tesseract v5.3.3.20231005
gtest-printers.cc
Go to the documentation of this file.
1// Copyright 2007, Google Inc.
2// All rights reserved.
3//
4// Redistribution and use in source and binary forms, with or without
5// modification, are permitted provided that the following conditions are
6// met:
7//
8// * Redistributions of source code must retain the above copyright
9// notice, this list of conditions and the following disclaimer.
10// * Redistributions in binary form must reproduce the above
11// copyright notice, this list of conditions and the following disclaimer
12// in the documentation and/or other materials provided with the
13// distribution.
14// * Neither the name of Google Inc. nor the names of its
15// contributors may be used to endorse or promote products derived from
16// this software without specific prior written permission.
17//
18// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
19// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
20// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
21// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
22// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
23// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
24// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
28// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29
30
31// Google Test - The Google C++ Testing and Mocking Framework
32//
33// This file implements a universal value printer that can print a
34// value of any type T:
35//
36// void ::testing::internal::UniversalPrinter<T>::Print(value, ostream_ptr);
37//
38// It uses the << operator when possible, and prints the bytes in the
39// object otherwise. A user can override its behavior for a class
40// type Foo by defining either operator<<(::std::ostream&, const Foo&)
41// or void PrintTo(const Foo&, ::std::ostream*) in the namespace that
42// defines Foo.
43
45
46#include <stdio.h>
47
48#include <cctype>
49#include <cstdint>
50#include <cwchar>
51#include <ostream> // NOLINT
52#include <string>
53#include <type_traits>
54
57
58namespace testing {
59
60namespace {
61
62using ::std::ostream;
63
64// Prints a segment of bytes in the given object.
69void PrintByteSegmentInObjectTo(const unsigned char* obj_bytes, size_t start,
70 size_t count, ostream* os) {
71 char text[5] = "";
72 for (size_t i = 0; i != count; i++) {
73 const size_t j = start + i;
74 if (i != 0) {
75 // Organizes the bytes into groups of 2 for easy parsing by
76 // human.
77 if ((j % 2) == 0)
78 *os << ' ';
79 else
80 *os << '-';
81 }
82 GTEST_SNPRINTF_(text, sizeof(text), "%02X", obj_bytes[j]);
83 *os << text;
84 }
85}
86
87// Prints the bytes in the given value to the given ostream.
88void PrintBytesInObjectToImpl(const unsigned char* obj_bytes, size_t count,
89 ostream* os) {
90 // Tells the user how big the object is.
91 *os << count << "-byte object <";
92
93 const size_t kThreshold = 132;
94 const size_t kChunkSize = 64;
95 // If the object size is bigger than kThreshold, we'll have to omit
96 // some details by printing only the first and the last kChunkSize
97 // bytes.
98 if (count < kThreshold) {
99 PrintByteSegmentInObjectTo(obj_bytes, 0, count, os);
100 } else {
101 PrintByteSegmentInObjectTo(obj_bytes, 0, kChunkSize, os);
102 *os << " ... ";
103 // Rounds up to 2-byte boundary.
104 const size_t resume_pos = (count - kChunkSize + 1)/2*2;
105 PrintByteSegmentInObjectTo(obj_bytes, resume_pos, count - resume_pos, os);
106 }
107 *os << ">";
108}
109
110// Helpers for widening a character to char32_t. Since the standard does not
111// specify if char / wchar_t is signed or unsigned, it is important to first
112// convert it to the unsigned type of the same width before widening it to
113// char32_t.
114template <typename CharType>
115char32_t ToChar32(CharType in) {
116 return static_cast<char32_t>(
117 static_cast<typename std::make_unsigned<CharType>::type>(in));
118}
119
120} // namespace
121
122namespace internal {
123
124// Delegates to PrintBytesInObjectToImpl() to print the bytes in the
125// given object. The delegation simplifies the implementation, which
126// uses the << operator and thus is easier done outside of the
127// ::testing::internal namespace, which contains a << operator that
128// sometimes conflicts with the one in STL.
129void PrintBytesInObjectTo(const unsigned char* obj_bytes, size_t count,
130 ostream* os) {
131 PrintBytesInObjectToImpl(obj_bytes, count, os);
132}
133
134// Depending on the value of a char (or wchar_t), we print it in one
135// of three formats:
136// - as is if it's a printable ASCII (e.g. 'a', '2', ' '),
137// - as a hexadecimal escape sequence (e.g. '\x7F'), or
138// - as a special escape sequence (e.g. '\r', '\n').
144
145// Returns true if c is a printable ASCII character. We test the
146// value of c directly instead of calling isprint(), which is buggy on
147// Windows Mobile.
148inline bool IsPrintableAscii(char32_t c) { return 0x20 <= c && c <= 0x7E; }
149
150// Prints c (of type char, char8_t, char16_t, char32_t, or wchar_t) as a
151// character literal without the quotes, escaping it when necessary; returns how
152// c was formatted.
153template <typename Char>
154static CharFormat PrintAsCharLiteralTo(Char c, ostream* os) {
155 const char32_t u_c = ToChar32(c);
156 switch (u_c) {
157 case L'\0':
158 *os << "\\0";
159 break;
160 case L'\'':
161 *os << "\\'";
162 break;
163 case L'\\':
164 *os << "\\\\";
165 break;
166 case L'\a':
167 *os << "\\a";
168 break;
169 case L'\b':
170 *os << "\\b";
171 break;
172 case L'\f':
173 *os << "\\f";
174 break;
175 case L'\n':
176 *os << "\\n";
177 break;
178 case L'\r':
179 *os << "\\r";
180 break;
181 case L'\t':
182 *os << "\\t";
183 break;
184 case L'\v':
185 *os << "\\v";
186 break;
187 default:
188 if (IsPrintableAscii(u_c)) {
189 *os << static_cast<char>(c);
190 return kAsIs;
191 } else {
192 ostream::fmtflags flags = os->flags();
193 *os << "\\x" << std::hex << std::uppercase << static_cast<int>(u_c);
194 os->flags(flags);
195 return kHexEscape;
196 }
197 }
198 return kSpecialEscape;
199}
200
201// Prints a char32_t c as if it's part of a string literal, escaping it when
202// necessary; returns how c was formatted.
203static CharFormat PrintAsStringLiteralTo(char32_t c, ostream* os) {
204 switch (c) {
205 case L'\'':
206 *os << "'";
207 return kAsIs;
208 case L'"':
209 *os << "\\\"";
210 return kSpecialEscape;
211 default:
212 return PrintAsCharLiteralTo(c, os);
213 }
214}
215
216static const char* GetCharWidthPrefix(char) {
217 return "";
218}
219
220static const char* GetCharWidthPrefix(signed char) {
221 return "";
222}
223
224static const char* GetCharWidthPrefix(unsigned char) {
225 return "";
226}
227
228#ifdef __cpp_char8_t
229static const char* GetCharWidthPrefix(char8_t) {
230 return "u8";
231}
232#endif
233
234static const char* GetCharWidthPrefix(char16_t) {
235 return "u";
236}
237
238static const char* GetCharWidthPrefix(char32_t) {
239 return "U";
240}
241
242static const char* GetCharWidthPrefix(wchar_t) {
243 return "L";
244}
245
246// Prints a char c as if it's part of a string literal, escaping it when
247// necessary; returns how c was formatted.
248static CharFormat PrintAsStringLiteralTo(char c, ostream* os) {
249 return PrintAsStringLiteralTo(ToChar32(c), os);
250}
251
252#ifdef __cpp_char8_t
253static CharFormat PrintAsStringLiteralTo(char8_t c, ostream* os) {
254 return PrintAsStringLiteralTo(ToChar32(c), os);
255}
256#endif
257
258static CharFormat PrintAsStringLiteralTo(char16_t c, ostream* os) {
259 return PrintAsStringLiteralTo(ToChar32(c), os);
260}
261
262static CharFormat PrintAsStringLiteralTo(wchar_t c, ostream* os) {
263 return PrintAsStringLiteralTo(ToChar32(c), os);
264}
265
266// Prints a character c (of type char, char8_t, char16_t, char32_t, or wchar_t)
267// and its code. '\0' is printed as "'\\0'", other unprintable characters are
268// also properly escaped using the standard C++ escape sequence.
269template <typename Char>
270void PrintCharAndCodeTo(Char c, ostream* os) {
271 // First, print c as a literal in the most readable form we can find.
272 *os << GetCharWidthPrefix(c) << "'";
273 const CharFormat format = PrintAsCharLiteralTo(c, os);
274 *os << "'";
275
276 // To aid user debugging, we also print c's code in decimal, unless
277 // it's 0 (in which case c was printed as '\\0', making the code
278 // obvious).
279 if (c == 0)
280 return;
281 *os << " (" << static_cast<int>(c);
282
283 // For more convenience, we print c's code again in hexadecimal,
284 // unless c was already printed in the form '\x##' or the code is in
285 // [1, 9].
286 if (format == kHexEscape || (1 <= c && c <= 9)) {
287 // Do nothing.
288 } else {
289 *os << ", 0x" << String::FormatHexInt(static_cast<int>(c));
290 }
291 *os << ")";
292}
293
294void PrintTo(unsigned char c, ::std::ostream* os) { PrintCharAndCodeTo(c, os); }
295void PrintTo(signed char c, ::std::ostream* os) { PrintCharAndCodeTo(c, os); }
296
297// Prints a wchar_t as a symbol if it is printable or as its internal
298// code otherwise and also as its code. L'\0' is printed as "L'\\0'".
299void PrintTo(wchar_t wc, ostream* os) { PrintCharAndCodeTo(wc, os); }
300
301// TODO(dcheng): Consider making this delegate to PrintCharAndCodeTo() as well.
302void PrintTo(char32_t c, ::std::ostream* os) {
303 *os << std::hex << "U+" << std::uppercase << std::setfill('0') << std::setw(4)
304 << static_cast<uint32_t>(c);
305}
306
307// Prints the given array of characters to the ostream. CharType must be either
308// char, char8_t, char16_t, char32_t, or wchar_t.
309// The array starts at begin, the length is len, it may include '\0' characters
310// and may not be NUL-terminated.
311template <typename CharType>
316static CharFormat PrintCharsAsStringTo(
317 const CharType* begin, size_t len, ostream* os) {
318 const char* const quote_prefix = GetCharWidthPrefix(*begin);
319 *os << quote_prefix << "\"";
320 bool is_previous_hex = false;
321 CharFormat print_format = kAsIs;
322 for (size_t index = 0; index < len; ++index) {
323 const CharType cur = begin[index];
324 if (is_previous_hex && IsXDigit(cur)) {
325 // Previous character is of '\x..' form and this character can be
326 // interpreted as another hexadecimal digit in its number. Break string to
327 // disambiguate.
328 *os << "\" " << quote_prefix << "\"";
329 }
330 is_previous_hex = PrintAsStringLiteralTo(cur, os) == kHexEscape;
331 // Remember if any characters required hex escaping.
332 if (is_previous_hex) {
333 print_format = kHexEscape;
334 }
335 }
336 *os << "\"";
337 return print_format;
338}
339
340// Prints a (const) char/wchar_t array of 'len' elements, starting at address
341// 'begin'. CharType must be either char or wchar_t.
342template <typename CharType>
347static void UniversalPrintCharArray(
348 const CharType* begin, size_t len, ostream* os) {
349 // The code
350 // const char kFoo[] = "foo";
351 // generates an array of 4, not 3, elements, with the last one being '\0'.
352 //
353 // Therefore when printing a char array, we don't print the last element if
354 // it's '\0', such that the output matches the string literal as it's
355 // written in the source code.
356 if (len > 0 && begin[len - 1] == '\0') {
357 PrintCharsAsStringTo(begin, len - 1, os);
358 return;
359 }
360
361 // If, however, the last element in the array is not '\0', e.g.
362 // const char kFoo[] = { 'f', 'o', 'o' };
363 // we must print the entire array. We also print a message to indicate
364 // that the array is not NUL-terminated.
365 PrintCharsAsStringTo(begin, len, os);
366 *os << " (no terminating NUL)";
367}
368
369// Prints a (const) char array of 'len' elements, starting at address 'begin'.
370void UniversalPrintArray(const char* begin, size_t len, ostream* os) {
371 UniversalPrintCharArray(begin, len, os);
372}
373
374#ifdef __cpp_char8_t
375// Prints a (const) char8_t array of 'len' elements, starting at address
376// 'begin'.
377void UniversalPrintArray(const char8_t* begin, size_t len, ostream* os) {
378 UniversalPrintCharArray(begin, len, os);
379}
380#endif
381
382// Prints a (const) char16_t array of 'len' elements, starting at address
383// 'begin'.
384void UniversalPrintArray(const char16_t* begin, size_t len, ostream* os) {
385 UniversalPrintCharArray(begin, len, os);
386}
387
388// Prints a (const) char32_t array of 'len' elements, starting at address
389// 'begin'.
390void UniversalPrintArray(const char32_t* begin, size_t len, ostream* os) {
391 UniversalPrintCharArray(begin, len, os);
392}
393
394// Prints a (const) wchar_t array of 'len' elements, starting at address
395// 'begin'.
396void UniversalPrintArray(const wchar_t* begin, size_t len, ostream* os) {
397 UniversalPrintCharArray(begin, len, os);
398}
399
400namespace {
401
402// Prints a null-terminated C-style string to the ostream.
403template <typename Char>
404void PrintCStringTo(const Char* s, ostream* os) {
405 if (s == nullptr) {
406 *os << "NULL";
407 } else {
408 *os << ImplicitCast_<const void*>(s) << " pointing to ";
409 PrintCharsAsStringTo(s, std::char_traits<Char>::length(s), os);
410 }
411}
412
413} // anonymous namespace
414
415void PrintTo(const char* s, ostream* os) { PrintCStringTo(s, os); }
416
417#ifdef __cpp_char8_t
418void PrintTo(const char8_t* s, ostream* os) { PrintCStringTo(s, os); }
419#endif
420
421void PrintTo(const char16_t* s, ostream* os) { PrintCStringTo(s, os); }
422
423void PrintTo(const char32_t* s, ostream* os) { PrintCStringTo(s, os); }
424
425// MSVC compiler can be configured to define whar_t as a typedef
426// of unsigned short. Defining an overload for const wchar_t* in that case
427// would cause pointers to unsigned shorts be printed as wide strings,
428// possibly accessing more memory than intended and causing invalid
429// memory accesses. MSVC defines _NATIVE_WCHAR_T_DEFINED symbol when
430// wchar_t is implemented as a native type.
431#if !defined(_MSC_VER) || defined(_NATIVE_WCHAR_T_DEFINED)
432// Prints the given wide C string to the ostream.
433void PrintTo(const wchar_t* s, ostream* os) { PrintCStringTo(s, os); }
434#endif // wchar_t is native
435
436namespace {
437
438bool ContainsUnprintableControlCodes(const char* str, size_t length) {
439 const unsigned char *s = reinterpret_cast<const unsigned char *>(str);
440
441 for (size_t i = 0; i < length; i++) {
442 unsigned char ch = *s++;
443 if (std::iscntrl(ch)) {
444 switch (ch) {
445 case '\t':
446 case '\n':
447 case '\r':
448 break;
449 default:
450 return true;
451 }
452 }
453 }
454 return false;
455}
456
457bool IsUTF8TrailByte(unsigned char t) { return 0x80 <= t && t<= 0xbf; }
458
459bool IsValidUTF8(const char* str, size_t length) {
460 const unsigned char *s = reinterpret_cast<const unsigned char *>(str);
461
462 for (size_t i = 0; i < length;) {
463 unsigned char lead = s[i++];
464
465 if (lead <= 0x7f) {
466 continue; // single-byte character (ASCII) 0..7F
467 }
468 if (lead < 0xc2) {
469 return false; // trail byte or non-shortest form
470 } else if (lead <= 0xdf && (i + 1) <= length && IsUTF8TrailByte(s[i])) {
471 ++i; // 2-byte character
472 } else if (0xe0 <= lead && lead <= 0xef && (i + 2) <= length &&
473 IsUTF8TrailByte(s[i]) &&
474 IsUTF8TrailByte(s[i + 1]) &&
475 // check for non-shortest form and surrogate
476 (lead != 0xe0 || s[i] >= 0xa0) &&
477 (lead != 0xed || s[i] < 0xa0)) {
478 i += 2; // 3-byte character
479 } else if (0xf0 <= lead && lead <= 0xf4 && (i + 3) <= length &&
480 IsUTF8TrailByte(s[i]) &&
481 IsUTF8TrailByte(s[i + 1]) &&
482 IsUTF8TrailByte(s[i + 2]) &&
483 // check for non-shortest form
484 (lead != 0xf0 || s[i] >= 0x90) &&
485 (lead != 0xf4 || s[i] < 0x90)) {
486 i += 3; // 4-byte character
487 } else {
488 return false;
489 }
490 }
491 return true;
492}
493
494void ConditionalPrintAsText(const char* str, size_t length, ostream* os) {
495 if (!ContainsUnprintableControlCodes(str, length) &&
496 IsValidUTF8(str, length)) {
497 *os << "\n As Text: \"" << str << "\"";
498 }
499}
500
501} // anonymous namespace
502
503void PrintStringTo(const ::std::string& s, ostream* os) {
504 if (PrintCharsAsStringTo(s.data(), s.size(), os) == kHexEscape) {
505 if (GTEST_FLAG(print_utf8)) {
506 ConditionalPrintAsText(s.data(), s.size(), os);
507 }
508 }
509}
510
511#ifdef __cpp_char8_t
512void PrintU8StringTo(const ::std::u8string& s, ostream* os) {
513 PrintCharsAsStringTo(s.data(), s.size(), os);
514}
515#endif
516
517void PrintU16StringTo(const ::std::u16string& s, ostream* os) {
518 PrintCharsAsStringTo(s.data(), s.size(), os);
519}
520
521void PrintU32StringTo(const ::std::u32string& s, ostream* os) {
522 PrintCharsAsStringTo(s.data(), s.size(), os);
523}
524
525#if GTEST_HAS_STD_WSTRING
526void PrintWideStringTo(const ::std::wstring& s, ostream* os) {
527 PrintCharsAsStringTo(s.data(), s.size(), os);
528}
529#endif // GTEST_HAS_STD_WSTRING
530
531} // namespace internal
532
533} // namespace testing
int * count
#define GTEST_ATTRIBUTE_NO_SANITIZE_MEMORY_
Definition: gtest-port.h:812
#define GTEST_ATTRIBUTE_NO_SANITIZE_HWADDRESS_
Definition: gtest-port.h:836
#define GTEST_FLAG(name)
Definition: gtest-port.h:2205
#define GTEST_ATTRIBUTE_NO_SANITIZE_THREAD_
Definition: gtest-port.h:848
#define GTEST_SNPRINTF_
Definition: gtest-port.h:2145
#define GTEST_ATTRIBUTE_NO_SANITIZE_ADDRESS_
Definition: gtest-port.h:824
bool IsPrintableAscii(char32_t c)
GTEST_API_ void PrintU32StringTo(const ::std::u32string &s, ::std::ostream *os)
GTEST_API_ void PrintStringTo(const ::std::string &s, ::std::ostream *os)
bool IsXDigit(char ch)
Definition: gtest-port.h:1940
GTEST_API_ void PrintBytesInObjectTo(const unsigned char *obj_bytes, size_t count, ::std::ostream *os)
GTEST_API_ void PrintU16StringTo(const ::std::u16string &s, ::std::ostream *os)
void UniversalPrintArray(const T *begin, size_t len, ::std::ostream *os)
void PrintTo(const T &value, ::std::ostream *os)
void PrintCharAndCodeTo(Char c, ostream *os)
type
Definition: upload.py:458
static std::string FormatHexInt(int value)
Definition: gtest.cc:2150