aboutsummaryrefslogtreecommitdiff
path: root/src/google/protobuf/compiler/cpp/cpp_helpers.cc
blob: 163dac0aa0bc429285cb1d5590e590a91b87f230 (plain) (blame)
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
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
// Protocol Buffers - Google's data interchange format
// Copyright 2008 Google Inc.  All rights reserved.
// https://developers.google.com/protocol-buffers/
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
//     * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//     * Redistributions in binary form must reproduce the above
// copyright notice, this list of conditions and the following disclaimer
// in the documentation and/or other materials provided with the
// distribution.
//     * Neither the name of Google Inc. nor the names of its
// contributors may be used to endorse or promote products derived from
// this software without specific prior written permission.
//
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

// Author: kenton@google.com (Kenton Varda)
//  Based on original Protocol Buffers design by
//  Sanjay Ghemawat, Jeff Dean, and others.

#include <google/protobuf/stubs/hash.h>
#include <limits>
#include <map>
#include <queue>
#include <vector>

#include <google/protobuf/stubs/logging.h>
#include <google/protobuf/stubs/common.h>
#include <google/protobuf/compiler/cpp/cpp_helpers.h>
#include <google/protobuf/io/printer.h>
#include <google/protobuf/io/zero_copy_stream.h>
#include <google/protobuf/stubs/strutil.h>
#include <google/protobuf/stubs/substitute.h>




namespace google {
namespace protobuf {
namespace compiler {
namespace cpp {

namespace {

static const char kAnyMessageName[] = "Any";
static const char kAnyProtoFile[] = "google/protobuf/any.proto";
static const char kGoogleProtobufPrefix[] = "google/protobuf/";

string DotsToUnderscores(const string& name) {
  return StringReplace(name, ".", "_", true);
}

string DotsToColons(const string& name) {
  return StringReplace(name, ".", "::", true);
}

const char* const kKeywordList[] = {
  "alignas", "alignof", "and", "and_eq", "asm", "auto", "bitand", "bitor",
  "bool", "break", "case", "catch", "char", "class", "compl", "const",
  "constexpr", "const_cast", "continue", "decltype", "default", "delete", "do",
  "double", "dynamic_cast", "else", "enum", "explicit", "export", "extern",
  "false", "float", "for", "friend", "goto", "if", "inline", "int", "long",
  "mutable", "namespace", "new", "noexcept", "not", "not_eq", "nullptr",
  "operator", "or", "or_eq", "private", "protected", "public", "register",
  "reinterpret_cast", "return", "short", "signed", "sizeof", "static",
  "static_assert", "static_cast", "struct", "switch", "template", "this",
  "thread_local", "throw", "true", "try", "typedef", "typeid", "typename",
  "union", "unsigned", "using", "virtual", "void", "volatile", "wchar_t",
  "while", "xor", "xor_eq"
};

hash_set<string> MakeKeywordsMap() {
  hash_set<string> result;
  for (int i = 0; i < GOOGLE_ARRAYSIZE(kKeywordList); i++) {
    result.insert(kKeywordList[i]);
  }
  return result;
}

hash_set<string> kKeywords = MakeKeywordsMap();

// Returns whether the provided descriptor has an extension. This includes its
// nested types.
bool HasExtension(const Descriptor* descriptor) {
  if (descriptor->extension_count() > 0) {
    return true;
  }
  for (int i = 0; i < descriptor->nested_type_count(); ++i) {
    if (HasExtension(descriptor->nested_type(i))) {
      return true;
    }
  }
  return false;
}

// Encode [0..63] as 'A'-'Z', 'a'-'z', '0'-'9', '_'
char Base63Char(int value) {
  GOOGLE_CHECK_GE(value, 0);
  if (value < 26) return 'A' + value;
  value -= 26;
  if (value < 26) return 'a' + value;
  value -= 26;
  if (value < 10) return '0' + value;
  GOOGLE_CHECK_EQ(value, 10);
  return '_';
}

// Given a c identifier has 63 legal characters we can't implement base64
// encoding. So we return the k least significant "digits" in base 63.
template <typename I>
string Base63(I n, int k) {
  string res;
  while (k-- > 0) {
    res += Base63Char(static_cast<int>(n % 63));
    n /= 63;
  }
  return res;
}

}  // namespace

string UnderscoresToCamelCase(const string& input, bool cap_next_letter) {
  string result;
  // Note:  I distrust ctype.h due to locales.
  for (int i = 0; i < input.size(); i++) {
    if ('a' <= input[i] && input[i] <= 'z') {
      if (cap_next_letter) {
        result += input[i] + ('A' - 'a');
      } else {
        result += input[i];
      }
      cap_next_letter = false;
    } else if ('A' <= input[i] && input[i] <= 'Z') {
      // Capital letters are left as-is.
      result += input[i];
      cap_next_letter = false;
    } else if ('0' <= input[i] && input[i] <= '9') {
      result += input[i];
      cap_next_letter = true;
    } else {
      cap_next_letter = true;
    }
  }
  return result;
}

const char kThickSeparator[] =
  "// ===================================================================\n";
const char kThinSeparator[] =
  "// -------------------------------------------------------------------\n";

bool CanInitializeByZeroing(const FieldDescriptor* field) {
  if (field->is_repeated() || field->is_extension()) return false;
  switch (field->cpp_type()) {
    case FieldDescriptor::CPPTYPE_ENUM:
      return field->default_value_enum()->number() == 0;
    case FieldDescriptor::CPPTYPE_INT32:
      return field->default_value_int32() == 0;
    case FieldDescriptor::CPPTYPE_INT64:
      return field->default_value_int64() == 0;
    case FieldDescriptor::CPPTYPE_UINT32:
      return field->default_value_uint32() == 0;
    case FieldDescriptor::CPPTYPE_UINT64:
      return field->default_value_uint64() == 0;
    case FieldDescriptor::CPPTYPE_FLOAT:
      return field->default_value_float() == 0;
    case FieldDescriptor::CPPTYPE_DOUBLE:
      return field->default_value_double() == 0;
    case FieldDescriptor::CPPTYPE_BOOL:
      return field->default_value_bool() == false;
    default:
      return false;
  }
}

string ClassName(const Descriptor* descriptor) {
  const Descriptor* parent = descriptor->containing_type();
  string res;
  if (parent) res += ClassName(parent) + "_";
  res += descriptor->name();
  if (IsMapEntryMessage(descriptor)) res += "_DoNotUse";
  return res;
}

string ClassName(const EnumDescriptor* enum_descriptor) {
  if (enum_descriptor->containing_type() == NULL) {
    return enum_descriptor->name();
  } else {
    return ClassName(enum_descriptor->containing_type()) + "_" +
           enum_descriptor->name();
  }
}

string Namespace(const string& package) {
  if (package.empty()) return "";
  return "::" + DotsToColons(package);
}

string DefaultInstanceName(const Descriptor* descriptor) {
  string prefix = descriptor->file()->package().empty() ? "" : "::";
  return prefix + DotsToColons(descriptor->file()->package()) + "::_" +
      ClassName(descriptor, false) + "_default_instance_";
}

string ReferenceFunctionName(const Descriptor* descriptor) {
  return QualifiedClassName(descriptor) + "_ReferenceStrong";
}

string SuperClassName(const Descriptor* descriptor, const Options& options) {
  return HasDescriptorMethods(descriptor->file(), options)
             ? "::google::protobuf::Message"
             : "::google::protobuf::MessageLite";
}

string FieldName(const FieldDescriptor* field) {
  string result = field->name();
  LowerString(&result);
  if (kKeywords.count(result) > 0) {
    result.append("_");
  }
  return result;
}

string EnumValueName(const EnumValueDescriptor* enum_value) {
  string result = enum_value->name();
  if (kKeywords.count(result) > 0) {
    result.append("_");
  }
  return result;
}

int EstimateAlignmentSize(const FieldDescriptor* field) {
  if (field == NULL) return 0;
  if (field->is_repeated()) return 8;
  switch (field->cpp_type()) {
    case FieldDescriptor::CPPTYPE_BOOL:
      return 1;

    case FieldDescriptor::CPPTYPE_INT32:
    case FieldDescriptor::CPPTYPE_UINT32:
    case FieldDescriptor::CPPTYPE_ENUM:
    case FieldDescriptor::CPPTYPE_FLOAT:
      return 4;

    case FieldDescriptor::CPPTYPE_INT64:
    case FieldDescriptor::CPPTYPE_UINT64:
    case FieldDescriptor::CPPTYPE_DOUBLE:
    case FieldDescriptor::CPPTYPE_STRING:
    case FieldDescriptor::CPPTYPE_MESSAGE:
      return 8;
  }
  GOOGLE_LOG(FATAL) << "Can't get here.";
  return -1;  // Make compiler happy.
}

string FieldConstantName(const FieldDescriptor *field) {
  string field_name = UnderscoresToCamelCase(field->name(), true);
  string result = "k" + field_name + "FieldNumber";

  if (!field->is_extension() &&
      field->containing_type()->FindFieldByCamelcaseName(
        field->camelcase_name()) != field) {
    // This field's camelcase name is not unique.  As a hack, add the field
    // number to the constant name.  This makes the constant rather useless,
    // but what can we do?
    result += "_" + SimpleItoa(field->number());
  }

  return result;
}

string FieldMessageTypeName(const FieldDescriptor* field) {
  // Note:  The Google-internal version of Protocol Buffers uses this function
  //   as a hook point for hacks to support legacy code.
  return ClassName(field->message_type(), true);
}

string StripProto(const string& filename) {
  if (HasSuffixString(filename, ".protodevel")) {
    return StripSuffixString(filename, ".protodevel");
  } else {
    return StripSuffixString(filename, ".proto");
  }
}

const char* PrimitiveTypeName(FieldDescriptor::CppType type) {
  switch (type) {
    case FieldDescriptor::CPPTYPE_INT32  : return "::google::protobuf::int32";
    case FieldDescriptor::CPPTYPE_INT64  : return "::google::protobuf::int64";
    case FieldDescriptor::CPPTYPE_UINT32 : return "::google::protobuf::uint32";
    case FieldDescriptor::CPPTYPE_UINT64 : return "::google::protobuf::uint64";
    case FieldDescriptor::CPPTYPE_DOUBLE : return "double";
    case FieldDescriptor::CPPTYPE_FLOAT  : return "float";
    case FieldDescriptor::CPPTYPE_BOOL   : return "bool";
    case FieldDescriptor::CPPTYPE_ENUM   : return "int";
    case FieldDescriptor::CPPTYPE_STRING : return "::std::string";
    case FieldDescriptor::CPPTYPE_MESSAGE: return NULL;

    // No default because we want the compiler to complain if any new
    // CppTypes are added.
  }

  GOOGLE_LOG(FATAL) << "Can't get here.";
  return NULL;
}

const char* DeclaredTypeMethodName(FieldDescriptor::Type type) {
  switch (type) {
    case FieldDescriptor::TYPE_INT32   : return "Int32";
    case FieldDescriptor::TYPE_INT64   : return "Int64";
    case FieldDescriptor::TYPE_UINT32  : return "UInt32";
    case FieldDescriptor::TYPE_UINT64  : return "UInt64";
    case FieldDescriptor::TYPE_SINT32  : return "SInt32";
    case FieldDescriptor::TYPE_SINT64  : return "SInt64";
    case FieldDescriptor::TYPE_FIXED32 : return "Fixed32";
    case FieldDescriptor::TYPE_FIXED64 : return "Fixed64";
    case FieldDescriptor::TYPE_SFIXED32: return "SFixed32";
    case FieldDescriptor::TYPE_SFIXED64: return "SFixed64";
    case FieldDescriptor::TYPE_FLOAT   : return "Float";
    case FieldDescriptor::TYPE_DOUBLE  : return "Double";

    case FieldDescriptor::TYPE_BOOL    : return "Bool";
    case FieldDescriptor::TYPE_ENUM    : return "Enum";

    case FieldDescriptor::TYPE_STRING  : return "String";
    case FieldDescriptor::TYPE_BYTES   : return "Bytes";
    case FieldDescriptor::TYPE_GROUP   : return "Group";
    case FieldDescriptor::TYPE_MESSAGE : return "Message";

    // No default because we want the compiler to complain if any new
    // types are added.
  }
  GOOGLE_LOG(FATAL) << "Can't get here.";
  return "";
}

string Int32ToString(int number) {
  // gcc rejects the decimal form of kint32min.
  if (number == kint32min) {
    GOOGLE_COMPILE_ASSERT(kint32min == (~0x7fffffff), kint32min_value_error);
    return "(~0x7fffffff)";
  } else {
    return SimpleItoa(number);
  }
}

string Int64ToString(int64 number) {
  // gcc rejects the decimal form of kint64min
  if (number == kint64min) {
    // Make sure we are in a 2's complement system.
    GOOGLE_COMPILE_ASSERT(kint64min == GOOGLE_LONGLONG(~0x7fffffffffffffff),
                   kint64min_value_error);
    return "GOOGLE_LONGLONG(~0x7fffffffffffffff)";
  }
  return "GOOGLE_LONGLONG(" + SimpleItoa(number) + ")";
}

string DefaultValue(const FieldDescriptor* field) {
  switch (field->cpp_type()) {
    case FieldDescriptor::CPPTYPE_INT32:
      return Int32ToString(field->default_value_int32());
    case FieldDescriptor::CPPTYPE_UINT32:
      return SimpleItoa(field->default_value_uint32()) + "u";
    case FieldDescriptor::CPPTYPE_INT64:
      return Int64ToString(field->default_value_int64());
    case FieldDescriptor::CPPTYPE_UINT64:
      return "GOOGLE_ULONGLONG(" + SimpleItoa(field->default_value_uint64())+ ")";
    case FieldDescriptor::CPPTYPE_DOUBLE: {
      double value = field->default_value_double();
      if (value == std::numeric_limits<double>::infinity()) {
        return "::google::protobuf::internal::Infinity()";
      } else if (value == -std::numeric_limits<double>::infinity()) {
        return "-::google::protobuf::internal::Infinity()";
      } else if (value != value) {
        return "::google::protobuf::internal::NaN()";
      } else {
        return SimpleDtoa(value);
      }
    }
    case FieldDescriptor::CPPTYPE_FLOAT:
      {
        float value = field->default_value_float();
        if (value == std::numeric_limits<float>::infinity()) {
          return "static_cast<float>(::google::protobuf::internal::Infinity())";
        } else if (value == -std::numeric_limits<float>::infinity()) {
          return "static_cast<float>(-::google::protobuf::internal::Infinity())";
        } else if (value != value) {
          return "static_cast<float>(::google::protobuf::internal::NaN())";
        } else {
          string float_value = SimpleFtoa(value);
          // If floating point value contains a period (.) or an exponent
          // (either E or e), then append suffix 'f' to make it a float
          // literal.
          if (float_value.find_first_of(".eE") != string::npos) {
            float_value.push_back('f');
          }
          return float_value;
        }
      }
    case FieldDescriptor::CPPTYPE_BOOL:
      return field->default_value_bool() ? "true" : "false";
    case FieldDescriptor::CPPTYPE_ENUM:
      // Lazy:  Generate a static_cast because we don't have a helper function
      //   that constructs the full name of an enum value.
      return strings::Substitute(
          "static_cast< $0 >($1)",
          ClassName(field->enum_type(), true),
          Int32ToString(field->default_value_enum()->number()));
    case FieldDescriptor::CPPTYPE_STRING:
      return "\"" + EscapeTrigraphs(
        CEscape(field->default_value_string())) +
        "\"";
    case FieldDescriptor::CPPTYPE_MESSAGE:
      return "*" + FieldMessageTypeName(field) +
             "::internal_default_instance()";
  }
  // Can't actually get here; make compiler happy.  (We could add a default
  // case above but then we wouldn't get the nice compiler warning when a
  // new type is added.)
  GOOGLE_LOG(FATAL) << "Can't get here.";
  return "";
}

// Convert a file name into a valid identifier.
string FilenameIdentifier(const string& filename) {
  string result;
  for (int i = 0; i < filename.size(); i++) {
    if (ascii_isalnum(filename[i])) {
      result.push_back(filename[i]);
    } else {
      // Not alphanumeric.  To avoid any possibility of name conflicts we
      // use the hex code for the character.
      StrAppend(&result, "_", strings::Hex(static_cast<uint8>(filename[i])));
    }
  }
  return result;
}

string FileLevelNamespace(const string& filename) {
  return "protobuf_" + FilenameIdentifier(filename);
}

// Return the qualified C++ name for a file level symbol.
string QualifiedFileLevelSymbol(const string& package, const string& name) {
  if (package.empty()) {
    return StrCat("::", name);
  }
  return StrCat("::", DotsToColons(package), "::", name);
}

// Escape C++ trigraphs by escaping question marks to \?
string EscapeTrigraphs(const string& to_escape) {
  return StringReplace(to_escape, "?", "\\?", true);
}

// Escaped function name to eliminate naming conflict.
string SafeFunctionName(const Descriptor* descriptor,
                        const FieldDescriptor* field,
                        const string& prefix) {
  // Do not use FieldName() since it will escape keywords.
  string name = field->name();
  LowerString(&name);
  string function_name = prefix + name;
  if (descriptor->FindFieldByName(function_name)) {
    // Single underscore will also make it conflicting with the private data
    // member. We use double underscore to escape function names.
    function_name.append("__");
  } else if (kKeywords.count(name) > 0) {
    // If the field name is a keyword, we append the underscore back to keep it
    // consistent with other function names.
    function_name.append("_");
  }
  return function_name;
}


static bool HasMapFields(const Descriptor* descriptor) {
  for (int i = 0; i < descriptor->field_count(); ++i) {
    if (descriptor->field(i)->is_map()) {
      return true;
    }
  }
  for (int i = 0; i < descriptor->nested_type_count(); ++i) {
    if (HasMapFields(descriptor->nested_type(i))) return true;
  }
  return false;
}

bool HasMapFields(const FileDescriptor* file) {
  for (int i = 0; i < file->message_type_count(); ++i) {
    if (HasMapFields(file->message_type(i))) return true;
  }
  return false;
}

static bool HasEnumDefinitions(const Descriptor* message_type) {
  if (message_type->enum_type_count() > 0) return true;
  for (int i = 0; i < message_type->nested_type_count(); ++i) {
    if (HasEnumDefinitions(message_type->nested_type(i))) return true;
  }
  return false;
}

bool HasEnumDefinitions(const FileDescriptor* file) {
  if (file->enum_type_count() > 0) return true;
  for (int i = 0; i < file->message_type_count(); ++i) {
    if (HasEnumDefinitions(file->message_type(i))) return true;
  }
  return false;
}

bool IsStringOrMessage(const FieldDescriptor* field) {
  switch (field->cpp_type()) {
    case FieldDescriptor::CPPTYPE_INT32:
    case FieldDescriptor::CPPTYPE_INT64:
    case FieldDescriptor::CPPTYPE_UINT32:
    case FieldDescriptor::CPPTYPE_UINT64:
    case FieldDescriptor::CPPTYPE_DOUBLE:
    case FieldDescriptor::CPPTYPE_FLOAT:
    case FieldDescriptor::CPPTYPE_BOOL:
    case FieldDescriptor::CPPTYPE_ENUM:
      return false;
    case FieldDescriptor::CPPTYPE_STRING:
    case FieldDescriptor::CPPTYPE_MESSAGE:
      return true;
  }

  GOOGLE_LOG(FATAL) << "Can't get here.";
  return false;
}

FieldOptions::CType EffectiveStringCType(const FieldDescriptor* field) {
  GOOGLE_DCHECK(field->cpp_type() == FieldDescriptor::CPPTYPE_STRING);
  // Open-source protobuf release only supports STRING ctype.
  return FieldOptions::STRING;

}

bool IsAnyMessage(const FileDescriptor* descriptor) {
  return descriptor->name() == kAnyProtoFile;
}

bool IsAnyMessage(const Descriptor* descriptor) {
  return descriptor->name() == kAnyMessageName &&
         descriptor->file()->name() == kAnyProtoFile;
}

bool IsWellKnownMessage(const FileDescriptor* descriptor) {
  return !descriptor->name().compare(0, 16, kGoogleProtobufPrefix);
}

enum Utf8CheckMode {
  STRICT = 0,  // Parsing will fail if non UTF-8 data is in string fields.
  VERIFY = 1,  // Only log an error but parsing will succeed.
  NONE = 2,  // No UTF-8 check.
};

// Which level of UTF-8 enforcemant is placed on this file.
static Utf8CheckMode GetUtf8CheckMode(const FieldDescriptor* field,
                                      const Options& options) {
  if (field->file()->syntax() == FileDescriptor::SYNTAX_PROTO3) {
    return STRICT;
  } else if (GetOptimizeFor(field->file(), options) !=
             FileOptions::LITE_RUNTIME) {
    return VERIFY;
  } else {
    return NONE;
  }
}

static void GenerateUtf8CheckCode(const FieldDescriptor* field,
                                  const Options& options, bool for_parse,
                                  const std::map<string, string>& variables,
                                  const char* parameters,
                                  const char* strict_function,
                                  const char* verify_function,
                                  io::Printer* printer) {
  switch (GetUtf8CheckMode(field, options)) {
    case STRICT: {
      if (for_parse) {
        printer->Print("DO_(");
      }
      printer->Print(
          "::google::protobuf::internal::WireFormatLite::$function$(\n",
          "function", strict_function);
      printer->Indent();
      printer->Print(variables, parameters);
      if (for_parse) {
        printer->Print("::google::protobuf::internal::WireFormatLite::PARSE,\n");
      } else {
        printer->Print("::google::protobuf::internal::WireFormatLite::SERIALIZE,\n");
      }
      printer->Print("\"$full_name$\")", "full_name", field->full_name());
      if (for_parse) {
        printer->Print(")");
      }
      printer->Print(";\n");
      printer->Outdent();
      break;
    }
    case VERIFY: {
      printer->Print(
          "::google::protobuf::internal::WireFormat::$function$(\n",
          "function", verify_function);
      printer->Indent();
      printer->Print(variables, parameters);
      if (for_parse) {
        printer->Print("::google::protobuf::internal::WireFormat::PARSE,\n");
      } else {
        printer->Print("::google::protobuf::internal::WireFormat::SERIALIZE,\n");
      }
      printer->Print("\"$full_name$\");\n", "full_name", field->full_name());
      printer->Outdent();
      break;
    }
    case NONE:
      break;
  }
}

void GenerateUtf8CheckCodeForString(const FieldDescriptor* field,
                                    const Options& options, bool for_parse,
                                    const std::map<string, string>& variables,
                                    const char* parameters,
                                    io::Printer* printer) {
  GenerateUtf8CheckCode(field, options, for_parse, variables, parameters,
                        "VerifyUtf8String", "VerifyUTF8StringNamedField",
                        printer);
}

void GenerateUtf8CheckCodeForCord(const FieldDescriptor* field,
                                  const Options& options, bool for_parse,
                                  const std::map<string, string>& variables,
                                  const char* parameters,
                                  io::Printer* printer) {
  GenerateUtf8CheckCode(field, options, for_parse, variables, parameters,
                        "VerifyUtf8Cord", "VerifyUTF8CordNamedField", printer);
}

namespace {

void Flatten(const Descriptor* descriptor,
             std::vector<const Descriptor*>* flatten) {
  for (int i = 0; i < descriptor->nested_type_count(); i++)
    Flatten(descriptor->nested_type(i), flatten);
  flatten->push_back(descriptor);
}

}  // namespace

void FlattenMessagesInFile(const FileDescriptor* file,
                           std::vector<const Descriptor*>* result) {
  for (int i = 0; i < file->message_type_count(); i++) {
    Flatten(file->message_type(i), result);
  }
}

bool HasWeakFields(const Descriptor* descriptor) {
  return false;
}

bool HasWeakFields(const FileDescriptor* file) {
  return false;
}

bool UsingImplicitWeakFields(const FileDescriptor* file,
                             const Options& options) {
  return options.lite_implicit_weak_fields &&
         GetOptimizeFor(file, options) == FileOptions::LITE_RUNTIME;
}

bool IsImplicitWeakField(const FieldDescriptor* field, const Options& options,
                         SCCAnalyzer* scc_analyzer) {
  return UsingImplicitWeakFields(field->file(), options) &&
         field->type() == FieldDescriptor::TYPE_MESSAGE &&
         !field->is_required() && !field->is_map() &&
         field->containing_oneof() == NULL &&
         !IsWellKnownMessage(field->message_type()->file()) &&
         // We do not support implicit weak fields between messages in the same
         // strongly-connected component.
         scc_analyzer->GetSCC(field->containing_type()) !=
             scc_analyzer->GetSCC(field->message_type());
}

struct CompareDescriptors {
  bool operator()(const Descriptor* a, const Descriptor* b) {
    return a->full_name() < b->full_name();
  }
};

SCCAnalyzer::NodeData SCCAnalyzer::DFS(const Descriptor* descriptor) {
  // Must not have visited already.
  GOOGLE_DCHECK_EQ(cache_.count(descriptor), 0);

  // Mark visited by inserting in map.
  NodeData& result = cache_[descriptor];
  // Initialize data structures.
  result.index = result.lowlink = index_++;
  stack_.push_back(descriptor);

  // Recurse the fields / nodes in graph
  for (int i = 0; i < descriptor->field_count(); i++) {
    const Descriptor* child = descriptor->field(i)->message_type();
    if (child) {
      if (cache_.count(child) == 0) {
        // unexplored node
        NodeData child_data = DFS(child);
        result.lowlink = std::min(result.lowlink, child_data.lowlink);
      } else {
        NodeData child_data = cache_[child];
        if (child_data.scc == NULL) {
          // Still in the stack_ so we found a back edge
          result.lowlink = std::min(result.lowlink, child_data.index);
        }
      }
    }
  }
  if (result.index == result.lowlink) {
    // This is the root of a strongly connected component
    SCC* scc = CreateSCC();
    while (true) {
      const Descriptor* scc_desc = stack_.back();
      scc->descriptors.push_back(scc_desc);
      // Remove from stack
      stack_.pop_back();
      cache_[scc_desc].scc = scc;

      if (scc_desc == descriptor) break;
    }

    // The order of descriptors is random and depends how this SCC was
    // discovered. In-order to ensure maximum stability we sort it by name.
    std::sort(scc->descriptors.begin(), scc->descriptors.end(),
              CompareDescriptors());
    AddChildren(scc);
  }
  return result;
}

void SCCAnalyzer::AddChildren(SCC* scc) {
  std::set<const SCC*> seen;
  for (int i = 0; i < scc->descriptors.size(); i++) {
    const Descriptor* descriptor = scc->descriptors[i];
    for (int j = 0; j < descriptor->field_count(); j++) {
      const Descriptor* child_msg = descriptor->field(j)->message_type();
      if (child_msg) {
        const SCC* child = GetSCC(child_msg);
        if (child == scc) continue;
        if (seen.insert(child).second) {
          scc->children.push_back(child);
        }
      }
    }
  }
}

MessageAnalysis SCCAnalyzer::GetSCCAnalysis(const SCC* scc) {
  if (analysis_cache_.count(scc)) return analysis_cache_[scc];
  MessageAnalysis result = MessageAnalysis();
  for (int i = 0; i < scc->descriptors.size(); i++) {
    const Descriptor* descriptor = scc->descriptors[i];
    if (descriptor->extension_range_count() > 0) {
      result.contains_extension = true;
    }
    for (int i = 0; i < descriptor->field_count(); i++) {
      const FieldDescriptor* field = descriptor->field(i);
      if (field->is_required()) {
        result.contains_required = true;
      }
      switch (field->type()) {
        case FieldDescriptor::TYPE_STRING:
        case FieldDescriptor::TYPE_BYTES: {
          if (field->options().ctype() == FieldOptions::CORD) {
            result.contains_cord = true;
          }
          break;
        }
        case FieldDescriptor::TYPE_GROUP:
        case FieldDescriptor::TYPE_MESSAGE: {
          const SCC* child = GetSCC(field->message_type());
          if (child != scc) {
            MessageAnalysis analysis = GetSCCAnalysis(child);
            result.contains_cord |= analysis.contains_cord;
            result.contains_extension |= analysis.contains_extension;
            if (!ShouldIgnoreRequiredFieldCheck(field, options_)) {
              result.contains_required |= analysis.contains_required;
            }
          } else {
            // This field points back into the same SCC hence the messages
            // in the SCC are recursive. Note if SCC contains more than two
            // nodes it has to be recursive, however this test also works for
            // a single node that is recursive.
            result.is_recursive = true;
          }
          break;
        }
        default:
          break;
      }
    }
  }
  // We deliberately only insert the result here. After we contracted the SCC
  // in the graph, the graph should be a DAG. Hence we shouldn't need to mark
  // nodes visited as we can never return to them. By inserting them here
  // we will go in an infinite loop if the SCC is not correct.
  return analysis_cache_[scc] = result;
}

void ListAllFields(const Descriptor* d,
                   std::vector<const FieldDescriptor*>* fields) {
  // Collect sub messages
  for (int i = 0; i < d->nested_type_count(); i++) {
    ListAllFields(d->nested_type(i), fields);
  }
  // Collect message level extensions.
  for (int i = 0; i < d->extension_count(); i++) {
    fields->push_back(d->extension(i));
  }
  // Add types of fields necessary
  for (int i = 0; i < d->field_count(); i++) {
    fields->push_back(d->field(i));
  }
}

void ListAllFields(const FileDescriptor* d,
                   std::vector<const FieldDescriptor*>* fields) {
  // Collect file level message.
  for (int i = 0; i < d->message_type_count(); i++) {
    ListAllFields(d->message_type(i), fields);
  }
  // Collect message level extensions.
  for (int i = 0; i < d->extension_count(); i++) {
    fields->push_back(d->extension(i));
  }
}

void ListAllTypesForServices(const FileDescriptor* fd,
                             std::vector<const Descriptor*>* types) {
  for (int i = 0; i < fd->service_count(); i++) {
    const ServiceDescriptor* sd = fd->service(i);
    for (int j = 0; j < sd->method_count(); j++) {
      const MethodDescriptor* method = sd->method(j);
      types->push_back(method->input_type());
      types->push_back(method->output_type());
    }
  }
}


}  // namespace cpp
}  // namespace compiler
}  // namespace protobuf
}  // namespace google