1// Protocol Buffers - Google's data interchange format
2// Copyright 2008 Google Inc.  All rights reserved.
3// http://code.google.com/p/protobuf/
4//
5// Redistribution and use in source and binary forms, with or without
6// modification, are permitted provided that the following conditions are
7// met:
8//
9//     * Redistributions of source code must retain the above copyright
10// notice, this list of conditions and the following disclaimer.
11//     * Redistributions in binary form must reproduce the above
12// copyright notice, this list of conditions and the following disclaimer
13// in the documentation and/or other materials provided with the
14// distribution.
15//     * Neither the name of Google Inc. nor the names of its
16// contributors may be used to endorse or promote products derived from
17// this software without specific prior written permission.
18//
19// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
20// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
21// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
22// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
23// OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
24// SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
25// LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
26// DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
27// THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
28// (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
29// OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
30
31// Author: jschorr@google.com (Joseph Schorr)
32//  Based on original Protocol Buffers design by
33//  Sanjay Ghemawat, Jeff Dean, and others.
34
35#include <math.h>
36#include <stdlib.h>
37#include <limits>
38
39#include <google/protobuf/text_format.h>
40#include <google/protobuf/io/zero_copy_stream_impl.h>
41#include <google/protobuf/io/tokenizer.h>
42#include <google/protobuf/unittest.pb.h>
43#include <google/protobuf/unittest_mset.pb.h>
44#include <google/protobuf/test_util.h>
45
46#include <google/protobuf/stubs/common.h>
47#include <google/protobuf/testing/file.h>
48#include <google/protobuf/testing/googletest.h>
49#include <gtest/gtest.h>
50#include <google/protobuf/stubs/strutil.h>
51#include <google/protobuf/stubs/substitute.h>
52
53namespace google {
54namespace protobuf {
55
56// Can't use an anonymous namespace here due to brokenness of Tru64 compiler.
57namespace text_format_unittest {
58
59inline bool IsNaN(double value) {
60  // NaN is never equal to anything, even itself.
61  return value != value;
62}
63
64// A basic string with different escapable characters for testing.
65const string kEscapeTestString =
66  "\"A string with ' characters \n and \r newlines and \t tabs and \001 "
67  "slashes \\ and  multiple   spaces";
68
69// A representation of the above string with all the characters escaped.
70const string kEscapeTestStringEscaped =
71  "\"\\\"A string with \\' characters \\n and \\r newlines "
72  "and \\t tabs and \\001 slashes \\\\ and  multiple   spaces\"";
73
74class TextFormatTest : public testing::Test {
75 public:
76  static void SetUpTestCase() {
77    File::ReadFileToStringOrDie(
78        TestSourceDir()
79        + "/google/protobuf/testdata/text_format_unittest_data.txt",
80        &static_proto_debug_string_);
81  }
82
83  TextFormatTest() : proto_debug_string_(static_proto_debug_string_) {}
84
85 protected:
86  // Debug string read from text_format_unittest_data.txt.
87  const string proto_debug_string_;
88  unittest::TestAllTypes proto_;
89
90 private:
91  static string static_proto_debug_string_;
92};
93string TextFormatTest::static_proto_debug_string_;
94
95class TextFormatExtensionsTest : public testing::Test {
96 public:
97  static void SetUpTestCase() {
98    File::ReadFileToStringOrDie(
99        TestSourceDir()
100        + "/google/protobuf/testdata/"
101          "text_format_unittest_extensions_data.txt",
102        &static_proto_debug_string_);
103  }
104
105  TextFormatExtensionsTest()
106      : proto_debug_string_(static_proto_debug_string_) {}
107
108 protected:
109  // Debug string read from text_format_unittest_data.txt.
110  const string proto_debug_string_;
111  unittest::TestAllExtensions proto_;
112
113 private:
114  static string static_proto_debug_string_;
115};
116string TextFormatExtensionsTest::static_proto_debug_string_;
117
118
119TEST_F(TextFormatTest, Basic) {
120  TestUtil::SetAllFields(&proto_);
121  EXPECT_EQ(proto_debug_string_, proto_.DebugString());
122}
123
124TEST_F(TextFormatExtensionsTest, Extensions) {
125  TestUtil::SetAllExtensions(&proto_);
126  EXPECT_EQ(proto_debug_string_, proto_.DebugString());
127}
128
129TEST_F(TextFormatTest, ShortDebugString) {
130  proto_.set_optional_int32(1);
131  proto_.set_optional_string("hello");
132  proto_.mutable_optional_nested_message()->set_bb(2);
133  proto_.mutable_optional_foreign_message();
134
135  EXPECT_EQ("optional_int32: 1 optional_string: \"hello\" "
136            "optional_nested_message { bb: 2 } "
137            "optional_foreign_message { }",
138            proto_.ShortDebugString());
139}
140
141TEST_F(TextFormatTest, ShortPrimitiveRepeateds) {
142  proto_.set_optional_int32(123);
143  proto_.add_repeated_int32(456);
144  proto_.add_repeated_int32(789);
145  proto_.add_repeated_string("foo");
146  proto_.add_repeated_string("bar");
147  proto_.add_repeated_nested_message()->set_bb(2);
148  proto_.add_repeated_nested_message()->set_bb(3);
149  proto_.add_repeated_nested_enum(unittest::TestAllTypes::FOO);
150  proto_.add_repeated_nested_enum(unittest::TestAllTypes::BAR);
151
152  TextFormat::Printer printer;
153  printer.SetUseShortRepeatedPrimitives(true);
154  string text;
155  printer.PrintToString(proto_, &text);
156
157  EXPECT_EQ("optional_int32: 123\n"
158            "repeated_int32: [456, 789]\n"
159            "repeated_string: \"foo\"\n"
160            "repeated_string: \"bar\"\n"
161            "repeated_nested_message {\n  bb: 2\n}\n"
162            "repeated_nested_message {\n  bb: 3\n}\n"
163            "repeated_nested_enum: [FOO, BAR]\n",
164            text);
165
166  // Try in single-line mode.
167  printer.SetSingleLineMode(true);
168  printer.PrintToString(proto_, &text);
169
170  EXPECT_EQ("optional_int32: 123 "
171            "repeated_int32: [456, 789] "
172            "repeated_string: \"foo\" "
173            "repeated_string: \"bar\" "
174            "repeated_nested_message { bb: 2 } "
175            "repeated_nested_message { bb: 3 } "
176            "repeated_nested_enum: [FOO, BAR] ",
177            text);
178}
179
180
181TEST_F(TextFormatTest, StringEscape) {
182  // Set the string value to test.
183  proto_.set_optional_string(kEscapeTestString);
184
185  // Get the DebugString from the proto.
186  string debug_string = proto_.DebugString();
187  string utf8_debug_string = proto_.Utf8DebugString();
188
189  // Hardcode a correct value to test against.
190  string correct_string = "optional_string: "
191      + kEscapeTestStringEscaped
192       + "\n";
193
194  // Compare.
195  EXPECT_EQ(correct_string, debug_string);
196  // UTF-8 string is the same as non-UTF-8 because
197  // the protocol buffer contains no UTF-8 text.
198  EXPECT_EQ(correct_string, utf8_debug_string);
199
200  string expected_short_debug_string = "optional_string: "
201      + kEscapeTestStringEscaped;
202  EXPECT_EQ(expected_short_debug_string, proto_.ShortDebugString());
203}
204
205TEST_F(TextFormatTest, Utf8DebugString) {
206  // Set the string value to test.
207  proto_.set_optional_string("\350\260\267\346\255\214");
208
209  // Get the DebugString from the proto.
210  string debug_string = proto_.DebugString();
211  string utf8_debug_string = proto_.Utf8DebugString();
212
213  // Hardcode a correct value to test against.
214  string correct_utf8_string = "optional_string: "
215      "\"\350\260\267\346\255\214\""
216      "\n";
217  string correct_string = "optional_string: "
218      "\"\\350\\260\\267\\346\\255\\214\""
219      "\n";
220
221  // Compare.
222  EXPECT_EQ(correct_utf8_string, utf8_debug_string);
223  EXPECT_EQ(correct_string, debug_string);
224}
225
226TEST_F(TextFormatTest, PrintUnknownFields) {
227  // Test printing of unknown fields in a message.
228
229  unittest::TestEmptyMessage message;
230  UnknownFieldSet* unknown_fields = message.mutable_unknown_fields();
231
232  unknown_fields->AddVarint(5, 1);
233  unknown_fields->AddFixed32(5, 2);
234  unknown_fields->AddFixed64(5, 3);
235  unknown_fields->AddLengthDelimited(5, "4");
236  unknown_fields->AddGroup(5)->AddVarint(10, 5);
237
238  unknown_fields->AddVarint(8, 1);
239  unknown_fields->AddVarint(8, 2);
240  unknown_fields->AddVarint(8, 3);
241
242  EXPECT_EQ(
243    "5: 1\n"
244    "5: 0x00000002\n"
245    "5: 0x0000000000000003\n"
246    "5: \"4\"\n"
247    "5 {\n"
248    "  10: 5\n"
249    "}\n"
250    "8: 1\n"
251    "8: 2\n"
252    "8: 3\n",
253    message.DebugString());
254}
255
256TEST_F(TextFormatTest, PrintUnknownMessage) {
257  // Test heuristic printing of messages in an UnknownFieldSet.
258
259  protobuf_unittest::TestAllTypes message;
260
261  // Cases which should not be interpreted as sub-messages.
262
263  // 'a' is a valid FIXED64 tag, so for the string to be parseable as a message
264  // it should be followed by 8 bytes.  Since this string only has two
265  // subsequent bytes, it should be treated as a string.
266  message.add_repeated_string("abc");
267
268  // 'd' happens to be a valid ENDGROUP tag.  So,
269  // UnknownFieldSet::MergeFromCodedStream() will successfully parse "def", but
270  // the ConsumedEntireMessage() check should fail.
271  message.add_repeated_string("def");
272
273  // A zero-length string should never be interpreted as a message even though
274  // it is technically valid as one.
275  message.add_repeated_string("");
276
277  // Case which should be interpreted as a sub-message.
278
279  // An actual nested message with content should always be interpreted as a
280  // nested message.
281  message.add_repeated_nested_message()->set_bb(123);
282
283  string data;
284  message.SerializeToString(&data);
285
286  string text;
287  UnknownFieldSet unknown_fields;
288  EXPECT_TRUE(unknown_fields.ParseFromString(data));
289  EXPECT_TRUE(TextFormat::PrintUnknownFieldsToString(unknown_fields, &text));
290  EXPECT_EQ(
291    "44: \"abc\"\n"
292    "44: \"def\"\n"
293    "44: \"\"\n"
294    "48 {\n"
295    "  1: 123\n"
296    "}\n",
297    text);
298}
299
300TEST_F(TextFormatTest, PrintMessageWithIndent) {
301  // Test adding an initial indent to printing.
302
303  protobuf_unittest::TestAllTypes message;
304
305  message.add_repeated_string("abc");
306  message.add_repeated_string("def");
307  message.add_repeated_nested_message()->set_bb(123);
308
309  string text;
310  TextFormat::Printer printer;
311  printer.SetInitialIndentLevel(1);
312  EXPECT_TRUE(printer.PrintToString(message, &text));
313  EXPECT_EQ(
314    "  repeated_string: \"abc\"\n"
315    "  repeated_string: \"def\"\n"
316    "  repeated_nested_message {\n"
317    "    bb: 123\n"
318    "  }\n",
319    text);
320}
321
322TEST_F(TextFormatTest, PrintMessageSingleLine) {
323  // Test printing a message on a single line.
324
325  protobuf_unittest::TestAllTypes message;
326
327  message.add_repeated_string("abc");
328  message.add_repeated_string("def");
329  message.add_repeated_nested_message()->set_bb(123);
330
331  string text;
332  TextFormat::Printer printer;
333  printer.SetInitialIndentLevel(1);
334  printer.SetSingleLineMode(true);
335  EXPECT_TRUE(printer.PrintToString(message, &text));
336  EXPECT_EQ(
337    "  repeated_string: \"abc\" repeated_string: \"def\" "
338    "repeated_nested_message { bb: 123 } ",
339    text);
340}
341
342TEST_F(TextFormatTest, ParseBasic) {
343  io::ArrayInputStream input_stream(proto_debug_string_.data(),
344                                    proto_debug_string_.size());
345  TextFormat::Parse(&input_stream, &proto_);
346  TestUtil::ExpectAllFieldsSet(proto_);
347}
348
349TEST_F(TextFormatExtensionsTest, ParseExtensions) {
350  io::ArrayInputStream input_stream(proto_debug_string_.data(),
351                                    proto_debug_string_.size());
352  TextFormat::Parse(&input_stream, &proto_);
353  TestUtil::ExpectAllExtensionsSet(proto_);
354}
355
356TEST_F(TextFormatTest, ParseStringEscape) {
357  // Create a parse string with escpaed characters in it.
358  string parse_string = "optional_string: "
359      + kEscapeTestStringEscaped
360      + "\n";
361
362  io::ArrayInputStream input_stream(parse_string.data(),
363                                    parse_string.size());
364  TextFormat::Parse(&input_stream, &proto_);
365
366  // Compare.
367  EXPECT_EQ(kEscapeTestString, proto_.optional_string());
368}
369
370TEST_F(TextFormatTest, ParseConcatenatedString) {
371  // Create a parse string with multiple parts on one line.
372  string parse_string = "optional_string: \"foo\" \"bar\"\n";
373
374  io::ArrayInputStream input_stream1(parse_string.data(),
375                                    parse_string.size());
376  TextFormat::Parse(&input_stream1, &proto_);
377
378  // Compare.
379  EXPECT_EQ("foobar", proto_.optional_string());
380
381  // Create a parse string with multiple parts on seperate lines.
382  parse_string = "optional_string: \"foo\"\n"
383                 "\"bar\"\n";
384
385  io::ArrayInputStream input_stream2(parse_string.data(),
386                                    parse_string.size());
387  TextFormat::Parse(&input_stream2, &proto_);
388
389  // Compare.
390  EXPECT_EQ("foobar", proto_.optional_string());
391}
392
393TEST_F(TextFormatTest, ParseFloatWithSuffix) {
394  // Test that we can parse a floating-point value with 'f' appended to the
395  // end.  This is needed for backwards-compatibility with proto1.
396
397  // Have it parse a float with the 'f' suffix.
398  string parse_string = "optional_float: 1.0f\n";
399
400  io::ArrayInputStream input_stream(parse_string.data(),
401                                    parse_string.size());
402
403  TextFormat::Parse(&input_stream, &proto_);
404
405  // Compare.
406  EXPECT_EQ(1.0, proto_.optional_float());
407}
408
409TEST_F(TextFormatTest, Comments) {
410  // Test that comments are ignored.
411
412  string parse_string = "optional_int32: 1  # a comment\n"
413                        "optional_int64: 2  # another comment";
414
415  io::ArrayInputStream input_stream(parse_string.data(),
416                                    parse_string.size());
417
418  TextFormat::Parse(&input_stream, &proto_);
419
420  // Compare.
421  EXPECT_EQ(1, proto_.optional_int32());
422  EXPECT_EQ(2, proto_.optional_int64());
423}
424
425TEST_F(TextFormatTest, OptionalColon) {
426  // Test that we can place a ':' after the field name of a nested message,
427  // even though we don't have to.
428
429  string parse_string = "optional_nested_message: { bb: 1}\n";
430
431  io::ArrayInputStream input_stream(parse_string.data(),
432                                    parse_string.size());
433
434  TextFormat::Parse(&input_stream, &proto_);
435
436  // Compare.
437  EXPECT_TRUE(proto_.has_optional_nested_message());
438  EXPECT_EQ(1, proto_.optional_nested_message().bb());
439}
440
441// Some platforms (e.g. Windows) insist on padding the exponent to three
442// digits when one or two would be just fine.
443static string RemoveRedundantZeros(string text) {
444  text = StringReplace(text, "e+0", "e+", true);
445  text = StringReplace(text, "e-0", "e-", true);
446  return text;
447}
448
449TEST_F(TextFormatTest, PrintExotic) {
450  unittest::TestAllTypes message;
451
452  // Note:  In C, a negative integer literal is actually the unary negation
453  //   operator being applied to a positive integer literal, and
454  //   9223372036854775808 is outside the range of int64.  However, it is not
455  //   outside the range of uint64.  Confusingly, this means that everything
456  //   works if we make the literal unsigned, even though we are negating it.
457  message.add_repeated_int64(-GOOGLE_ULONGLONG(9223372036854775808));
458  message.add_repeated_uint64(GOOGLE_ULONGLONG(18446744073709551615));
459  message.add_repeated_double(123.456);
460  message.add_repeated_double(1.23e21);
461  message.add_repeated_double(1.23e-18);
462  message.add_repeated_double(std::numeric_limits<double>::infinity());
463  message.add_repeated_double(-std::numeric_limits<double>::infinity());
464  message.add_repeated_double(std::numeric_limits<double>::quiet_NaN());
465  message.add_repeated_string(string("\000\001\a\b\f\n\r\t\v\\\'\"", 12));
466
467  // Fun story:  We used to use 1.23e22 instead of 1.23e21 above, but this
468  //   seemed to trigger an odd case on MinGW/GCC 3.4.5 where GCC's parsing of
469  //   the value differed from strtod()'s parsing.  That is to say, the
470  //   following assertion fails on MinGW:
471  //     assert(1.23e22 == strtod("1.23e22", NULL));
472  //   As a result, SimpleDtoa() would print the value as
473  //   "1.2300000000000001e+22" to make sure strtod() produce the exact same
474  //   result.  Our goal is to test runtime parsing, not compile-time parsing,
475  //   so this wasn't our problem.  It was found that using 1.23e21 did not
476  //   have this problem, so we switched to that instead.
477
478  EXPECT_EQ(
479    "repeated_int64: -9223372036854775808\n"
480    "repeated_uint64: 18446744073709551615\n"
481    "repeated_double: 123.456\n"
482    "repeated_double: 1.23e+21\n"
483    "repeated_double: 1.23e-18\n"
484    "repeated_double: inf\n"
485    "repeated_double: -inf\n"
486    "repeated_double: nan\n"
487    "repeated_string: \"\\000\\001\\007\\010\\014\\n\\r\\t\\013\\\\\\'\\\"\"\n",
488    RemoveRedundantZeros(message.DebugString()));
489}
490
491TEST_F(TextFormatTest, PrintFloatPrecision) {
492  unittest::TestAllTypes message;
493
494  message.add_repeated_float(1.2);
495  message.add_repeated_float(1.23);
496  message.add_repeated_float(1.234);
497  message.add_repeated_float(1.2345);
498  message.add_repeated_float(1.23456);
499  message.add_repeated_float(1.2e10);
500  message.add_repeated_float(1.23e10);
501  message.add_repeated_float(1.234e10);
502  message.add_repeated_float(1.2345e10);
503  message.add_repeated_float(1.23456e10);
504  message.add_repeated_double(1.2);
505  message.add_repeated_double(1.23);
506  message.add_repeated_double(1.234);
507  message.add_repeated_double(1.2345);
508  message.add_repeated_double(1.23456);
509  message.add_repeated_double(1.234567);
510  message.add_repeated_double(1.2345678);
511  message.add_repeated_double(1.23456789);
512  message.add_repeated_double(1.234567898);
513  message.add_repeated_double(1.2345678987);
514  message.add_repeated_double(1.23456789876);
515  message.add_repeated_double(1.234567898765);
516  message.add_repeated_double(1.2345678987654);
517  message.add_repeated_double(1.23456789876543);
518  message.add_repeated_double(1.2e100);
519  message.add_repeated_double(1.23e100);
520  message.add_repeated_double(1.234e100);
521  message.add_repeated_double(1.2345e100);
522  message.add_repeated_double(1.23456e100);
523  message.add_repeated_double(1.234567e100);
524  message.add_repeated_double(1.2345678e100);
525  message.add_repeated_double(1.23456789e100);
526  message.add_repeated_double(1.234567898e100);
527  message.add_repeated_double(1.2345678987e100);
528  message.add_repeated_double(1.23456789876e100);
529  message.add_repeated_double(1.234567898765e100);
530  message.add_repeated_double(1.2345678987654e100);
531  message.add_repeated_double(1.23456789876543e100);
532
533  EXPECT_EQ(
534    "repeated_float: 1.2\n"
535    "repeated_float: 1.23\n"
536    "repeated_float: 1.234\n"
537    "repeated_float: 1.2345\n"
538    "repeated_float: 1.23456\n"
539    "repeated_float: 1.2e+10\n"
540    "repeated_float: 1.23e+10\n"
541    "repeated_float: 1.234e+10\n"
542    "repeated_float: 1.2345e+10\n"
543    "repeated_float: 1.23456e+10\n"
544    "repeated_double: 1.2\n"
545    "repeated_double: 1.23\n"
546    "repeated_double: 1.234\n"
547    "repeated_double: 1.2345\n"
548    "repeated_double: 1.23456\n"
549    "repeated_double: 1.234567\n"
550    "repeated_double: 1.2345678\n"
551    "repeated_double: 1.23456789\n"
552    "repeated_double: 1.234567898\n"
553    "repeated_double: 1.2345678987\n"
554    "repeated_double: 1.23456789876\n"
555    "repeated_double: 1.234567898765\n"
556    "repeated_double: 1.2345678987654\n"
557    "repeated_double: 1.23456789876543\n"
558    "repeated_double: 1.2e+100\n"
559    "repeated_double: 1.23e+100\n"
560    "repeated_double: 1.234e+100\n"
561    "repeated_double: 1.2345e+100\n"
562    "repeated_double: 1.23456e+100\n"
563    "repeated_double: 1.234567e+100\n"
564    "repeated_double: 1.2345678e+100\n"
565    "repeated_double: 1.23456789e+100\n"
566    "repeated_double: 1.234567898e+100\n"
567    "repeated_double: 1.2345678987e+100\n"
568    "repeated_double: 1.23456789876e+100\n"
569    "repeated_double: 1.234567898765e+100\n"
570    "repeated_double: 1.2345678987654e+100\n"
571    "repeated_double: 1.23456789876543e+100\n",
572    RemoveRedundantZeros(message.DebugString()));
573}
574
575
576TEST_F(TextFormatTest, AllowPartial) {
577  unittest::TestRequired message;
578  TextFormat::Parser parser;
579  parser.AllowPartialMessage(true);
580  EXPECT_TRUE(parser.ParseFromString("a: 1", &message));
581  EXPECT_EQ(1, message.a());
582  EXPECT_FALSE(message.has_b());
583  EXPECT_FALSE(message.has_c());
584}
585
586TEST_F(TextFormatTest, ParseExotic) {
587  unittest::TestAllTypes message;
588  ASSERT_TRUE(TextFormat::ParseFromString(
589    "repeated_int32: -1\n"
590    "repeated_int32: -2147483648\n"
591    "repeated_int64: -1\n"
592    "repeated_int64: -9223372036854775808\n"
593    "repeated_uint32: 4294967295\n"
594    "repeated_uint32: 2147483648\n"
595    "repeated_uint64: 18446744073709551615\n"
596    "repeated_uint64: 9223372036854775808\n"
597    "repeated_double: 123.0\n"
598    "repeated_double: 123.5\n"
599    "repeated_double: 0.125\n"
600    "repeated_double: 1.23E17\n"
601    "repeated_double: 1.235E+22\n"
602    "repeated_double: 1.235e-18\n"
603    "repeated_double: 123.456789\n"
604    "repeated_double: inf\n"
605    "repeated_double: Infinity\n"
606    "repeated_double: -inf\n"
607    "repeated_double: -Infinity\n"
608    "repeated_double: nan\n"
609    "repeated_double: NaN\n"
610    "repeated_string: \"\\000\\001\\a\\b\\f\\n\\r\\t\\v\\\\\\'\\\"\"\n",
611    &message));
612
613  ASSERT_EQ(2, message.repeated_int32_size());
614  EXPECT_EQ(-1, message.repeated_int32(0));
615  // Note:  In C, a negative integer literal is actually the unary negation
616  //   operator being applied to a positive integer literal, and 2147483648 is
617  //   outside the range of int32.  However, it is not outside the range of
618  //   uint32.  Confusingly, this means that everything works if we make the
619  //   literal unsigned, even though we are negating it.
620  EXPECT_EQ(-2147483648u, message.repeated_int32(1));
621
622  ASSERT_EQ(2, message.repeated_int64_size());
623  EXPECT_EQ(-1, message.repeated_int64(0));
624  // Note:  In C, a negative integer literal is actually the unary negation
625  //   operator being applied to a positive integer literal, and
626  //   9223372036854775808 is outside the range of int64.  However, it is not
627  //   outside the range of uint64.  Confusingly, this means that everything
628  //   works if we make the literal unsigned, even though we are negating it.
629  EXPECT_EQ(-GOOGLE_ULONGLONG(9223372036854775808), message.repeated_int64(1));
630
631  ASSERT_EQ(2, message.repeated_uint32_size());
632  EXPECT_EQ(4294967295u, message.repeated_uint32(0));
633  EXPECT_EQ(2147483648u, message.repeated_uint32(1));
634
635  ASSERT_EQ(2, message.repeated_uint64_size());
636  EXPECT_EQ(GOOGLE_ULONGLONG(18446744073709551615), message.repeated_uint64(0));
637  EXPECT_EQ(GOOGLE_ULONGLONG(9223372036854775808), message.repeated_uint64(1));
638
639  ASSERT_EQ(13, message.repeated_double_size());
640  EXPECT_EQ(123.0     , message.repeated_double(0));
641  EXPECT_EQ(123.5     , message.repeated_double(1));
642  EXPECT_EQ(0.125     , message.repeated_double(2));
643  EXPECT_EQ(1.23E17   , message.repeated_double(3));
644  EXPECT_EQ(1.235E22  , message.repeated_double(4));
645  EXPECT_EQ(1.235E-18 , message.repeated_double(5));
646  EXPECT_EQ(123.456789, message.repeated_double(6));
647  EXPECT_EQ(message.repeated_double(7), numeric_limits<double>::infinity());
648  EXPECT_EQ(message.repeated_double(8), numeric_limits<double>::infinity());
649  EXPECT_EQ(message.repeated_double(9), -numeric_limits<double>::infinity());
650  EXPECT_EQ(message.repeated_double(10), -numeric_limits<double>::infinity());
651  EXPECT_TRUE(IsNaN(message.repeated_double(11)));
652  EXPECT_TRUE(IsNaN(message.repeated_double(12)));
653
654  // Note:  Since these string literals have \0's in them, we must explicitly
655  //   pass their sizes to string's constructor.
656  ASSERT_EQ(1, message.repeated_string_size());
657  EXPECT_EQ(string("\000\001\a\b\f\n\r\t\v\\\'\"", 12),
658            message.repeated_string(0));
659}
660
661class TextFormatParserTest : public testing::Test {
662 protected:
663  void ExpectFailure(const string& input, const string& message, int line,
664                     int col) {
665    scoped_ptr<unittest::TestAllTypes> proto(new unittest::TestAllTypes);
666    ExpectFailure(input, message, line, col, proto.get());
667  }
668
669  void ExpectFailure(const string& input, const string& message, int line,
670                     int col, Message* proto) {
671    ExpectMessage(input, message, line, col, proto, false);
672  }
673
674  void ExpectMessage(const string& input, const string& message, int line,
675                     int col, Message* proto, bool expected_result) {
676    TextFormat::Parser parser;
677    MockErrorCollector error_collector;
678    parser.RecordErrorsTo(&error_collector);
679    EXPECT_EQ(parser.ParseFromString(input, proto), expected_result);
680    EXPECT_EQ(SimpleItoa(line) + ":" + SimpleItoa(col) + ": " + message + "\n",
681              error_collector.text_);
682  }
683
684  // An error collector which simply concatenates all its errors into a big
685  // block of text which can be checked.
686  class MockErrorCollector : public io::ErrorCollector {
687   public:
688    MockErrorCollector() {}
689    ~MockErrorCollector() {}
690
691    string text_;
692
693    // implements ErrorCollector -------------------------------------
694    void AddError(int line, int column, const string& message) {
695      strings::SubstituteAndAppend(&text_, "$0:$1: $2\n",
696                                   line + 1, column + 1, message);
697    }
698
699    void AddWarning(int line, int column, const string& message) {
700      AddError(line, column, "WARNING:" + message);
701    }
702  };
703};
704
705TEST_F(TextFormatParserTest, ParseFieldValueFromString) {
706  scoped_ptr<unittest::TestAllTypes> message(new unittest::TestAllTypes);
707  const Descriptor* d = message->GetDescriptor();
708
709#define EXPECT_FIELD(name, value, valuestring) \
710  EXPECT_TRUE(TextFormat::ParseFieldValueFromString( \
711    valuestring, d->FindFieldByName("optional_" #name), message.get())); \
712  EXPECT_EQ(value, message->optional_##name()); \
713  EXPECT_TRUE(message->has_optional_##name());
714
715#define EXPECT_FLOAT_FIELD(name, value, valuestring) \
716  EXPECT_TRUE(TextFormat::ParseFieldValueFromString( \
717    valuestring, d->FindFieldByName("optional_" #name), message.get())); \
718  EXPECT_FLOAT_EQ(value, message->optional_##name()); \
719  EXPECT_TRUE(message->has_optional_##name());
720
721#define EXPECT_DOUBLE_FIELD(name, value, valuestring) \
722  EXPECT_TRUE(TextFormat::ParseFieldValueFromString( \
723    valuestring, d->FindFieldByName("optional_" #name), message.get())); \
724  EXPECT_DOUBLE_EQ(value, message->optional_##name()); \
725  EXPECT_TRUE(message->has_optional_##name());
726
727#define EXPECT_INVALID(name, valuestring) \
728  EXPECT_FALSE(TextFormat::ParseFieldValueFromString( \
729    valuestring, d->FindFieldByName("optional_" #name), message.get()));
730
731  // int32
732  EXPECT_FIELD(int32, 1, "1");
733  EXPECT_FIELD(int32, -1, "-1");
734  EXPECT_FIELD(int32, 0x1234, "0x1234");
735  EXPECT_INVALID(int32, "a");
736  EXPECT_INVALID(int32, "999999999999999999999999999999999999");
737  EXPECT_INVALID(int32, "1,2");
738
739  // int64
740  EXPECT_FIELD(int64, 1, "1");
741  EXPECT_FIELD(int64, -1, "-1");
742  EXPECT_FIELD(int64, 0x1234567812345678LL, "0x1234567812345678");
743  EXPECT_INVALID(int64, "a");
744  EXPECT_INVALID(int64, "999999999999999999999999999999999999");
745  EXPECT_INVALID(int64, "1,2");
746
747  // uint64
748  EXPECT_FIELD(uint64, 1, "1");
749  EXPECT_FIELD(uint64, 0xf234567812345678ULL, "0xf234567812345678");
750  EXPECT_INVALID(uint64, "-1");
751  EXPECT_INVALID(uint64, "a");
752  EXPECT_INVALID(uint64, "999999999999999999999999999999999999");
753  EXPECT_INVALID(uint64, "1,2");
754
755  // fixed32
756  EXPECT_FIELD(fixed32, 1, "1");
757  EXPECT_FIELD(fixed32, 0x12345678, "0x12345678");
758  EXPECT_INVALID(fixed32, "-1");
759  EXPECT_INVALID(fixed32, "a");
760  EXPECT_INVALID(fixed32, "999999999999999999999999999999999999");
761  EXPECT_INVALID(fixed32, "1,2");
762
763  // fixed64
764  EXPECT_FIELD(fixed64, 1, "1");
765  EXPECT_FIELD(fixed64, 0x1234567812345678ULL, "0x1234567812345678");
766  EXPECT_INVALID(fixed64, "-1");
767  EXPECT_INVALID(fixed64, "a");
768  EXPECT_INVALID(fixed64, "999999999999999999999999999999999999");
769  EXPECT_INVALID(fixed64, "1,2");
770
771  // bool
772  EXPECT_FIELD(bool, true, "true");
773  EXPECT_FIELD(bool, false, "false");
774  EXPECT_INVALID(bool, "1");
775  EXPECT_INVALID(bool, "on");
776  EXPECT_INVALID(bool, "a");
777  EXPECT_INVALID(bool, "True");
778
779  // float
780  EXPECT_FIELD(float, 1, "1");
781  EXPECT_FLOAT_FIELD(float, 1.5, "1.5");
782  EXPECT_FLOAT_FIELD(float, 1.5e3, "1.5e3");
783  EXPECT_FLOAT_FIELD(float, -4.55, "-4.55");
784  EXPECT_INVALID(float, "a");
785  EXPECT_INVALID(float, "1,2");
786
787  // double
788  EXPECT_FIELD(double, 1, "1");
789  EXPECT_FIELD(double, -1, "-1");
790  EXPECT_DOUBLE_FIELD(double, 2.3, "2.3");
791  EXPECT_DOUBLE_FIELD(double, 3e5, "3e5");
792  EXPECT_INVALID(double, "a");
793  EXPECT_INVALID(double, "1,2");
794
795  // string
796  EXPECT_FIELD(string, "hello", "\"hello\"");
797  EXPECT_FIELD(string, "-1.87", "'-1.87'");
798  EXPECT_INVALID(string, "hello");  // without quote for value
799
800  // enum
801  EXPECT_FIELD(nested_enum, unittest::TestAllTypes::BAR, "BAR");
802  EXPECT_INVALID(nested_enum, "1");  // number not supported
803  EXPECT_INVALID(nested_enum, "FOOBAR");
804
805  // message
806  EXPECT_TRUE(TextFormat::ParseFieldValueFromString(
807    "<bb:12>", d->FindFieldByName("optional_nested_message"), message.get()));
808  EXPECT_EQ(12, message->optional_nested_message().bb()); \
809  EXPECT_TRUE(message->has_optional_nested_message());
810  EXPECT_INVALID(nested_message, "any");
811
812#undef EXPECT_FIELD
813#undef EXPECT_FLOAT_FIELD
814#undef EXPECT_DOUBLE_FIELD
815#undef EXPECT_INVALID
816}
817
818
819TEST_F(TextFormatParserTest, InvalidToken) {
820  ExpectFailure("optional_bool: true\n-5\n", "Expected identifier.",
821                2, 1);
822
823  ExpectFailure("optional_bool: true;\n", "Expected identifier.", 1, 20);
824  ExpectFailure("\"some string\"", "Expected identifier.", 1, 1);
825}
826
827TEST_F(TextFormatParserTest, InvalidFieldName) {
828  ExpectFailure(
829      "invalid_field: somevalue\n",
830      "Message type \"protobuf_unittest.TestAllTypes\" has no field named "
831      "\"invalid_field\".",
832      1, 14);
833}
834
835TEST_F(TextFormatParserTest, InvalidCapitalization) {
836  // We require that group names be exactly as they appear in the .proto.
837  ExpectFailure(
838      "optionalgroup {\na: 15\n}\n",
839      "Message type \"protobuf_unittest.TestAllTypes\" has no field named "
840      "\"optionalgroup\".",
841      1, 15);
842  ExpectFailure(
843      "OPTIONALgroup {\na: 15\n}\n",
844      "Message type \"protobuf_unittest.TestAllTypes\" has no field named "
845      "\"OPTIONALgroup\".",
846      1, 15);
847  ExpectFailure(
848      "Optional_Double: 10.0\n",
849      "Message type \"protobuf_unittest.TestAllTypes\" has no field named "
850      "\"Optional_Double\".",
851      1, 16);
852}
853
854TEST_F(TextFormatParserTest, InvalidFieldValues) {
855  // Invalid values for a double/float field.
856  ExpectFailure("optional_double: \"hello\"\n", "Expected double.", 1, 18);
857  ExpectFailure("optional_double: true\n", "Expected double.", 1, 18);
858  ExpectFailure("optional_double: !\n", "Expected double.", 1, 18);
859  ExpectFailure("optional_double {\n  \n}\n", "Expected \":\", found \"{\".",
860                1, 17);
861
862  // Invalid values for a signed integer field.
863  ExpectFailure("optional_int32: \"hello\"\n", "Expected integer.", 1, 17);
864  ExpectFailure("optional_int32: true\n", "Expected integer.", 1, 17);
865  ExpectFailure("optional_int32: 4.5\n", "Expected integer.", 1, 17);
866  ExpectFailure("optional_int32: !\n", "Expected integer.", 1, 17);
867  ExpectFailure("optional_int32 {\n \n}\n", "Expected \":\", found \"{\".",
868                1, 16);
869  ExpectFailure("optional_int32: 0x80000000\n",
870                "Integer out of range.", 1, 17);
871  ExpectFailure("optional_int32: -0x80000001\n",
872                "Integer out of range.", 1, 18);
873  ExpectFailure("optional_int64: 0x8000000000000000\n",
874                "Integer out of range.", 1, 17);
875  ExpectFailure("optional_int64: -0x8000000000000001\n",
876                "Integer out of range.", 1, 18);
877
878  // Invalid values for an unsigned integer field.
879  ExpectFailure("optional_uint64: \"hello\"\n", "Expected integer.", 1, 18);
880  ExpectFailure("optional_uint64: true\n", "Expected integer.", 1, 18);
881  ExpectFailure("optional_uint64: 4.5\n", "Expected integer.", 1, 18);
882  ExpectFailure("optional_uint64: -5\n", "Expected integer.", 1, 18);
883  ExpectFailure("optional_uint64: !\n", "Expected integer.", 1, 18);
884  ExpectFailure("optional_uint64 {\n \n}\n", "Expected \":\", found \"{\".",
885                1, 17);
886  ExpectFailure("optional_uint32: 0x100000000\n",
887                "Integer out of range.", 1, 18);
888  ExpectFailure("optional_uint64: 0x10000000000000000\n",
889                "Integer out of range.", 1, 18);
890
891  // Invalid values for a boolean field.
892  ExpectFailure("optional_bool: \"hello\"\n", "Expected identifier.", 1, 16);
893  ExpectFailure("optional_bool: 5\n", "Expected identifier.", 1, 16);
894  ExpectFailure("optional_bool: -7.5\n", "Expected identifier.", 1, 16);
895  ExpectFailure("optional_bool: !\n", "Expected identifier.", 1, 16);
896
897  ExpectFailure(
898      "optional_bool: meh\n",
899      "Invalid value for boolean field \"optional_bool\". Value: \"meh\".",
900      2, 1);
901
902  ExpectFailure("optional_bool {\n \n}\n", "Expected \":\", found \"{\".",
903                1, 15);
904
905  // Invalid values for a string field.
906  ExpectFailure("optional_string: true\n", "Expected string.", 1, 18);
907  ExpectFailure("optional_string: 5\n", "Expected string.", 1, 18);
908  ExpectFailure("optional_string: -7.5\n", "Expected string.", 1, 18);
909  ExpectFailure("optional_string: !\n", "Expected string.", 1, 18);
910  ExpectFailure("optional_string {\n \n}\n", "Expected \":\", found \"{\".",
911                1, 17);
912
913  // Invalid values for an enumeration field.
914  ExpectFailure("optional_nested_enum: \"hello\"\n", "Expected identifier.",
915                1, 23);
916
917  ExpectFailure("optional_nested_enum: 5\n", "Expected identifier.", 1, 23);
918  ExpectFailure("optional_nested_enum: -7.5\n", "Expected identifier.", 1, 23);
919  ExpectFailure("optional_nested_enum: !\n", "Expected identifier.", 1, 23);
920
921  ExpectFailure(
922      "optional_nested_enum: grah\n",
923      "Unknown enumeration value of \"grah\" for field "
924      "\"optional_nested_enum\".", 2, 1);
925
926  ExpectFailure(
927      "optional_nested_enum {\n \n}\n",
928      "Expected \":\", found \"{\".", 1, 22);
929}
930
931TEST_F(TextFormatParserTest, MessageDelimeters) {
932  // Non-matching delimeters.
933  ExpectFailure("OptionalGroup <\n \n}\n", "Expected \">\", found \"}\".",
934                3, 1);
935
936  // Invalid delimeters.
937  ExpectFailure("OptionalGroup [\n \n]\n", "Expected \"{\", found \"[\".",
938                1, 15);
939
940  // Unending message.
941  ExpectFailure("optional_nested_message {\n \nbb: 118\n",
942                "Expected identifier.",
943                4, 1);
944}
945
946TEST_F(TextFormatParserTest, UnknownExtension) {
947  // Non-matching delimeters.
948  ExpectFailure("[blahblah]: 123",
949                "Extension \"blahblah\" is not defined or is not an "
950                "extension of \"protobuf_unittest.TestAllTypes\".",
951                1, 11);
952}
953
954TEST_F(TextFormatParserTest, MissingRequired) {
955  unittest::TestRequired message;
956  ExpectFailure("a: 1",
957                "Message missing required fields: b, c",
958                0, 1, &message);
959}
960
961TEST_F(TextFormatParserTest, ParseDuplicateRequired) {
962  unittest::TestRequired message;
963  ExpectFailure("a: 1 b: 2 c: 3 a: 1",
964                "Non-repeated field \"a\" is specified multiple times.",
965                1, 17, &message);
966}
967
968TEST_F(TextFormatParserTest, ParseDuplicateOptional) {
969  unittest::ForeignMessage message;
970  ExpectFailure("c: 1 c: 2",
971                "Non-repeated field \"c\" is specified multiple times.",
972                1, 7, &message);
973}
974
975TEST_F(TextFormatParserTest, MergeDuplicateRequired) {
976  unittest::TestRequired message;
977  TextFormat::Parser parser;
978  EXPECT_TRUE(parser.MergeFromString("a: 1 b: 2 c: 3 a: 4", &message));
979  EXPECT_EQ(4, message.a());
980}
981
982TEST_F(TextFormatParserTest, MergeDuplicateOptional) {
983  unittest::ForeignMessage message;
984  TextFormat::Parser parser;
985  EXPECT_TRUE(parser.MergeFromString("c: 1 c: 2", &message));
986  EXPECT_EQ(2, message.c());
987}
988
989TEST_F(TextFormatParserTest, PrintErrorsToStderr) {
990  vector<string> errors;
991
992  {
993    ScopedMemoryLog log;
994    unittest::TestAllTypes proto;
995    EXPECT_FALSE(TextFormat::ParseFromString("no_such_field: 1", &proto));
996    errors = log.GetMessages(ERROR);
997  }
998
999  ASSERT_EQ(1, errors.size());
1000  EXPECT_EQ("Error parsing text-format protobuf_unittest.TestAllTypes: "
1001            "1:14: Message type \"protobuf_unittest.TestAllTypes\" has no field "
1002            "named \"no_such_field\".",
1003            errors[0]);
1004}
1005
1006TEST_F(TextFormatParserTest, FailsOnTokenizationError) {
1007  vector<string> errors;
1008
1009  {
1010    ScopedMemoryLog log;
1011    unittest::TestAllTypes proto;
1012    EXPECT_FALSE(TextFormat::ParseFromString("\020", &proto));
1013    errors = log.GetMessages(ERROR);
1014  }
1015
1016  ASSERT_EQ(1, errors.size());
1017  EXPECT_EQ("Error parsing text-format protobuf_unittest.TestAllTypes: "
1018            "1:1: Invalid control characters encountered in text.",
1019            errors[0]);
1020}
1021
1022TEST_F(TextFormatParserTest, ParseDeprecatedField) {
1023  unittest::TestDeprecatedFields message;
1024  ExpectMessage("deprecated_int32: 42",
1025                "WARNING:text format contains deprecated field "
1026                "\"deprecated_int32\"", 1, 21, &message, true);
1027}
1028
1029class TextFormatMessageSetTest : public testing::Test {
1030 protected:
1031  static const char proto_debug_string_[];
1032};
1033const char TextFormatMessageSetTest::proto_debug_string_[] =
1034"message_set {\n"
1035"  [protobuf_unittest.TestMessageSetExtension1] {\n"
1036"    i: 23\n"
1037"  }\n"
1038"  [protobuf_unittest.TestMessageSetExtension2] {\n"
1039"    str: \"foo\"\n"
1040"  }\n"
1041"}\n";
1042
1043
1044TEST_F(TextFormatMessageSetTest, Serialize) {
1045  protobuf_unittest::TestMessageSetContainer proto;
1046  protobuf_unittest::TestMessageSetExtension1* item_a =
1047    proto.mutable_message_set()->MutableExtension(
1048      protobuf_unittest::TestMessageSetExtension1::message_set_extension);
1049  item_a->set_i(23);
1050  protobuf_unittest::TestMessageSetExtension2* item_b =
1051    proto.mutable_message_set()->MutableExtension(
1052      protobuf_unittest::TestMessageSetExtension2::message_set_extension);
1053  item_b->set_str("foo");
1054  EXPECT_EQ(proto_debug_string_, proto.DebugString());
1055}
1056
1057TEST_F(TextFormatMessageSetTest, Deserialize) {
1058  protobuf_unittest::TestMessageSetContainer proto;
1059  ASSERT_TRUE(TextFormat::ParseFromString(proto_debug_string_, &proto));
1060  EXPECT_EQ(23, proto.message_set().GetExtension(
1061    protobuf_unittest::TestMessageSetExtension1::message_set_extension).i());
1062  EXPECT_EQ("foo", proto.message_set().GetExtension(
1063    protobuf_unittest::TestMessageSetExtension2::message_set_extension).str());
1064
1065  // Ensure that these are the only entries present.
1066  vector<const FieldDescriptor*> descriptors;
1067  proto.message_set().GetReflection()->ListFields(
1068    proto.message_set(), &descriptors);
1069  EXPECT_EQ(2, descriptors.size());
1070}
1071
1072}  // namespace text_format_unittest
1073}  // namespace protobuf
1074}  // namespace google
1075