items_[cur_it].argN_;
if(argN == format_item_t::argN_ignored)
continue;
if(argN ==format_item_t::argN_no_posit)
ordered_args=false;
else if(argN == format_item_t::argN_tabulation) special_things=true;
else if(argN > max_argN) max_argN = argN;
++num_items;
++cur_it;
} // loop on %'s
BOOST_ASSERT(cur_it == num_items);
// store the final piece of string
string_t & piece = (cur_it==0) ? prefix_ : items_[cur_it-1].appendix_;
piece += buf.substr(i0);
if( !ordered_args)
{
if(max_argN >= 0 ) // dont mix positional with non-positionnal directives
{
if(exceptions() & io::bad_format_string_bit)
boost::throw_exception(io::bad_format_string());
// else do nothing. => positionnal arguments are processed as non-positionnal
}
// set things like it would have been with positional directives :
int non_ordered_items = 0;
for(int i=0; i< num_items; ++i)
if(items_[i].argN_ == format_item_t::argN_no_posit)
{
items_[i].argN_ = non_ordered_items;
++non_ordered_items;
}
max_argN = non_ordered_items-1;
}
// C: set some member data :
items_.resize(num_items);
if(special_things) style_ |= special_needs;
num_args_ = max_argN + 1;
if(ordered_args) style_ |= ordered;
else style_ &= ~ordered;
}
} // namespace boost
#endif // BOOST_FORMAT_PARSING_HPP
|