diff options
Diffstat (limited to 'src/s3select/rapidjson/test/unittest/readertest.cpp')
-rw-r--r-- | src/s3select/rapidjson/test/unittest/readertest.cpp | 2370 |
1 files changed, 2370 insertions, 0 deletions
diff --git a/src/s3select/rapidjson/test/unittest/readertest.cpp b/src/s3select/rapidjson/test/unittest/readertest.cpp new file mode 100644 index 000000000..f828dbbe2 --- /dev/null +++ b/src/s3select/rapidjson/test/unittest/readertest.cpp @@ -0,0 +1,2370 @@ +// Tencent is pleased to support the open source community by making RapidJSON available. +// +// Copyright (C) 2015 THL A29 Limited, a Tencent company, and Milo Yip. +// +// Licensed under the MIT License (the "License"); you may not use this file except +// in compliance with the License. You may obtain a copy of the License at +// +// http://opensource.org/licenses/MIT +// +// Unless required by applicable law or agreed to in writing, software distributed +// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR +// CONDITIONS OF ANY KIND, either express or implied. See the License for the +// specific language governing permissions and limitations under the License. + +#include "unittest.h" + +#include "rapidjson/reader.h" +#include "rapidjson/internal/dtoa.h" +#include "rapidjson/internal/itoa.h" +#include "rapidjson/memorystream.h" + +#include <limits> + +using namespace rapidjson; + +RAPIDJSON_DIAG_PUSH +#ifdef __GNUC__ +RAPIDJSON_DIAG_OFF(effc++) +RAPIDJSON_DIAG_OFF(float-equal) +RAPIDJSON_DIAG_OFF(missing-noreturn) +#if __GNUC__ >= 7 +RAPIDJSON_DIAG_OFF(dangling-else) +#endif +#endif // __GNUC__ + +#ifdef __clang__ +RAPIDJSON_DIAG_OFF(variadic-macros) +RAPIDJSON_DIAG_OFF(c++98-compat-pedantic) +#endif + +template<bool expect> +struct ParseBoolHandler : BaseReaderHandler<UTF8<>, ParseBoolHandler<expect> > { + ParseBoolHandler() : step_(0) {} + bool Default() { ADD_FAILURE(); return false; } + // gcc 4.8.x generates warning in EXPECT_EQ(bool, bool) on this gtest version. + // Workaround with EXPECT_TRUE(). + bool Bool(bool b) { /*EXPECT_EQ(expect, b); */EXPECT_TRUE(expect == b); ++step_; return true; } + + unsigned step_; +}; + +TEST(Reader, ParseTrue) { + StringStream s("true"); + ParseBoolHandler<true> h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(1u, h.step_); +} + +TEST(Reader, ParseFalse) { + StringStream s("false"); + ParseBoolHandler<false> h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(1u, h.step_); +} + +struct ParseIntHandler : BaseReaderHandler<UTF8<>, ParseIntHandler> { + ParseIntHandler() : step_(0), actual_() {} + bool Default() { ADD_FAILURE(); return false; } + bool Int(int i) { actual_ = i; step_++; return true; } + + unsigned step_; + int actual_; +}; + +struct ParseUintHandler : BaseReaderHandler<UTF8<>, ParseUintHandler> { + ParseUintHandler() : step_(0), actual_() {} + bool Default() { ADD_FAILURE(); return false; } + bool Uint(unsigned i) { actual_ = i; step_++; return true; } + + unsigned step_; + unsigned actual_; +}; + +struct ParseInt64Handler : BaseReaderHandler<UTF8<>, ParseInt64Handler> { + ParseInt64Handler() : step_(0), actual_() {} + bool Default() { ADD_FAILURE(); return false; } + bool Int64(int64_t i) { actual_ = i; step_++; return true; } + + unsigned step_; + int64_t actual_; +}; + +struct ParseUint64Handler : BaseReaderHandler<UTF8<>, ParseUint64Handler> { + ParseUint64Handler() : step_(0), actual_() {} + bool Default() { ADD_FAILURE(); return false; } + bool Uint64(uint64_t i) { actual_ = i; step_++; return true; } + + unsigned step_; + uint64_t actual_; +}; + +struct ParseDoubleHandler : BaseReaderHandler<UTF8<>, ParseDoubleHandler> { + ParseDoubleHandler() : step_(0), actual_() {} + bool Default() { ADD_FAILURE(); return false; } + bool Double(double d) { actual_ = d; step_++; return true; } + + unsigned step_; + double actual_; +}; + +TEST(Reader, ParseNumber_Integer) { +#define TEST_INTEGER(Handler, str, x) \ + { \ + StringStream s(str); \ + Handler h; \ + Reader reader; \ + reader.Parse(s, h); \ + EXPECT_EQ(1u, h.step_); \ + EXPECT_EQ(x, h.actual_); \ + } + + TEST_INTEGER(ParseUintHandler, "0", 0u); + TEST_INTEGER(ParseUintHandler, "123", 123u); + TEST_INTEGER(ParseUintHandler, "2147483648", 2147483648u); // 2^31 - 1 (cannot be stored in int) + TEST_INTEGER(ParseUintHandler, "4294967295", 4294967295u); + + TEST_INTEGER(ParseIntHandler, "-123", -123); + TEST_INTEGER(ParseIntHandler, "-2147483648", static_cast<int32_t>(0x80000000)); // -2^31 (min of int) + + TEST_INTEGER(ParseUint64Handler, "4294967296", RAPIDJSON_UINT64_C2(1, 0)); // 2^32 (max of unsigned + 1, force to use uint64_t) + TEST_INTEGER(ParseUint64Handler, "18446744073709551615", RAPIDJSON_UINT64_C2(0xFFFFFFFF, 0xFFFFFFFF)); // 2^64 - 1 (max of uint64_t) + + TEST_INTEGER(ParseInt64Handler, "-2147483649", static_cast<int64_t>(RAPIDJSON_UINT64_C2(0xFFFFFFFF, 0x7FFFFFFF))); // -2^31 -1 (min of int - 1, force to use int64_t) + TEST_INTEGER(ParseInt64Handler, "-9223372036854775808", static_cast<int64_t>(RAPIDJSON_UINT64_C2(0x80000000, 0x00000000))); // -2^63 (min of int64_t) + + // Random test for uint32_t/int32_t + { + union { + uint32_t u; + int32_t i; + }u; + Random r; + + for (unsigned i = 0; i < 100000; i++) { + u.u = r(); + + char buffer[32]; + *internal::u32toa(u.u, buffer) = '\0'; + TEST_INTEGER(ParseUintHandler, buffer, u.u); + + if (u.i < 0) { + *internal::i32toa(u.i, buffer) = '\0'; + TEST_INTEGER(ParseIntHandler, buffer, u.i); + } + } + } + + // Random test for uint64_t/int64_t + { + union { + uint64_t u; + int64_t i; + }u; + Random r; + + for (unsigned i = 0; i < 100000; i++) { + u.u = uint64_t(r()) << 32; + u.u |= r(); + + char buffer[32]; + if (u.u > uint64_t(4294967295u)) { + *internal::u64toa(u.u, buffer) = '\0'; + TEST_INTEGER(ParseUint64Handler, buffer, u.u); + } + + if (u.i < -int64_t(2147483648u)) { + *internal::i64toa(u.i, buffer) = '\0'; + TEST_INTEGER(ParseInt64Handler, buffer, u.i); + } + } + } +#undef TEST_INTEGER +} + +template<bool fullPrecision> +static void TestParseDouble() { +#define TEST_DOUBLE(fullPrecision, str, x) \ + { \ + StringStream s(str); \ + ParseDoubleHandler h; \ + Reader reader; \ + ASSERT_EQ(kParseErrorNone, reader.Parse<fullPrecision ? kParseFullPrecisionFlag : 0>(s, h).Code()); \ + EXPECT_EQ(1u, h.step_); \ + internal::Double e(x), a(h.actual_); \ + if (fullPrecision) { \ + EXPECT_EQ(e.Uint64Value(), a.Uint64Value()); \ + if (e.Uint64Value() != a.Uint64Value()) \ + printf(" String: %s\n Actual: %.17g\nExpected: %.17g\n", str, h.actual_, x); \ + } \ + else { \ + EXPECT_EQ(e.Sign(), a.Sign()); /* for 0.0 != -0.0 */ \ + EXPECT_DOUBLE_EQ(x, h.actual_); \ + } \ + } + + TEST_DOUBLE(fullPrecision, "0.0", 0.0); + TEST_DOUBLE(fullPrecision, "-0.0", -0.0); // For checking issue #289 + TEST_DOUBLE(fullPrecision, "0e100", 0.0); // For checking issue #1249 + TEST_DOUBLE(fullPrecision, "1.0", 1.0); + TEST_DOUBLE(fullPrecision, "-1.0", -1.0); + TEST_DOUBLE(fullPrecision, "1.5", 1.5); + TEST_DOUBLE(fullPrecision, "-1.5", -1.5); + TEST_DOUBLE(fullPrecision, "3.1416", 3.1416); + TEST_DOUBLE(fullPrecision, "1E10", 1E10); + TEST_DOUBLE(fullPrecision, "1e10", 1e10); + TEST_DOUBLE(fullPrecision, "1E+10", 1E+10); + TEST_DOUBLE(fullPrecision, "1E-10", 1E-10); + TEST_DOUBLE(fullPrecision, "-1E10", -1E10); + TEST_DOUBLE(fullPrecision, "-1e10", -1e10); + TEST_DOUBLE(fullPrecision, "-1E+10", -1E+10); + TEST_DOUBLE(fullPrecision, "-1E-10", -1E-10); + TEST_DOUBLE(fullPrecision, "1.234E+10", 1.234E+10); + TEST_DOUBLE(fullPrecision, "1.234E-10", 1.234E-10); + TEST_DOUBLE(fullPrecision, "1.79769e+308", 1.79769e+308); + TEST_DOUBLE(fullPrecision, "2.22507e-308", 2.22507e-308); + TEST_DOUBLE(fullPrecision, "-1.79769e+308", -1.79769e+308); + TEST_DOUBLE(fullPrecision, "-2.22507e-308", -2.22507e-308); + TEST_DOUBLE(fullPrecision, "4.9406564584124654e-324", 4.9406564584124654e-324); // minimum denormal + TEST_DOUBLE(fullPrecision, "2.2250738585072009e-308", 2.2250738585072009e-308); // Max subnormal double + TEST_DOUBLE(fullPrecision, "2.2250738585072014e-308", 2.2250738585072014e-308); // Min normal positive double + TEST_DOUBLE(fullPrecision, "1.7976931348623157e+308", 1.7976931348623157e+308); // Max double + TEST_DOUBLE(fullPrecision, "1e-10000", 0.0); // must underflow + TEST_DOUBLE(fullPrecision, "18446744073709551616", 18446744073709551616.0); // 2^64 (max of uint64_t + 1, force to use double) + TEST_DOUBLE(fullPrecision, "-9223372036854775809", -9223372036854775809.0); // -2^63 - 1(min of int64_t + 1, force to use double) + TEST_DOUBLE(fullPrecision, "0.9868011474609375", 0.9868011474609375); // https://github.com/Tencent/rapidjson/issues/120 + TEST_DOUBLE(fullPrecision, "123e34", 123e34); // Fast Path Cases In Disguise + TEST_DOUBLE(fullPrecision, "45913141877270640000.0", 45913141877270640000.0); + TEST_DOUBLE(fullPrecision, "2.2250738585072011e-308", 2.2250738585072011e-308); // http://www.exploringbinary.com/php-hangs-on-numeric-value-2-2250738585072011e-308/ + TEST_DOUBLE(fullPrecision, "1e-00011111111111", 0.0); // Issue #313 + TEST_DOUBLE(fullPrecision, "-1e-00011111111111", -0.0); + TEST_DOUBLE(fullPrecision, "1e-214748363", 0.0); // Maximum supported negative exponent + TEST_DOUBLE(fullPrecision, "1e-214748364", 0.0); + TEST_DOUBLE(fullPrecision, "1e-21474836311", 0.0); + TEST_DOUBLE(fullPrecision, "1.00000000001e-2147483638", 0.0); + TEST_DOUBLE(fullPrecision, "0.017976931348623157e+310", 1.7976931348623157e+308); // Max double in another form + TEST_DOUBLE(fullPrecision, "128.74836467836484838364836483643636483648e-336", 0.0); // Issue #1251 + + // Since + // abs((2^-1022 - 2^-1074) - 2.2250738585072012e-308) = 3.109754131239141401123495768877590405345064751974375599... x 10^-324 + // abs((2^-1022) - 2.2250738585072012e-308) = 1.830902327173324040642192159804623318305533274168872044... x 10 ^ -324 + // So 2.2250738585072012e-308 should round to 2^-1022 = 2.2250738585072014e-308 + TEST_DOUBLE(fullPrecision, "2.2250738585072012e-308", 2.2250738585072014e-308); // http://www.exploringbinary.com/java-hangs-when-converting-2-2250738585072012e-308/ + + // More closer to normal/subnormal boundary + // boundary = 2^-1022 - 2^-1075 = 2.225073858507201136057409796709131975934819546351645648... x 10^-308 + TEST_DOUBLE(fullPrecision, "2.22507385850720113605740979670913197593481954635164564e-308", 2.2250738585072009e-308); + TEST_DOUBLE(fullPrecision, "2.22507385850720113605740979670913197593481954635164565e-308", 2.2250738585072014e-308); + + // 1.0 is in (1.0 - 2^-54, 1.0 + 2^-53) + // 1.0 - 2^-54 = 0.999999999999999944488848768742172978818416595458984375 + TEST_DOUBLE(fullPrecision, "0.999999999999999944488848768742172978818416595458984375", 1.0); // round to even + TEST_DOUBLE(fullPrecision, "0.999999999999999944488848768742172978818416595458984374", 0.99999999999999989); // previous double + TEST_DOUBLE(fullPrecision, "0.999999999999999944488848768742172978818416595458984376", 1.0); // next double + // 1.0 + 2^-53 = 1.00000000000000011102230246251565404236316680908203125 + TEST_DOUBLE(fullPrecision, "1.00000000000000011102230246251565404236316680908203125", 1.0); // round to even + TEST_DOUBLE(fullPrecision, "1.00000000000000011102230246251565404236316680908203124", 1.0); // previous double + TEST_DOUBLE(fullPrecision, "1.00000000000000011102230246251565404236316680908203126", 1.00000000000000022); // next double + + // Numbers from https://github.com/floitsch/double-conversion/blob/master/test/cctest/test-strtod.cc + + TEST_DOUBLE(fullPrecision, "72057594037927928.0", 72057594037927928.0); + TEST_DOUBLE(fullPrecision, "72057594037927936.0", 72057594037927936.0); + TEST_DOUBLE(fullPrecision, "72057594037927932.0", 72057594037927936.0); + TEST_DOUBLE(fullPrecision, "7205759403792793199999e-5", 72057594037927928.0); + TEST_DOUBLE(fullPrecision, "7205759403792793200001e-5", 72057594037927936.0); + + TEST_DOUBLE(fullPrecision, "9223372036854774784.0", 9223372036854774784.0); + TEST_DOUBLE(fullPrecision, "9223372036854775808.0", 9223372036854775808.0); + TEST_DOUBLE(fullPrecision, "9223372036854775296.0", 9223372036854775808.0); + TEST_DOUBLE(fullPrecision, "922337203685477529599999e-5", 9223372036854774784.0); + TEST_DOUBLE(fullPrecision, "922337203685477529600001e-5", 9223372036854775808.0); + + TEST_DOUBLE(fullPrecision, "10141204801825834086073718800384", 10141204801825834086073718800384.0); + TEST_DOUBLE(fullPrecision, "10141204801825835211973625643008", 10141204801825835211973625643008.0); + TEST_DOUBLE(fullPrecision, "10141204801825834649023672221696", 10141204801825835211973625643008.0); + TEST_DOUBLE(fullPrecision, "1014120480182583464902367222169599999e-5", 10141204801825834086073718800384.0); + TEST_DOUBLE(fullPrecision, "1014120480182583464902367222169600001e-5", 10141204801825835211973625643008.0); + + TEST_DOUBLE(fullPrecision, "5708990770823838890407843763683279797179383808", 5708990770823838890407843763683279797179383808.0); + TEST_DOUBLE(fullPrecision, "5708990770823839524233143877797980545530986496", 5708990770823839524233143877797980545530986496.0); + TEST_DOUBLE(fullPrecision, "5708990770823839207320493820740630171355185152", 5708990770823839524233143877797980545530986496.0); + TEST_DOUBLE(fullPrecision, "5708990770823839207320493820740630171355185151999e-3", 5708990770823838890407843763683279797179383808.0); + TEST_DOUBLE(fullPrecision, "5708990770823839207320493820740630171355185152001e-3", 5708990770823839524233143877797980545530986496.0); + + { + char n1e308[310]; // '1' followed by 308 '0' + n1e308[0] = '1'; + for (int i = 1; i < 309; i++) + n1e308[i] = '0'; + n1e308[309] = '\0'; + TEST_DOUBLE(fullPrecision, n1e308, 1E308); + } + + // Cover trimming + TEST_DOUBLE(fullPrecision, +"2.22507385850720113605740979670913197593481954635164564802342610972482222202107694551652952390813508" +"7914149158913039621106870086438694594645527657207407820621743379988141063267329253552286881372149012" +"9811224514518898490572223072852551331557550159143974763979834118019993239625482890171070818506906306" +"6665599493827577257201576306269066333264756530000924588831643303777979186961204949739037782970490505" +"1080609940730262937128958950003583799967207254304360284078895771796150945516748243471030702609144621" +"5722898802581825451803257070188608721131280795122334262883686223215037756666225039825343359745688844" +"2390026549819838548794829220689472168983109969836584681402285424333066033985088644580400103493397042" +"7567186443383770486037861622771738545623065874679014086723327636718751234567890123456789012345678901" +"e-308", + 2.2250738585072014e-308); + + { + static const unsigned count = 100; // Tested with 1000000 locally + Random r; + Reader reader; // Reusing reader to prevent heap allocation + + // Exhaustively test different exponents with random significant + for (uint64_t exp = 0; exp < 2047; exp++) { + ; + for (unsigned i = 0; i < count; i++) { + // Need to call r() in two statements for cross-platform coherent sequence. + uint64_t u = (exp << 52) | uint64_t(r() & 0x000FFFFF) << 32; + u |= uint64_t(r()); + internal::Double d = internal::Double(u); + + char buffer[32]; + *internal::dtoa(d.Value(), buffer) = '\0'; + + StringStream s(buffer); + ParseDoubleHandler h; + ASSERT_EQ(kParseErrorNone, reader.Parse<fullPrecision ? kParseFullPrecisionFlag : 0>(s, h).Code()); + EXPECT_EQ(1u, h.step_); + internal::Double a(h.actual_); + if (fullPrecision) { + EXPECT_EQ(d.Uint64Value(), a.Uint64Value()); + if (d.Uint64Value() != a.Uint64Value()) + printf(" String: %s\n Actual: %.17g\nExpected: %.17g\n", buffer, h.actual_, d.Value()); + } + else { + EXPECT_EQ(d.Sign(), a.Sign()); // for 0.0 != -0.0 + EXPECT_DOUBLE_EQ(d.Value(), h.actual_); + } + } + } + } + + // Issue #340 + TEST_DOUBLE(fullPrecision, "7.450580596923828e-9", 7.450580596923828e-9); + { + internal::Double d(1.0); + for (int i = 0; i < 324; i++) { + char buffer[32]; + *internal::dtoa(d.Value(), buffer) = '\0'; + + StringStream s(buffer); + ParseDoubleHandler h; + Reader reader; + ASSERT_EQ(kParseErrorNone, reader.Parse<fullPrecision ? kParseFullPrecisionFlag : 0>(s, h).Code()); + EXPECT_EQ(1u, h.step_); + internal::Double a(h.actual_); + if (fullPrecision) { + EXPECT_EQ(d.Uint64Value(), a.Uint64Value()); + if (d.Uint64Value() != a.Uint64Value()) + printf(" String: %s\n Actual: %.17g\nExpected: %.17g\n", buffer, h.actual_, d.Value()); + } + else { + EXPECT_EQ(d.Sign(), a.Sign()); // for 0.0 != -0.0 + EXPECT_DOUBLE_EQ(d.Value(), h.actual_); + } + + + d = d.Value() * 0.5; + } + } + + // Issue 1249 + TEST_DOUBLE(fullPrecision, "0e100", 0.0); + + // Issue 1251 + TEST_DOUBLE(fullPrecision, "128.74836467836484838364836483643636483648e-336", 0.0); + + // Issue 1256 + TEST_DOUBLE(fullPrecision, + "6223372036854775296.1701512723685473547372536854755293372036854685477" + "529752233737201701512337200972013723685473123372036872036854236854737" + "247372368372367752975258547752975254729752547372368737201701512354737" + "83723677529752585477247372368372368547354737253685475529752", + 6223372036854775808.0); + +#if 0 + // Test (length + exponent) overflow + TEST_DOUBLE(fullPrecision, "0e+2147483647", 0.0); + TEST_DOUBLE(fullPrecision, "0e-2147483648", 0.0); + TEST_DOUBLE(fullPrecision, "1e-2147483648", 0.0); + TEST_DOUBLE(fullPrecision, "0e+9223372036854775807", 0.0); + TEST_DOUBLE(fullPrecision, "0e-9223372036854775808", 0.0); +#endif + + if (fullPrecision) + { + TEST_DOUBLE(fullPrecision, "1e-325", 0.0); + TEST_DOUBLE(fullPrecision, "1e-324", 0.0); + TEST_DOUBLE(fullPrecision, "2e-324", 0.0); + TEST_DOUBLE(fullPrecision, "2.4703282292062327e-324", 0.0); + TEST_DOUBLE(fullPrecision, "2.4703282292062328e-324", 5e-324); + TEST_DOUBLE(fullPrecision, "2.48e-324",5e-324); + TEST_DOUBLE(fullPrecision, "2.5e-324", 5e-324); + + // Slightly above max-normal + TEST_DOUBLE(fullPrecision, "1.7976931348623158e+308", 1.7976931348623158e+308); + + TEST_DOUBLE(fullPrecision, + "17976931348623157081452742373170435679807056752584499659891747680315726" + "07800285387605895586327668781715404589535143824642343213268894641827684" + "67546703537516986049910576551282076245490090389328944075868508455133942" + "30458323690322294816580855933212334827479782620414472316873817718091929" + "9881250404026184124858368", + (std::numeric_limits<double>::max)()); + + TEST_DOUBLE(fullPrecision, + "243546080556034731077856379609316893158278902575447060151047" + "212703405344938119816206067372775299130836050315842578309818" + "316450894337978612745889730079163798234256495613858256849283" + "467066859489192118352020514036083287319232435355752493038825" + "828481044358810649108367633313557305310641892225870327827273" + "41408256.000000", + 2.4354608055603473e+307); + // 9007199254740991 * 2^971 (max normal) + TEST_DOUBLE(fullPrecision, + "1.797693134862315708145274237317043567980705675258449965989174768031572607800285" + "38760589558632766878171540458953514382464234321326889464182768467546703537516986" + "04991057655128207624549009038932894407586850845513394230458323690322294816580855" + "9332123348274797826204144723168738177180919299881250404026184124858368e+308", + 1.797693134862315708e+308 // 0x1.fffffffffffffp1023 + ); +#if 0 + // TODO: + // Should work at least in full-precision mode... + TEST_DOUBLE(fullPrecision, + "0.00000000000000000000000000000000000000000000000000000000000" + "0000000000000000000000000000000000000000000000000000000000000" + "0000000000000000000000000000000000000000000000000000000000000" + "0000000000000000000000000000000000000000000000000000000000000" + "0000000000000000000000000000000000000000000000000000000000000" + "0000000000000000000024703282292062327208828439643411068618252" + "9901307162382212792841250337753635104375932649918180817996189" + "8982823477228588654633283551779698981993873980053909390631503" + "5659515570226392290858392449105184435931802849936536152500319" + "3704576782492193656236698636584807570015857692699037063119282" + "7955855133292783433840935197801553124659726357957462276646527" + "2827220056374006485499977096599470454020828166226237857393450" + "7363390079677619305775067401763246736009689513405355374585166" + "6113422376667860416215968046191446729184030053005753084904876" + "5391711386591646239524912623653881879636239373280423891018672" + "3484976682350898633885879256283027559956575244555072551893136" + "9083625477918694866799496832404970582102851318545139621383772" + "2826145437693412532098591327667236328125", + 0.0); +#endif + // 9007199254740991 * 2^-1074 = (2^53 - 1) * 2^-1074 + TEST_DOUBLE(fullPrecision, + "4.450147717014402272114819593418263951869639092703291296046852219449644444042153" + "89103305904781627017582829831782607924221374017287738918929105531441481564124348" + "67599762821265346585071045737627442980259622449029037796981144446145705102663115" + "10031828794952795966823603998647925096578034214163701381261333311989876551545144" + "03152612538132666529513060001849177663286607555958373922409899478075565940981010" + "21612198814605258742579179000071675999344145086087205681577915435923018910334964" + "86942061405218289243144579760516365090360651414037721744226256159024466852576737" + "24464300755133324500796506867194913776884780053099639677097589658441378944337966" + "21993967316936280457084866613206797017728916080020698679408551343728867675409720" + "757232455434770912461317493580281734466552734375e-308", + 4.450147717014402272e-308 // 0x1.fffffffffffffp-1022 + ); + // 9007199254740990 * 2^-1074 + TEST_DOUBLE(fullPrecision, + "4.450147717014401778049173752171719775300846224481918930987049605124880018456471" + "39035755177760751831052846195619008686241717547743167145836439860405887584484471" + "19639655002484083577939142623582164522087943959208000909794783876158397872163051" + "22622675229968408654350206725478309956546318828765627255022767720818849892988457" + "26333908582101604036318532842699932130356061901518261174396928478121372742040102" + "17446565569357687263889031732270082446958029584739170416643195242132750803227473" + "16608838720742955671061336566907126801014814608027120593609275183716632624844904" + "31985250929886016737037234388448352929102742708402644340627409931664203093081360" + "70794835812045179006047003875039546061891526346421705014598610179523165038319441" + "51446491086954182492263498716056346893310546875e-308", + 4.450147717014401778e-308 // 0x1.ffffffffffffep-1022 + ); + // half way between the two numbers above. + // round to nearest even. + TEST_DOUBLE(fullPrecision, + "4.450147717014402025081996672794991863585242658592605113516950912287262231249312" + "64069530541271189424317838013700808305231545782515453032382772695923684574304409" + "93619708911874715081505094180604803751173783204118519353387964161152051487413083" + "16327252012460602310586905362063117526562176521464664318142050516404363222266800" + "64743260560117135282915796422274554896821334728738317548403413978098469341510556" + "19529382191981473003234105366170879223151087335413188049110555339027884856781219" + "01775450062980622457102958163711745945687733011032421168917765671370549738710820" + "78224775842509670618916870627821633352993761380751142008862499795052791018709663" + "46394401564490729731565935244123171539810221213221201847003580761626016356864581" + "1358486831521563686919762403704226016998291015625e-308", + 4.450147717014401778e-308 // 0x1.ffffffffffffep-1022 + ); + TEST_DOUBLE(fullPrecision, + "4.450147717014402025081996672794991863585242658592605113516950912287262231249312" + "64069530541271189424317838013700808305231545782515453032382772695923684574304409" + "93619708911874715081505094180604803751173783204118519353387964161152051487413083" + "16327252012460602310586905362063117526562176521464664318142050516404363222266800" + "64743260560117135282915796422274554896821334728738317548403413978098469341510556" + "19529382191981473003234105366170879223151087335413188049110555339027884856781219" + "01775450062980622457102958163711745945687733011032421168917765671370549738710820" + "78224775842509670618916870627821633352993761380751142008862499795052791018709663" + "46394401564490729731565935244123171539810221213221201847003580761626016356864581" + "13584868315215636869197624037042260169982910156250000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000e-308", + 4.450147717014401778e-308 // 0x1.ffffffffffffep-1022 + ); +#if 0 + // ... round up + // TODO: + // Should work at least in full-precision mode... + TEST_DOUBLE(fullPrecision, + "4.450147717014402025081996672794991863585242658592605113516950912287262231249312" + "64069530541271189424317838013700808305231545782515453032382772695923684574304409" + "93619708911874715081505094180604803751173783204118519353387964161152051487413083" + "16327252012460602310586905362063117526562176521464664318142050516404363222266800" + "64743260560117135282915796422274554896821334728738317548403413978098469341510556" + "19529382191981473003234105366170879223151087335413188049110555339027884856781219" + "01775450062980622457102958163711745945687733011032421168917765671370549738710820" + "78224775842509670618916870627821633352993761380751142008862499795052791018709663" + "46394401564490729731565935244123171539810221213221201847003580761626016356864581" + "13584868315215636869197624037042260169982910156250000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000001e-308", + 4.450147717014402272e-308 // 0x1.fffffffffffffp-1022 + ); +#endif + // ... round down + TEST_DOUBLE(fullPrecision, + "4.450147717014402025081996672794991863585242658592605113516950912287262231249312" + "64069530541271189424317838013700808305231545782515453032382772695923684574304409" + "93619708911874715081505094180604803751173783204118519353387964161152051487413083" + "16327252012460602310586905362063117526562176521464664318142050516404363222266800" + "64743260560117135282915796422274554896821334728738317548403413978098469341510556" + "19529382191981473003234105366170879223151087335413188049110555339027884856781219" + "01775450062980622457102958163711745945687733011032421168917765671370549738710820" + "78224775842509670618916870627821633352993761380751142008862499795052791018709663" + "46394401564490729731565935244123171539810221213221201847003580761626016356864581" + "13584868315215636869197624037042260169982910156249999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999e-308", + 4.450147717014401778e-308 // 0x1.ffffffffffffep-1022 + ); + // Slightly below half way between max-normal and infinity. + // Should round down. + TEST_DOUBLE(fullPrecision, + "1.797693134862315807937289714053034150799341327100378269361737789804449682927647" + "50946649017977587207096330286416692887910946555547851940402630657488671505820681" + "90890200070838367627385484581771153176447573027006985557136695962284291481986083" + "49364752927190741684443655107043427115596995080930428801779041744977919999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999" + "99999999999999999999999999999999999999999999999999999999999999999999999999999999e+308", + 1.797693134862315708e+308 // 0x1.fffffffffffffp1023 + ); + } + +#undef TEST_DOUBLE +} + +TEST(Reader, ParseNumber_NormalPrecisionDouble) { + TestParseDouble<false>(); +} + +TEST(Reader, ParseNumber_FullPrecisionDouble) { + TestParseDouble<true>(); +} + +TEST(Reader, ParseNumber_NormalPrecisionError) { + static unsigned count = 1000000; + Random r; + + double ulpSum = 0.0; + double ulpMax = 0.0; + for (unsigned i = 0; i < count; i++) { + internal::Double e, a; + do { + // Need to call r() in two statements for cross-platform coherent sequence. + uint64_t u = uint64_t(r()) << 32; + u |= uint64_t(r()); + e = u; + } while (e.IsNan() || e.IsInf() || !e.IsNormal()); + + char buffer[32]; + *internal::dtoa(e.Value(), buffer) = '\0'; + + StringStream s(buffer); + ParseDoubleHandler h; + Reader reader; + ASSERT_EQ(kParseErrorNone, reader.Parse(s, h).Code()); + EXPECT_EQ(1u, h.step_); + + a = h.actual_; + uint64_t bias1 = e.ToBias(); + uint64_t bias2 = a.ToBias(); + double ulp = static_cast<double>(bias1 >= bias2 ? bias1 - bias2 : bias2 - bias1); + ulpMax = (std::max)(ulpMax, ulp); + ulpSum += ulp; + } + printf("ULP Average = %g, Max = %g \n", ulpSum / count, ulpMax); +} + +template<bool fullPrecision> +static void TestParseNumberError() { +#define TEST_NUMBER_ERROR(errorCode, str, errorOffset, streamPos) \ + { \ + char buffer[2048]; \ + ASSERT_LT(std::strlen(str), 2048u); \ + sprintf(buffer, "%s", str); \ + InsituStringStream s(buffer); \ + BaseReaderHandler<> h; \ + Reader reader; \ + EXPECT_FALSE(reader.Parse<fullPrecision ? kParseFullPrecisionFlag : 0>(s, h)); \ + EXPECT_EQ(errorCode, reader.GetParseErrorCode());\ + EXPECT_EQ(errorOffset, reader.GetErrorOffset());\ + EXPECT_EQ(streamPos, s.Tell());\ + } + + // Number too big to be stored in double. + { + char n1e309[311]; // '1' followed by 309 '0' + n1e309[0] = '1'; + for (int i = 1; i < 310; i++) + n1e309[i] = '0'; + n1e309[310] = '\0'; + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, n1e309, 0u, 310u); + } + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e309", 0u, 5u); + + // Miss fraction part in number. + TEST_NUMBER_ERROR(kParseErrorNumberMissFraction, "1.", 2u, 2u); + TEST_NUMBER_ERROR(kParseErrorNumberMissFraction, "1.a", 2u, 2u); + + // Miss exponent in number. + TEST_NUMBER_ERROR(kParseErrorNumberMissExponent, "1e", 2u, 2u); + TEST_NUMBER_ERROR(kParseErrorNumberMissExponent, "1e_", 2u, 2u); + + // Issue 849 + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1.8e308", 0u, 7u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "5e308", 0u, 5u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e309", 0u, 5u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1.0e310", 0u, 7u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1.00e310", 0u, 8u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "-1.8e308", 0u, 8u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "-1e309", 0u, 6u); + + // Issue 1253 + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "2e308", 0u, 5u); + + // Issue 1259 + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, + "88474320368547737236837236775298547354737253685475547552933720368546854775297525" + "29337203685468547770151233720097201372368547312337203687203685423685123372036872" + "03685473724737236837236775297525854775297525472975254737236873720170151235473783" + "7236737247372368772473723683723456789012E66", 0u, 283u); + +#if 0 + // Test (length + exponent) overflow + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e+2147483647", 0u, 13u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e+9223372036854775807", 0u, 22u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e+10000", 0u, 8u); + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, "1e+50000", 0u, 8u); +#endif + + // 9007199254740992 * 2^971 ("infinity") + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, + "1.797693134862315907729305190789024733617976978942306572734300811577326758055009" + "63132708477322407536021120113879871393357658789768814416622492847430639474124377" + "76789342486548527630221960124609411945308295208500576883815068234246288147391311" + "0540827237163350510684586298239947245938479716304835356329624224137216e+308", 0u, 315u); + + // TODO: + // These tests (currently) fail in normal-precision mode + if (fullPrecision) + { + // Half way between max-normal and infinity + // Should round to infinity in nearest-even mode. + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, + "1.797693134862315807937289714053034150799341327100378269361737789804449682927647" + "50946649017977587207096330286416692887910946555547851940402630657488671505820681" + "90890200070838367627385484581771153176447573027006985557136695962284291481986083" + "49364752927190741684443655107043427115596995080930428801779041744977920000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000e+308", 0u, 1125u); + // ...round up + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, + "1.797693134862315807937289714053034150799341327100378269361737789804449682927647" + "50946649017977587207096330286416692887910946555547851940402630657488671505820681" + "90890200070838367627385484581771153176447573027006985557136695962284291481986083" + "49364752927190741684443655107043427115596995080930428801779041744977920000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000001e+308", 0u, 1205u); + } + + TEST_NUMBER_ERROR(kParseErrorNumberTooBig, + "10000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "00000000000000000000000000000000000000000000000000000000000000000000000000000000" + "0000000000000000000000000000000000000000000000000000000000000000000001", 0u, 310u); + +#undef TEST_NUMBER_ERROR +} + +TEST(Reader, ParseNumberError_NormalPrecisionDouble) { + TestParseNumberError<false>(); +} + +TEST(Reader, ParseNumberError_FullPrecisionDouble) { + TestParseNumberError<true>(); +} + +template <typename Encoding> +struct ParseStringHandler : BaseReaderHandler<Encoding, ParseStringHandler<Encoding> > { + ParseStringHandler() : str_(0), length_(0), copy_() {} + ~ParseStringHandler() { EXPECT_TRUE(str_ != 0); if (copy_) free(const_cast<typename Encoding::Ch*>(str_)); } + + ParseStringHandler(const ParseStringHandler&); + ParseStringHandler& operator=(const ParseStringHandler&); + + bool Default() { ADD_FAILURE(); return false; } + bool String(const typename Encoding::Ch* str, size_t length, bool copy) { + EXPECT_EQ(0, str_); + if (copy) { + str_ = static_cast<typename Encoding::Ch*>(malloc((length + 1) * sizeof(typename Encoding::Ch))); + memcpy(const_cast<typename Encoding::Ch*>(str_), str, (length + 1) * sizeof(typename Encoding::Ch)); + } + else + str_ = str; + length_ = length; + copy_ = copy; + return true; + } + + const typename Encoding::Ch* str_; + size_t length_; + bool copy_; +}; + +TEST(Reader, ParseString) { +#define TEST_STRING(Encoding, e, x) \ + { \ + Encoding::Ch* buffer = StrDup(x); \ + GenericInsituStringStream<Encoding> is(buffer); \ + ParseStringHandler<Encoding> h; \ + GenericReader<Encoding, Encoding> reader; \ + reader.Parse<kParseInsituFlag | kParseValidateEncodingFlag>(is, h); \ + EXPECT_EQ(0, StrCmp<Encoding::Ch>(e, h.str_)); \ + EXPECT_EQ(StrLen(e), h.length_); \ + free(buffer); \ + GenericStringStream<Encoding> s(x); \ + ParseStringHandler<Encoding> h2; \ + GenericReader<Encoding, Encoding> reader2; \ + reader2.Parse(s, h2); \ + EXPECT_EQ(0, StrCmp<Encoding::Ch>(e, h2.str_)); \ + EXPECT_EQ(StrLen(e), h2.length_); \ + } + + // String constant L"\xXX" can only specify character code in bytes, which is not endianness-neutral. + // And old compiler does not support u"" and U"" string literal. So here specify string literal by array of Ch. + // In addition, GCC 4.8 generates -Wnarrowing warnings when character code >= 128 are assigned to signed integer types. + // Therefore, utype is added for declaring unsigned array, and then cast it to Encoding::Ch. +#define ARRAY(...) { __VA_ARGS__ } +#define TEST_STRINGARRAY(Encoding, utype, array, x) \ + { \ + static const utype ue[] = array; \ + static const Encoding::Ch* e = reinterpret_cast<const Encoding::Ch *>(&ue[0]); \ + TEST_STRING(Encoding, e, x); \ + } + +#define TEST_STRINGARRAY2(Encoding, utype, earray, xarray) \ + { \ + static const utype ue[] = earray; \ + static const utype xe[] = xarray; \ + static const Encoding::Ch* e = reinterpret_cast<const Encoding::Ch *>(&ue[0]); \ + static const Encoding::Ch* x = reinterpret_cast<const Encoding::Ch *>(&xe[0]); \ + TEST_STRING(Encoding, e, x); \ + } + + TEST_STRING(UTF8<>, "", "\"\""); + TEST_STRING(UTF8<>, "Hello", "\"Hello\""); + TEST_STRING(UTF8<>, "Hello\nWorld", "\"Hello\\nWorld\""); + TEST_STRING(UTF8<>, "\"\\/\b\f\n\r\t", "\"\\\"\\\\/\\b\\f\\n\\r\\t\""); + TEST_STRING(UTF8<>, "\x24", "\"\\u0024\""); // Dollar sign U+0024 + TEST_STRING(UTF8<>, "\xC2\xA2", "\"\\u00A2\""); // Cents sign U+00A2 + TEST_STRING(UTF8<>, "\xE2\x82\xAC", "\"\\u20AC\""); // Euro sign U+20AC + TEST_STRING(UTF8<>, "\xF0\x9D\x84\x9E", "\"\\uD834\\uDD1E\""); // G clef sign U+1D11E + + // UTF16 + TEST_STRING(UTF16<>, L"", L"\"\""); + TEST_STRING(UTF16<>, L"Hello", L"\"Hello\""); + TEST_STRING(UTF16<>, L"Hello\nWorld", L"\"Hello\\nWorld\""); + TEST_STRING(UTF16<>, L"\"\\/\b\f\n\r\t", L"\"\\\"\\\\/\\b\\f\\n\\r\\t\""); + TEST_STRINGARRAY(UTF16<>, wchar_t, ARRAY(0x0024, 0x0000), L"\"\\u0024\""); + TEST_STRINGARRAY(UTF16<>, wchar_t, ARRAY(0x00A2, 0x0000), L"\"\\u00A2\""); // Cents sign U+00A2 + TEST_STRINGARRAY(UTF16<>, wchar_t, ARRAY(0x20AC, 0x0000), L"\"\\u20AC\""); // Euro sign U+20AC + TEST_STRINGARRAY(UTF16<>, wchar_t, ARRAY(0xD834, 0xDD1E, 0x0000), L"\"\\uD834\\uDD1E\""); // G clef sign U+1D11E + + // UTF32 + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY('\0'), ARRAY('\"', '\"', '\0')); + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY('H', 'e', 'l', 'l', 'o', '\0'), ARRAY('\"', 'H', 'e', 'l', 'l', 'o', '\"', '\0')); + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY('H', 'e', 'l', 'l', 'o', '\n', 'W', 'o', 'r', 'l', 'd', '\0'), ARRAY('\"', 'H', 'e', 'l', 'l', 'o', '\\', 'n', 'W', 'o', 'r', 'l', 'd', '\"', '\0')); + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY('\"', '\\', '/', '\b', '\f', '\n', '\r', '\t', '\0'), ARRAY('\"', '\\', '\"', '\\', '\\', '/', '\\', 'b', '\\', 'f', '\\', 'n', '\\', 'r', '\\', 't', '\"', '\0')); + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY(0x00024, 0x0000), ARRAY('\"', '\\', 'u', '0', '0', '2', '4', '\"', '\0')); + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY(0x000A2, 0x0000), ARRAY('\"', '\\', 'u', '0', '0', 'A', '2', '\"', '\0')); // Cents sign U+00A2 + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY(0x020AC, 0x0000), ARRAY('\"', '\\', 'u', '2', '0', 'A', 'C', '\"', '\0')); // Euro sign U+20AC + TEST_STRINGARRAY2(UTF32<>, unsigned, ARRAY(0x1D11E, 0x0000), ARRAY('\"', '\\', 'u', 'D', '8', '3', '4', '\\', 'u', 'D', 'D', '1', 'E', '\"', '\0')); // G clef sign U+1D11E + +#undef TEST_STRINGARRAY +#undef ARRAY +#undef TEST_STRING + + // Support of null character in string + { + StringStream s("\"Hello\\u0000World\""); + const char e[] = "Hello\0World"; + ParseStringHandler<UTF8<> > h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(0, memcmp(e, h.str_, h.length_ + 1)); + EXPECT_EQ(11u, h.length_); + } +} + +TEST(Reader, ParseString_Transcoding) { + const char* x = "\"Hello\""; + const wchar_t* e = L"Hello"; + GenericStringStream<UTF8<> > is(x); + GenericReader<UTF8<>, UTF16<> > reader; + ParseStringHandler<UTF16<> > h; + reader.Parse(is, h); + EXPECT_EQ(0, StrCmp<UTF16<>::Ch>(e, h.str_)); + EXPECT_EQ(StrLen(e), h.length_); +} + +TEST(Reader, ParseString_TranscodingWithValidation) { + const char* x = "\"Hello\""; + const wchar_t* e = L"Hello"; + GenericStringStream<UTF8<> > is(x); + GenericReader<UTF8<>, UTF16<> > reader; + ParseStringHandler<UTF16<> > h; + reader.Parse<kParseValidateEncodingFlag>(is, h); + EXPECT_EQ(0, StrCmp<UTF16<>::Ch>(e, h.str_)); + EXPECT_EQ(StrLen(e), h.length_); +} + +TEST(Reader, ParseString_NonDestructive) { + StringStream s("\"Hello\\nWorld\""); + ParseStringHandler<UTF8<> > h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(0, StrCmp("Hello\nWorld", h.str_)); + EXPECT_EQ(11u, h.length_); +} + +template <typename Encoding> +ParseErrorCode TestString(const typename Encoding::Ch* str) { + GenericStringStream<Encoding> s(str); + BaseReaderHandler<Encoding> h; + GenericReader<Encoding, Encoding> reader; + reader.template Parse<kParseValidateEncodingFlag>(s, h); + return reader.GetParseErrorCode(); +} + +TEST(Reader, ParseString_Error) { +#define TEST_STRING_ERROR(errorCode, str, errorOffset, streamPos)\ +{\ + GenericStringStream<UTF8<> > s(str);\ + BaseReaderHandler<UTF8<> > h;\ + GenericReader<UTF8<> , UTF8<> > reader;\ + reader.Parse<kParseValidateEncodingFlag>(s, h);\ + EXPECT_EQ(errorCode, reader.GetParseErrorCode());\ + EXPECT_EQ(errorOffset, reader.GetErrorOffset());\ + EXPECT_EQ(streamPos, s.Tell());\ +} + +#define ARRAY(...) { __VA_ARGS__ } +#define TEST_STRINGENCODING_ERROR(Encoding, TargetEncoding, utype, array) \ + { \ + static const utype ue[] = array; \ + static const Encoding::Ch* e = reinterpret_cast<const Encoding::Ch *>(&ue[0]); \ + EXPECT_EQ(kParseErrorStringInvalidEncoding, TestString<Encoding>(e));\ + /* decode error */\ + GenericStringStream<Encoding> s(e);\ + BaseReaderHandler<TargetEncoding> h;\ + GenericReader<Encoding, TargetEncoding> reader;\ + reader.Parse(s, h);\ + EXPECT_EQ(kParseErrorStringInvalidEncoding, reader.GetParseErrorCode());\ + } + + // Invalid escape character in string. + TEST_STRING_ERROR(kParseErrorStringEscapeInvalid, "[\"\\a\"]", 2u, 3u); + + // Incorrect hex digit after \\u escape in string. + TEST_STRING_ERROR(kParseErrorStringUnicodeEscapeInvalidHex, "[\"\\uABCG\"]", 2u, 7u); + + // Quotation in \\u escape in string (Issue #288) + TEST_STRING_ERROR(kParseErrorStringUnicodeEscapeInvalidHex, "[\"\\uaaa\"]", 2u, 7u); + TEST_STRING_ERROR(kParseErrorStringUnicodeEscapeInvalidHex, "[\"\\uD800\\uFFF\"]", 2u, 13u); + + // The surrogate pair in string is invalid. + TEST_STRING_ERROR(kParseErrorStringUnicodeSurrogateInvalid, "[\"\\uD800X\"]", 2u, 8u); + TEST_STRING_ERROR(kParseErrorStringUnicodeSurrogateInvalid, "[\"\\uD800\\uFFFF\"]", 2u, 14u); + + // Single low surrogate pair in string is invalid. + TEST_STRING_ERROR(kParseErrorStringUnicodeSurrogateInvalid, "[\"\\udc4d\"]", 2u, 8u); + + // Missing a closing quotation mark in string. + TEST_STRING_ERROR(kParseErrorStringMissQuotationMark, "[\"Test]", 7u, 7u); + + // http://www.cl.cam.ac.uk/~mgk25/ucs/examples/UTF-8-test.txt + + // 3 Malformed sequences + + // 3.1 Unexpected continuation bytes + { + char e[] = { '[', '\"', 0, '\"', ']', '\0' }; + for (unsigned char c = 0x80u; c <= 0xBFu; c++) { + e[2] = static_cast<char>(c); + ParseErrorCode error = TestString<UTF8<> >(e); + EXPECT_EQ(kParseErrorStringInvalidEncoding, error); + if (error != kParseErrorStringInvalidEncoding) + std::cout << static_cast<unsigned>(c) << std::endl; + } + } + + // 3.2 Lonely start characters, 3.5 Impossible bytes + { + char e[] = { '[', '\"', 0, ' ', '\"', ']', '\0' }; + for (unsigned c = 0xC0u; c <= 0xFFu; c++) { + e[2] = static_cast<char>(c); + unsigned streamPos; + if (c <= 0xC1u) + streamPos = 3; // 0xC0 - 0xC1 + else if (c <= 0xDFu) + streamPos = 4; // 0xC2 - 0xDF + else if (c <= 0xEFu) + streamPos = 5; // 0xE0 - 0xEF + else if (c <= 0xF4u) + streamPos = 6; // 0xF0 - 0xF4 + else + streamPos = 3; // 0xF5 - 0xFF + TEST_STRING_ERROR(kParseErrorStringInvalidEncoding, e, 2u, streamPos); + } + } + + // 4 Overlong sequences + + // 4.1 Examples of an overlong ASCII character + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xC0u, 0xAFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xE0u, 0x80u, 0xAFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xF0u, 0x80u, 0x80u, 0xAFu, '\"', ']', '\0')); + + // 4.2 Maximum overlong sequences + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xC1u, 0xBFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xE0u, 0x9Fu, 0xBFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xF0u, 0x8Fu, 0xBFu, 0xBFu, '\"', ']', '\0')); + + // 4.3 Overlong representation of the NUL character + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xC0u, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xE0u, 0x80u, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xF0u, 0x80u, 0x80u, 0x80u, '\"', ']', '\0')); + + // 5 Illegal code positions + + // 5.1 Single UTF-16 surrogates + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xA0u, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xADu, 0xBFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xAEu, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xAFu, 0xBFu, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xB0u, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xBEu, 0x80u, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF8<>, UTF16<>, unsigned char, ARRAY('[', '\"', 0xEDu, 0xBFu, 0xBFu, '\"', ']', '\0')); + + // Malform UTF-16 sequences + TEST_STRINGENCODING_ERROR(UTF16<>, UTF8<>, wchar_t, ARRAY('[', '\"', 0xDC00, 0xDC00, '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(UTF16<>, UTF8<>, wchar_t, ARRAY('[', '\"', 0xD800, 0xD800, '\"', ']', '\0')); + + // Malform UTF-32 sequence + TEST_STRINGENCODING_ERROR(UTF32<>, UTF8<>, unsigned, ARRAY('[', '\"', 0x110000, '\"', ']', '\0')); + + // Malform ASCII sequence + TEST_STRINGENCODING_ERROR(ASCII<>, UTF8<>, char, ARRAY('[', '\"', char(0x80u), '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(ASCII<>, UTF8<>, char, ARRAY('[', '\"', char(0x01u), '\"', ']', '\0')); + TEST_STRINGENCODING_ERROR(ASCII<>, UTF8<>, char, ARRAY('[', '\"', char(0x1Cu), '\"', ']', '\0')); + +#undef ARRAY +#undef TEST_STRINGARRAY_ERROR +} + +template <unsigned count> +struct ParseArrayHandler : BaseReaderHandler<UTF8<>, ParseArrayHandler<count> > { + ParseArrayHandler() : step_(0) {} + + bool Default() { ADD_FAILURE(); return false; } + bool Uint(unsigned i) { EXPECT_EQ(step_, i); step_++; return true; } + bool StartArray() { EXPECT_EQ(0u, step_); step_++; return true; } + bool EndArray(SizeType) { step_++; return true; } + + unsigned step_; +}; + +TEST(Reader, ParseEmptyArray) { + char *json = StrDup("[ ] "); + InsituStringStream s(json); + ParseArrayHandler<0> h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(2u, h.step_); + free(json); +} + +TEST(Reader, ParseArray) { + char *json = StrDup("[1, 2, 3, 4]"); + InsituStringStream s(json); + ParseArrayHandler<4> h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(6u, h.step_); + free(json); +} + +TEST(Reader, ParseArray_Error) { +#define TEST_ARRAY_ERROR(errorCode, str, errorOffset) \ + { \ + unsigned streamPos = errorOffset; \ + char buffer[1001]; \ + strncpy(buffer, str, 1000); \ + InsituStringStream s(buffer); \ + BaseReaderHandler<> h; \ + GenericReader<UTF8<>, UTF8<>, CrtAllocator> reader; \ + EXPECT_FALSE(reader.Parse(s, h)); \ + EXPECT_EQ(errorCode, reader.GetParseErrorCode());\ + EXPECT_EQ(errorOffset, reader.GetErrorOffset());\ + EXPECT_EQ(streamPos, s.Tell());\ + } + + // Missing a comma or ']' after an array element. + TEST_ARRAY_ERROR(kParseErrorArrayMissCommaOrSquareBracket, "[1", 2u); + TEST_ARRAY_ERROR(kParseErrorArrayMissCommaOrSquareBracket, "[1}", 2u); + TEST_ARRAY_ERROR(kParseErrorArrayMissCommaOrSquareBracket, "[1 2]", 3u); + + // Array cannot have a trailing comma (without kParseTrailingCommasFlag); + // a value must follow a comma + TEST_ARRAY_ERROR(kParseErrorValueInvalid, "[1,]", 3u); + +#undef TEST_ARRAY_ERROR +} + +struct ParseObjectHandler : BaseReaderHandler<UTF8<>, ParseObjectHandler> { + ParseObjectHandler() : step_(0) {} + + bool Default() { ADD_FAILURE(); return false; } + bool Null() { EXPECT_EQ(8u, step_); step_++; return true; } + bool Bool(bool b) { + switch(step_) { + case 4: EXPECT_TRUE(b); step_++; return true; + case 6: EXPECT_FALSE(b); step_++; return true; + default: ADD_FAILURE(); return false; + } + } + bool Int(int i) { + switch(step_) { + case 10: EXPECT_EQ(123, i); step_++; return true; + case 15: EXPECT_EQ(1, i); step_++; return true; + case 16: EXPECT_EQ(2, i); step_++; return true; + case 17: EXPECT_EQ(3, i); step_++; return true; + default: ADD_FAILURE(); return false; + } + } + bool Uint(unsigned i) { return Int(static_cast<int>(i)); } + bool Double(double d) { EXPECT_EQ(12u, step_); EXPECT_DOUBLE_EQ(3.1416, d); step_++; return true; } + bool String(const char* str, size_t, bool) { + switch(step_) { + case 1: EXPECT_STREQ("hello", str); step_++; return true; + case 2: EXPECT_STREQ("world", str); step_++; return true; + case 3: EXPECT_STREQ("t", str); step_++; return true; + case 5: EXPECT_STREQ("f", str); step_++; return true; + case 7: EXPECT_STREQ("n", str); step_++; return true; + case 9: EXPECT_STREQ("i", str); step_++; return true; + case 11: EXPECT_STREQ("pi", str); step_++; return true; + case 13: EXPECT_STREQ("a", str); step_++; return true; + default: ADD_FAILURE(); return false; + } + } + bool StartObject() { EXPECT_EQ(0u, step_); step_++; return true; } + bool EndObject(SizeType memberCount) { EXPECT_EQ(19u, step_); EXPECT_EQ(7u, memberCount); step_++; return true; } + bool StartArray() { EXPECT_EQ(14u, step_); step_++; return true; } + bool EndArray(SizeType elementCount) { EXPECT_EQ(18u, step_); EXPECT_EQ(3u, elementCount); step_++; return true; } + + unsigned step_; +}; + +TEST(Reader, ParseObject) { + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] } "; + + // Insitu + { + char* json2 = StrDup(json); + InsituStringStream s(json2); + ParseObjectHandler h; + Reader reader; + reader.Parse<kParseInsituFlag>(s, h); + EXPECT_EQ(20u, h.step_); + free(json2); + } + + // Normal + { + StringStream s(json); + ParseObjectHandler h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(20u, h.step_); + } +} + +struct ParseEmptyObjectHandler : BaseReaderHandler<UTF8<>, ParseEmptyObjectHandler> { + ParseEmptyObjectHandler() : step_(0) {} + + bool Default() { ADD_FAILURE(); return false; } + bool StartObject() { EXPECT_EQ(0u, step_); step_++; return true; } + bool EndObject(SizeType) { EXPECT_EQ(1u, step_); step_++; return true; } + + unsigned step_; +}; + +TEST(Reader, Parse_EmptyObject) { + StringStream s("{ } "); + ParseEmptyObjectHandler h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(2u, h.step_); +} + +struct ParseMultipleRootHandler : BaseReaderHandler<UTF8<>, ParseMultipleRootHandler> { + ParseMultipleRootHandler() : step_(0) {} + + bool Default() { ADD_FAILURE(); return false; } + bool StartObject() { EXPECT_EQ(0u, step_); step_++; return true; } + bool EndObject(SizeType) { EXPECT_EQ(1u, step_); step_++; return true; } + bool StartArray() { EXPECT_EQ(2u, step_); step_++; return true; } + bool EndArray(SizeType) { EXPECT_EQ(3u, step_); step_++; return true; } + + unsigned step_; +}; + +template <unsigned parseFlags> +void TestMultipleRoot() { + StringStream s("{}[] a"); + ParseMultipleRootHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<parseFlags>(s, h)); + EXPECT_EQ(2u, h.step_); + EXPECT_TRUE(reader.Parse<parseFlags>(s, h)); + EXPECT_EQ(4u, h.step_); + EXPECT_EQ(' ', s.Take()); + EXPECT_EQ('a', s.Take()); +} + +TEST(Reader, Parse_MultipleRoot) { + TestMultipleRoot<kParseStopWhenDoneFlag>(); +} + +TEST(Reader, ParseIterative_MultipleRoot) { + TestMultipleRoot<kParseIterativeFlag | kParseStopWhenDoneFlag>(); +} + +template <unsigned parseFlags> +void TestInsituMultipleRoot() { + char* buffer = strdup("{}[] a"); + InsituStringStream s(buffer); + ParseMultipleRootHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag | parseFlags>(s, h)); + EXPECT_EQ(2u, h.step_); + EXPECT_TRUE(reader.Parse<kParseInsituFlag | parseFlags>(s, h)); + EXPECT_EQ(4u, h.step_); + EXPECT_EQ(' ', s.Take()); + EXPECT_EQ('a', s.Take()); + free(buffer); +} + +TEST(Reader, ParseInsitu_MultipleRoot) { + TestInsituMultipleRoot<kParseStopWhenDoneFlag>(); +} + +TEST(Reader, ParseInsituIterative_MultipleRoot) { + TestInsituMultipleRoot<kParseIterativeFlag | kParseStopWhenDoneFlag>(); +} + +#define TEST_ERROR(errorCode, str, errorOffset) \ + { \ + unsigned streamPos = errorOffset; \ + char buffer[1001]; \ + strncpy(buffer, str, 1000); \ + InsituStringStream s(buffer); \ + BaseReaderHandler<> h; \ + Reader reader; \ + EXPECT_FALSE(reader.Parse(s, h)); \ + EXPECT_EQ(errorCode, reader.GetParseErrorCode());\ + EXPECT_EQ(errorOffset, reader.GetErrorOffset());\ + EXPECT_EQ(streamPos, s.Tell());\ + } + +TEST(Reader, ParseDocument_Error) { + // The document is empty. + TEST_ERROR(kParseErrorDocumentEmpty, "", 0u); + TEST_ERROR(kParseErrorDocumentEmpty, " ", 1u); + TEST_ERROR(kParseErrorDocumentEmpty, " \n", 2u); + + // The document root must not follow by other values. + TEST_ERROR(kParseErrorDocumentRootNotSingular, "[] 0", 3u); + TEST_ERROR(kParseErrorDocumentRootNotSingular, "{} 0", 3u); + TEST_ERROR(kParseErrorDocumentRootNotSingular, "null []", 5u); + TEST_ERROR(kParseErrorDocumentRootNotSingular, "0 {}", 2u); +} + +TEST(Reader, ParseValue_Error) { + // Invalid value. + TEST_ERROR(kParseErrorValueInvalid, "nulL", 3u); + TEST_ERROR(kParseErrorValueInvalid, "truE", 3u); + TEST_ERROR(kParseErrorValueInvalid, "falsE", 4u); + TEST_ERROR(kParseErrorValueInvalid, "a]", 0u); + TEST_ERROR(kParseErrorValueInvalid, ".1", 0u); +} + +TEST(Reader, ParseObject_Error) { + // Missing a name for object member. + TEST_ERROR(kParseErrorObjectMissName, "{1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{null:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{true:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{false:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{1:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{[]:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{{}:1}", 1u); + TEST_ERROR(kParseErrorObjectMissName, "{xyz:1}", 1u); + + // Missing a colon after a name of object member. + TEST_ERROR(kParseErrorObjectMissColon, "{\"a\" 1}", 5u); + TEST_ERROR(kParseErrorObjectMissColon, "{\"a\",1}", 4u); + + // Must be a comma or '}' after an object member + TEST_ERROR(kParseErrorObjectMissCommaOrCurlyBracket, "{\"a\":1]", 6u); + + // Object cannot have a trailing comma (without kParseTrailingCommasFlag); + // an object member name must follow a comma + TEST_ERROR(kParseErrorObjectMissName, "{\"a\":1,}", 7u); + + // This tests that MemoryStream is checking the length in Peek(). + { + MemoryStream ms("{\"a\"", 1); + BaseReaderHandler<> h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseStopWhenDoneFlag>(ms, h)); + EXPECT_EQ(kParseErrorObjectMissName, reader.GetParseErrorCode()); + } +} + +#undef TEST_ERROR + +TEST(Reader, SkipWhitespace) { + StringStream ss(" A \t\tB\n \n\nC\r\r \rD \t\n\r E"); + const char* expected = "ABCDE"; + for (size_t i = 0; i < 5; i++) { + SkipWhitespace(ss); + EXPECT_EQ(expected[i], ss.Take()); + } +} + +// Test implementing a stream without copy stream optimization. +// Clone from GenericStringStream except that copy constructor is disabled. +template <typename Encoding> +class CustomStringStream { +public: + typedef typename Encoding::Ch Ch; + + CustomStringStream(const Ch *src) : src_(src), head_(src) {} + + Ch Peek() const { return *src_; } + Ch Take() { return *src_++; } + size_t Tell() const { return static_cast<size_t>(src_ - head_); } + + Ch* PutBegin() { RAPIDJSON_ASSERT(false); return 0; } + void Put(Ch) { RAPIDJSON_ASSERT(false); } + void Flush() { RAPIDJSON_ASSERT(false); } + size_t PutEnd(Ch*) { RAPIDJSON_ASSERT(false); return 0; } + +private: + // Prohibit copy constructor & assignment operator. + CustomStringStream(const CustomStringStream&); + CustomStringStream& operator=(const CustomStringStream&); + + const Ch* src_; //!< Current read position. + const Ch* head_; //!< Original head of the string. +}; + +// If the following code is compiled, it should generate compilation error as predicted. +// Because CustomStringStream<> is not copyable via making copy constructor private. +#if 0 +namespace rapidjson { + +template <typename Encoding> +struct StreamTraits<CustomStringStream<Encoding> > { + enum { copyOptimization = 1 }; +}; + +} // namespace rapidjson +#endif + +TEST(Reader, CustomStringStream) { + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] } "; + CustomStringStream<UTF8<char> > s(json); + ParseObjectHandler h; + Reader reader; + reader.Parse(s, h); + EXPECT_EQ(20u, h.step_); +} + +#include <sstream> + +class IStreamWrapper { +public: + typedef char Ch; + + IStreamWrapper(std::istream& is) : is_(is) {} + + Ch Peek() const { + int c = is_.peek(); + return c == std::char_traits<char>::eof() ? '\0' : static_cast<Ch>(c); + } + + Ch Take() { + int c = is_.get(); + return c == std::char_traits<char>::eof() ? '\0' : static_cast<Ch>(c); + } + + size_t Tell() const { return static_cast<size_t>(is_.tellg()); } + + Ch* PutBegin() { assert(false); return 0; } + void Put(Ch) { assert(false); } + void Flush() { assert(false); } + size_t PutEnd(Ch*) { assert(false); return 0; } + +private: + IStreamWrapper(const IStreamWrapper&); + IStreamWrapper& operator=(const IStreamWrapper&); + + std::istream& is_; +}; + +class WIStreamWrapper { +public: + typedef wchar_t Ch; + + WIStreamWrapper(std::wistream& is) : is_(is) {} + + Ch Peek() const { + unsigned c = is_.peek(); + return c == std::char_traits<wchar_t>::eof() ? Ch('\0') : static_cast<Ch>(c); + } + + Ch Take() { + unsigned c = is_.get(); + return c == std::char_traits<wchar_t>::eof() ? Ch('\0') : static_cast<Ch>(c); + } + + size_t Tell() const { return static_cast<size_t>(is_.tellg()); } + + Ch* PutBegin() { assert(false); return 0; } + void Put(Ch) { assert(false); } + void Flush() { assert(false); } + size_t PutEnd(Ch*) { assert(false); return 0; } + +private: + WIStreamWrapper(const WIStreamWrapper&); + WIStreamWrapper& operator=(const WIStreamWrapper&); + + std::wistream& is_; +}; + +TEST(Reader, Parse_IStreamWrapper_StringStream) { + const char* json = "[1,2,3,4]"; + + std::stringstream ss(json); + IStreamWrapper is(ss); + + Reader reader; + ParseArrayHandler<4> h; + reader.Parse(is, h); + EXPECT_FALSE(reader.HasParseError()); +} + +// Test iterative parsing. + +#define TESTERRORHANDLING(text, errorCode, offset)\ +{\ + unsigned streamPos = offset; \ + StringStream json(text); \ + BaseReaderHandler<> handler; \ + Reader reader; \ + reader.Parse<kParseIterativeFlag>(json, handler); \ + EXPECT_TRUE(reader.HasParseError()); \ + EXPECT_EQ(errorCode, reader.GetParseErrorCode()); \ + EXPECT_EQ(offset, reader.GetErrorOffset()); \ + EXPECT_EQ(streamPos, json.Tell()); \ +} + +TEST(Reader, IterativeParsing_ErrorHandling) { + TESTERRORHANDLING("{\"a\": a}", kParseErrorValueInvalid, 6u); + + TESTERRORHANDLING("", kParseErrorDocumentEmpty, 0u); + TESTERRORHANDLING("{}{}", kParseErrorDocumentRootNotSingular, 2u); + + TESTERRORHANDLING("{1}", kParseErrorObjectMissName, 1u); + TESTERRORHANDLING("{\"a\", 1}", kParseErrorObjectMissColon, 4u); + TESTERRORHANDLING("{\"a\"}", kParseErrorObjectMissColon, 4u); + TESTERRORHANDLING("{\"a\": 1", kParseErrorObjectMissCommaOrCurlyBracket, 7u); + TESTERRORHANDLING("[1 2 3]", kParseErrorArrayMissCommaOrSquareBracket, 3u); + TESTERRORHANDLING("{\"a: 1", kParseErrorStringMissQuotationMark, 6u); + TESTERRORHANDLING("{\"a\":}", kParseErrorValueInvalid, 5u); + TESTERRORHANDLING("{\"a\":]", kParseErrorValueInvalid, 5u); + TESTERRORHANDLING("[1,2,}", kParseErrorValueInvalid, 5u); + TESTERRORHANDLING("[}]", kParseErrorValueInvalid, 1u); + TESTERRORHANDLING("[,]", kParseErrorValueInvalid, 1u); + TESTERRORHANDLING("[1,,]", kParseErrorValueInvalid, 3u); + + // Trailing commas are not allowed without kParseTrailingCommasFlag + TESTERRORHANDLING("{\"a\": 1,}", kParseErrorObjectMissName, 8u); + TESTERRORHANDLING("[1,2,3,]", kParseErrorValueInvalid, 7u); + + // Any JSON value can be a valid root element in RFC7159. + TESTERRORHANDLING("\"ab", kParseErrorStringMissQuotationMark, 3u); + TESTERRORHANDLING("truE", kParseErrorValueInvalid, 3u); + TESTERRORHANDLING("False", kParseErrorValueInvalid, 0u); + TESTERRORHANDLING("true, false", kParseErrorDocumentRootNotSingular, 4u); + TESTERRORHANDLING("false, false", kParseErrorDocumentRootNotSingular, 5u); + TESTERRORHANDLING("nulL", kParseErrorValueInvalid, 3u); + TESTERRORHANDLING("null , null", kParseErrorDocumentRootNotSingular, 5u); + TESTERRORHANDLING("1a", kParseErrorDocumentRootNotSingular, 1u); +} + +template<typename Encoding = UTF8<> > +struct IterativeParsingReaderHandler { + typedef typename Encoding::Ch Ch; + + const static uint32_t LOG_NULL = 0x10000000; + const static uint32_t LOG_BOOL = 0x20000000; + const static uint32_t LOG_INT = 0x30000000; + const static uint32_t LOG_UINT = 0x40000000; + const static uint32_t LOG_INT64 = 0x50000000; + const static uint32_t LOG_UINT64 = 0x60000000; + const static uint32_t LOG_DOUBLE = 0x70000000; + const static uint32_t LOG_STRING = 0x80000000; + const static uint32_t LOG_STARTOBJECT = 0x90000000; + const static uint32_t LOG_KEY = 0xA0000000; + const static uint32_t LOG_ENDOBJECT = 0xB0000000; + const static uint32_t LOG_STARTARRAY = 0xC0000000; + const static uint32_t LOG_ENDARRAY = 0xD0000000; + + const static size_t LogCapacity = 256; + uint32_t Logs[LogCapacity]; + size_t LogCount; + + IterativeParsingReaderHandler() : LogCount(0) { + } + + bool Null() { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_NULL; return true; } + + bool Bool(bool) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_BOOL; return true; } + + bool Int(int) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_INT; return true; } + + bool Uint(unsigned) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_INT; return true; } + + bool Int64(int64_t) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_INT64; return true; } + + bool Uint64(uint64_t) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_UINT64; return true; } + + bool Double(double) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_DOUBLE; return true; } + + bool RawNumber(const Ch*, SizeType, bool) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_STRING; return true; } + + bool String(const Ch*, SizeType, bool) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_STRING; return true; } + + bool StartObject() { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_STARTOBJECT; return true; } + + bool Key (const Ch*, SizeType, bool) { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_KEY; return true; } + + bool EndObject(SizeType c) { + RAPIDJSON_ASSERT(LogCount < LogCapacity); + RAPIDJSON_ASSERT((static_cast<uint32_t>(c) & 0xF0000000) == 0); + Logs[LogCount++] = LOG_ENDOBJECT | static_cast<uint32_t>(c); + return true; + } + + bool StartArray() { RAPIDJSON_ASSERT(LogCount < LogCapacity); Logs[LogCount++] = LOG_STARTARRAY; return true; } + + bool EndArray(SizeType c) { + RAPIDJSON_ASSERT(LogCount < LogCapacity); + RAPIDJSON_ASSERT((static_cast<uint32_t>(c) & 0xF0000000) == 0); + Logs[LogCount++] = LOG_ENDARRAY | static_cast<uint32_t>(c); + return true; + } +}; + +TEST(Reader, IterativeParsing_General) { + { + StringStream is("[1, {\"k\": [1, 2]}, null, false, true, \"string\", 1.2]"); + Reader reader; + IterativeParsingReaderHandler<> handler; + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_FALSE(r.IsError()); + EXPECT_FALSE(reader.HasParseError()); + + uint32_t e[] = { + handler.LOG_STARTARRAY, + handler.LOG_INT, + handler.LOG_STARTOBJECT, + handler.LOG_KEY, + handler.LOG_STARTARRAY, + handler.LOG_INT, + handler.LOG_INT, + handler.LOG_ENDARRAY | 2, + handler.LOG_ENDOBJECT | 1, + handler.LOG_NULL, + handler.LOG_BOOL, + handler.LOG_BOOL, + handler.LOG_STRING, + handler.LOG_DOUBLE, + handler.LOG_ENDARRAY | 7 + }; + + EXPECT_EQ(sizeof(e) / sizeof(int), handler.LogCount); + + for (size_t i = 0; i < handler.LogCount; ++i) { + EXPECT_EQ(e[i], handler.Logs[i]) << "i = " << i; + } + } +} + +TEST(Reader, IterativeParsing_Count) { + { + StringStream is("[{}, {\"k\": 1}, [1], []]"); + Reader reader; + IterativeParsingReaderHandler<> handler; + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_FALSE(r.IsError()); + EXPECT_FALSE(reader.HasParseError()); + + uint32_t e[] = { + handler.LOG_STARTARRAY, + handler.LOG_STARTOBJECT, + handler.LOG_ENDOBJECT | 0, + handler.LOG_STARTOBJECT, + handler.LOG_KEY, + handler.LOG_INT, + handler.LOG_ENDOBJECT | 1, + handler.LOG_STARTARRAY, + handler.LOG_INT, + handler.LOG_ENDARRAY | 1, + handler.LOG_STARTARRAY, + handler.LOG_ENDARRAY | 0, + handler.LOG_ENDARRAY | 4 + }; + + EXPECT_EQ(sizeof(e) / sizeof(int), handler.LogCount); + + for (size_t i = 0; i < handler.LogCount; ++i) { + EXPECT_EQ(e[i], handler.Logs[i]) << "i = " << i; + } + } +} + +TEST(Reader, IterativePullParsing_General) { + { + IterativeParsingReaderHandler<> handler; + uint32_t e[] = { + handler.LOG_STARTARRAY, + handler.LOG_INT, + handler.LOG_STARTOBJECT, + handler.LOG_KEY, + handler.LOG_STARTARRAY, + handler.LOG_INT, + handler.LOG_INT, + handler.LOG_ENDARRAY | 2, + handler.LOG_ENDOBJECT | 1, + handler.LOG_NULL, + handler.LOG_BOOL, + handler.LOG_BOOL, + handler.LOG_STRING, + handler.LOG_DOUBLE, + handler.LOG_ENDARRAY | 7 + }; + + StringStream is("[1, {\"k\": [1, 2]}, null, false, true, \"string\", 1.2]"); + Reader reader; + + reader.IterativeParseInit(); + while (!reader.IterativeParseComplete()) { + size_t oldLogCount = handler.LogCount; + EXPECT_TRUE(oldLogCount < sizeof(e) / sizeof(int)) << "overrun"; + + EXPECT_TRUE(reader.IterativeParseNext<kParseDefaultFlags>(is, handler)) << "parse fail"; + EXPECT_EQ(handler.LogCount, oldLogCount + 1) << "handler should be invoked exactly once each time"; + EXPECT_EQ(e[oldLogCount], handler.Logs[oldLogCount]) << "wrong event returned"; + } + + EXPECT_FALSE(reader.HasParseError()); + EXPECT_EQ(sizeof(e) / sizeof(int), handler.LogCount) << "handler invoked wrong number of times"; + + // The handler should not be invoked when the JSON has been fully read, but it should not fail + size_t oldLogCount = handler.LogCount; + EXPECT_TRUE(reader.IterativeParseNext<kParseDefaultFlags>(is, handler)) << "parse-next past complete is allowed"; + EXPECT_EQ(handler.LogCount, oldLogCount) << "parse-next past complete should not invoke handler"; + EXPECT_FALSE(reader.HasParseError()) << "parse-next past complete should not generate parse error"; + } +} + +// Test iterative parsing on kParseErrorTermination. +struct HandlerTerminateAtStartObject : public IterativeParsingReaderHandler<> { + bool StartObject() { return false; } +}; + +struct HandlerTerminateAtStartArray : public IterativeParsingReaderHandler<> { + bool StartArray() { return false; } +}; + +struct HandlerTerminateAtEndObject : public IterativeParsingReaderHandler<> { + bool EndObject(SizeType) { return false; } +}; + +struct HandlerTerminateAtEndArray : public IterativeParsingReaderHandler<> { + bool EndArray(SizeType) { return false; } +}; + +TEST(Reader, IterativeParsing_ShortCircuit) { + { + HandlerTerminateAtStartObject handler; + Reader reader; + StringStream is("[1, {}]"); + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(4u, r.Offset()); + } + + { + HandlerTerminateAtStartArray handler; + Reader reader; + StringStream is("{\"a\": []}"); + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(6u, r.Offset()); + } + + { + HandlerTerminateAtEndObject handler; + Reader reader; + StringStream is("[1, {}]"); + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(5u, r.Offset()); + } + + { + HandlerTerminateAtEndArray handler; + Reader reader; + StringStream is("{\"a\": []}"); + + ParseResult r = reader.Parse<kParseIterativeFlag>(is, handler); + + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(7u, r.Offset()); + } +} + +// For covering BaseReaderHandler default functions +TEST(Reader, BaseReaderHandler_Default) { + BaseReaderHandler<> h; + Reader reader; + StringStream is("[null, true, -1, 1, -1234567890123456789, 1234567890123456789, 3.14, \"s\", { \"a\" : 1 }]"); + EXPECT_TRUE(reader.Parse(is, h)); +} + +template <int e> +struct TerminateHandler { + bool Null() { return e != 0; } + bool Bool(bool) { return e != 1; } + bool Int(int) { return e != 2; } + bool Uint(unsigned) { return e != 3; } + bool Int64(int64_t) { return e != 4; } + bool Uint64(uint64_t) { return e != 5; } + bool Double(double) { return e != 6; } + bool RawNumber(const char*, SizeType, bool) { return e != 7; } + bool String(const char*, SizeType, bool) { return e != 8; } + bool StartObject() { return e != 9; } + bool Key(const char*, SizeType, bool) { return e != 10; } + bool EndObject(SizeType) { return e != 11; } + bool StartArray() { return e != 12; } + bool EndArray(SizeType) { return e != 13; } +}; + +#define TEST_TERMINATION(e, json)\ +{\ + Reader reader;\ + TerminateHandler<e> h;\ + StringStream is(json);\ + EXPECT_FALSE(reader.Parse(is, h));\ + EXPECT_EQ(kParseErrorTermination, reader.GetParseErrorCode());\ +} + +TEST(Reader, ParseTerminationByHandler) { + TEST_TERMINATION(0, "[null"); + TEST_TERMINATION(1, "[true"); + TEST_TERMINATION(1, "[false"); + TEST_TERMINATION(2, "[-1"); + TEST_TERMINATION(3, "[1"); + TEST_TERMINATION(4, "[-1234567890123456789"); + TEST_TERMINATION(5, "[1234567890123456789"); + TEST_TERMINATION(6, "[0.5]"); + // RawNumber() is never called + TEST_TERMINATION(8, "[\"a\""); + TEST_TERMINATION(9, "[{"); + TEST_TERMINATION(10, "[{\"a\""); + TEST_TERMINATION(11, "[{}"); + TEST_TERMINATION(11, "[{\"a\":1}"); // non-empty object + TEST_TERMINATION(12, "{\"a\":["); + TEST_TERMINATION(13, "{\"a\":[]"); + TEST_TERMINATION(13, "{\"a\":[1]"); // non-empty array +} + +TEST(Reader, ParseComments) { + const char* json = + "// Here is a one-line comment.\n" + "{// And here's another one\n" + " /*And here's an in-line one.*/\"hello\" : \"world\"," + " \"t\" :/* And one with '*' symbol*/true ," + "/* A multiline comment\n" + " goes here*/" + " \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3]" + "}/*And the last one to be sure */"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(20u, h.step_); +} + +TEST(Reader, ParseEmptyInlineComment) { + const char* json = "{/**/\"hello\" : \"world\", \"t\" : true, \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(20u, h.step_); +} + +TEST(Reader, ParseEmptyOnelineComment) { + const char* json = "{//\n\"hello\" : \"world\", \"t\" : true, \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(20u, h.step_); +} + +TEST(Reader, ParseMultipleCommentsInARow) { + const char* json = + "{/* first comment *//* second */\n" + "/* third */ /*fourth*/// last one\n" + "\"hello\" : \"world\", \"t\" : true, \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(20u, h.step_); +} + +TEST(Reader, InlineCommentsAreDisabledByDefault) { + { + const char* json = "{/* Inline comment. */\"hello\" : \"world\", \"t\" : true, \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseDefaultFlags>(s, h)); + } + + { + const char* json = + "{\"hello\" : /* Multiline comment starts here\n" + " continues here\n" + " and ends here */\"world\", \"t\" :true , \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseDefaultFlags>(s, h)); + } +} + +TEST(Reader, OnelineCommentsAreDisabledByDefault) { + const char* json = "{// One-line comment\n\"hello\" : \"world\", \"t\" : true , \"f\" : false, \"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3] }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseDefaultFlags>(s, h)); +} + +TEST(Reader, EofAfterOneLineComment) { + const char* json = "{\"hello\" : \"world\" // EOF is here -->\0 \n}"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(kParseErrorObjectMissCommaOrCurlyBracket, reader.GetParseErrorCode()); +} + +TEST(Reader, IncompleteMultilineComment) { + const char* json = "{\"hello\" : \"world\" /* EOF is here -->\0 */}"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(kParseErrorUnspecificSyntaxError, reader.GetParseErrorCode()); +} + +TEST(Reader, IncompleteMultilineComment2) { + const char* json = "{\"hello\" : \"world\" /* *\0 */}"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(kParseErrorUnspecificSyntaxError, reader.GetParseErrorCode()); +} + +TEST(Reader, UnrecognizedComment) { + const char* json = "{\"hello\" : \"world\" /! }"; + + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_FALSE(reader.Parse<kParseCommentsFlag>(s, h)); + EXPECT_EQ(kParseErrorUnspecificSyntaxError, reader.GetParseErrorCode()); +} + +struct NumbersAsStringsHandler { + bool Null() { return true; } + bool Bool(bool) { return true; } + bool Int(int) { return true; } + bool Uint(unsigned) { return true; } + bool Int64(int64_t) { return true; } + bool Uint64(uint64_t) { return true; } + bool Double(double) { return true; } + // 'str' is not null-terminated + bool RawNumber(const char* str, SizeType length, bool) { + EXPECT_TRUE(str != 0); + EXPECT_TRUE(expected_len_ == length); + EXPECT_TRUE(strncmp(str, expected_, length) == 0); + return true; + } + bool String(const char*, SizeType, bool) { return true; } + bool StartObject() { return true; } + bool Key(const char*, SizeType, bool) { return true; } + bool EndObject(SizeType) { return true; } + bool StartArray() { return true; } + bool EndArray(SizeType) { return true; } + + NumbersAsStringsHandler(const char* expected) + : expected_(expected) + , expected_len_(strlen(expected)) {} + + const char* expected_; + size_t expected_len_; +}; + +TEST(Reader, NumbersAsStrings) { + { + const char* json = "{ \"pi\": 3.1416 } "; + StringStream s(json); + NumbersAsStringsHandler h("3.1416"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + char* json = StrDup("{ \"pi\": 3.1416 } "); + InsituStringStream s(json); + NumbersAsStringsHandler h("3.1416"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag|kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const char* json = "{ \"gigabyte\": 1.0e9 } "; + StringStream s(json); + NumbersAsStringsHandler h("1.0e9"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + char* json = StrDup("{ \"gigabyte\": 1.0e9 } "); + InsituStringStream s(json); + NumbersAsStringsHandler h("1.0e9"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag|kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const char* json = "{ \"pi\": 314.159e-2 } "; + StringStream s(json); + NumbersAsStringsHandler h("314.159e-2"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + char* json = StrDup("{ \"gigabyte\": 314.159e-2 } "); + InsituStringStream s(json); + NumbersAsStringsHandler h("314.159e-2"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag|kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const char* json = "{ \"negative\": -1.54321 } "; + StringStream s(json); + NumbersAsStringsHandler h("-1.54321"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + char* json = StrDup("{ \"negative\": -1.54321 } "); + InsituStringStream s(json); + NumbersAsStringsHandler h("-1.54321"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag|kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const char* json = "{ \"pi\": 314.159e-2 } "; + std::stringstream ss(json); + IStreamWrapper s(ss); + NumbersAsStringsHandler h("314.159e-2"); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + char n1e319[321]; // '1' followed by 319 '0' + n1e319[0] = '1'; + for (int i = 1; i < 320; i++) + n1e319[i] = '0'; + n1e319[320] = '\0'; + StringStream s(n1e319); + NumbersAsStringsHandler h(n1e319); + Reader reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } +} + +struct NumbersAsStringsHandlerWChar_t { + bool Null() { return true; } + bool Bool(bool) { return true; } + bool Int(int) { return true; } + bool Uint(unsigned) { return true; } + bool Int64(int64_t) { return true; } + bool Uint64(uint64_t) { return true; } + bool Double(double) { return true; } + // 'str' is not null-terminated + bool RawNumber(const wchar_t* str, SizeType length, bool) { + EXPECT_TRUE(str != 0); + EXPECT_TRUE(expected_len_ == length); + EXPECT_TRUE(wcsncmp(str, expected_, length) == 0); + return true; + } + bool String(const wchar_t*, SizeType, bool) { return true; } + bool StartObject() { return true; } + bool Key(const wchar_t*, SizeType, bool) { return true; } + bool EndObject(SizeType) { return true; } + bool StartArray() { return true; } + bool EndArray(SizeType) { return true; } + + NumbersAsStringsHandlerWChar_t(const wchar_t* expected) + : expected_(expected) + , expected_len_(wcslen(expected)) {} + + const wchar_t* expected_; + size_t expected_len_; +}; + +TEST(Reader, NumbersAsStringsWChar_t) { + { + const wchar_t* json = L"{ \"pi\": 3.1416 } "; + GenericStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"3.1416"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + wchar_t* json = StrDup(L"{ \"pi\": 3.1416 } "); + GenericInsituStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"3.1416"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag | kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const wchar_t* json = L"{ \"gigabyte\": 1.0e9 } "; + GenericStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"1.0e9"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + wchar_t* json = StrDup(L"{ \"gigabyte\": 1.0e9 } "); + GenericInsituStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"1.0e9"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag | kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const wchar_t* json = L"{ \"pi\": 314.159e-2 } "; + GenericStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"314.159e-2"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + wchar_t* json = StrDup(L"{ \"gigabyte\": 314.159e-2 } "); + GenericInsituStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"314.159e-2"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag | kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const wchar_t* json = L"{ \"negative\": -1.54321 } "; + GenericStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"-1.54321"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + wchar_t* json = StrDup(L"{ \"negative\": -1.54321 } "); + GenericInsituStringStream<UTF16<> > s(json); + NumbersAsStringsHandlerWChar_t h(L"-1.54321"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseInsituFlag | kParseNumbersAsStringsFlag>(s, h)); + free(json); + } + { + const wchar_t* json = L"{ \"pi\": 314.159e-2 } "; + std::wstringstream ss(json); + WIStreamWrapper s(ss); + NumbersAsStringsHandlerWChar_t h(L"314.159e-2"); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } + { + wchar_t n1e319[321]; // '1' followed by 319 '0' + n1e319[0] = L'1'; + for(int i = 1; i < 320; i++) + n1e319[i] = L'0'; + n1e319[320] = L'\0'; + GenericStringStream<UTF16<> > s(n1e319); + NumbersAsStringsHandlerWChar_t h(n1e319); + GenericReader<UTF16<>, UTF16<> > reader; + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag>(s, h)); + } +} + +template <unsigned extraFlags> +void TestTrailingCommas() { + { + StringStream s("[1,2,3,]"); + ParseArrayHandler<3> h; + Reader reader; + EXPECT_TRUE(reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h)); + EXPECT_EQ(5u, h.step_); + } + { + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false," + "\"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3],}"; + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h)); + EXPECT_EQ(20u, h.step_); + } + { + // whitespace around trailing commas + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false," + "\"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3\n,\n]\n,\n} "; + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h)); + EXPECT_EQ(20u, h.step_); + } + { + // comments around trailing commas + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false, \"n\": null," + "\"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3/*test*/,/*test*/]/*test*/,/*test*/}"; + StringStream s(json); + ParseObjectHandler h; + Reader reader; + EXPECT_TRUE(reader.Parse<extraFlags|kParseTrailingCommasFlag|kParseCommentsFlag>(s, h)); + EXPECT_EQ(20u, h.step_); + } +} + +TEST(Reader, TrailingCommas) { + TestTrailingCommas<kParseNoFlags>(); +} + +TEST(Reader, TrailingCommasIterative) { + TestTrailingCommas<kParseIterativeFlag>(); +} + +template <unsigned extraFlags> +void TestMultipleTrailingCommaErrors() { + // only a single trailing comma is allowed. + { + StringStream s("[1,2,3,,]"); + ParseArrayHandler<3> h; + Reader reader; + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorValueInvalid, r.Code()); + EXPECT_EQ(7u, r.Offset()); + } + { + const char* json = "{ \"hello\" : \"world\", \"t\" : true , \"f\" : false," + "\"n\": null, \"i\":123, \"pi\": 3.1416, \"a\":[1, 2, 3,],,}"; + StringStream s(json); + ParseObjectHandler h; + Reader reader; + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorObjectMissName, r.Code()); + EXPECT_EQ(95u, r.Offset()); + } +} + +TEST(Reader, MultipleTrailingCommaErrors) { + TestMultipleTrailingCommaErrors<kParseNoFlags>(); +} + +TEST(Reader, MultipleTrailingCommaErrorsIterative) { + TestMultipleTrailingCommaErrors<kParseIterativeFlag>(); +} + +template <unsigned extraFlags> +void TestEmptyExceptForCommaErrors() { + // not allowed even with trailing commas enabled; the + // trailing comma must follow a value. + { + StringStream s("[,]"); + ParseArrayHandler<3> h; + Reader reader; + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorValueInvalid, r.Code()); + EXPECT_EQ(1u, r.Offset()); + } + { + StringStream s("{,}"); + ParseObjectHandler h; + Reader reader; + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorObjectMissName, r.Code()); + EXPECT_EQ(1u, r.Offset()); + } +} + +TEST(Reader, EmptyExceptForCommaErrors) { + TestEmptyExceptForCommaErrors<kParseNoFlags>(); +} + +TEST(Reader, EmptyExceptForCommaErrorsIterative) { + TestEmptyExceptForCommaErrors<kParseIterativeFlag>(); +} + +template <unsigned extraFlags> +void TestTrailingCommaHandlerTermination() { + { + HandlerTerminateAtEndArray h; + Reader reader; + StringStream s("[1,2,3,]"); + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(7u, r.Offset()); + } + { + HandlerTerminateAtEndObject h; + Reader reader; + StringStream s("{\"t\": true, \"f\": false,}"); + ParseResult r = reader.Parse<extraFlags|kParseTrailingCommasFlag>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorTermination, r.Code()); + EXPECT_EQ(23u, r.Offset()); + } +} + +TEST(Reader, TrailingCommaHandlerTermination) { + TestTrailingCommaHandlerTermination<kParseNoFlags>(); +} + +TEST(Reader, TrailingCommaHandlerTerminationIterative) { + TestTrailingCommaHandlerTermination<kParseIterativeFlag>(); +} + +TEST(Reader, ParseNanAndInfinity) { +#define TEST_NAN_INF(str, x) \ + { \ + { \ + StringStream s(str); \ + ParseDoubleHandler h; \ + Reader reader; \ + ASSERT_EQ(kParseErrorNone, reader.Parse<kParseNanAndInfFlag>(s, h).Code()); \ + EXPECT_EQ(1u, h.step_); \ + internal::Double e(x), a(h.actual_); \ + EXPECT_EQ(e.IsNan(), a.IsNan()); \ + EXPECT_EQ(e.IsInf(), a.IsInf()); \ + if (!e.IsNan()) \ + EXPECT_EQ(e.Sign(), a.Sign()); \ + } \ + { \ + const char* json = "{ \"naninfdouble\": " str " } "; \ + StringStream s(json); \ + NumbersAsStringsHandler h(str); \ + Reader reader; \ + EXPECT_TRUE(reader.Parse<kParseNumbersAsStringsFlag|kParseNanAndInfFlag>(s, h)); \ + } \ + { \ + char* json = StrDup("{ \"naninfdouble\": " str " } "); \ + InsituStringStream s(json); \ + NumbersAsStringsHandler h(str); \ + Reader reader; \ + EXPECT_TRUE(reader.Parse<kParseInsituFlag|kParseNumbersAsStringsFlag|kParseNanAndInfFlag>(s, h)); \ + free(json); \ + } \ + } +#define TEST_NAN_INF_ERROR(errorCode, str, errorOffset) \ + { \ + unsigned streamPos = errorOffset; \ + char buffer[1001]; \ + strncpy(buffer, str, 1000); \ + InsituStringStream s(buffer); \ + BaseReaderHandler<> h; \ + Reader reader; \ + EXPECT_FALSE(reader.Parse<kParseNanAndInfFlag>(s, h)); \ + EXPECT_EQ(errorCode, reader.GetParseErrorCode());\ + EXPECT_EQ(errorOffset, reader.GetErrorOffset());\ + EXPECT_EQ(streamPos, s.Tell());\ + } + + double nan = std::numeric_limits<double>::quiet_NaN(); + double inf = std::numeric_limits<double>::infinity(); + + TEST_NAN_INF("NaN", nan); + TEST_NAN_INF("-NaN", nan); + TEST_NAN_INF("Inf", inf); + TEST_NAN_INF("Infinity", inf); + TEST_NAN_INF("-Inf", -inf); + TEST_NAN_INF("-Infinity", -inf); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "NInf", 1u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "NaInf", 2u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "INan", 1u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "InNan", 2u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "nan", 1u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "-nan", 1u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "NAN", 1u); + TEST_NAN_INF_ERROR(kParseErrorValueInvalid, "-Infinty", 6u); + +#undef TEST_NAN_INF_ERROR +#undef TEST_NAN_INF +} + +TEST(Reader, EscapedApostrophe) { + const char json[] = " { \"foo\": \"bar\\'buzz\" } "; + + BaseReaderHandler<> h; + + { + StringStream s(json); + Reader reader; + ParseResult r = reader.Parse<kParseNoFlags>(s, h); + EXPECT_TRUE(reader.HasParseError()); + EXPECT_EQ(kParseErrorStringEscapeInvalid, r.Code()); + EXPECT_EQ(14u, r.Offset()); + } + + { + StringStream s(json); + Reader reader; + ParseResult r = reader.Parse<kParseEscapedApostropheFlag>(s, h); + EXPECT_FALSE(reader.HasParseError()); + EXPECT_EQ(kParseErrorNone, r.Code()); + EXPECT_EQ(0u, r.Offset()); + } +} + +RAPIDJSON_DIAG_POP |