summaryrefslogtreecommitdiff
path: root/zen/string_tools.h
blob: 0a24ab2ab7588e32b44c056c14445049286f13a1 (plain)
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
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
// *****************************************************************************
// * This file is part of the FreeFileSync project. It is distributed under    *
// * GNU General Public License: http://www.gnu.org/licenses/gpl-3.0           *
// * Copyright (C) Zenju (zenju AT freefilesync DOT org) - All Rights Reserved *
// *****************************************************************************

#ifndef STRING_TOOLS_H_213458973046
#define STRING_TOOLS_H_213458973046

#include <cctype>  //isspace
#include <cwctype> //iswspace
#include <cstdio>  //sprintf
#include <cwchar>  //swprintf
#include <algorithm>
#include <cassert>
#include <vector>
#include <sstream>
#include "stl_tools.h"
#include "string_traits.h"


//enhance arbitray string class with useful non-member functions:
namespace zen
{
template <class Char> bool isWhiteSpace(Char c);
template <class Char> bool isDigit     (Char c); //not exactly the same as "std::isdigit" -> we consider '0'-'9' only!
template <class Char> bool isHexDigit  (Char c);
template <class Char> bool isAsciiAlpha(Char c);
template <class Char> Char asciiToLower(Char c);

//case-sensitive comparison (compile-time correctness:  use different number of arguments as STL comparison predicates!)
struct CmpBinary { template <class Char> int operator()(const Char* lhs, size_t lhsLen, const Char* rhs, size_t rhsLen) const; };

//basic case-insensitive comparison (considering A-Z only!)
struct CmpAsciiNoCase { template <class Char> int operator()(const Char* lhs, size_t lhsLen, const Char* rhs, size_t rhsLen) const; };

struct LessAsciiNoCase
{
    template <class S> //don't support heterogenous input! => use as container predicate only!
    bool operator()(const S& lhs, const S& rhs) const { return CmpAsciiNoCase()(strBegin(lhs), strLength(lhs), strBegin(rhs), strLength(rhs)) < 0; }
};

//both S and T can be strings or char/wchar_t arrays or simple char/wchar_t
template <class S, class T> bool contains(const S& str, const T& term);

template <class S, class T>                 bool startsWith(const S& str, const T& prefix);
template <class S, class T, class Function> bool startsWith(const S& str, const T& prefix,  Function cmpStringFun);

template <class S, class T>                 bool endsWith  (const S& str, const T& postfix);
template <class S, class T, class Function> bool endsWith  (const S& str, const T& postfix, Function cmpStringFun);

template <class S, class T>                 bool strEqual(const S& lhs, const T& rhs);
template <class S, class T, class Function> bool strEqual(const S& lhs, const T& rhs, Function cmpStringFun);

enum FailureReturnVal
{
    IF_MISSING_RETURN_ALL,
    IF_MISSING_RETURN_NONE
};

template <class S, class T> S afterLast  (const S& str, const T& term, FailureReturnVal rv);
template <class S, class T> S beforeLast (const S& str, const T& term, FailureReturnVal rv);
template <class S, class T> S afterFirst (const S& str, const T& term, FailureReturnVal rv);
template <class S, class T> S beforeFirst(const S& str, const T& term, FailureReturnVal rv);

enum class SplitType
{
    ALLOW_EMPTY,
    SKIP_EMPTY
};
template <class S, class T> std::vector<S> split(const S& str, const T& delimiter, SplitType st);

template <class S>                 S    trimCpy(S  str, bool fromLeft = true, bool fromRight = true);
template <class S>                 void trim   (S& str, bool fromLeft = true, bool fromRight = true);
template <class S, class Function> void trim(S& str, bool fromLeft, bool fromRight, Function trimThisChar);

template <class S, class T, class U> void replace   (      S& str, const T& oldTerm, const U& newTerm, bool replaceAll = true);
template <class S, class T, class U> S    replaceCpy(const S& str, const T& oldTerm, const U& newTerm, bool replaceAll = true);

//high-performance conversion between numbers and strings
template <class S,   class Num> S   numberTo(const Num& number);
template <class Num, class S>   Num stringTo(const S&   str);

std::pair<char, char> hexify  (unsigned char c, bool upperCase = true);
char                  unhexify(char high, char low);

template <class S, class T, class Num> S printNumber(const T& format, const Num& number); //format a single number using std::snprintf()

//string to string conversion: converts string-like type into char-compatible target string class
template <class T, class S> T copyStringTo(S&& str);
















//---------------------- implementation ----------------------
template <> inline
bool isWhiteSpace(char c)
{
    assert(c != 0); //std C++ does not consider 0 as white space
    //caveat 1: std::isspace() takes an int, but expects an unsigned char
    //caveat 2: some parts of UTF-8 chars are erroneously seen as whitespace, e.g. the a0 from "\xec\x8b\xa0" (MSVC)
    return static_cast<unsigned char>(c) < 128 &&
           std::isspace(static_cast<unsigned char>(c)) != 0;
}

template <> inline
bool isWhiteSpace(wchar_t c)
{
    assert(c != 0); //std C++ does not consider 0 as white space
    return std::iswspace(c) != 0;
}


template <class Char> inline
bool isDigit(Char c) //similar to implmenetation of std::isdigit()!
{
    static_assert(IsSameType<Char, char>::value || IsSameType<Char, wchar_t>::value, "");
    return static_cast<Char>('0') <= c && c <= static_cast<Char>('9');
}


template <class Char> inline
bool isHexDigit(Char c)
{
    static_assert(IsSameType<Char, char>::value || IsSameType<Char, wchar_t>::value, "");
    return (static_cast<Char>('0') <= c && c <= static_cast<Char>('9')) ||
           (static_cast<Char>('A') <= c && c <= static_cast<Char>('F')) ||
           (static_cast<Char>('a') <= c && c <= static_cast<Char>('f'));
}


template <class Char> inline
bool isAsciiAlpha(Char c)
{
    static_assert(IsSameType<Char, char>::value || IsSameType<Char, wchar_t>::value, "");
    return (static_cast<Char>('A') <= c && c <= static_cast<Char>('Z')) ||
           (static_cast<Char>('a') <= c && c <= static_cast<Char>('z'));
}


template <class Char> inline
Char asciiToLower(Char c)
{
    if (static_cast<Char>('A') <= c && c <= static_cast<Char>('Z'))
        return static_cast<Char>(c - static_cast<Char>('A') + static_cast<Char>('a'));
    return c;
}


template <class S, class T, class Function> inline
bool startsWith(const S& str, const T& prefix, Function cmpStringFun)
{
    const size_t pfLen = strLength(prefix);
    if (strLength(str) < pfLen)
        return false;

    return cmpStringFun(strBegin(str),    pfLen,
                        strBegin(prefix), pfLen) == 0;
}


template <class S, class T, class Function> inline
bool endsWith(const S& str, const T& postfix, Function cmpStringFun)
{
    const size_t strLen = strLength(str);
    const size_t pfLen  = strLength(postfix);
    if (strLen < pfLen)
        return false;

    return cmpStringFun(strBegin(str) + strLen - pfLen, pfLen,
                        strBegin(postfix), pfLen) == 0;
}


template <class S, class T, class Function> inline
bool strEqual(const S& lhs, const T& rhs, Function cmpStringFun)
{
    return cmpStringFun(strBegin(lhs), strLength(lhs), strBegin(rhs), strLength(rhs)) == 0;
}


template <class S, class T> inline bool startsWith(const S& str, const T& prefix ) { return startsWith(str, prefix,  CmpBinary()); }
template <class S, class T> inline bool endsWith  (const S& str, const T& postfix) { return endsWith  (str, postfix, CmpBinary()); }
template <class S, class T> inline bool strEqual  (const S& lhs, const T& rhs    ) { return strEqual  (lhs, rhs,     CmpBinary()); }


template <class S, class T> inline
bool contains(const S& str, const T& term)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t strLen  = strLength(str);
    const size_t termLen = strLength(term);
    if (strLen < termLen)
        return false;

    const auto* const strFirst  = strBegin(str);
    const auto* const strLast   = strFirst + strLen;
    const auto* const termFirst = strBegin(term);

    return std::search(strFirst, strLast,
                       termFirst, termFirst + termLen) != strLast;
}


template <class S, class T> inline
S afterLast(const S& str, const T& term, FailureReturnVal rv)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t termLen = strLength(term);
    assert(termLen > 0);

    const auto* const strFirst  = strBegin(str);
    const auto* const strLast   = strFirst + strLength(str);
    const auto* const termFirst = strBegin(term);

    const auto* it = search_last(strFirst, strLast,
                                 termFirst, termFirst + termLen);
    if (it == strLast)
        return rv == IF_MISSING_RETURN_ALL ? str : S();

    it += termLen;
    return S(it, strLast - it);
}


template <class S, class T> inline
S beforeLast(const S& str, const T& term, FailureReturnVal rv)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t termLen = strLength(term);
    assert(termLen > 0);

    const auto* const strFirst  = strBegin(str);
    const auto* const strLast   = strFirst + strLength(str);
    const auto* const termFirst = strBegin(term);

    const auto* it = search_last(strFirst, strLast,
                                 termFirst, termFirst + termLen);
    if (it == strLast)
        return rv == IF_MISSING_RETURN_ALL ? str : S();

    return S(strFirst, it - strFirst);
}


template <class S, class T> inline
S afterFirst(const S& str, const T& term, FailureReturnVal rv)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t termLen = strLength(term);
    assert(termLen > 0);

    const auto* const strFirst  = strBegin(str);
    const auto* const strLast   = strFirst + strLength(str);
    const auto* const termFirst = strBegin(term);

    const auto* it = std::search(strFirst, strLast,
                                 termFirst, termFirst + termLen);
    if (it == strLast)
        return rv == IF_MISSING_RETURN_ALL ? str : S();

    it += termLen;
    return S(it, strLast - it);
}


template <class S, class T> inline
S beforeFirst(const S& str, const T& term, FailureReturnVal rv)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t termLen = strLength(term);
    assert(termLen > 0);

    const auto* const strFirst  = strBegin(str);
    const auto* const strLast   = strFirst + strLength(str);
    const auto* const termFirst = strBegin(term);

    auto it = std::search(strFirst, strLast,
                          termFirst,  termFirst  + termLen);
    if (it == strLast)
        return rv == IF_MISSING_RETURN_ALL ? str : S();

    return S(strFirst, it - strFirst);
}


template <class S, class T> inline
std::vector<S> split(const S& str, const T& delimiter, SplitType st)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    const size_t delimLen = strLength(delimiter);
    assert(delimLen > 0);
    if (delimLen == 0)
    {
        if (str.empty() && st == SplitType::SKIP_EMPTY)
            return {};
        return { str };
    }

    const auto* const delimFirst = strBegin(delimiter);
    const auto* const delimLast  = delimFirst + delimLen;

    const auto* blockStart    = strBegin(str);
    const auto* const strLast = blockStart + strLength(str);

    std::vector<S> output;
    for (;;)
    {
        const auto* const blockEnd = std::search(blockStart, strLast,
                                                 delimFirst, delimLast);
        if (blockStart != blockEnd || st == SplitType::ALLOW_EMPTY)
            output.emplace_back(blockStart, blockEnd - blockStart);

        if (blockEnd == strLast)
            return output;
        blockStart = blockEnd + delimLen;
    }
}


namespace impl
{
ZEN_INIT_DETECT_MEMBER(append);

//either call operator+=(S(str, len)) or append(str, len)
template <class S, class InputIterator> inline
typename EnableIf<HasMember_append<S>::value>::Type stringAppend(S& str, InputIterator first, InputIterator last) { str.append(first, last);  }

template <class S, class InputIterator> inline
typename EnableIf<!HasMember_append<S>::value>::Type stringAppend(S& str, InputIterator first, InputIterator last) { str += S(first, last); }
}


template <class S, class T, class U> inline
S replaceCpy(const S& str, const T& oldTerm, const U& newTerm, bool replaceAll)
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    static_assert(IsSameType<typename GetCharType<T>::Type, typename GetCharType<U>::Type>::value, "");
    const size_t oldLen = strLength(oldTerm);
    if (oldLen == 0)
        return str;

    const auto* const oldBegin = strBegin(oldTerm);
    const auto* const oldEnd   = oldBegin + oldLen;

    const auto* const newBegin = strBegin(newTerm);
    const auto* const newEnd   = newBegin + strLength(newTerm);

    S output;

    for (auto it = str.begin();;)
    {
        const auto itFound = std::search(it, str.end(),
                                         oldBegin, oldEnd);
        if (itFound == str.end() && it == str.begin())
            return str; //optimize "oldTerm not found": return ref-counted copy

        impl::stringAppend(output, it, itFound);
        if (itFound == str.end())
            return output;

        impl::stringAppend(output, newBegin, newEnd);
        it = itFound + oldLen;

        if (!replaceAll)
        {
            impl::stringAppend(output, it, str.end());
            return output;
        }
    }
}


template <class S, class T, class U> inline
void replace(S& str, const T& oldTerm, const U& newTerm, bool replaceAll)
{
    str = replaceCpy(str, oldTerm, newTerm, replaceAll);
}


template <class S, class Function> inline
void trim(S& str, bool fromLeft, bool fromRight, Function trimThisChar)
{
    assert(fromLeft || fromRight);

    const auto* const oldBegin = strBegin(str);
    const auto*       newBegin = oldBegin;
    const auto*       newEnd   = oldBegin + strLength(str);

    if (fromRight)
        while (newBegin != newEnd && trimThisChar(newEnd[-1]))
            --newEnd;

    if (fromLeft)
        while (newBegin != newEnd && trimThisChar(*newBegin))
            ++newBegin;

    if (newBegin != oldBegin)
        str = S(newBegin, newEnd - newBegin); //minor inefficiency: in case "str" is not shared, we could save an allocation and do a memory move only
    else
        str.resize(newEnd - newBegin);
}


template <class S> inline
void trim(S& str, bool fromLeft, bool fromRight)
{
    using CharType = typename GetCharType<S>::Type;
    trim(str, fromLeft, fromRight, [](CharType c) { return isWhiteSpace(c); });
}


template <class S> inline
S trimCpy(S str, bool fromLeft, bool fromRight)
{
    //implementing trimCpy() in terms of trim(), instead of the other way round, avoids memory allocations when trimming from right!
    trim(str, fromLeft, fromRight);
    return std::move(str); //"str" is an l-value parameter => no copy elision!
}


namespace impl
{
template <class S, class T>
struct CopyStringToString
{
    T copy(const S& src) const { return T(strBegin(src), strLength(src)); }
};

template <class T>
struct CopyStringToString<T, T> //perf: we don't need a deep copy if string types match
{
    template <class S>
    T copy(S&& str) const { return std::forward<S>(str); }
};

inline int strcmpWithNulls(const char*    ptr1, const char*    ptr2, size_t num) { return std::memcmp (ptr1, ptr2, num); }
inline int strcmpWithNulls(const wchar_t* ptr1, const wchar_t* ptr2, size_t num) { return std::wmemcmp(ptr1, ptr2, num); }
}

template <class T, class S> inline
T copyStringTo(S&& str) { return impl::CopyStringToString<std::decay_t<S>, T>().copy(std::forward<S>(str)); }


template <class Char> inline
int CmpBinary::operator()(const Char* lhs, size_t lhsLen, const Char* rhs, size_t rhsLen) const
{
    //support embedded 0, unlike strncmp/wcsncmp!
    const int rv = impl::strcmpWithNulls(lhs, rhs, std::min(lhsLen, rhsLen));
    if (rv != 0)
        return rv;
    return static_cast<int>(lhsLen) - static_cast<int>(rhsLen);
}


template <class Char> inline
int CmpAsciiNoCase::operator()(const Char* lhs, size_t lhsLen, const Char* rhs, size_t rhsLen) const
{
    const auto* const lhsLast = lhs + std::min(lhsLen, rhsLen);
    while (lhs != lhsLast)
    {
        const Char charL = asciiToLower(*lhs++); //ordering: lower-case chars have higher code points than uppper-case
        const Char charR = asciiToLower(*rhs++); //
        if (charL != charR)
            return static_cast<unsigned int>(charL) - static_cast<unsigned int>(charR); //unsigned char-comparison is the convention!
        //unsigned underflow is well-defined!
    }
    return static_cast<int>(lhsLen) - static_cast<int>(rhsLen);
}


namespace impl
{
template <class Num> inline
int saferPrintf(char* buffer, size_t bufferSize, const char* format, const Num& number) //there is no such thing as a "safe" printf ;)
{
    return std::snprintf(buffer, bufferSize, format, number); //C99: returns number of chars written if successful, < 0 or >= bufferSize on failure
}

template <class Num> inline
int saferPrintf(wchar_t* buffer, size_t bufferSize, const wchar_t* format, const Num& number)
{
    return std::swprintf(buffer, bufferSize, format, number); //C99: returns number of chars written if successful, < 0 on failure (including buffer too small)
}
}

template <class S, class T, class Num> inline
S printNumber(const T& format, const Num& number) //format a single number using ::sprintf
{
    static_assert(IsSameType<typename GetCharType<S>::Type, typename GetCharType<T>::Type>::value, "");
    using CharType = typename GetCharType<S>::Type;

    const int BUFFER_SIZE = 128;
    CharType buffer[BUFFER_SIZE]; //zero-initialize?
    const int charsWritten = impl::saferPrintf(buffer, BUFFER_SIZE, strBegin(format), number);

    return 0 < charsWritten && charsWritten < BUFFER_SIZE ? S(buffer, charsWritten) : S();
}


namespace impl
{
enum NumberType
{
    NUM_TYPE_SIGNED_INT,
    NUM_TYPE_UNSIGNED_INT,
    NUM_TYPE_FLOATING_POINT,
    NUM_TYPE_OTHER,
};


template <class S, class Num> inline
S numberTo(const Num& number, Int2Type<NUM_TYPE_OTHER>) //default number to string conversion using streams: convenient, but SLOW, SLOW, SLOW!!!! (~ factor of 20)
{
    using CharType = typename GetCharType<S>::Type;

    std::basic_ostringstream<CharType> ss;
    ss << number;
    return copyStringTo<S>(ss.str());
}


template <class S, class Num> inline S floatToString(const Num& number, char   ) { return printNumber<S>( "%g", static_cast<double>(number)); }
template <class S, class Num> inline S floatToString(const Num& number, wchar_t) { return printNumber<S>(L"%g", static_cast<double>(number)); }

template <class S, class Num> inline
S numberTo(const Num& number, Int2Type<NUM_TYPE_FLOATING_POINT>)
{
    return floatToString<S>(number, typename GetCharType<S>::Type());
}


/*
perf: integer to string: (executed 10 mio. times)
    std::stringstream - 14796 ms
    std::sprintf      -  3086 ms
    formatInteger     -   778 ms
*/

template <class OutputIterator, class Num> inline
void formatNegativeInteger(Num n, OutputIterator& it)
{
    assert(n < 0);
    using CharType = typename std::iterator_traits<OutputIterator>::value_type;
    do
    {
        const Num tmp = n / 10;
        *--it = static_cast<CharType>('0' + (tmp * 10 - n)); //8% faster than using modulus operator!
        n = tmp;
    }
    while (n != 0);

    *--it = static_cast<CharType>('-');
}

template <class OutputIterator, class Num> inline
void formatPositiveInteger(Num n, OutputIterator& it)
{
    assert(n >= 0);
    using CharType = typename std::iterator_traits<OutputIterator>::value_type;
    do
    {
        const Num tmp = n / 10;
        *--it = static_cast<CharType>('0' + (n - tmp * 10)); //8% faster than using modulus operator!
        n = tmp;
    }
    while (n != 0);
}


template <class S, class Num> inline
S numberTo(const Num& number, Int2Type<NUM_TYPE_SIGNED_INT>)
{
    using CharType = typename GetCharType<S>::Type;
    CharType buffer[2 + sizeof(Num) * 241 / 100]; //zero-initialize?
    //it's generally faster to use a buffer than to rely on String::operator+=() (in)efficiency
    //required chars (+ sign char): 1 + ceil(ln_10(256^sizeof(n) / 2 + 1))    -> divide by 2 for signed half-range; second +1 since one half starts with 1!
    // <= 1 + ceil(ln_10(256^sizeof(n))) =~ 1 + ceil(sizeof(n) * 2.4082) <= 2 + floor(sizeof(n) * 2.41)

    //caveat: consider INT_MIN: technically -INT_MIN == INT_MIN
    auto it = std::end(buffer);
    if (number < 0)
        formatNegativeInteger(number, it);
    else
        formatPositiveInteger(number, it);
    assert(it >= std::begin(buffer));

    return S(&*it, std::end(buffer) - it);
}


template <class S, class Num> inline
S numberTo(const Num& number, Int2Type<NUM_TYPE_UNSIGNED_INT>)
{
    using CharType = typename GetCharType<S>::Type;
    CharType buffer[1 + sizeof(Num) * 241 / 100]; //zero-initialize?
    //required chars: ceil(ln_10(256^sizeof(n))) =~ ceil(sizeof(n) * 2.4082) <= 1 + floor(sizeof(n) * 2.41)

    auto it = std::end(buffer);
    formatPositiveInteger(number, it);
    assert(it >= std::begin(buffer));

    return S(&*it, std::end(buffer) - it);
}

//--------------------------------------------------------------------------------

template <class Num, class S> inline
Num stringTo(const S& str, Int2Type<NUM_TYPE_OTHER>) //default string to number conversion using streams: convenient, but SLOW
{
    using CharType = typename GetCharType<S>::Type;
    Num number = 0;
    std::basic_istringstream<CharType>(copyStringTo<std::basic_string<CharType>>(str)) >> number;
    return number;
}


template <class Num> inline Num stringToFloat(const char*    str) { return std::strtod(str, nullptr); }
template <class Num> inline Num stringToFloat(const wchar_t* str) { return std::wcstod(str, nullptr); }

template <class Num, class S> inline
Num stringTo(const S& str, Int2Type<NUM_TYPE_FLOATING_POINT>)
{
    return stringToFloat<Num>(strBegin(str));
}

template <class Num, class S>
Num extractInteger(const S& str, bool& hasMinusSign) //very fast conversion to integers: slightly faster than std::atoi, but more importantly: generic
{
    using CharType = typename GetCharType<S>::Type;

    const CharType* first = strBegin(str);
    const CharType* last  = first + strLength(str);

    while (first != last && isWhiteSpace(*first)) //skip leading whitespace
        ++first;

    //handle minus sign
    hasMinusSign = false;
    if (first != last)
    {
        if (*first == static_cast<CharType>('-'))
        {
            hasMinusSign = true;
            ++first;
        }
        else if (*first == static_cast<CharType>('+'))
            ++first;
    }

    Num number = 0;
    for (const CharType* it = first; it != last; ++it)
    {
        const CharType c = *it;
        if (static_cast<CharType>('0') <= c && c <= static_cast<CharType>('9'))
        {
            number *= 10;
            number += c - static_cast<CharType>('0');
        }
        else //rest of string should contain whitespace only, it's NOT a bug if there is something else!
            break; //assert(std::all_of(iter, last, &isWhiteSpace<CharType>)); -> this is NO assert situation
    }
    return number;
}


template <class Num, class S> inline
Num stringTo(const S& str, Int2Type<NUM_TYPE_SIGNED_INT>)
{
    bool hasMinusSign = false; //handle minus sign
    const Num number = extractInteger<Num>(str, hasMinusSign);
    return hasMinusSign ? -number : number;
}


template <class Num, class S> inline
Num stringTo(const S& str, Int2Type<NUM_TYPE_UNSIGNED_INT>) //very fast conversion to integers: slightly faster than std::atoi, but more importantly: generic
{
    bool hasMinusSign = false; //handle minus sign
    const Num number = extractInteger<Num>(str, hasMinusSign);
    if (hasMinusSign)
    {
        assert(false);
        return 0U;
    }
    return number;
}
}


template <class S, class Num> inline
S numberTo(const Num& number)
{
    using TypeTag = Int2Type<
                    IsSignedInt  <Num>::value ? impl::NUM_TYPE_SIGNED_INT :
                    IsUnsignedInt<Num>::value ? impl::NUM_TYPE_UNSIGNED_INT :
                    IsFloat      <Num>::value ? impl::NUM_TYPE_FLOATING_POINT :
                    impl::NUM_TYPE_OTHER>;

    return impl::numberTo<S>(number, TypeTag());
}


template <class Num, class S> inline
Num stringTo(const S& str)
{
    using TypeTag = Int2Type<
                    IsSignedInt  <Num>::value ? impl::NUM_TYPE_SIGNED_INT :
                    IsUnsignedInt<Num>::value ? impl::NUM_TYPE_UNSIGNED_INT :
                    IsFloat      <Num>::value ? impl::NUM_TYPE_FLOATING_POINT :
                    impl::NUM_TYPE_OTHER>;

    return impl::stringTo<Num>(str, TypeTag());
}


inline //hexify beats "printNumber<std::string>("%02X", c)" by a nice factor of 3!
std::pair<char, char> hexify(unsigned char c, bool upperCase)
{
    auto hexifyDigit = [upperCase](int num) -> char //input [0, 15], output 0-9, A-F
    {
        assert(0 <= num&& num <= 15);  //guaranteed by design below!
        if (num <= 9)
            return static_cast<char>('0' + num); //no signed/unsigned char problem here!

        if (upperCase)
            return static_cast<char>('A' + (num - 10));
        else
            return static_cast<char>('a' + (num - 10));
    };
    return std::make_pair(hexifyDigit(c / 16), hexifyDigit(c % 16));
}


inline //unhexify beats "::sscanf(&it[3], "%02X", &tmp)" by a factor of 3000 for ~250000 calls!!!
char unhexify(char high, char low)
{
    auto unhexifyDigit = [](char hex) -> int //input 0-9, a-f, A-F; output range: [0, 15]
    {
        if ('0' <= hex && hex <= '9') //no signed/unsigned char problem here!
            return hex - '0';
        else if ('A' <= hex && hex <= 'F')
            return (hex - 'A') + 10;
        else if ('a' <= hex && hex <= 'f')
            return (hex - 'a') + 10;
        assert(false);
        return 0;
    };
    return static_cast<unsigned char>(16 * unhexifyDigit(high) + unhexifyDigit(low)); //[!] convert to unsigned char first, then to char (which may be signed)
}
}

#endif //STRING_TOOLS_H_213458973046
bgstack15