From 76e017a1b8a042bcb682b68821cc33f29788cd9a Mon Sep 17 00:00:00 2001 From: Florian Frank Date: Sat, 7 Nov 2009 23:57:25 +0100 Subject: simplified build structure, tweak bs iterations --- Rakefile | 40 +- benchmarks/generator2_benchmark.rb | 8 +- benchmarks/generator_benchmark.rb | 8 +- benchmarks/parser2_benchmark.rb | 8 +- benchmarks/parser_benchmark.rb | 8 +- ext/json/ext/extconf_generator.rb | 18 + ext/json/ext/extconf_parser.rb | 14 + ext/json/ext/generator.c | 1256 +++++++++++++++++++++++ ext/json/ext/generator.h | 178 ++++ ext/json/ext/generator/extconf.rb | 18 - ext/json/ext/generator/generator.c | 1256 ----------------------- ext/json/ext/generator/generator.h | 178 ---- ext/json/ext/parser.c | 1914 ++++++++++++++++++++++++++++++++++++ ext/json/ext/parser.h | 79 ++ ext/json/ext/parser.rl | 771 +++++++++++++++ ext/json/ext/parser/extconf.rb | 14 - ext/json/ext/parser/parser.c | 1914 ------------------------------------ ext/json/ext/parser/parser.h | 79 -- ext/json/ext/parser/parser.rl | 771 --------------- 19 files changed, 4265 insertions(+), 4267 deletions(-) create mode 100644 ext/json/ext/extconf_generator.rb create mode 100644 ext/json/ext/extconf_parser.rb create mode 100644 ext/json/ext/generator.c create mode 100644 ext/json/ext/generator.h delete mode 100644 ext/json/ext/generator/extconf.rb delete mode 100644 ext/json/ext/generator/generator.c delete mode 100644 ext/json/ext/generator/generator.h create mode 100644 ext/json/ext/parser.c create mode 100644 ext/json/ext/parser.h create mode 100644 ext/json/ext/parser.rl delete mode 100644 ext/json/ext/parser/extconf.rb delete mode 100644 ext/json/ext/parser/parser.c delete mode 100644 ext/json/ext/parser/parser.h delete mode 100644 ext/json/ext/parser/parser.rl diff --git a/Rakefile b/Rakefile index eead037..9fc0e76 100644 --- a/Rakefile +++ b/Rakefile @@ -9,31 +9,29 @@ rescue LoadError puts "WARNING: rake-compiler is not installed. You will not be able to build the json gem until you install it." end -require 'rake/clean' -CLOBBER.include Dir['benchmarks/data/*.{dat,log}'] - require 'rbconfig' include Config +require 'rake/clean' +CLOBBER.include Dir['benchmarks/data/*.{dat,log}'] +CLEAN.include FileList['diagrams/*.*'], 'doc', 'coverage', 'tmp', + FileList["ext/**/{Makefile,mkmf.log}"], + FileList["{ext,lib}/**/*.{so,bundle,#{CONFIG['DLEXT']},o,obj,pdb,lib,manifest,exp,def}"] + MAKE = ENV['MAKE'] || %w[gmake make].find { |c| system(c, '-v') } PKG_NAME = 'json' PKG_TITLE = 'JSON Implementation for Ruby' PKG_VERSION = File.read('VERSION').chomp PKG_FILES = FileList["**/*"].exclude(/CVS|pkg|tmp|coverage|Makefile|\.nfs\./).exclude(/\.(so|bundle|o|#{CONFIG['DLEXT']})$/) EXT_ROOT_DIR = 'ext/json/ext' -EXT_PARSER_DIR = "#{EXT_ROOT_DIR}/parser" EXT_PARSER_DL = "#{EXT_ROOT_DIR}/parser.#{CONFIG['DLEXT']}" -EXT_PARSER_SRC = "#{EXT_PARSER_DIR}/parser.c" +EXT_PARSER_SRC = "#{EXT_ROOT_DIR}/parser.c" PKG_FILES << EXT_PARSER_SRC -EXT_GENERATOR_DIR = "#{EXT_ROOT_DIR}/generator" EXT_GENERATOR_DL = "#{EXT_ROOT_DIR}/generator.#{CONFIG['DLEXT']}" -EXT_GENERATOR_SRC = "#{EXT_GENERATOR_DIR}/generator.c" +EXT_GENERATOR_SRC = "#{EXT_ROOT_DIR}/generator.c" RAGEL_CODEGEN = %w[rlcodegen rlgen-cd ragel].find { |c| system(c, '-v') } RAGEL_DOTGEN = %w[rlgen-dot rlgen-cd ragel].find { |c| system(c, '-v') } -RAGEL_PATH = "#{EXT_PARSER_DIR}/parser.rl" -CLEAN.include FileList['diagrams/*.*'], 'doc', 'coverage', 'tmp', - FileList["ext/**/{Makefile,mkmf.log}"], - FileList["{ext,lib}/**/*.{so,bundle,#{CONFIG['DLEXT']},o,obj,pdb,lib,manifest,exp,def}"] +RAGEL_PATH = "#{EXT_ROOT_DIR}/parser.rl" def myruby(*args, &block) @myruby ||= File.join(CONFIG['bindir'], CONFIG['ruby_install_name']) @@ -75,19 +73,17 @@ desc "Compiling extension" task :compile_ext => [ EXT_PARSER_DL, EXT_GENERATOR_DL ] file EXT_PARSER_DL => EXT_PARSER_SRC do - cd EXT_PARSER_DIR do - myruby 'extconf.rb' + cd EXT_ROOT_DIR do + myruby 'extconf_parser.rb' sh MAKE end - cp "#{EXT_PARSER_DIR}/parser.#{CONFIG['DLEXT']}", EXT_ROOT_DIR end file EXT_GENERATOR_DL => EXT_GENERATOR_SRC do - cd EXT_GENERATOR_DIR do - myruby 'extconf.rb' + cd EXT_ROOT_DIR do + myruby 'extconf_generator.rb' sh MAKE end - cp "#{EXT_GENERATOR_DIR}/generator.#{CONFIG['DLEXT']}", EXT_ROOT_DIR end desc "Generate parser with ragel" @@ -98,7 +94,7 @@ task :ragel_clean do end file EXT_PARSER_SRC => RAGEL_PATH do - cd EXT_PARSER_DIR do + cd EXT_ROOT_DIR do if RAGEL_CODEGEN == 'ragel' sh "ragel parser.rl -G2 -o parser.c" else @@ -217,7 +213,7 @@ if defined?(Gem) and defined?(Rake::GemPackageTask) and defined?(Rake::Extension s.files = PKG_FILES - s.extensions = FileList['ext/**/extconf.rb'] + s.extensions = FileList['ext/**/extconf_*.rb'] s.require_path = EXT_ROOT_DIR s.require_paths << 'ext' @@ -246,19 +242,21 @@ if defined?(Gem) and defined?(Rake::GemPackageTask) and defined?(Rake::Extension Rake::ExtensionTask.new do |ext| ext.name = 'parser' + ext.config_script = 'extconf_parser.rb' ext.gem_spec = spec_ext ext.cross_compile = true ext.cross_platform = 'i386-mswin32' - ext.ext_dir = 'ext/json/ext/parser' + ext.ext_dir = 'ext/json/ext' ext.lib_dir = 'lib/json/ext' end Rake::ExtensionTask.new do |ext| ext.name = 'generator' + ext.config_script = 'extconf_generator.rb' ext.gem_spec = spec_ext ext.cross_compile = true ext.cross_platform = 'i386-mswin32' - ext.ext_dir = 'ext/json/ext/generator' + ext.ext_dir = 'ext/json/ext' ext.lib_dir = 'lib/json/ext' end end diff --git a/benchmarks/generator2_benchmark.rb b/benchmarks/generator2_benchmark.rb index f253881..7b9fa74 100755 --- a/benchmarks/generator2_benchmark.rb +++ b/benchmarks/generator2_benchmark.rb @@ -65,7 +65,7 @@ class Generator2BenchmarkExt < Bullshit::RepeatCase include JSONGeneratorCommon warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false @@ -91,7 +91,7 @@ class Generator2BenchmarkPure < Bullshit::RepeatCase include JSONGeneratorCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -116,7 +116,7 @@ class Generator2BenchmarkRails < Bullshit::RepeatCase include Generator2BenchmarkCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -147,7 +147,7 @@ class Generator2BenchmarkYajl < Bullshit::RepeatCase include Generator2BenchmarkCommon warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false diff --git a/benchmarks/generator_benchmark.rb b/benchmarks/generator_benchmark.rb index 281f443..9d6aec3 100755 --- a/benchmarks/generator_benchmark.rb +++ b/benchmarks/generator_benchmark.rb @@ -67,7 +67,7 @@ class GeneratorBenchmarkExt < Bullshit::RepeatCase include JSONGeneratorCommon warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false @@ -93,7 +93,7 @@ class GeneratorBenchmarkPure < Bullshit::RepeatCase include JSONGeneratorCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -118,7 +118,7 @@ class GeneratorBenchmarkRails < Bullshit::RepeatCase include GeneratorBenchmarkCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -149,7 +149,7 @@ class GeneratorBenchmarkYajl < Bullshit::RepeatCase include GeneratorBenchmarkCommon warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false diff --git a/benchmarks/parser2_benchmark.rb b/benchmarks/parser2_benchmark.rb index b0a5b7e..bc80772 100755 --- a/benchmarks/parser2_benchmark.rb +++ b/benchmarks/parser2_benchmark.rb @@ -39,7 +39,7 @@ class Parser2BenchmarkExt < Bullshit::RepeatCase include Parser2BenchmarkCommon warmup yes - iterations 1000 + iterations 4000 truncate_data do enabled false @@ -70,7 +70,7 @@ class Parser2BenchmarkPure < Bullshit::RepeatCase include Parser2BenchmarkCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -134,7 +134,7 @@ end class Parser2BenchmarkRails < Bullshit::RepeatCase warmup yes - iterations 1000 + iterations 500 truncate_data do alpha_level 0.05 @@ -170,7 +170,7 @@ end class Parser2BenchmarkYajl < Bullshit::RepeatCase warmup yes - iterations 1000 + iterations 4000 truncate_data do alpha_level 0.05 diff --git a/benchmarks/parser_benchmark.rb b/benchmarks/parser_benchmark.rb index 5fcef0a..acb8ea4 100755 --- a/benchmarks/parser_benchmark.rb +++ b/benchmarks/parser_benchmark.rb @@ -41,7 +41,7 @@ class ParserBenchmarkExt < Bullshit::RepeatCase include ParserBenchmarkCommon warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false @@ -72,7 +72,7 @@ class ParserBenchmarkPure < Bullshit::RepeatCase include ParserBenchmarkCommon warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -138,7 +138,7 @@ end class ParserBenchmarkRails < Bullshit::RepeatCase warmup yes - iterations 1000 + iterations 500 truncate_data do enabled false @@ -175,7 +175,7 @@ end class ParserBenchmarkYajl < Bullshit::RepeatCase warmup yes - iterations 1000 + iterations 8000 truncate_data do enabled false diff --git a/ext/json/ext/extconf_generator.rb b/ext/json/ext/extconf_generator.rb new file mode 100644 index 0000000..33a5625 --- /dev/null +++ b/ext/json/ext/extconf_generator.rb @@ -0,0 +1,18 @@ +require 'mkmf' +require 'rbconfig' + +unless $CFLAGS.gsub!(/ -O[\dsz]?/, ' -O3') + $CFLAGS << ' -O3' +end +if CONFIG['CC'] =~ /gcc/ + $CFLAGS << ' -Wall' + #$CFLAGS.gsub!(/ -O[\dsz]?/, ' -O0 -ggdb') +end +if RUBY_VERSION >= '1.9' + $CFLAGS << ' -DRUBY_19' +end + +have_header("ruby/st.h") || have_header("st.h") +have_header("ruby/re.h") || have_header("re.h") +have_header("ruby/encoding.h") +create_makefile 'generator' diff --git a/ext/json/ext/extconf_parser.rb b/ext/json/ext/extconf_parser.rb new file mode 100644 index 0000000..9662e9a --- /dev/null +++ b/ext/json/ext/extconf_parser.rb @@ -0,0 +1,14 @@ +require 'mkmf' +require 'rbconfig' + +unless $CFLAGS.gsub!(/ -O[\dsz]?/, ' -O3') + $CFLAGS << ' -O3' +end +if CONFIG['CC'] =~ /gcc/ + $CFLAGS << ' -Wall' + #$CFLAGS.gsub!(/ -O[\dsz]?/, ' -O0 -ggdb') +end + +have_header("ruby/st.h") || have_header("st.h") +have_header("re.h") +create_makefile 'parser' diff --git a/ext/json/ext/generator.c b/ext/json/ext/generator.c new file mode 100644 index 0000000..1db22a7 --- /dev/null +++ b/ext/json/ext/generator.c @@ -0,0 +1,1256 @@ +#include "generator.h" + +#ifdef HAVE_RUBY_ENCODING_H +static VALUE CEncoding_UTF_8; +static ID i_encoding, i_encode; +#endif + +static VALUE mJSON, mExt, mGenerator, cState, mGeneratorMethods, mObject, + mHash, mArray, mInteger, mFloat, mString, mString_Extend, + mTrueClass, mFalseClass, mNilClass, eGeneratorError, + eNestingError, CRegexp_MULTILINE; + +static ID i_to_s, i_to_json, i_new, i_indent, i_space, i_space_before, + i_object_nl, i_array_nl, i_max_nesting, i_allow_nan, i_ascii_only, + i_pack, i_unpack, i_create_id, i_extend; + +/* + * Copyright 2001-2004 Unicode, Inc. + * + * Disclaimer + * + * This source code is provided as is by Unicode, Inc. No claims are + * made as to fitness for any particular purpose. No warranties of any + * kind are expressed or implied. The recipient agrees to determine + * applicability of information provided. If this file has been + * purchased on magnetic or optical media from Unicode, Inc., the + * sole remedy for any claim will be exchange of defective media + * within 90 days of receipt. + * + * Limitations on Rights to Redistribute This Code + * + * Unicode, Inc. hereby grants the right to freely use the information + * supplied in this file in the creation of products supporting the + * Unicode Standard, and to make copies of this file in any form + * for internal or external distribution as long as this notice + * remains attached. + */ + +/* + * Index into the table below with the first byte of a UTF-8 sequence to + * get the number of trailing bytes that are supposed to follow it. + * Note that *legal* UTF-8 values can't have 4 or 5-bytes. The table is + * left as-is for anyone who may want to do such conversion, which was + * allowed in earlier algorithms. + */ +static const char trailingBytesForUTF8[256] = { + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, + 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, + 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5 +}; + +/* + * Magic values subtracted from a buffer value during UTF8 conversion. + * This table contains as many values as there might be trailing bytes + * in a UTF-8 sequence. + */ +static const UTF32 offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL, + 0x03C82080UL, 0xFA082080UL, 0x82082080UL }; + +/* + * Utility routine to tell whether a sequence of bytes is legal UTF-8. + * This must be called with the length pre-determined by the first byte. + * If not calling this from ConvertUTF8to*, then the length can be set by: + * length = trailingBytesForUTF8[*source]+1; + * and the sequence is illegal right away if there aren't that many bytes + * available. + * If presented with a length > 4, this returns 0. The Unicode + * definition of UTF-8 goes up to 4-byte sequences. + */ +static unsigned char isLegalUTF8(const UTF8 *source, int length) +{ + UTF8 a; + const UTF8 *srcptr = source+length; + switch (length) { + default: return 0; + /* Everything else falls through when "1"... */ + case 4: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return 0; + case 3: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return 0; + case 2: if ((a = (*--srcptr)) > 0xBF) return 0; + + switch (*source) { + /* no fall-through in this inner switch */ + case 0xE0: if (a < 0xA0) return 0; break; + case 0xED: if (a > 0x9F) return 0; break; + case 0xF0: if (a < 0x90) return 0; break; + case 0xF4: if (a > 0x8F) return 0; break; + default: if (a < 0x80) return 0; + } + + case 1: if (*source >= 0x80 && *source < 0xC2) return 0; + } + if (*source > 0xF4) return 0; + return 1; +} + +/* Escapes the UTF16 character and stores the result in the buffer buf. */ +static void unicode_escape(char *buf, UTF16 character) +{ + const char *digits = "0123456789abcdef"; + + buf[2] = digits[character >> 12]; + buf[3] = digits[(character >> 8) & 0xf]; + buf[4] = digits[(character >> 4) & 0xf]; + buf[5] = digits[character & 0xf]; +} + +/* Escapes the UTF16 character and stores the result in the buffer buf, then + * the buffer buf іs appended to the FBuffer buffer. */ +static void unicode_escape_to_buffer(FBuffer *buffer, char buf[6], UTF16 + character) +{ + unicode_escape(buf, character); + fbuffer_append(buffer, buf, 6); +} + +/* Converts string to a JSON string in FBuffer buffer, where all but the ASCII + * and control characters are JSON escaped. */ +static void convert_UTF8_to_JSON_ASCII(FBuffer *buffer, VALUE string) +{ + const UTF8 *source = (UTF8 *) RSTRING_PTR(string); + const UTF8 *sourceEnd = source + RSTRING_LEN(string); + char buf[6] = { '\\', 'u' }; + + while (source < sourceEnd) { + UTF32 ch = 0; + unsigned short extraBytesToRead = trailingBytesForUTF8[*source]; + if (source + extraBytesToRead >= sourceEnd) { + rb_raise(rb_path2class("JSON::GeneratorError"), + "partial character in source, but hit end"); + } + if (!isLegalUTF8(source, extraBytesToRead+1)) { + rb_raise(rb_path2class("JSON::GeneratorError"), + "source sequence is illegal/malformed utf-8"); + } + /* + * The cases all fall through. See "Note A" below. + */ + switch (extraBytesToRead) { + case 5: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */ + case 4: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */ + case 3: ch += *source++; ch <<= 6; + case 2: ch += *source++; ch <<= 6; + case 1: ch += *source++; ch <<= 6; + case 0: ch += *source++; + } + ch -= offsetsFromUTF8[extraBytesToRead]; + + if (ch <= UNI_MAX_BMP) { /* Target is a character <= 0xFFFF */ + /* UTF-16 surrogate values are illegal in UTF-32 */ + if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) { +#if UNI_STRICT_CONVERSION + source -= (extraBytesToRead+1); /* return to the illegal value itself */ + rb_raise(rb_path2class("JSON::GeneratorError"), + "source sequence is illegal/malformed utf-8"); +#else + unicode_escape_to_buffer(buffer, buf, UNI_REPLACEMENT_CHAR); +#endif + } else { + /* normal case */ + if (ch >= 0x20 && ch <= 0x7f) { + switch (ch) { + case '\\': + fbuffer_append(buffer, "\\\\", 2); + break; + case '"': + fbuffer_append(buffer, "\\\"", 2); + break; + default: + fbuffer_append_char(buffer, ch); + break; + } + } else { + switch (ch) { + case '\n': + fbuffer_append(buffer, "\\n", 2); + break; + case '\r': + fbuffer_append(buffer, "\\r", 2); + break; + case '\t': + fbuffer_append(buffer, "\\t", 2); + break; + case '\f': + fbuffer_append(buffer, "\\f", 2); + break; + case '\b': + fbuffer_append(buffer, "\\b", 2); + break; + default: + unicode_escape_to_buffer(buffer, buf, (UTF16) ch); + break; + } + } + } + } else if (ch > UNI_MAX_UTF16) { +#if UNI_STRICT_CONVERSION + source -= (extraBytesToRead+1); /* return to the start */ + rb_raise(rb_path2class("JSON::GeneratorError"), + "source sequence is illegal/malformed utf8"); +#else + unicode_escape_to_buffer(buffer, buf, UNI_REPLACEMENT_CHAR); +#endif + } else { + /* target is a character in range 0xFFFF - 0x10FFFF. */ + ch -= halfBase; + unicode_escape_to_buffer(buffer, buf, (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START)); + unicode_escape_to_buffer(buffer, buf, (UTF16)((ch & halfMask) + UNI_SUR_LOW_START)); + } + } +} + +/* Converts string to a JSON string in FBuffer buffer, where only the + * characters required by the JSON standard are JSON escaped. The remaining + * characters (should be UTF8) are just passed through and appended to the + * result. */ +static void convert_UTF8_to_JSON(FBuffer *buffer, VALUE string) +{ + const char *ptr = RSTRING_PTR(string), *p; + int len = RSTRING_LEN(string), start = 0, end = 0; + const char *escape = NULL; + int escape_len; + unsigned char c; + char buf[6] = { '\\', 'u' }; + + for (start = 0, end = 0; end < len;) { + p = ptr + end; + c = (unsigned char) *p; + if (c < 0x20) { + switch (c) { + case '\n': + escape = "\\n"; + escape_len = 2; + break; + case '\r': + escape = "\\r"; + escape_len = 2; + break; + case '\t': + escape = "\\t"; + escape_len = 2; + break; + case '\f': + escape = "\\f"; + escape_len = 2; + break; + case '\b': + escape = "\\b"; + escape_len = 2; + break; + default: + unicode_escape(buf, (UTF16) *p); + escape = buf; + escape_len = 6; + break; + } + } else { + switch (c) { + case '\\': + escape = "\\\\"; + escape_len = 2; + break; + case '"': + escape = "\\\""; + escape_len = 2; + break; + default: + end++; + continue; + break; + } + } + fbuffer_append(buffer, ptr + start, end - start); + fbuffer_append(buffer, escape, escape_len); + start = ++end; + escape = NULL; + } + fbuffer_append(buffer, ptr + start, end - start); +} + +/* fbuffer implementation */ + +static FBuffer *fbuffer_alloc() +{ + FBuffer *fb = ALLOC(FBuffer); + memset((void *) fb, 0, sizeof(FBuffer)); + fb->initial_length = FBUFFER_INITIAL_LENGTH; + return fb; +} + +static FBuffer *fbuffer_alloc_with_length(unsigned int initial_length) +{ + assert(initial_length > 0); + FBuffer *fb = ALLOC(FBuffer); + memset((void *) fb, 0, sizeof(FBuffer)); + fb->initial_length = initial_length; + return fb; +} + + +static void fbuffer_free(FBuffer *fb) +{ + if (fb->ptr) ruby_xfree(fb->ptr); + ruby_xfree(fb); +} + +static void fbuffer_free_only_buffer(FBuffer *fb) +{ + ruby_xfree(fb); +} + +static void fbuffer_clear(FBuffer *fb) +{ + fb->len = 0; +} + +static void fbuffer_inc_capa(FBuffer *fb, unsigned int requested) +{ + unsigned int required; + + if (!fb->ptr) { + fb->ptr = ALLOC_N(char, fb->initial_length); + fb->capa = fb->initial_length; + } + + for (required = fb->capa; requested > required - fb->len; required <<= 1); + + if (required > fb->capa) { + fb->ptr = (char *) REALLOC_N((long*) fb->ptr, char, required); + fb->capa = required; + } +} + +static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned int len) +{ + if (len > 0) { + fbuffer_inc_capa(fb, len); + memcpy(fb->ptr + fb->len, newstr, len); + fb->len += len; + } +} + +static void fbuffer_append_char(FBuffer *fb, char newchr) +{ + fbuffer_inc_capa(fb, 1); + *(fb->ptr + fb->len) = newchr; + fb->len++; +} + +/* + * Document-module: JSON::Ext::Generator + * + * This is the JSON generator implemented as a C extension. It can be + * configured to be used by setting + * + * JSON.generator = JSON::Ext::Generator + * + * with the method generator= in JSON. + * + */ + +/* + * call-seq: to_json(state = nil, depth = 0) + * + * Returns a JSON string containing a JSON object, that is generated from + * this Hash instance. + * _state_ is a JSON::State object, that can also be used to configure the + * produced JSON string output further. + * _depth_ is used to find out nesting depth, to indent accordingly. + */ +static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(state = nil, depth = 0) + * + * Returns a JSON string containing a JSON array, that is generated from + * this Array instance. + * _state_ is a JSON::State object, that can also be used to configure the + * produced JSON string output further. + * _depth_ is used to find out nesting depth, to indent accordingly. + */ +static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self) { + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(*) + * + * Returns a JSON string representation for this Integer number. + */ +static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(*) + * + * Returns a JSON string representation for this Float number. + */ +static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: String.included(modul) + * + * Extends _modul_ with the String::Extend module. + */ +static VALUE mString_included_s(VALUE self, VALUE modul) { + VALUE result = rb_funcall(modul, i_extend, 1, mString_Extend); + return result; +} + +/* + * call-seq: to_json(*) + * + * This string should be encoded with UTF-8 A call to this method + * returns a JSON string encoded with UTF16 big endian characters as + * \u????. + */ +static VALUE mString_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json_raw_object() + * + * This method creates a raw object hash, that can be nested into + * other data structures and will be generated as a raw string. This + * method should be used, if you want to convert raw strings to JSON + * instead of UTF-8 strings, e. g. binary data. + */ +static VALUE mString_to_json_raw_object(VALUE self) +{ + VALUE ary; + VALUE result = rb_hash_new(); + rb_hash_aset(result, rb_funcall(mJSON, i_create_id, 0), rb_class_name(rb_obj_class(self))); + ary = rb_funcall(self, i_unpack, 1, rb_str_new2("C*")); + rb_hash_aset(result, rb_str_new2("raw"), ary); + return result; +} + +/* + * call-seq: to_json_raw(*args) + * + * This method creates a JSON text from the result of a call to + * to_json_raw_object of this String. + */ +static VALUE mString_to_json_raw(int argc, VALUE *argv, VALUE self) +{ + VALUE obj = mString_to_json_raw_object(self); + Check_Type(obj, T_HASH); + return mHash_to_json(argc, argv, obj); +} + +/* + * call-seq: json_create(o) + * + * Raw Strings are JSON Objects (the raw bytes are stored in an array for the + * key "raw"). The Ruby String can be created by this module method. + */ +static VALUE mString_Extend_json_create(VALUE self, VALUE o) +{ + VALUE ary; + Check_Type(o, T_HASH); + ary = rb_hash_aref(o, rb_str_new2("raw")); + return rb_funcall(ary, i_pack, 1, rb_str_new2("C*")); +} + +/* + * call-seq: to_json(*) + * + * Returns a JSON string for true: 'true'. + */ +static VALUE mTrueClass_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(*) + * + * Returns a JSON string for false: 'false'. + */ +static VALUE mFalseClass_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(*) + * + */ +static VALUE mNilClass_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + rb_scan_args(argc, argv, "02", &state, &depth); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, self, depth); +} + +/* + * call-seq: to_json(*) + * + * Converts this object to a string (calling #to_s), converts + * it to a JSON string, and returns the result. This is a fallback, if no + * special method #to_json was defined for some object. + */ +static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self) +{ + VALUE state, depth; + VALUE string = rb_funcall(self, i_to_s, 0); + rb_scan_args(argc, argv, "02", &state, &depth); + Check_Type(string, T_STRING); + state = cState_from_state_s(cState, state); + return cState_partial_generate(state, string, depth); +} + +static void State_free(JSON_Generator_State *state) +{ + if (state->indent) ruby_xfree(state->indent); + if (state->space) ruby_xfree(state->space); + if (state->space_before) ruby_xfree(state->space_before); + if (state->object_nl) ruby_xfree(state->object_nl); + if (state->array_nl) ruby_xfree(state->array_nl); + if (state->array_delim) fbuffer_free(state->array_delim); + if (state->object_delim) fbuffer_free(state->object_delim); + if (state->object_delim2) fbuffer_free(state->object_delim2); + ruby_xfree(state); +} + +static JSON_Generator_State *State_allocate() +{ + JSON_Generator_State *state = ALLOC(JSON_Generator_State); + return state; +} + +static VALUE cState_s_allocate(VALUE klass) +{ + JSON_Generator_State *state = State_allocate(); + return Data_Wrap_Struct(klass, NULL, State_free, state); +} + +/* + * call-seq: configure(opts) + * + * Configure this State instance with the Hash _opts_, and return + * itself. + */ +static VALUE cState_configure(VALUE self, VALUE opts) +{ + VALUE tmp; + GET_STATE(self); + tmp = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); + if (NIL_P(tmp)) tmp = rb_convert_type(opts, T_HASH, "Hash", "to_h"); + if (NIL_P(tmp)) { + rb_raise(rb_eArgError, "opts has to be hash like or convertable into a hash"); + } + opts = tmp; + tmp = rb_hash_aref(opts, ID2SYM(i_indent)); + if (RTEST(tmp)) { + Check_Type(tmp, T_STRING); + state->indent = strdup(RSTRING_PTR(tmp)); + state->indent_len = strlen(state->indent); + } + tmp = rb_hash_aref(opts, ID2SYM(i_space)); + if (RTEST(tmp)) { + Check_Type(tmp, T_STRING); + state->space = strdup(RSTRING_PTR(tmp)); + state->space_len = strlen(state->space); + } + tmp = rb_hash_aref(opts, ID2SYM(i_space_before)); + if (RTEST(tmp)) { + Check_Type(tmp, T_STRING); + state->space_before = strdup(RSTRING_PTR(tmp)); + state->space_before_len = strlen(state->space_before); + } + tmp = rb_hash_aref(opts, ID2SYM(i_array_nl)); + if (RTEST(tmp)) { + Check_Type(tmp, T_STRING); + state->array_nl = strdup(RSTRING_PTR(tmp)); + state->array_nl_len = strlen(state->array_nl); + } + tmp = rb_hash_aref(opts, ID2SYM(i_object_nl)); + if (RTEST(tmp)) { + Check_Type(tmp, T_STRING); + state->object_nl = strdup(RSTRING_PTR(tmp)); + state->object_nl_len = strlen(state->object_nl); + } + tmp = ID2SYM(i_max_nesting); + state->max_nesting = 19; + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE max_nesting = rb_hash_aref(opts, tmp); + if (RTEST(max_nesting)) { + Check_Type(max_nesting, T_FIXNUM); + state->max_nesting = FIX2LONG(max_nesting); + } else { + state->max_nesting = 0; + } + } + tmp = rb_hash_aref(opts, ID2SYM(i_allow_nan)); + state->allow_nan = RTEST(tmp); + tmp = rb_hash_aref(opts, ID2SYM(i_ascii_only)); + state->ascii_only = RTEST(tmp); + return self; +} + +/* + * call-seq: to_h + * + * Returns the configuration instance variables as a hash, that can be + * passed to the configure method. + */ +static VALUE cState_to_h(VALUE self) +{ + VALUE result = rb_hash_new(); + GET_STATE(self); + rb_hash_aset(result, ID2SYM(i_indent), rb_str_new2(state->indent)); + rb_hash_aset(result, ID2SYM(i_space), rb_str_new2(state->space)); + rb_hash_aset(result, ID2SYM(i_space_before), rb_str_new2(state->space_before)); + rb_hash_aset(result, ID2SYM(i_object_nl), rb_str_new2(state->object_nl)); + rb_hash_aset(result, ID2SYM(i_array_nl), rb_str_new2(state->array_nl)); + rb_hash_aset(result, ID2SYM(i_allow_nan), state->allow_nan ? Qtrue : Qfalse); + rb_hash_aset(result, ID2SYM(i_ascii_only), state->ascii_only ? Qtrue : Qfalse); + rb_hash_aset(result, ID2SYM(i_max_nesting), LONG2FIX(state->max_nesting)); + return result; +} + +/* + * The fbuffer2rstring breaks encapsulation of Ruby's String datatype to avoid + * calling memcpy while creating a RString from a c string. This is rather + * hackish code, I am not sure if it's a good idea to keep it. + */ +#ifdef RUBY_19 +#define STR_NOEMBED FL_USER1 + +#define STR_SET_EMBED_LEN(str, n) do { \ + long tmp_n = (n);\ + RBASIC(str)->flags &= ~RSTRING_EMBED_LEN_MASK;\ + RBASIC(str)->flags |= (tmp_n) << RSTRING_EMBED_LEN_SHIFT;\ +} while (0) + +#define STR_SET_NOEMBED(str) do {\ + FL_SET(str, STR_NOEMBED);\ + STR_SET_EMBED_LEN(str, 0);\ +} while (0) + +static VALUE fbuffer2rstring(FBuffer *buffer) +{ + NEWOBJ(str, struct RString); + OBJSETUP(str, rb_cString, T_STRING); + + str->as.heap.ptr = FBUFFER_PTR(buffer); + str->as.heap.len = FBUFFER_LEN(buffer); + str->as.heap.aux.capa = FBUFFER_CAPA(buffer); + STR_SET_NOEMBED(str); + + return (VALUE) str; +} +#else +static VALUE fbuffer2rstring(FBuffer *buffer) +{ + NEWOBJ(str, struct RString); + OBJSETUP(str, rb_cString, T_STRING); + + str->ptr = FBUFFER_PTR(buffer); + str->len = FBUFFER_LEN(buffer); + str->aux.capa = FBUFFER_CAPA(buffer); + + return (VALUE) str; +} +#endif + +static void generate_json(FBuffer *buffer, VALUE Vstate, JSON_Generator_State *state, VALUE obj, long depth) +{ + VALUE tmp; + switch (TYPE(obj)) { + case T_HASH: + { + char *object_nl = state->object_nl; + long object_nl_len = state->object_nl_len; + char *indent = state->indent; + long indent_len = state->indent_len; + long max_nesting = state->max_nesting; + char *delim = FBUFFER_PTR(state->object_delim); + long delim_len = FBUFFER_LEN(state->object_delim); + char *delim2 = FBUFFER_PTR(state->object_delim2); + long delim2_len = FBUFFER_LEN(state->object_delim2); + int i, j; + depth++; + if (max_nesting != 0 && depth > max_nesting) { + fbuffer_free(buffer); + rb_raise(eNestingError, "nesting of %ld is too deep", depth); + } + fbuffer_append_char(buffer, '{'); + VALUE keys = rb_funcall(obj, rb_intern("keys"), 0); + VALUE key, key_to_s; + for(i = 0; i < RARRAY_LEN(keys); i++) { + if (i > 0) fbuffer_append(buffer, delim, delim_len); + if (object_nl) { + fbuffer_append(buffer, object_nl, object_nl_len); + } + if (indent) { + for (j = 0; j < depth; j++) { + fbuffer_append(buffer, indent, indent_len); + } + } + key = rb_ary_entry(keys, i); + key_to_s = rb_funcall(key, i_to_s, 0); + Check_Type(key_to_s, T_STRING); + generate_json(buffer, Vstate, state, key_to_s, depth); + fbuffer_append(buffer, delim2, delim2_len); + generate_json(buffer, Vstate, state, rb_hash_aref(obj, key), depth); + } + depth--; + if (object_nl) { + fbuffer_append(buffer, object_nl, object_nl_len); + if (indent) { + for (j = 0; j < depth; j++) { + fbuffer_append(buffer, indent, indent_len); + } + } + } + fbuffer_append_char(buffer, '}'); + } + break; + case T_ARRAY: + { + char *array_nl = state->array_nl; + long array_nl_len = state->array_nl_len; + char *indent = state->indent; + long indent_len = state->indent_len; + long max_nesting = state->max_nesting; + char *delim = FBUFFER_PTR(state->array_delim); + long delim_len = FBUFFER_LEN(state->array_delim); + int i, j; + depth++; + if (max_nesting != 0 && depth > max_nesting) { + fbuffer_free(buffer); + rb_raise(eNestingError, "nesting of %ld is too deep", depth); + } + fbuffer_append_char(buffer, '['); + if (array_nl) fbuffer_append(buffer, array_nl, array_nl_len); + for(i = 0; i < RARRAY_LEN(obj); i++) { + if (i > 0) fbuffer_append(buffer, delim, delim_len); + if (indent) { + for (j = 0; j < depth; j++) { + fbuffer_append(buffer, indent, indent_len); + } + } + generate_json(buffer, Vstate, state, rb_ary_entry(obj, i), depth); + } + depth--; + if (array_nl) { + fbuffer_append(buffer, array_nl, array_nl_len); + if (indent) { + for (j = 0; j < depth; j++) { + fbuffer_append(buffer, indent, indent_len); + } + } + } + fbuffer_append_char(buffer, ']'); + } + break; + case T_STRING: + fbuffer_append_char(buffer, '"'); +#ifdef HAVE_RUBY_ENCODING_H + obj = rb_funcall(obj, i_encode, 1, CEncoding_UTF_8); +#endif + if (state->ascii_only) { + convert_UTF8_to_JSON_ASCII(buffer, obj); + } else { + convert_UTF8_to_JSON(buffer, obj); + } + fbuffer_append_char(buffer, '"'); + break; + case T_NIL: + fbuffer_append(buffer, "null", 4); + break; + case T_FALSE: + fbuffer_append(buffer, "false", 5); + break; + case T_TRUE: + fbuffer_append(buffer, "true", 4); + break; + case T_FIXNUM: + case T_BIGNUM: + tmp = rb_funcall(obj, i_to_s, 0); + fbuffer_append(buffer, RSTRING_PAIR(tmp)); + break; + case T_FLOAT: + { + char allow_nan = state->allow_nan; + double value = RFLOAT_VALUE(obj); + tmp = rb_funcall(obj, i_to_s, 0); + if (!allow_nan) { + if (isinf(value)) { + fbuffer_free(buffer); + rb_raise(eGeneratorError, "%u: %s not allowed in JSON", __LINE__, StringValueCStr(tmp)); + } else if (isnan(value)) { + fbuffer_free(buffer); + rb_raise(eGeneratorError, "%u: %s not allowed in JSON", __LINE__, StringValueCStr(tmp)); + } + } + fbuffer_append(buffer, RSTRING_PAIR(tmp)); + } + break; + default: + if (rb_respond_to(obj, i_to_json)) { + tmp = rb_funcall(obj, i_to_json, 2, Vstate, INT2FIX(depth + 1)); + Check_Type(tmp, T_STRING); + fbuffer_append(buffer, RSTRING_PAIR(tmp)); + } else { + tmp = rb_funcall(obj, i_to_s, 0); + Check_Type(tmp, T_STRING); + generate_json(buffer, Vstate, state, tmp, depth + 1); + } + break; + } +} + +/* + * call-seq: partial_generate(obj) + * + * Generates a part of a JSON document from object +obj+ and returns the + * result. + */ +static VALUE cState_partial_generate(VALUE self, VALUE obj, VALUE depth) +{ + VALUE result; + FBuffer *buffer = fbuffer_alloc(); + GET_STATE(self); + + if (state->object_delim) { + fbuffer_clear(state->object_delim); + } else { + state->object_delim = fbuffer_alloc_with_length(16); + } + fbuffer_append_char(state->object_delim, ','); + if (state->object_delim2) { + fbuffer_clear(state->object_delim2); + } else { + state->object_delim2 = fbuffer_alloc_with_length(16); + } + fbuffer_append_char(state->object_delim2, ':'); + if (state->space) fbuffer_append(state->object_delim2, state->space, state->space_len); + + if (state->array_delim) { + fbuffer_clear(state->array_delim); + } else { + state->array_delim = fbuffer_alloc_with_length(16); + } + fbuffer_append_char(state->array_delim, ','); + if (state->array_nl) fbuffer_append(state->array_delim, state->array_nl, state->array_nl_len); + + generate_json(buffer, self, state, obj, NIL_P(depth) ? 0 : FIX2INT(depth)); + result = fbuffer2rstring(buffer); + fbuffer_free_only_buffer(buffer); + FORCE_UTF8(result); + return result; +} + +/* + * call-seq: generate(obj) + * + * Generates a valid JSON document from object +obj+ and returns the + * result. If no valid JSON document can be created this method raises a + * GeneratorError exception. + */ +static VALUE cState_generate(VALUE self, VALUE obj) +{ + VALUE result = cState_partial_generate(self, obj, Qnil); + VALUE re, args[2]; + args[0] = rb_str_new2("\\A\\s*(?:\\[.*\\]|\\{.*\\})\\s*\\Z"); + args[1] = CRegexp_MULTILINE; + re = rb_class_new_instance(2, args, rb_cRegexp); + if (NIL_P(rb_reg_match(re, result))) { + rb_raise(eGeneratorError, "only generation of JSON objects or arrays allowed"); + } + return result; +} + +/* + * call-seq: new(opts = {}) + * + * Instantiates a new State object, configured by _opts_. + * + * _opts_ can have the following keys: + * + * * *indent*: a string used to indent levels (default: ''), + * * *space*: a string that is put after, a : or , delimiter (default: ''), + * * *space_before*: a string that is put before a : pair delimiter (default: ''), + * * *object_nl*: a string that is put at the end of a JSON object (default: ''), + * * *array_nl*: a string that is put at the end of a JSON array (default: ''), + * * *allow_nan*: true if NaN, Infinity, and -Infinity should be + * generated, otherwise an exception is thrown, if these values are + * encountered. This options defaults to false. + */ +static VALUE cState_initialize(int argc, VALUE *argv, VALUE self) +{ + VALUE opts; + GET_STATE(self); + MEMZERO(state, JSON_Generator_State, 1); + state->max_nesting = 19; + rb_scan_args(argc, argv, "01", &opts); + if (!NIL_P(opts)) cState_configure(self, opts); + return self; +} + +/* + * call-seq: from_state(opts) + * + * Creates a State object from _opts_, which ought to be Hash to create a + * new State instance configured by _opts_, something else to create an + * unconfigured instance. If _opts_ is a State object, it is just returned. + */ +static VALUE cState_from_state_s(VALUE self, VALUE opts) +{ + if (rb_obj_is_kind_of(opts, self)) { + return opts; + } else if (rb_obj_is_kind_of(opts, rb_cHash)) { + return rb_funcall(self, i_new, 1, opts); + } else { + return rb_funcall(self, i_new, 0); + } +} + +/* + * call-seq: indent() + * + * This string is used to indent levels in the JSON text. + */ +static VALUE cState_indent(VALUE self) +{ + GET_STATE(self); + return state->indent ? rb_str_new2(state->indent) : rb_str_new2(""); +} + +/* + * call-seq: indent=(indent) + * + * This string is used to indent levels in the JSON text. + */ +static VALUE cState_indent_set(VALUE self, VALUE indent) +{ + GET_STATE(self); + Check_Type(indent, T_STRING); + if (RSTRING_LEN(indent) == 0) { + if (state->indent) { + ruby_xfree(state->indent); + state->indent = NULL; + } + } else { + if (state->indent) ruby_xfree(state->indent); + state->indent = strdup(RSTRING_PTR(indent)); + } + return Qnil; +} + +/* + * call-seq: space() + * + * This string is used to insert a space between the tokens in a JSON + * string. + */ +static VALUE cState_space(VALUE self) +{ + GET_STATE(self); + return state->space ? rb_str_new2(state->space) : rb_str_new2(""); +} + +/* + * call-seq: space=(space) + * + * This string is used to insert a space between the tokens in a JSON + * string. + */ +static VALUE cState_space_set(VALUE self, VALUE space) +{ + GET_STATE(self); + Check_Type(space, T_STRING); + if (RSTRING_LEN(space) == 0) { + if (state->space) { + ruby_xfree(state->space); + state->space = NULL; + } + } else { + if (state->space) ruby_xfree(state->space); + state->space = strdup(RSTRING_PTR(space)); + } + return Qnil; +} + +/* + * call-seq: space_before() + * + * This string is used to insert a space before the ':' in JSON objects. + */ +static VALUE cState_space_before(VALUE self) +{ + GET_STATE(self); + return state->space_before ? rb_str_new2(state->space_before) : rb_str_new2(""); +} + +/* + * call-seq: space_before=(space_before) + * + * This string is used to insert a space before the ':' in JSON objects. + */ +static VALUE cState_space_before_set(VALUE self, VALUE space_before) +{ + GET_STATE(self); + Check_Type(space_before, T_STRING); + if (RSTRING_LEN(space_before) == 0) { + if (state->space_before) { + ruby_xfree(state->space_before); + state->space_before = NULL; + } + } else { + if (state->space_before) ruby_xfree(state->space_before); + state->space_before = strdup(RSTRING_PTR(space_before)); + } + return Qnil; +} + +/* + * call-seq: object_nl() + * + * This string is put at the end of a line that holds a JSON object (or + * Hash). + */ +static VALUE cState_object_nl(VALUE self) +{ + GET_STATE(self); + return state->object_nl ? rb_str_new2(state->object_nl) : rb_str_new2(""); +} + +/* + * call-seq: object_nl=(object_nl) + * + * This string is put at the end of a line that holds a JSON object (or + * Hash). + */ +static VALUE cState_object_nl_set(VALUE self, VALUE object_nl) +{ + GET_STATE(self); + Check_Type(object_nl, T_STRING); + if (RSTRING_LEN(object_nl) == 0) { + if (state->object_nl) { + ruby_xfree(state->object_nl); + state->object_nl = NULL; + } + } else { + if (state->object_nl) ruby_xfree(state->object_nl); + state->object_nl = strdup(RSTRING_PTR(object_nl)); + } + return Qnil; +} + +/* + * call-seq: array_nl() + * + * This string is put at the end of a line that holds a JSON array. + */ +static VALUE cState_array_nl(VALUE self) +{ + GET_STATE(self); + return state->array_nl ? rb_str_new2(state->array_nl) : rb_str_new2(""); +} + +/* + * call-seq: array_nl=(array_nl) + * + * This string is put at the end of a line that holds a JSON array. + */ +static VALUE cState_array_nl_set(VALUE self, VALUE array_nl) +{ + GET_STATE(self); + Check_Type(array_nl, T_STRING); + if (RSTRING_LEN(array_nl) == 0) { + if (state->array_nl) { + ruby_xfree(state->array_nl); + state->array_nl = NULL; + } + } else { + if (state->array_nl) ruby_xfree(state->array_nl); + state->array_nl = strdup(RSTRING_PTR(array_nl)); + } + return Qnil; +} + +/* + * call-seq: max_nesting + * + * This integer returns the maximum level of data structure nesting in + * the generated JSON, max_nesting = 0 if no maximum is checked. + */ +static VALUE cState_max_nesting(VALUE self) +{ + GET_STATE(self); + return LONG2FIX(state->max_nesting); +} + +/* + * call-seq: max_nesting=(depth) + * + * This sets the maximum level of data structure nesting in the generated JSON + * to the integer depth, max_nesting = 0 if no maximum should be checked. + */ +static VALUE cState_max_nesting_set(VALUE self, VALUE depth) +{ + GET_STATE(self); + Check_Type(depth, T_FIXNUM); + return state->max_nesting = FIX2LONG(depth); +} + +/* + * call-seq: allow_nan? + * + * Returns true, if NaN, Infinity, and -Infinity should be generated, otherwise + * returns false. + */ +static VALUE cState_allow_nan_p(VALUE self) +{ + GET_STATE(self); + return state->allow_nan ? Qtrue : Qfalse; +} + +/* + * call-seq: ascii_only? + * + * Returns true, if NaN, Infinity, and -Infinity should be generated, otherwise + * returns false. + */ +static VALUE cState_ascii_only_p(VALUE self) +{ + GET_STATE(self); + return state->ascii_only ? Qtrue : Qfalse; +} + +/* + * + */ +void Init_generator() +{ + rb_require("json/common"); + + mJSON = rb_define_module("JSON"); + mExt = rb_define_module_under(mJSON, "Ext"); + mGenerator = rb_define_module_under(mExt, "Generator"); + + eGeneratorError = rb_path2class("JSON::GeneratorError"); + eNestingError = rb_path2class("JSON::NestingError"); + + cState = rb_define_class_under(mGenerator, "State", rb_cObject); + rb_define_alloc_func(cState, cState_s_allocate); + rb_define_singleton_method(cState, "from_state", cState_from_state_s, 1); + rb_define_method(cState, "initialize", cState_initialize, -1); + rb_define_method(cState, "indent", cState_indent, 0); + rb_define_method(cState, "indent=", cState_indent_set, 1); + rb_define_method(cState, "space", cState_space, 0); + rb_define_method(cState, "space=", cState_space_set, 1); + rb_define_method(cState, "space_before", cState_space_before, 0); + rb_define_method(cState, "space_before=", cState_space_before_set, 1); + rb_define_method(cState, "object_nl", cState_object_nl, 0); + rb_define_method(cState, "object_nl=", cState_object_nl_set, 1); + rb_define_method(cState, "array_nl", cState_array_nl, 0); + rb_define_method(cState, "array_nl=", cState_array_nl_set, 1); + rb_define_method(cState, "max_nesting", cState_max_nesting, 0); + rb_define_method(cState, "max_nesting=", cState_max_nesting_set, 1); + rb_define_method(cState, "allow_nan?", cState_allow_nan_p, 0); + rb_define_method(cState, "ascii_only?", cState_ascii_only_p, 0); + rb_define_method(cState, "configure", cState_configure, 1); + rb_define_method(cState, "to_h", cState_to_h, 0); + rb_define_method(cState, "generate", cState_generate, 1); + rb_define_method(cState, "partial_generate", cState_partial_generate, 1); + + mGeneratorMethods = rb_define_module_under(mGenerator, "GeneratorMethods"); + mObject = rb_define_module_under(mGeneratorMethods, "Object"); + rb_define_method(mObject, "to_json", mObject_to_json, -1); + mHash = rb_define_module_under(mGeneratorMethods, "Hash"); + rb_define_method(mHash, "to_json", mHash_to_json, -1); + mArray = rb_define_module_under(mGeneratorMethods, "Array"); + rb_define_method(mArray, "to_json", mArray_to_json, -1); + mInteger = rb_define_module_under(mGeneratorMethods, "Integer"); + rb_define_method(mInteger, "to_json", mInteger_to_json, -1); + mFloat = rb_define_module_under(mGeneratorMethods, "Float"); + rb_define_method(mFloat, "to_json", mFloat_to_json, -1); + mString = rb_define_module_under(mGeneratorMethods, "String"); + rb_define_singleton_method(mString, "included", mString_included_s, 1); + rb_define_method(mString, "to_json", mString_to_json, -1); + rb_define_method(mString, "to_json_raw", mString_to_json_raw, -1); + rb_define_method(mString, "to_json_raw_object", mString_to_json_raw_object, 0); + mString_Extend = rb_define_module_under(mString, "Extend"); + rb_define_method(mString_Extend, "json_create", mString_Extend_json_create, 1); + mTrueClass = rb_define_module_under(mGeneratorMethods, "TrueClass"); + rb_define_method(mTrueClass, "to_json", mTrueClass_to_json, -1); + mFalseClass = rb_define_module_under(mGeneratorMethods, "FalseClass"); + rb_define_method(mFalseClass, "to_json", mFalseClass_to_json, -1); + mNilClass = rb_define_module_under(mGeneratorMethods, "NilClass"); + rb_define_method(mNilClass, "to_json", mNilClass_to_json, -1); + + CRegexp_MULTILINE = rb_const_get(rb_cRegexp, rb_intern("MULTILINE")); + i_to_s = rb_intern("to_s"); + i_to_json = rb_intern("to_json"); + i_new = rb_intern("new"); + i_indent = rb_intern("indent"); + i_space = rb_intern("space"); + i_space_before = rb_intern("space_before"); + i_object_nl = rb_intern("object_nl"); + i_array_nl = rb_intern("array_nl"); + i_max_nesting = rb_intern("max_nesting"); + i_allow_nan = rb_intern("allow_nan"); + i_ascii_only = rb_intern("ascii_only"); + i_pack = rb_intern("pack"); + i_unpack = rb_intern("unpack"); + i_create_id = rb_intern("create_id"); + i_extend = rb_intern("extend"); +#ifdef HAVE_RUBY_ENCODING_H + CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); + i_encoding = rb_intern("encoding"); + i_encode = rb_intern("encode"); +#endif +} diff --git a/ext/json/ext/generator.h b/ext/json/ext/generator.h new file mode 100644 index 0000000..5b5a28c --- /dev/null +++ b/ext/json/ext/generator.h @@ -0,0 +1,178 @@ +#ifndef _GENERATOR_H_ +#define _GENERATOR_H_ + +#include +#include +#include + +#include "ruby.h" + +#if HAVE_RUBY_ST_H +#include "ruby/st.h" +#endif + +#if HAVE_ST_H +#include "st.h" +#endif + +#if HAVE_RUBY_RE_H +#include "ruby/re.h" +#endif + +#if HAVE_RE_H +#include "re.h" +#endif + +#ifdef HAVE_RUBY_ENCODING_H +#include "ruby/encoding.h" +#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding()) +#else +#define FORCE_UTF8(obj) +#endif + +#ifndef RHASH_TBL +#define RHASH_TBL(hsh) (RHASH(hsh)->tbl) +#endif + +#ifndef RHASH_SIZE +#define RHASH_SIZE(hsh) (RHASH(hsh)->tbl->num_entries) +#endif + +#ifndef RFLOAT_VALUE +#define RFLOAT_VALUE(val) (RFLOAT(val)->value) +#endif + +#ifndef RARRAY_PTR +#define RARRAY_PTR(ARRAY) RARRAY(ARRAY)->ptr +#endif +#ifndef RARRAY_LEN +#define RARRAY_LEN(ARRAY) RARRAY(ARRAY)->len +#endif +#ifndef RSTRING_PTR +#define RSTRING_PTR(string) RSTRING(string)->ptr +#endif +#ifndef RSTRING_LEN +#define RSTRING_LEN(string) RSTRING(string)->len +#endif + +#define RSTRING_PAIR(string) RSTRING_PTR(string), RSTRING_LEN(string) + +/* fbuffer implementation */ + +typedef struct FBufferStruct { + unsigned int initial_length; + char *ptr; + unsigned int len; + unsigned int capa; +} FBuffer; + +#define FBUFFER_INITIAL_LENGTH 4096 + +#define FBUFFER_PTR(fb) (fb->ptr) +#define FBUFFER_LEN(fb) (fb->len) +#define FBUFFER_CAPA(fb) (fb->capa) +#define FBUFFER_PAIR(fb) FBUFFER_PTR(fb), FBUFFER_LEN(fb) + +static FBuffer *fbuffer_alloc(); +static FBuffer *fbuffer_alloc_with_length(unsigned initial_length); +static void fbuffer_free(FBuffer *fb); +static void fbuffer_free_only_buffer(FBuffer *fb); +static void fbuffer_clear(FBuffer *fb); +static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned int len); +static void fbuffer_append_char(FBuffer *fb, char newchr); + +/* unicode defintions */ + +#define UNI_STRICT_CONVERSION 1 + +typedef unsigned long UTF32; /* at least 32 bits */ +typedef unsigned short UTF16; /* at least 16 bits */ +typedef unsigned char UTF8; /* typically 8 bits */ + +#define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD +#define UNI_MAX_BMP (UTF32)0x0000FFFF +#define UNI_MAX_UTF16 (UTF32)0x0010FFFF +#define UNI_MAX_UTF32 (UTF32)0x7FFFFFFF +#define UNI_MAX_LEGAL_UTF32 (UTF32)0x0010FFFF + +#define UNI_SUR_HIGH_START (UTF32)0xD800 +#define UNI_SUR_HIGH_END (UTF32)0xDBFF +#define UNI_SUR_LOW_START (UTF32)0xDC00 +#define UNI_SUR_LOW_END (UTF32)0xDFFF + +static const int halfShift = 10; /* used for shifting by 10 bits */ + +static const UTF32 halfBase = 0x0010000UL; +static const UTF32 halfMask = 0x3FFUL; + +static unsigned char isLegalUTF8(const UTF8 *source, int length); +static void unicode_escape(char *buf, UTF16 character); +static void unicode_escape_to_buffer(FBuffer *buffer, char buf[6], UTF16 character); +static void convert_UTF8_to_JSON_ASCII(FBuffer *buffer, VALUE string); +static void convert_UTF8_to_JSON(FBuffer *buffer, VALUE string); + +/* ruby api and some helpers */ + +typedef struct JSON_Generator_StateStruct { + char *indent; + long indent_len; + char *space; + long space_len; + char *space_before; + long space_before_len; + char *object_nl; + long object_nl_len; + char *array_nl; + long array_nl_len; + FBuffer *array_delim; + FBuffer *object_delim; + FBuffer *object_delim2; + long max_nesting; + char allow_nan; + char ascii_only; +} JSON_Generator_State; + +#define GET_STATE(self) \ + JSON_Generator_State *state; \ + Data_Get_Struct(self, JSON_Generator_State, state) + +static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mString_included_s(VALUE self, VALUE modul); +static VALUE mString_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mString_to_json_raw_object(VALUE self); +static VALUE mString_to_json_raw(int argc, VALUE *argv, VALUE self); +static VALUE mString_Extend_json_create(VALUE self, VALUE o); +static VALUE mTrueClass_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mFalseClass_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mNilClass_to_json(int argc, VALUE *argv, VALUE self); +static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self); +static void State_free(JSON_Generator_State *state); +static JSON_Generator_State *State_allocate(); +static VALUE cState_s_allocate(VALUE klass); +static VALUE cState_configure(VALUE self, VALUE opts); +static VALUE cState_to_h(VALUE self); +static VALUE fbuffer2rstring(FBuffer *buffer); +static void generate_json(FBuffer *buffer, VALUE Vstate, JSON_Generator_State *state, VALUE obj, long depth); +static VALUE cState_partial_generate(VALUE self, VALUE obj, VALUE depth); +static VALUE cState_generate(VALUE self, VALUE obj); +static VALUE cState_initialize(int argc, VALUE *argv, VALUE self); +static VALUE cState_from_state_s(VALUE self, VALUE opts); +static VALUE cState_indent(VALUE self); +static VALUE cState_indent_set(VALUE self, VALUE indent); +static VALUE cState_space(VALUE self); +static VALUE cState_space_set(VALUE self, VALUE space); +static VALUE cState_space_before(VALUE self); +static VALUE cState_space_before_set(VALUE self, VALUE space_before); +static VALUE cState_object_nl(VALUE self); +static VALUE cState_object_nl_set(VALUE self, VALUE object_nl); +static VALUE cState_array_nl(VALUE self); +static VALUE cState_array_nl_set(VALUE self, VALUE array_nl); +static VALUE cState_max_nesting(VALUE self); +static VALUE cState_max_nesting_set(VALUE self, VALUE depth); +static VALUE cState_allow_nan_p(VALUE self); +static VALUE cState_ascii_only_p(VALUE self); + +#endif diff --git a/ext/json/ext/generator/extconf.rb b/ext/json/ext/generator/extconf.rb deleted file mode 100644 index 33a5625..0000000 --- a/ext/json/ext/generator/extconf.rb +++ /dev/null @@ -1,18 +0,0 @@ -require 'mkmf' -require 'rbconfig' - -unless $CFLAGS.gsub!(/ -O[\dsz]?/, ' -O3') - $CFLAGS << ' -O3' -end -if CONFIG['CC'] =~ /gcc/ - $CFLAGS << ' -Wall' - #$CFLAGS.gsub!(/ -O[\dsz]?/, ' -O0 -ggdb') -end -if RUBY_VERSION >= '1.9' - $CFLAGS << ' -DRUBY_19' -end - -have_header("ruby/st.h") || have_header("st.h") -have_header("ruby/re.h") || have_header("re.h") -have_header("ruby/encoding.h") -create_makefile 'generator' diff --git a/ext/json/ext/generator/generator.c b/ext/json/ext/generator/generator.c deleted file mode 100644 index 1db22a7..0000000 --- a/ext/json/ext/generator/generator.c +++ /dev/null @@ -1,1256 +0,0 @@ -#include "generator.h" - -#ifdef HAVE_RUBY_ENCODING_H -static VALUE CEncoding_UTF_8; -static ID i_encoding, i_encode; -#endif - -static VALUE mJSON, mExt, mGenerator, cState, mGeneratorMethods, mObject, - mHash, mArray, mInteger, mFloat, mString, mString_Extend, - mTrueClass, mFalseClass, mNilClass, eGeneratorError, - eNestingError, CRegexp_MULTILINE; - -static ID i_to_s, i_to_json, i_new, i_indent, i_space, i_space_before, - i_object_nl, i_array_nl, i_max_nesting, i_allow_nan, i_ascii_only, - i_pack, i_unpack, i_create_id, i_extend; - -/* - * Copyright 2001-2004 Unicode, Inc. - * - * Disclaimer - * - * This source code is provided as is by Unicode, Inc. No claims are - * made as to fitness for any particular purpose. No warranties of any - * kind are expressed or implied. The recipient agrees to determine - * applicability of information provided. If this file has been - * purchased on magnetic or optical media from Unicode, Inc., the - * sole remedy for any claim will be exchange of defective media - * within 90 days of receipt. - * - * Limitations on Rights to Redistribute This Code - * - * Unicode, Inc. hereby grants the right to freely use the information - * supplied in this file in the creation of products supporting the - * Unicode Standard, and to make copies of this file in any form - * for internal or external distribution as long as this notice - * remains attached. - */ - -/* - * Index into the table below with the first byte of a UTF-8 sequence to - * get the number of trailing bytes that are supposed to follow it. - * Note that *legal* UTF-8 values can't have 4 or 5-bytes. The table is - * left as-is for anyone who may want to do such conversion, which was - * allowed in earlier algorithms. - */ -static const char trailingBytesForUTF8[256] = { - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0, - 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1, - 2,2,2,2,2,2,2,2,2,2,2,2,2,2,2,2, 3,3,3,3,3,3,3,3,4,4,4,4,5,5,5,5 -}; - -/* - * Magic values subtracted from a buffer value during UTF8 conversion. - * This table contains as many values as there might be trailing bytes - * in a UTF-8 sequence. - */ -static const UTF32 offsetsFromUTF8[6] = { 0x00000000UL, 0x00003080UL, 0x000E2080UL, - 0x03C82080UL, 0xFA082080UL, 0x82082080UL }; - -/* - * Utility routine to tell whether a sequence of bytes is legal UTF-8. - * This must be called with the length pre-determined by the first byte. - * If not calling this from ConvertUTF8to*, then the length can be set by: - * length = trailingBytesForUTF8[*source]+1; - * and the sequence is illegal right away if there aren't that many bytes - * available. - * If presented with a length > 4, this returns 0. The Unicode - * definition of UTF-8 goes up to 4-byte sequences. - */ -static unsigned char isLegalUTF8(const UTF8 *source, int length) -{ - UTF8 a; - const UTF8 *srcptr = source+length; - switch (length) { - default: return 0; - /* Everything else falls through when "1"... */ - case 4: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return 0; - case 3: if ((a = (*--srcptr)) < 0x80 || a > 0xBF) return 0; - case 2: if ((a = (*--srcptr)) > 0xBF) return 0; - - switch (*source) { - /* no fall-through in this inner switch */ - case 0xE0: if (a < 0xA0) return 0; break; - case 0xED: if (a > 0x9F) return 0; break; - case 0xF0: if (a < 0x90) return 0; break; - case 0xF4: if (a > 0x8F) return 0; break; - default: if (a < 0x80) return 0; - } - - case 1: if (*source >= 0x80 && *source < 0xC2) return 0; - } - if (*source > 0xF4) return 0; - return 1; -} - -/* Escapes the UTF16 character and stores the result in the buffer buf. */ -static void unicode_escape(char *buf, UTF16 character) -{ - const char *digits = "0123456789abcdef"; - - buf[2] = digits[character >> 12]; - buf[3] = digits[(character >> 8) & 0xf]; - buf[4] = digits[(character >> 4) & 0xf]; - buf[5] = digits[character & 0xf]; -} - -/* Escapes the UTF16 character and stores the result in the buffer buf, then - * the buffer buf іs appended to the FBuffer buffer. */ -static void unicode_escape_to_buffer(FBuffer *buffer, char buf[6], UTF16 - character) -{ - unicode_escape(buf, character); - fbuffer_append(buffer, buf, 6); -} - -/* Converts string to a JSON string in FBuffer buffer, where all but the ASCII - * and control characters are JSON escaped. */ -static void convert_UTF8_to_JSON_ASCII(FBuffer *buffer, VALUE string) -{ - const UTF8 *source = (UTF8 *) RSTRING_PTR(string); - const UTF8 *sourceEnd = source + RSTRING_LEN(string); - char buf[6] = { '\\', 'u' }; - - while (source < sourceEnd) { - UTF32 ch = 0; - unsigned short extraBytesToRead = trailingBytesForUTF8[*source]; - if (source + extraBytesToRead >= sourceEnd) { - rb_raise(rb_path2class("JSON::GeneratorError"), - "partial character in source, but hit end"); - } - if (!isLegalUTF8(source, extraBytesToRead+1)) { - rb_raise(rb_path2class("JSON::GeneratorError"), - "source sequence is illegal/malformed utf-8"); - } - /* - * The cases all fall through. See "Note A" below. - */ - switch (extraBytesToRead) { - case 5: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */ - case 4: ch += *source++; ch <<= 6; /* remember, illegal UTF-8 */ - case 3: ch += *source++; ch <<= 6; - case 2: ch += *source++; ch <<= 6; - case 1: ch += *source++; ch <<= 6; - case 0: ch += *source++; - } - ch -= offsetsFromUTF8[extraBytesToRead]; - - if (ch <= UNI_MAX_BMP) { /* Target is a character <= 0xFFFF */ - /* UTF-16 surrogate values are illegal in UTF-32 */ - if (ch >= UNI_SUR_HIGH_START && ch <= UNI_SUR_LOW_END) { -#if UNI_STRICT_CONVERSION - source -= (extraBytesToRead+1); /* return to the illegal value itself */ - rb_raise(rb_path2class("JSON::GeneratorError"), - "source sequence is illegal/malformed utf-8"); -#else - unicode_escape_to_buffer(buffer, buf, UNI_REPLACEMENT_CHAR); -#endif - } else { - /* normal case */ - if (ch >= 0x20 && ch <= 0x7f) { - switch (ch) { - case '\\': - fbuffer_append(buffer, "\\\\", 2); - break; - case '"': - fbuffer_append(buffer, "\\\"", 2); - break; - default: - fbuffer_append_char(buffer, ch); - break; - } - } else { - switch (ch) { - case '\n': - fbuffer_append(buffer, "\\n", 2); - break; - case '\r': - fbuffer_append(buffer, "\\r", 2); - break; - case '\t': - fbuffer_append(buffer, "\\t", 2); - break; - case '\f': - fbuffer_append(buffer, "\\f", 2); - break; - case '\b': - fbuffer_append(buffer, "\\b", 2); - break; - default: - unicode_escape_to_buffer(buffer, buf, (UTF16) ch); - break; - } - } - } - } else if (ch > UNI_MAX_UTF16) { -#if UNI_STRICT_CONVERSION - source -= (extraBytesToRead+1); /* return to the start */ - rb_raise(rb_path2class("JSON::GeneratorError"), - "source sequence is illegal/malformed utf8"); -#else - unicode_escape_to_buffer(buffer, buf, UNI_REPLACEMENT_CHAR); -#endif - } else { - /* target is a character in range 0xFFFF - 0x10FFFF. */ - ch -= halfBase; - unicode_escape_to_buffer(buffer, buf, (UTF16)((ch >> halfShift) + UNI_SUR_HIGH_START)); - unicode_escape_to_buffer(buffer, buf, (UTF16)((ch & halfMask) + UNI_SUR_LOW_START)); - } - } -} - -/* Converts string to a JSON string in FBuffer buffer, where only the - * characters required by the JSON standard are JSON escaped. The remaining - * characters (should be UTF8) are just passed through and appended to the - * result. */ -static void convert_UTF8_to_JSON(FBuffer *buffer, VALUE string) -{ - const char *ptr = RSTRING_PTR(string), *p; - int len = RSTRING_LEN(string), start = 0, end = 0; - const char *escape = NULL; - int escape_len; - unsigned char c; - char buf[6] = { '\\', 'u' }; - - for (start = 0, end = 0; end < len;) { - p = ptr + end; - c = (unsigned char) *p; - if (c < 0x20) { - switch (c) { - case '\n': - escape = "\\n"; - escape_len = 2; - break; - case '\r': - escape = "\\r"; - escape_len = 2; - break; - case '\t': - escape = "\\t"; - escape_len = 2; - break; - case '\f': - escape = "\\f"; - escape_len = 2; - break; - case '\b': - escape = "\\b"; - escape_len = 2; - break; - default: - unicode_escape(buf, (UTF16) *p); - escape = buf; - escape_len = 6; - break; - } - } else { - switch (c) { - case '\\': - escape = "\\\\"; - escape_len = 2; - break; - case '"': - escape = "\\\""; - escape_len = 2; - break; - default: - end++; - continue; - break; - } - } - fbuffer_append(buffer, ptr + start, end - start); - fbuffer_append(buffer, escape, escape_len); - start = ++end; - escape = NULL; - } - fbuffer_append(buffer, ptr + start, end - start); -} - -/* fbuffer implementation */ - -static FBuffer *fbuffer_alloc() -{ - FBuffer *fb = ALLOC(FBuffer); - memset((void *) fb, 0, sizeof(FBuffer)); - fb->initial_length = FBUFFER_INITIAL_LENGTH; - return fb; -} - -static FBuffer *fbuffer_alloc_with_length(unsigned int initial_length) -{ - assert(initial_length > 0); - FBuffer *fb = ALLOC(FBuffer); - memset((void *) fb, 0, sizeof(FBuffer)); - fb->initial_length = initial_length; - return fb; -} - - -static void fbuffer_free(FBuffer *fb) -{ - if (fb->ptr) ruby_xfree(fb->ptr); - ruby_xfree(fb); -} - -static void fbuffer_free_only_buffer(FBuffer *fb) -{ - ruby_xfree(fb); -} - -static void fbuffer_clear(FBuffer *fb) -{ - fb->len = 0; -} - -static void fbuffer_inc_capa(FBuffer *fb, unsigned int requested) -{ - unsigned int required; - - if (!fb->ptr) { - fb->ptr = ALLOC_N(char, fb->initial_length); - fb->capa = fb->initial_length; - } - - for (required = fb->capa; requested > required - fb->len; required <<= 1); - - if (required > fb->capa) { - fb->ptr = (char *) REALLOC_N((long*) fb->ptr, char, required); - fb->capa = required; - } -} - -static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned int len) -{ - if (len > 0) { - fbuffer_inc_capa(fb, len); - memcpy(fb->ptr + fb->len, newstr, len); - fb->len += len; - } -} - -static void fbuffer_append_char(FBuffer *fb, char newchr) -{ - fbuffer_inc_capa(fb, 1); - *(fb->ptr + fb->len) = newchr; - fb->len++; -} - -/* - * Document-module: JSON::Ext::Generator - * - * This is the JSON generator implemented as a C extension. It can be - * configured to be used by setting - * - * JSON.generator = JSON::Ext::Generator - * - * with the method generator= in JSON. - * - */ - -/* - * call-seq: to_json(state = nil, depth = 0) - * - * Returns a JSON string containing a JSON object, that is generated from - * this Hash instance. - * _state_ is a JSON::State object, that can also be used to configure the - * produced JSON string output further. - * _depth_ is used to find out nesting depth, to indent accordingly. - */ -static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(state = nil, depth = 0) - * - * Returns a JSON string containing a JSON array, that is generated from - * this Array instance. - * _state_ is a JSON::State object, that can also be used to configure the - * produced JSON string output further. - * _depth_ is used to find out nesting depth, to indent accordingly. - */ -static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self) { - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(*) - * - * Returns a JSON string representation for this Integer number. - */ -static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(*) - * - * Returns a JSON string representation for this Float number. - */ -static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: String.included(modul) - * - * Extends _modul_ with the String::Extend module. - */ -static VALUE mString_included_s(VALUE self, VALUE modul) { - VALUE result = rb_funcall(modul, i_extend, 1, mString_Extend); - return result; -} - -/* - * call-seq: to_json(*) - * - * This string should be encoded with UTF-8 A call to this method - * returns a JSON string encoded with UTF16 big endian characters as - * \u????. - */ -static VALUE mString_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json_raw_object() - * - * This method creates a raw object hash, that can be nested into - * other data structures and will be generated as a raw string. This - * method should be used, if you want to convert raw strings to JSON - * instead of UTF-8 strings, e. g. binary data. - */ -static VALUE mString_to_json_raw_object(VALUE self) -{ - VALUE ary; - VALUE result = rb_hash_new(); - rb_hash_aset(result, rb_funcall(mJSON, i_create_id, 0), rb_class_name(rb_obj_class(self))); - ary = rb_funcall(self, i_unpack, 1, rb_str_new2("C*")); - rb_hash_aset(result, rb_str_new2("raw"), ary); - return result; -} - -/* - * call-seq: to_json_raw(*args) - * - * This method creates a JSON text from the result of a call to - * to_json_raw_object of this String. - */ -static VALUE mString_to_json_raw(int argc, VALUE *argv, VALUE self) -{ - VALUE obj = mString_to_json_raw_object(self); - Check_Type(obj, T_HASH); - return mHash_to_json(argc, argv, obj); -} - -/* - * call-seq: json_create(o) - * - * Raw Strings are JSON Objects (the raw bytes are stored in an array for the - * key "raw"). The Ruby String can be created by this module method. - */ -static VALUE mString_Extend_json_create(VALUE self, VALUE o) -{ - VALUE ary; - Check_Type(o, T_HASH); - ary = rb_hash_aref(o, rb_str_new2("raw")); - return rb_funcall(ary, i_pack, 1, rb_str_new2("C*")); -} - -/* - * call-seq: to_json(*) - * - * Returns a JSON string for true: 'true'. - */ -static VALUE mTrueClass_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(*) - * - * Returns a JSON string for false: 'false'. - */ -static VALUE mFalseClass_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(*) - * - */ -static VALUE mNilClass_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - rb_scan_args(argc, argv, "02", &state, &depth); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, self, depth); -} - -/* - * call-seq: to_json(*) - * - * Converts this object to a string (calling #to_s), converts - * it to a JSON string, and returns the result. This is a fallback, if no - * special method #to_json was defined for some object. - */ -static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self) -{ - VALUE state, depth; - VALUE string = rb_funcall(self, i_to_s, 0); - rb_scan_args(argc, argv, "02", &state, &depth); - Check_Type(string, T_STRING); - state = cState_from_state_s(cState, state); - return cState_partial_generate(state, string, depth); -} - -static void State_free(JSON_Generator_State *state) -{ - if (state->indent) ruby_xfree(state->indent); - if (state->space) ruby_xfree(state->space); - if (state->space_before) ruby_xfree(state->space_before); - if (state->object_nl) ruby_xfree(state->object_nl); - if (state->array_nl) ruby_xfree(state->array_nl); - if (state->array_delim) fbuffer_free(state->array_delim); - if (state->object_delim) fbuffer_free(state->object_delim); - if (state->object_delim2) fbuffer_free(state->object_delim2); - ruby_xfree(state); -} - -static JSON_Generator_State *State_allocate() -{ - JSON_Generator_State *state = ALLOC(JSON_Generator_State); - return state; -} - -static VALUE cState_s_allocate(VALUE klass) -{ - JSON_Generator_State *state = State_allocate(); - return Data_Wrap_Struct(klass, NULL, State_free, state); -} - -/* - * call-seq: configure(opts) - * - * Configure this State instance with the Hash _opts_, and return - * itself. - */ -static VALUE cState_configure(VALUE self, VALUE opts) -{ - VALUE tmp; - GET_STATE(self); - tmp = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); - if (NIL_P(tmp)) tmp = rb_convert_type(opts, T_HASH, "Hash", "to_h"); - if (NIL_P(tmp)) { - rb_raise(rb_eArgError, "opts has to be hash like or convertable into a hash"); - } - opts = tmp; - tmp = rb_hash_aref(opts, ID2SYM(i_indent)); - if (RTEST(tmp)) { - Check_Type(tmp, T_STRING); - state->indent = strdup(RSTRING_PTR(tmp)); - state->indent_len = strlen(state->indent); - } - tmp = rb_hash_aref(opts, ID2SYM(i_space)); - if (RTEST(tmp)) { - Check_Type(tmp, T_STRING); - state->space = strdup(RSTRING_PTR(tmp)); - state->space_len = strlen(state->space); - } - tmp = rb_hash_aref(opts, ID2SYM(i_space_before)); - if (RTEST(tmp)) { - Check_Type(tmp, T_STRING); - state->space_before = strdup(RSTRING_PTR(tmp)); - state->space_before_len = strlen(state->space_before); - } - tmp = rb_hash_aref(opts, ID2SYM(i_array_nl)); - if (RTEST(tmp)) { - Check_Type(tmp, T_STRING); - state->array_nl = strdup(RSTRING_PTR(tmp)); - state->array_nl_len = strlen(state->array_nl); - } - tmp = rb_hash_aref(opts, ID2SYM(i_object_nl)); - if (RTEST(tmp)) { - Check_Type(tmp, T_STRING); - state->object_nl = strdup(RSTRING_PTR(tmp)); - state->object_nl_len = strlen(state->object_nl); - } - tmp = ID2SYM(i_max_nesting); - state->max_nesting = 19; - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE max_nesting = rb_hash_aref(opts, tmp); - if (RTEST(max_nesting)) { - Check_Type(max_nesting, T_FIXNUM); - state->max_nesting = FIX2LONG(max_nesting); - } else { - state->max_nesting = 0; - } - } - tmp = rb_hash_aref(opts, ID2SYM(i_allow_nan)); - state->allow_nan = RTEST(tmp); - tmp = rb_hash_aref(opts, ID2SYM(i_ascii_only)); - state->ascii_only = RTEST(tmp); - return self; -} - -/* - * call-seq: to_h - * - * Returns the configuration instance variables as a hash, that can be - * passed to the configure method. - */ -static VALUE cState_to_h(VALUE self) -{ - VALUE result = rb_hash_new(); - GET_STATE(self); - rb_hash_aset(result, ID2SYM(i_indent), rb_str_new2(state->indent)); - rb_hash_aset(result, ID2SYM(i_space), rb_str_new2(state->space)); - rb_hash_aset(result, ID2SYM(i_space_before), rb_str_new2(state->space_before)); - rb_hash_aset(result, ID2SYM(i_object_nl), rb_str_new2(state->object_nl)); - rb_hash_aset(result, ID2SYM(i_array_nl), rb_str_new2(state->array_nl)); - rb_hash_aset(result, ID2SYM(i_allow_nan), state->allow_nan ? Qtrue : Qfalse); - rb_hash_aset(result, ID2SYM(i_ascii_only), state->ascii_only ? Qtrue : Qfalse); - rb_hash_aset(result, ID2SYM(i_max_nesting), LONG2FIX(state->max_nesting)); - return result; -} - -/* - * The fbuffer2rstring breaks encapsulation of Ruby's String datatype to avoid - * calling memcpy while creating a RString from a c string. This is rather - * hackish code, I am not sure if it's a good idea to keep it. - */ -#ifdef RUBY_19 -#define STR_NOEMBED FL_USER1 - -#define STR_SET_EMBED_LEN(str, n) do { \ - long tmp_n = (n);\ - RBASIC(str)->flags &= ~RSTRING_EMBED_LEN_MASK;\ - RBASIC(str)->flags |= (tmp_n) << RSTRING_EMBED_LEN_SHIFT;\ -} while (0) - -#define STR_SET_NOEMBED(str) do {\ - FL_SET(str, STR_NOEMBED);\ - STR_SET_EMBED_LEN(str, 0);\ -} while (0) - -static VALUE fbuffer2rstring(FBuffer *buffer) -{ - NEWOBJ(str, struct RString); - OBJSETUP(str, rb_cString, T_STRING); - - str->as.heap.ptr = FBUFFER_PTR(buffer); - str->as.heap.len = FBUFFER_LEN(buffer); - str->as.heap.aux.capa = FBUFFER_CAPA(buffer); - STR_SET_NOEMBED(str); - - return (VALUE) str; -} -#else -static VALUE fbuffer2rstring(FBuffer *buffer) -{ - NEWOBJ(str, struct RString); - OBJSETUP(str, rb_cString, T_STRING); - - str->ptr = FBUFFER_PTR(buffer); - str->len = FBUFFER_LEN(buffer); - str->aux.capa = FBUFFER_CAPA(buffer); - - return (VALUE) str; -} -#endif - -static void generate_json(FBuffer *buffer, VALUE Vstate, JSON_Generator_State *state, VALUE obj, long depth) -{ - VALUE tmp; - switch (TYPE(obj)) { - case T_HASH: - { - char *object_nl = state->object_nl; - long object_nl_len = state->object_nl_len; - char *indent = state->indent; - long indent_len = state->indent_len; - long max_nesting = state->max_nesting; - char *delim = FBUFFER_PTR(state->object_delim); - long delim_len = FBUFFER_LEN(state->object_delim); - char *delim2 = FBUFFER_PTR(state->object_delim2); - long delim2_len = FBUFFER_LEN(state->object_delim2); - int i, j; - depth++; - if (max_nesting != 0 && depth > max_nesting) { - fbuffer_free(buffer); - rb_raise(eNestingError, "nesting of %ld is too deep", depth); - } - fbuffer_append_char(buffer, '{'); - VALUE keys = rb_funcall(obj, rb_intern("keys"), 0); - VALUE key, key_to_s; - for(i = 0; i < RARRAY_LEN(keys); i++) { - if (i > 0) fbuffer_append(buffer, delim, delim_len); - if (object_nl) { - fbuffer_append(buffer, object_nl, object_nl_len); - } - if (indent) { - for (j = 0; j < depth; j++) { - fbuffer_append(buffer, indent, indent_len); - } - } - key = rb_ary_entry(keys, i); - key_to_s = rb_funcall(key, i_to_s, 0); - Check_Type(key_to_s, T_STRING); - generate_json(buffer, Vstate, state, key_to_s, depth); - fbuffer_append(buffer, delim2, delim2_len); - generate_json(buffer, Vstate, state, rb_hash_aref(obj, key), depth); - } - depth--; - if (object_nl) { - fbuffer_append(buffer, object_nl, object_nl_len); - if (indent) { - for (j = 0; j < depth; j++) { - fbuffer_append(buffer, indent, indent_len); - } - } - } - fbuffer_append_char(buffer, '}'); - } - break; - case T_ARRAY: - { - char *array_nl = state->array_nl; - long array_nl_len = state->array_nl_len; - char *indent = state->indent; - long indent_len = state->indent_len; - long max_nesting = state->max_nesting; - char *delim = FBUFFER_PTR(state->array_delim); - long delim_len = FBUFFER_LEN(state->array_delim); - int i, j; - depth++; - if (max_nesting != 0 && depth > max_nesting) { - fbuffer_free(buffer); - rb_raise(eNestingError, "nesting of %ld is too deep", depth); - } - fbuffer_append_char(buffer, '['); - if (array_nl) fbuffer_append(buffer, array_nl, array_nl_len); - for(i = 0; i < RARRAY_LEN(obj); i++) { - if (i > 0) fbuffer_append(buffer, delim, delim_len); - if (indent) { - for (j = 0; j < depth; j++) { - fbuffer_append(buffer, indent, indent_len); - } - } - generate_json(buffer, Vstate, state, rb_ary_entry(obj, i), depth); - } - depth--; - if (array_nl) { - fbuffer_append(buffer, array_nl, array_nl_len); - if (indent) { - for (j = 0; j < depth; j++) { - fbuffer_append(buffer, indent, indent_len); - } - } - } - fbuffer_append_char(buffer, ']'); - } - break; - case T_STRING: - fbuffer_append_char(buffer, '"'); -#ifdef HAVE_RUBY_ENCODING_H - obj = rb_funcall(obj, i_encode, 1, CEncoding_UTF_8); -#endif - if (state->ascii_only) { - convert_UTF8_to_JSON_ASCII(buffer, obj); - } else { - convert_UTF8_to_JSON(buffer, obj); - } - fbuffer_append_char(buffer, '"'); - break; - case T_NIL: - fbuffer_append(buffer, "null", 4); - break; - case T_FALSE: - fbuffer_append(buffer, "false", 5); - break; - case T_TRUE: - fbuffer_append(buffer, "true", 4); - break; - case T_FIXNUM: - case T_BIGNUM: - tmp = rb_funcall(obj, i_to_s, 0); - fbuffer_append(buffer, RSTRING_PAIR(tmp)); - break; - case T_FLOAT: - { - char allow_nan = state->allow_nan; - double value = RFLOAT_VALUE(obj); - tmp = rb_funcall(obj, i_to_s, 0); - if (!allow_nan) { - if (isinf(value)) { - fbuffer_free(buffer); - rb_raise(eGeneratorError, "%u: %s not allowed in JSON", __LINE__, StringValueCStr(tmp)); - } else if (isnan(value)) { - fbuffer_free(buffer); - rb_raise(eGeneratorError, "%u: %s not allowed in JSON", __LINE__, StringValueCStr(tmp)); - } - } - fbuffer_append(buffer, RSTRING_PAIR(tmp)); - } - break; - default: - if (rb_respond_to(obj, i_to_json)) { - tmp = rb_funcall(obj, i_to_json, 2, Vstate, INT2FIX(depth + 1)); - Check_Type(tmp, T_STRING); - fbuffer_append(buffer, RSTRING_PAIR(tmp)); - } else { - tmp = rb_funcall(obj, i_to_s, 0); - Check_Type(tmp, T_STRING); - generate_json(buffer, Vstate, state, tmp, depth + 1); - } - break; - } -} - -/* - * call-seq: partial_generate(obj) - * - * Generates a part of a JSON document from object +obj+ and returns the - * result. - */ -static VALUE cState_partial_generate(VALUE self, VALUE obj, VALUE depth) -{ - VALUE result; - FBuffer *buffer = fbuffer_alloc(); - GET_STATE(self); - - if (state->object_delim) { - fbuffer_clear(state->object_delim); - } else { - state->object_delim = fbuffer_alloc_with_length(16); - } - fbuffer_append_char(state->object_delim, ','); - if (state->object_delim2) { - fbuffer_clear(state->object_delim2); - } else { - state->object_delim2 = fbuffer_alloc_with_length(16); - } - fbuffer_append_char(state->object_delim2, ':'); - if (state->space) fbuffer_append(state->object_delim2, state->space, state->space_len); - - if (state->array_delim) { - fbuffer_clear(state->array_delim); - } else { - state->array_delim = fbuffer_alloc_with_length(16); - } - fbuffer_append_char(state->array_delim, ','); - if (state->array_nl) fbuffer_append(state->array_delim, state->array_nl, state->array_nl_len); - - generate_json(buffer, self, state, obj, NIL_P(depth) ? 0 : FIX2INT(depth)); - result = fbuffer2rstring(buffer); - fbuffer_free_only_buffer(buffer); - FORCE_UTF8(result); - return result; -} - -/* - * call-seq: generate(obj) - * - * Generates a valid JSON document from object +obj+ and returns the - * result. If no valid JSON document can be created this method raises a - * GeneratorError exception. - */ -static VALUE cState_generate(VALUE self, VALUE obj) -{ - VALUE result = cState_partial_generate(self, obj, Qnil); - VALUE re, args[2]; - args[0] = rb_str_new2("\\A\\s*(?:\\[.*\\]|\\{.*\\})\\s*\\Z"); - args[1] = CRegexp_MULTILINE; - re = rb_class_new_instance(2, args, rb_cRegexp); - if (NIL_P(rb_reg_match(re, result))) { - rb_raise(eGeneratorError, "only generation of JSON objects or arrays allowed"); - } - return result; -} - -/* - * call-seq: new(opts = {}) - * - * Instantiates a new State object, configured by _opts_. - * - * _opts_ can have the following keys: - * - * * *indent*: a string used to indent levels (default: ''), - * * *space*: a string that is put after, a : or , delimiter (default: ''), - * * *space_before*: a string that is put before a : pair delimiter (default: ''), - * * *object_nl*: a string that is put at the end of a JSON object (default: ''), - * * *array_nl*: a string that is put at the end of a JSON array (default: ''), - * * *allow_nan*: true if NaN, Infinity, and -Infinity should be - * generated, otherwise an exception is thrown, if these values are - * encountered. This options defaults to false. - */ -static VALUE cState_initialize(int argc, VALUE *argv, VALUE self) -{ - VALUE opts; - GET_STATE(self); - MEMZERO(state, JSON_Generator_State, 1); - state->max_nesting = 19; - rb_scan_args(argc, argv, "01", &opts); - if (!NIL_P(opts)) cState_configure(self, opts); - return self; -} - -/* - * call-seq: from_state(opts) - * - * Creates a State object from _opts_, which ought to be Hash to create a - * new State instance configured by _opts_, something else to create an - * unconfigured instance. If _opts_ is a State object, it is just returned. - */ -static VALUE cState_from_state_s(VALUE self, VALUE opts) -{ - if (rb_obj_is_kind_of(opts, self)) { - return opts; - } else if (rb_obj_is_kind_of(opts, rb_cHash)) { - return rb_funcall(self, i_new, 1, opts); - } else { - return rb_funcall(self, i_new, 0); - } -} - -/* - * call-seq: indent() - * - * This string is used to indent levels in the JSON text. - */ -static VALUE cState_indent(VALUE self) -{ - GET_STATE(self); - return state->indent ? rb_str_new2(state->indent) : rb_str_new2(""); -} - -/* - * call-seq: indent=(indent) - * - * This string is used to indent levels in the JSON text. - */ -static VALUE cState_indent_set(VALUE self, VALUE indent) -{ - GET_STATE(self); - Check_Type(indent, T_STRING); - if (RSTRING_LEN(indent) == 0) { - if (state->indent) { - ruby_xfree(state->indent); - state->indent = NULL; - } - } else { - if (state->indent) ruby_xfree(state->indent); - state->indent = strdup(RSTRING_PTR(indent)); - } - return Qnil; -} - -/* - * call-seq: space() - * - * This string is used to insert a space between the tokens in a JSON - * string. - */ -static VALUE cState_space(VALUE self) -{ - GET_STATE(self); - return state->space ? rb_str_new2(state->space) : rb_str_new2(""); -} - -/* - * call-seq: space=(space) - * - * This string is used to insert a space between the tokens in a JSON - * string. - */ -static VALUE cState_space_set(VALUE self, VALUE space) -{ - GET_STATE(self); - Check_Type(space, T_STRING); - if (RSTRING_LEN(space) == 0) { - if (state->space) { - ruby_xfree(state->space); - state->space = NULL; - } - } else { - if (state->space) ruby_xfree(state->space); - state->space = strdup(RSTRING_PTR(space)); - } - return Qnil; -} - -/* - * call-seq: space_before() - * - * This string is used to insert a space before the ':' in JSON objects. - */ -static VALUE cState_space_before(VALUE self) -{ - GET_STATE(self); - return state->space_before ? rb_str_new2(state->space_before) : rb_str_new2(""); -} - -/* - * call-seq: space_before=(space_before) - * - * This string is used to insert a space before the ':' in JSON objects. - */ -static VALUE cState_space_before_set(VALUE self, VALUE space_before) -{ - GET_STATE(self); - Check_Type(space_before, T_STRING); - if (RSTRING_LEN(space_before) == 0) { - if (state->space_before) { - ruby_xfree(state->space_before); - state->space_before = NULL; - } - } else { - if (state->space_before) ruby_xfree(state->space_before); - state->space_before = strdup(RSTRING_PTR(space_before)); - } - return Qnil; -} - -/* - * call-seq: object_nl() - * - * This string is put at the end of a line that holds a JSON object (or - * Hash). - */ -static VALUE cState_object_nl(VALUE self) -{ - GET_STATE(self); - return state->object_nl ? rb_str_new2(state->object_nl) : rb_str_new2(""); -} - -/* - * call-seq: object_nl=(object_nl) - * - * This string is put at the end of a line that holds a JSON object (or - * Hash). - */ -static VALUE cState_object_nl_set(VALUE self, VALUE object_nl) -{ - GET_STATE(self); - Check_Type(object_nl, T_STRING); - if (RSTRING_LEN(object_nl) == 0) { - if (state->object_nl) { - ruby_xfree(state->object_nl); - state->object_nl = NULL; - } - } else { - if (state->object_nl) ruby_xfree(state->object_nl); - state->object_nl = strdup(RSTRING_PTR(object_nl)); - } - return Qnil; -} - -/* - * call-seq: array_nl() - * - * This string is put at the end of a line that holds a JSON array. - */ -static VALUE cState_array_nl(VALUE self) -{ - GET_STATE(self); - return state->array_nl ? rb_str_new2(state->array_nl) : rb_str_new2(""); -} - -/* - * call-seq: array_nl=(array_nl) - * - * This string is put at the end of a line that holds a JSON array. - */ -static VALUE cState_array_nl_set(VALUE self, VALUE array_nl) -{ - GET_STATE(self); - Check_Type(array_nl, T_STRING); - if (RSTRING_LEN(array_nl) == 0) { - if (state->array_nl) { - ruby_xfree(state->array_nl); - state->array_nl = NULL; - } - } else { - if (state->array_nl) ruby_xfree(state->array_nl); - state->array_nl = strdup(RSTRING_PTR(array_nl)); - } - return Qnil; -} - -/* - * call-seq: max_nesting - * - * This integer returns the maximum level of data structure nesting in - * the generated JSON, max_nesting = 0 if no maximum is checked. - */ -static VALUE cState_max_nesting(VALUE self) -{ - GET_STATE(self); - return LONG2FIX(state->max_nesting); -} - -/* - * call-seq: max_nesting=(depth) - * - * This sets the maximum level of data structure nesting in the generated JSON - * to the integer depth, max_nesting = 0 if no maximum should be checked. - */ -static VALUE cState_max_nesting_set(VALUE self, VALUE depth) -{ - GET_STATE(self); - Check_Type(depth, T_FIXNUM); - return state->max_nesting = FIX2LONG(depth); -} - -/* - * call-seq: allow_nan? - * - * Returns true, if NaN, Infinity, and -Infinity should be generated, otherwise - * returns false. - */ -static VALUE cState_allow_nan_p(VALUE self) -{ - GET_STATE(self); - return state->allow_nan ? Qtrue : Qfalse; -} - -/* - * call-seq: ascii_only? - * - * Returns true, if NaN, Infinity, and -Infinity should be generated, otherwise - * returns false. - */ -static VALUE cState_ascii_only_p(VALUE self) -{ - GET_STATE(self); - return state->ascii_only ? Qtrue : Qfalse; -} - -/* - * - */ -void Init_generator() -{ - rb_require("json/common"); - - mJSON = rb_define_module("JSON"); - mExt = rb_define_module_under(mJSON, "Ext"); - mGenerator = rb_define_module_under(mExt, "Generator"); - - eGeneratorError = rb_path2class("JSON::GeneratorError"); - eNestingError = rb_path2class("JSON::NestingError"); - - cState = rb_define_class_under(mGenerator, "State", rb_cObject); - rb_define_alloc_func(cState, cState_s_allocate); - rb_define_singleton_method(cState, "from_state", cState_from_state_s, 1); - rb_define_method(cState, "initialize", cState_initialize, -1); - rb_define_method(cState, "indent", cState_indent, 0); - rb_define_method(cState, "indent=", cState_indent_set, 1); - rb_define_method(cState, "space", cState_space, 0); - rb_define_method(cState, "space=", cState_space_set, 1); - rb_define_method(cState, "space_before", cState_space_before, 0); - rb_define_method(cState, "space_before=", cState_space_before_set, 1); - rb_define_method(cState, "object_nl", cState_object_nl, 0); - rb_define_method(cState, "object_nl=", cState_object_nl_set, 1); - rb_define_method(cState, "array_nl", cState_array_nl, 0); - rb_define_method(cState, "array_nl=", cState_array_nl_set, 1); - rb_define_method(cState, "max_nesting", cState_max_nesting, 0); - rb_define_method(cState, "max_nesting=", cState_max_nesting_set, 1); - rb_define_method(cState, "allow_nan?", cState_allow_nan_p, 0); - rb_define_method(cState, "ascii_only?", cState_ascii_only_p, 0); - rb_define_method(cState, "configure", cState_configure, 1); - rb_define_method(cState, "to_h", cState_to_h, 0); - rb_define_method(cState, "generate", cState_generate, 1); - rb_define_method(cState, "partial_generate", cState_partial_generate, 1); - - mGeneratorMethods = rb_define_module_under(mGenerator, "GeneratorMethods"); - mObject = rb_define_module_under(mGeneratorMethods, "Object"); - rb_define_method(mObject, "to_json", mObject_to_json, -1); - mHash = rb_define_module_under(mGeneratorMethods, "Hash"); - rb_define_method(mHash, "to_json", mHash_to_json, -1); - mArray = rb_define_module_under(mGeneratorMethods, "Array"); - rb_define_method(mArray, "to_json", mArray_to_json, -1); - mInteger = rb_define_module_under(mGeneratorMethods, "Integer"); - rb_define_method(mInteger, "to_json", mInteger_to_json, -1); - mFloat = rb_define_module_under(mGeneratorMethods, "Float"); - rb_define_method(mFloat, "to_json", mFloat_to_json, -1); - mString = rb_define_module_under(mGeneratorMethods, "String"); - rb_define_singleton_method(mString, "included", mString_included_s, 1); - rb_define_method(mString, "to_json", mString_to_json, -1); - rb_define_method(mString, "to_json_raw", mString_to_json_raw, -1); - rb_define_method(mString, "to_json_raw_object", mString_to_json_raw_object, 0); - mString_Extend = rb_define_module_under(mString, "Extend"); - rb_define_method(mString_Extend, "json_create", mString_Extend_json_create, 1); - mTrueClass = rb_define_module_under(mGeneratorMethods, "TrueClass"); - rb_define_method(mTrueClass, "to_json", mTrueClass_to_json, -1); - mFalseClass = rb_define_module_under(mGeneratorMethods, "FalseClass"); - rb_define_method(mFalseClass, "to_json", mFalseClass_to_json, -1); - mNilClass = rb_define_module_under(mGeneratorMethods, "NilClass"); - rb_define_method(mNilClass, "to_json", mNilClass_to_json, -1); - - CRegexp_MULTILINE = rb_const_get(rb_cRegexp, rb_intern("MULTILINE")); - i_to_s = rb_intern("to_s"); - i_to_json = rb_intern("to_json"); - i_new = rb_intern("new"); - i_indent = rb_intern("indent"); - i_space = rb_intern("space"); - i_space_before = rb_intern("space_before"); - i_object_nl = rb_intern("object_nl"); - i_array_nl = rb_intern("array_nl"); - i_max_nesting = rb_intern("max_nesting"); - i_allow_nan = rb_intern("allow_nan"); - i_ascii_only = rb_intern("ascii_only"); - i_pack = rb_intern("pack"); - i_unpack = rb_intern("unpack"); - i_create_id = rb_intern("create_id"); - i_extend = rb_intern("extend"); -#ifdef HAVE_RUBY_ENCODING_H - CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); - i_encoding = rb_intern("encoding"); - i_encode = rb_intern("encode"); -#endif -} diff --git a/ext/json/ext/generator/generator.h b/ext/json/ext/generator/generator.h deleted file mode 100644 index 5b5a28c..0000000 --- a/ext/json/ext/generator/generator.h +++ /dev/null @@ -1,178 +0,0 @@ -#ifndef _GENERATOR_H_ -#define _GENERATOR_H_ - -#include -#include -#include - -#include "ruby.h" - -#if HAVE_RUBY_ST_H -#include "ruby/st.h" -#endif - -#if HAVE_ST_H -#include "st.h" -#endif - -#if HAVE_RUBY_RE_H -#include "ruby/re.h" -#endif - -#if HAVE_RE_H -#include "re.h" -#endif - -#ifdef HAVE_RUBY_ENCODING_H -#include "ruby/encoding.h" -#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding()) -#else -#define FORCE_UTF8(obj) -#endif - -#ifndef RHASH_TBL -#define RHASH_TBL(hsh) (RHASH(hsh)->tbl) -#endif - -#ifndef RHASH_SIZE -#define RHASH_SIZE(hsh) (RHASH(hsh)->tbl->num_entries) -#endif - -#ifndef RFLOAT_VALUE -#define RFLOAT_VALUE(val) (RFLOAT(val)->value) -#endif - -#ifndef RARRAY_PTR -#define RARRAY_PTR(ARRAY) RARRAY(ARRAY)->ptr -#endif -#ifndef RARRAY_LEN -#define RARRAY_LEN(ARRAY) RARRAY(ARRAY)->len -#endif -#ifndef RSTRING_PTR -#define RSTRING_PTR(string) RSTRING(string)->ptr -#endif -#ifndef RSTRING_LEN -#define RSTRING_LEN(string) RSTRING(string)->len -#endif - -#define RSTRING_PAIR(string) RSTRING_PTR(string), RSTRING_LEN(string) - -/* fbuffer implementation */ - -typedef struct FBufferStruct { - unsigned int initial_length; - char *ptr; - unsigned int len; - unsigned int capa; -} FBuffer; - -#define FBUFFER_INITIAL_LENGTH 4096 - -#define FBUFFER_PTR(fb) (fb->ptr) -#define FBUFFER_LEN(fb) (fb->len) -#define FBUFFER_CAPA(fb) (fb->capa) -#define FBUFFER_PAIR(fb) FBUFFER_PTR(fb), FBUFFER_LEN(fb) - -static FBuffer *fbuffer_alloc(); -static FBuffer *fbuffer_alloc_with_length(unsigned initial_length); -static void fbuffer_free(FBuffer *fb); -static void fbuffer_free_only_buffer(FBuffer *fb); -static void fbuffer_clear(FBuffer *fb); -static void fbuffer_append(FBuffer *fb, const char *newstr, unsigned int len); -static void fbuffer_append_char(FBuffer *fb, char newchr); - -/* unicode defintions */ - -#define UNI_STRICT_CONVERSION 1 - -typedef unsigned long UTF32; /* at least 32 bits */ -typedef unsigned short UTF16; /* at least 16 bits */ -typedef unsigned char UTF8; /* typically 8 bits */ - -#define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD -#define UNI_MAX_BMP (UTF32)0x0000FFFF -#define UNI_MAX_UTF16 (UTF32)0x0010FFFF -#define UNI_MAX_UTF32 (UTF32)0x7FFFFFFF -#define UNI_MAX_LEGAL_UTF32 (UTF32)0x0010FFFF - -#define UNI_SUR_HIGH_START (UTF32)0xD800 -#define UNI_SUR_HIGH_END (UTF32)0xDBFF -#define UNI_SUR_LOW_START (UTF32)0xDC00 -#define UNI_SUR_LOW_END (UTF32)0xDFFF - -static const int halfShift = 10; /* used for shifting by 10 bits */ - -static const UTF32 halfBase = 0x0010000UL; -static const UTF32 halfMask = 0x3FFUL; - -static unsigned char isLegalUTF8(const UTF8 *source, int length); -static void unicode_escape(char *buf, UTF16 character); -static void unicode_escape_to_buffer(FBuffer *buffer, char buf[6], UTF16 character); -static void convert_UTF8_to_JSON_ASCII(FBuffer *buffer, VALUE string); -static void convert_UTF8_to_JSON(FBuffer *buffer, VALUE string); - -/* ruby api and some helpers */ - -typedef struct JSON_Generator_StateStruct { - char *indent; - long indent_len; - char *space; - long space_len; - char *space_before; - long space_before_len; - char *object_nl; - long object_nl_len; - char *array_nl; - long array_nl_len; - FBuffer *array_delim; - FBuffer *object_delim; - FBuffer *object_delim2; - long max_nesting; - char allow_nan; - char ascii_only; -} JSON_Generator_State; - -#define GET_STATE(self) \ - JSON_Generator_State *state; \ - Data_Get_Struct(self, JSON_Generator_State, state) - -static VALUE mHash_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mArray_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mInteger_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mFloat_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mString_included_s(VALUE self, VALUE modul); -static VALUE mString_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mString_to_json_raw_object(VALUE self); -static VALUE mString_to_json_raw(int argc, VALUE *argv, VALUE self); -static VALUE mString_Extend_json_create(VALUE self, VALUE o); -static VALUE mTrueClass_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mFalseClass_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mNilClass_to_json(int argc, VALUE *argv, VALUE self); -static VALUE mObject_to_json(int argc, VALUE *argv, VALUE self); -static void State_free(JSON_Generator_State *state); -static JSON_Generator_State *State_allocate(); -static VALUE cState_s_allocate(VALUE klass); -static VALUE cState_configure(VALUE self, VALUE opts); -static VALUE cState_to_h(VALUE self); -static VALUE fbuffer2rstring(FBuffer *buffer); -static void generate_json(FBuffer *buffer, VALUE Vstate, JSON_Generator_State *state, VALUE obj, long depth); -static VALUE cState_partial_generate(VALUE self, VALUE obj, VALUE depth); -static VALUE cState_generate(VALUE self, VALUE obj); -static VALUE cState_initialize(int argc, VALUE *argv, VALUE self); -static VALUE cState_from_state_s(VALUE self, VALUE opts); -static VALUE cState_indent(VALUE self); -static VALUE cState_indent_set(VALUE self, VALUE indent); -static VALUE cState_space(VALUE self); -static VALUE cState_space_set(VALUE self, VALUE space); -static VALUE cState_space_before(VALUE self); -static VALUE cState_space_before_set(VALUE self, VALUE space_before); -static VALUE cState_object_nl(VALUE self); -static VALUE cState_object_nl_set(VALUE self, VALUE object_nl); -static VALUE cState_array_nl(VALUE self); -static VALUE cState_array_nl_set(VALUE self, VALUE array_nl); -static VALUE cState_max_nesting(VALUE self); -static VALUE cState_max_nesting_set(VALUE self, VALUE depth); -static VALUE cState_allow_nan_p(VALUE self); -static VALUE cState_ascii_only_p(VALUE self); - -#endif diff --git a/ext/json/ext/parser.c b/ext/json/ext/parser.c new file mode 100644 index 0000000..0122571 --- /dev/null +++ b/ext/json/ext/parser.c @@ -0,0 +1,1914 @@ + +#line 1 "parser.rl" +#include "parser.h" + +/* unicode */ + +static const char digit_values[256] = { + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, + -1, -1, -1, -1, -1, -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1 +}; + +static UTF32 unescape_unicode(const unsigned char *p) +{ + char b; + UTF32 result = 0; + b = digit_values[p[0]]; + if (b < 0) return UNI_REPLACEMENT_CHAR; + result = (result << 4) | b; + b = digit_values[p[1]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + b = digit_values[p[2]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + b = digit_values[p[3]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + return result; +} + +static int convert_UTF32_to_UTF8(char *buf, UTF32 ch) +{ + int len = 1; + if (ch <= 0x7F) { + buf[0] = (char) ch; + } else if (ch <= 0x07FF) { + buf[0] = (char) ((ch >> 6) | 0xC0); + buf[1] = (char) ((ch & 0x3F) | 0x80); + len++; + } else if (ch <= 0xFFFF) { + buf[0] = (char) ((ch >> 12) | 0xE0); + buf[1] = (char) (((ch >> 6) & 0x3F) | 0x80); + buf[2] = (char) ((ch & 0x3F) | 0x80); + len += 2; + } else if (ch <= 0x1fffff) { + buf[0] =(char) ((ch >> 18) | 0xF0); + buf[1] =(char) (((ch >> 12) & 0x3F) | 0x80); + buf[2] =(char) (((ch >> 6) & 0x3F) | 0x80); + buf[3] =(char) ((ch & 0x3F) | 0x80); + len += 3; + } else { + buf[0] = '?'; + } + return len; +} + + +#ifdef HAVE_RUBY_ENCODING_H +static VALUE CEncoding_ASCII_8BIT, CEncoding_UTF_8, CEncoding_UTF_16BE, + CEncoding_UTF_16LE, CEncoding_UTF_32BE, CEncoding_UTF_32LE; +static ID i_encoding, i_encode, i_encode_bang, i_force_encoding; +#else +static ID i_iconv; +#endif + +static VALUE mJSON, mExt, cParser, eParserError, eNestingError; +static VALUE CNaN, CInfinity, CMinusInfinity; + +static ID i_json_creatable_p, i_json_create, i_create_id, i_create_additions, + i_chr, i_max_nesting, i_allow_nan, i_object_class, i_array_class; + + +#line 108 "parser.rl" + + + +#line 90 "parser.c" +static const int JSON_object_start = 1; +static const int JSON_object_first_final = 27; +static const int JSON_object_error = 0; + +static const int JSON_object_en_main = 1; + + +#line 141 "parser.rl" + + +static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + VALUE last_name = Qnil; + VALUE object_class = json->object_class; + + if (json->max_nesting && json->current_nesting > json->max_nesting) { + rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); + } + + *result = NIL_P(object_class) ? rb_hash_new() : rb_class_new_instance(0, 0, object_class); + + +#line 114 "parser.c" + { + cs = JSON_object_start; + } + +#line 156 "parser.rl" + +#line 121 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + if ( (*p) == 123 ) + goto st2; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + switch( (*p) ) { + case 13: goto st2; + case 32: goto st2; + case 34: goto tr2; + case 47: goto st23; + case 125: goto tr4; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st2; + goto st0; +tr2: +#line 127 "parser.rl" + { + char *np = JSON_parse_string(json, p, pe, &last_name); + if (np == NULL) { p--; {p++; cs = 3; goto _out;} } else {p = (( np))-1;} + } + goto st3; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: +#line 159 "parser.c" + switch( (*p) ) { + case 13: goto st3; + case 32: goto st3; + case 47: goto st4; + case 58: goto st8; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st3; + goto st0; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + switch( (*p) ) { + case 42: goto st5; + case 47: goto st7; + } + goto st0; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + if ( (*p) == 42 ) + goto st6; + goto st5; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + switch( (*p) ) { + case 42: goto st6; + case 47: goto st3; + } + goto st5; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( (*p) == 10 ) + goto st3; + goto st7; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + switch( (*p) ) { + case 13: goto st8; + case 32: goto st8; + case 34: goto tr11; + case 45: goto tr11; + case 47: goto st19; + case 73: goto tr11; + case 78: goto tr11; + case 91: goto tr11; + case 102: goto tr11; + case 110: goto tr11; + case 116: goto tr11; + case 123: goto tr11; + } + if ( (*p) > 10 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr11; + } else if ( (*p) >= 9 ) + goto st8; + goto st0; +tr11: +#line 116 "parser.rl" + { + VALUE v = Qnil; + char *np = JSON_parse_value(json, p, pe, &v); + if (np == NULL) { + p--; {p++; cs = 9; goto _out;} + } else { + rb_hash_aset(*result, last_name, v); + {p = (( np))-1;} + } + } + goto st9; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: +#line 242 "parser.c" + switch( (*p) ) { + case 13: goto st9; + case 32: goto st9; + case 44: goto st10; + case 47: goto st15; + case 125: goto tr4; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st9; + goto st0; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: + switch( (*p) ) { + case 13: goto st10; + case 32: goto st10; + case 34: goto tr2; + case 47: goto st11; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st10; + goto st0; +st11: + if ( ++p == pe ) + goto _test_eof11; +case 11: + switch( (*p) ) { + case 42: goto st12; + case 47: goto st14; + } + goto st0; +st12: + if ( ++p == pe ) + goto _test_eof12; +case 12: + if ( (*p) == 42 ) + goto st13; + goto st12; +st13: + if ( ++p == pe ) + goto _test_eof13; +case 13: + switch( (*p) ) { + case 42: goto st13; + case 47: goto st10; + } + goto st12; +st14: + if ( ++p == pe ) + goto _test_eof14; +case 14: + if ( (*p) == 10 ) + goto st10; + goto st14; +st15: + if ( ++p == pe ) + goto _test_eof15; +case 15: + switch( (*p) ) { + case 42: goto st16; + case 47: goto st18; + } + goto st0; +st16: + if ( ++p == pe ) + goto _test_eof16; +case 16: + if ( (*p) == 42 ) + goto st17; + goto st16; +st17: + if ( ++p == pe ) + goto _test_eof17; +case 17: + switch( (*p) ) { + case 42: goto st17; + case 47: goto st9; + } + goto st16; +st18: + if ( ++p == pe ) + goto _test_eof18; +case 18: + if ( (*p) == 10 ) + goto st9; + goto st18; +tr4: +#line 132 "parser.rl" + { p--; {p++; cs = 27; goto _out;} } + goto st27; +st27: + if ( ++p == pe ) + goto _test_eof27; +case 27: +#line 338 "parser.c" + goto st0; +st19: + if ( ++p == pe ) + goto _test_eof19; +case 19: + switch( (*p) ) { + case 42: goto st20; + case 47: goto st22; + } + goto st0; +st20: + if ( ++p == pe ) + goto _test_eof20; +case 20: + if ( (*p) == 42 ) + goto st21; + goto st20; +st21: + if ( ++p == pe ) + goto _test_eof21; +case 21: + switch( (*p) ) { + case 42: goto st21; + case 47: goto st8; + } + goto st20; +st22: + if ( ++p == pe ) + goto _test_eof22; +case 22: + if ( (*p) == 10 ) + goto st8; + goto st22; +st23: + if ( ++p == pe ) + goto _test_eof23; +case 23: + switch( (*p) ) { + case 42: goto st24; + case 47: goto st26; + } + goto st0; +st24: + if ( ++p == pe ) + goto _test_eof24; +case 24: + if ( (*p) == 42 ) + goto st25; + goto st24; +st25: + if ( ++p == pe ) + goto _test_eof25; +case 25: + switch( (*p) ) { + case 42: goto st25; + case 47: goto st2; + } + goto st24; +st26: + if ( ++p == pe ) + goto _test_eof26; +case 26: + if ( (*p) == 10 ) + goto st2; + goto st26; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof11: cs = 11; goto _test_eof; + _test_eof12: cs = 12; goto _test_eof; + _test_eof13: cs = 13; goto _test_eof; + _test_eof14: cs = 14; goto _test_eof; + _test_eof15: cs = 15; goto _test_eof; + _test_eof16: cs = 16; goto _test_eof; + _test_eof17: cs = 17; goto _test_eof; + _test_eof18: cs = 18; goto _test_eof; + _test_eof27: cs = 27; goto _test_eof; + _test_eof19: cs = 19; goto _test_eof; + _test_eof20: cs = 20; goto _test_eof; + _test_eof21: cs = 21; goto _test_eof; + _test_eof22: cs = 22; goto _test_eof; + _test_eof23: cs = 23; goto _test_eof; + _test_eof24: cs = 24; goto _test_eof; + _test_eof25: cs = 25; goto _test_eof; + _test_eof26: cs = 26; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 157 "parser.rl" + + if (cs >= JSON_object_first_final) { + if (RTEST(json->create_id)) { + VALUE klassname = rb_hash_aref(*result, json->create_id); + if (!NIL_P(klassname)) { + VALUE klass = rb_path2class(StringValueCStr(klassname)); + if RTEST(rb_funcall(klass, i_json_creatable_p, 0)) { + *result = rb_funcall(klass, i_json_create, 1, *result); + } + } + } + return p + 1; + } else { + return NULL; + } +} + + +#line 455 "parser.c" +static const int JSON_value_start = 1; +static const int JSON_value_first_final = 21; +static const int JSON_value_error = 0; + +static const int JSON_value_en_main = 1; + + +#line 255 "parser.rl" + + +static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + +#line 471 "parser.c" + { + cs = JSON_value_start; + } + +#line 262 "parser.rl" + +#line 478 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + switch( (*p) ) { + case 34: goto tr0; + case 45: goto tr2; + case 73: goto st2; + case 78: goto st9; + case 91: goto tr5; + case 102: goto st11; + case 110: goto st15; + case 116: goto st18; + case 123: goto tr9; + } + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr2; + goto st0; +st0: +cs = 0; + goto _out; +tr0: +#line 203 "parser.rl" + { + char *np = JSON_parse_string(json, p, pe, result); + if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} + } + goto st21; +tr2: +#line 208 "parser.rl" + { + char *np; + if(pe > p + 9 && !strncmp(MinusInfinity, p, 9)) { + if (json->allow_nan) { + *result = CMinusInfinity; + {p = (( p + 10))-1;} + p--; {p++; cs = 21; goto _out;} + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } + } + np = JSON_parse_float(json, p, pe, result); + if (np != NULL) {p = (( np))-1;} + np = JSON_parse_integer(json, p, pe, result); + if (np != NULL) {p = (( np))-1;} + p--; {p++; cs = 21; goto _out;} + } + goto st21; +tr5: +#line 226 "parser.rl" + { + char *np; + json->current_nesting++; + np = JSON_parse_array(json, p, pe, result); + json->current_nesting--; + if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} + } + goto st21; +tr9: +#line 234 "parser.rl" + { + char *np; + json->current_nesting++; + np = JSON_parse_object(json, p, pe, result); + json->current_nesting--; + if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} + } + goto st21; +tr16: +#line 196 "parser.rl" + { + if (json->allow_nan) { + *result = CInfinity; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 8); + } + } + goto st21; +tr18: +#line 189 "parser.rl" + { + if (json->allow_nan) { + *result = CNaN; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 2); + } + } + goto st21; +tr22: +#line 183 "parser.rl" + { + *result = Qfalse; + } + goto st21; +tr25: +#line 180 "parser.rl" + { + *result = Qnil; + } + goto st21; +tr28: +#line 186 "parser.rl" + { + *result = Qtrue; + } + goto st21; +st21: + if ( ++p == pe ) + goto _test_eof21; +case 21: +#line 242 "parser.rl" + { p--; {p++; cs = 21; goto _out;} } +#line 593 "parser.c" + goto st0; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + if ( (*p) == 110 ) + goto st3; + goto st0; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: + if ( (*p) == 102 ) + goto st4; + goto st0; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + if ( (*p) == 105 ) + goto st5; + goto st0; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + if ( (*p) == 110 ) + goto st6; + goto st0; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + if ( (*p) == 105 ) + goto st7; + goto st0; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( (*p) == 116 ) + goto st8; + goto st0; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + if ( (*p) == 121 ) + goto tr16; + goto st0; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: + if ( (*p) == 97 ) + goto st10; + goto st0; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: + if ( (*p) == 78 ) + goto tr18; + goto st0; +st11: + if ( ++p == pe ) + goto _test_eof11; +case 11: + if ( (*p) == 97 ) + goto st12; + goto st0; +st12: + if ( ++p == pe ) + goto _test_eof12; +case 12: + if ( (*p) == 108 ) + goto st13; + goto st0; +st13: + if ( ++p == pe ) + goto _test_eof13; +case 13: + if ( (*p) == 115 ) + goto st14; + goto st0; +st14: + if ( ++p == pe ) + goto _test_eof14; +case 14: + if ( (*p) == 101 ) + goto tr22; + goto st0; +st15: + if ( ++p == pe ) + goto _test_eof15; +case 15: + if ( (*p) == 117 ) + goto st16; + goto st0; +st16: + if ( ++p == pe ) + goto _test_eof16; +case 16: + if ( (*p) == 108 ) + goto st17; + goto st0; +st17: + if ( ++p == pe ) + goto _test_eof17; +case 17: + if ( (*p) == 108 ) + goto tr25; + goto st0; +st18: + if ( ++p == pe ) + goto _test_eof18; +case 18: + if ( (*p) == 114 ) + goto st19; + goto st0; +st19: + if ( ++p == pe ) + goto _test_eof19; +case 19: + if ( (*p) == 117 ) + goto st20; + goto st0; +st20: + if ( ++p == pe ) + goto _test_eof20; +case 20: + if ( (*p) == 101 ) + goto tr28; + goto st0; + } + _test_eof21: cs = 21; goto _test_eof; + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof11: cs = 11; goto _test_eof; + _test_eof12: cs = 12; goto _test_eof; + _test_eof13: cs = 13; goto _test_eof; + _test_eof14: cs = 14; goto _test_eof; + _test_eof15: cs = 15; goto _test_eof; + _test_eof16: cs = 16; goto _test_eof; + _test_eof17: cs = 17; goto _test_eof; + _test_eof18: cs = 18; goto _test_eof; + _test_eof19: cs = 19; goto _test_eof; + _test_eof20: cs = 20; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 263 "parser.rl" + + if (cs >= JSON_value_first_final) { + return p; + } else { + return NULL; + } +} + + +#line 764 "parser.c" +static const int JSON_integer_start = 1; +static const int JSON_integer_first_final = 5; +static const int JSON_integer_error = 0; + +static const int JSON_integer_en_main = 1; + + +#line 279 "parser.rl" + + +static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + +#line 780 "parser.c" + { + cs = JSON_integer_start; + } + +#line 286 "parser.rl" + json->memo = p; + +#line 788 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + switch( (*p) ) { + case 45: goto st2; + case 48: goto st3; + } + if ( 49 <= (*p) && (*p) <= 57 ) + goto st4; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + if ( (*p) == 48 ) + goto st3; + if ( 49 <= (*p) && (*p) <= 57 ) + goto st4; + goto st0; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st0; + goto tr4; +tr4: +#line 276 "parser.rl" + { p--; {p++; cs = 5; goto _out;} } + goto st5; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: +#line 829 "parser.c" + goto st0; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st4; + goto tr4; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 288 "parser.rl" + + if (cs >= JSON_integer_first_final) { + long len = p - json->memo; + *result = rb_Integer(rb_str_new(json->memo, len)); + return p + 1; + } else { + return NULL; + } +} + + +#line 860 "parser.c" +static const int JSON_float_start = 1; +static const int JSON_float_first_final = 10; +static const int JSON_float_error = 0; + +static const int JSON_float_en_main = 1; + + +#line 310 "parser.rl" + + +static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + +#line 876 "parser.c" + { + cs = JSON_float_start; + } + +#line 317 "parser.rl" + json->memo = p; + +#line 884 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + switch( (*p) ) { + case 45: goto st2; + case 48: goto st3; + } + if ( 49 <= (*p) && (*p) <= 57 ) + goto st9; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + if ( (*p) == 48 ) + goto st3; + if ( 49 <= (*p) && (*p) <= 57 ) + goto st9; + goto st0; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: + switch( (*p) ) { + case 46: goto st4; + case 69: goto st6; + case 101: goto st6; + } + goto st0; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st5; + goto st0; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + switch( (*p) ) { + case 69: goto st6; + case 101: goto st6; + } + if ( (*p) > 46 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st5; + } else if ( (*p) >= 45 ) + goto st0; + goto tr7; +tr7: +#line 304 "parser.rl" + { p--; {p++; cs = 10; goto _out;} } + goto st10; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: +#line 949 "parser.c" + goto st0; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + switch( (*p) ) { + case 43: goto st7; + case 45: goto st7; + } + if ( 48 <= (*p) && (*p) <= 57 ) + goto st8; + goto st0; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( 48 <= (*p) && (*p) <= 57 ) + goto st8; + goto st0; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + switch( (*p) ) { + case 69: goto st0; + case 101: goto st0; + } + if ( (*p) > 46 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st8; + } else if ( (*p) >= 45 ) + goto st0; + goto tr7; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: + switch( (*p) ) { + case 46: goto st4; + case 69: goto st6; + case 101: goto st6; + } + if ( 48 <= (*p) && (*p) <= 57 ) + goto st9; + goto st0; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 319 "parser.rl" + + if (cs >= JSON_float_first_final) { + long len = p - json->memo; + *result = rb_Float(rb_str_new(json->memo, len)); + return p + 1; + } else { + return NULL; + } +} + + + +#line 1023 "parser.c" +static const int JSON_array_start = 1; +static const int JSON_array_first_final = 17; +static const int JSON_array_error = 0; + +static const int JSON_array_en_main = 1; + + +#line 355 "parser.rl" + + +static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + VALUE array_class = json->array_class; + + if (json->max_nesting && json->current_nesting > json->max_nesting) { + rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); + } + *result = NIL_P(array_class) ? rb_ary_new() : rb_class_new_instance(0, 0, array_class); + + +#line 1045 "parser.c" + { + cs = JSON_array_start; + } + +#line 368 "parser.rl" + +#line 1052 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + if ( (*p) == 91 ) + goto st2; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + switch( (*p) ) { + case 13: goto st2; + case 32: goto st2; + case 34: goto tr2; + case 45: goto tr2; + case 47: goto st13; + case 73: goto tr2; + case 78: goto tr2; + case 91: goto tr2; + case 93: goto tr4; + case 102: goto tr2; + case 110: goto tr2; + case 116: goto tr2; + case 123: goto tr2; + } + if ( (*p) > 10 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr2; + } else if ( (*p) >= 9 ) + goto st2; + goto st0; +tr2: +#line 336 "parser.rl" + { + VALUE v = Qnil; + char *np = JSON_parse_value(json, p, pe, &v); + if (np == NULL) { + p--; {p++; cs = 3; goto _out;} + } else { + rb_ary_push(*result, v); + {p = (( np))-1;} + } + } + goto st3; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: +#line 1107 "parser.c" + switch( (*p) ) { + case 13: goto st3; + case 32: goto st3; + case 44: goto st4; + case 47: goto st9; + case 93: goto tr4; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st3; + goto st0; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + switch( (*p) ) { + case 13: goto st4; + case 32: goto st4; + case 34: goto tr2; + case 45: goto tr2; + case 47: goto st5; + case 73: goto tr2; + case 78: goto tr2; + case 91: goto tr2; + case 102: goto tr2; + case 110: goto tr2; + case 116: goto tr2; + case 123: goto tr2; + } + if ( (*p) > 10 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto tr2; + } else if ( (*p) >= 9 ) + goto st4; + goto st0; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + switch( (*p) ) { + case 42: goto st6; + case 47: goto st8; + } + goto st0; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + if ( (*p) == 42 ) + goto st7; + goto st6; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + switch( (*p) ) { + case 42: goto st7; + case 47: goto st4; + } + goto st6; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + if ( (*p) == 10 ) + goto st4; + goto st8; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: + switch( (*p) ) { + case 42: goto st10; + case 47: goto st12; + } + goto st0; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: + if ( (*p) == 42 ) + goto st11; + goto st10; +st11: + if ( ++p == pe ) + goto _test_eof11; +case 11: + switch( (*p) ) { + case 42: goto st11; + case 47: goto st3; + } + goto st10; +st12: + if ( ++p == pe ) + goto _test_eof12; +case 12: + if ( (*p) == 10 ) + goto st3; + goto st12; +tr4: +#line 347 "parser.rl" + { p--; {p++; cs = 17; goto _out;} } + goto st17; +st17: + if ( ++p == pe ) + goto _test_eof17; +case 17: +#line 1214 "parser.c" + goto st0; +st13: + if ( ++p == pe ) + goto _test_eof13; +case 13: + switch( (*p) ) { + case 42: goto st14; + case 47: goto st16; + } + goto st0; +st14: + if ( ++p == pe ) + goto _test_eof14; +case 14: + if ( (*p) == 42 ) + goto st15; + goto st14; +st15: + if ( ++p == pe ) + goto _test_eof15; +case 15: + switch( (*p) ) { + case 42: goto st15; + case 47: goto st2; + } + goto st14; +st16: + if ( ++p == pe ) + goto _test_eof16; +case 16: + if ( (*p) == 10 ) + goto st2; + goto st16; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof11: cs = 11; goto _test_eof; + _test_eof12: cs = 12; goto _test_eof; + _test_eof17: cs = 17; goto _test_eof; + _test_eof13: cs = 13; goto _test_eof; + _test_eof14: cs = 14; goto _test_eof; + _test_eof15: cs = 15; goto _test_eof; + _test_eof16: cs = 16; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 369 "parser.rl" + + if(cs >= JSON_array_first_final) { + return p + 1; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } +} + +static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd) +{ + char *p = string, *pe = string, *unescape; + int unescape_len; + + while (pe < stringEnd) { + if (*pe == '\\') { + unescape = (char *) "?"; + unescape_len = 1; + if (pe > p) rb_str_buf_cat(result, p, pe - p); + switch (*++pe) { + case 'n': + unescape = (char *) "\n"; + break; + case 'r': + unescape = (char *) "\r"; + break; + case 't': + unescape = (char *) "\t"; + break; + case '"': + unescape = (char *) "\""; + break; + case '\\': + unescape = (char *) "\\"; + break; + case 'b': + unescape = (char *) "\b"; + break; + case 'f': + unescape = (char *) "\f"; + break; + case 'u': + if (pe > stringEnd - 4) { + return Qnil; + } else { + char buf[4]; + UTF32 ch = unescape_unicode((unsigned char *) ++pe); + pe += 3; + if (UNI_SUR_HIGH_START == (ch & 0xFC00)) { + pe++; + if (pe > stringEnd - 6) return Qnil; + if (pe[0] == '\\' && pe[1] == 'u') { + UTF32 sur = unescape_unicode((unsigned char *) pe + 2); + ch = (((ch & 0x3F) << 10) | ((((ch >> 6) & 0xF) + 1) << 16) + | (sur & 0x3FF)); + pe += 5; + } else { + unescape = (char *) "?"; + break; + } + } + unescape_len = convert_UTF32_to_UTF8(buf, ch); + unescape = buf; + } + break; + default: + p = pe; + continue; + } + rb_str_buf_cat(result, unescape, unescape_len); + p = ++pe; + } else { + pe++; + } + } + rb_str_buf_cat(result, p, pe - p); + return result; +} + + +#line 1350 "parser.c" +static const int JSON_string_start = 1; +static const int JSON_string_first_final = 8; +static const int JSON_string_error = 0; + +static const int JSON_string_en_main = 1; + + +#line 467 "parser.rl" + + +static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + *result = rb_str_buf_new(0); + +#line 1367 "parser.c" + { + cs = JSON_string_start; + } + +#line 475 "parser.rl" + json->memo = p; + +#line 1375 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +case 1: + if ( (*p) == 34 ) + goto st2; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + switch( (*p) ) { + case 34: goto tr2; + case 92: goto st3; + } + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st2; +tr2: +#line 453 "parser.rl" + { + *result = json_string_unescape(*result, json->memo + 1, p); + if (NIL_P(*result)) { + p--; + {p++; cs = 8; goto _out;} + } else { + FORCE_UTF8(*result); + {p = (( p + 1))-1;} + } + } +#line 464 "parser.rl" + { p--; {p++; cs = 8; goto _out;} } + goto st8; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: +#line 1418 "parser.c" + goto st0; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: + if ( (*p) == 117 ) + goto st4; + if ( 0 <= (*p) && (*p) <= 31 ) + goto st0; + goto st2; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st5; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st5; + } else + goto st5; + goto st0; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st6; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st6; + } else + goto st6; + goto st0; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st7; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st7; + } else + goto st7; + goto st0; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( (*p) < 65 ) { + if ( 48 <= (*p) && (*p) <= 57 ) + goto st2; + } else if ( (*p) > 70 ) { + if ( 97 <= (*p) && (*p) <= 102 ) + goto st2; + } else + goto st2; + goto st0; + } + _test_eof2: cs = 2; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 477 "parser.rl" + + if (cs >= JSON_string_first_final) { + return p + 1; + } else { + return NULL; + } +} + + + +#line 1505 "parser.c" +static const int JSON_start = 1; +static const int JSON_first_final = 10; +static const int JSON_error = 0; + +static const int JSON_en_main = 1; + + +#line 511 "parser.rl" + + +/* + * Document-class: JSON::Ext::Parser + * + * This is the JSON parser implemented as a C extension. It can be configured + * to be used by setting + * + * JSON.parser = JSON::Ext::Parser + * + * with the method parser= in JSON. + * + */ + +static VALUE convert_encoding(VALUE source) +{ + char *ptr = RSTRING_PTR(source); + long len = RSTRING_LEN(source); + if (len < 2) { + rb_raise(eParserError, "A JSON text must at least contain two octets!"); + } +#ifdef HAVE_RUBY_ENCODING_H + { + VALUE encoding = rb_funcall(source, i_encoding, 0); + if (encoding == CEncoding_ASCII_8BIT) { + if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32BE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16BE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32LE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16LE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else { + FORCE_UTF8(source); + } + } else { + source = rb_funcall(source, i_encode, 1, CEncoding_UTF_8); + } + } +#else + if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32be"), source); + } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16be"), source); + } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32le"), source); + } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16le"), source); + } +#endif + return source; +} + +/* + * call-seq: new(source, opts => {}) + * + * Creates a new JSON::Ext::Parser instance for the string _source_. + * + * Creates a new JSON::Ext::Parser instance for the string _source_. + * + * It will be configured by the _opts_ hash. _opts_ can have the following + * keys: + * + * _opts_ can have the following keys: + * * *max_nesting*: The maximum depth of nesting allowed in the parsed data + * structures. Disable depth checking with :max_nesting => false|nil|0, it + * defaults to 19. + * * *allow_nan*: If set to true, allow NaN, Infinity and -Infinity in + * defiance of RFC 4627 to be parsed by the Parser. This option defaults to + * false. + * * *create_additions*: If set to false, the Parser doesn't create + * additions even if a matchin class and create_id was found. This option + * defaults to true. + * * *object_class*: Defaults to Hash + * * *array_class*: Defaults to Array + */ +static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self) +{ + char *ptr; + long len; + VALUE source, opts; + GET_PARSER; + rb_scan_args(argc, argv, "11", &source, &opts); + source = convert_encoding(StringValue(source)); + ptr = RSTRING_PTR(source); + len = RSTRING_LEN(source); + if (!NIL_P(opts)) { + opts = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); + if (NIL_P(opts)) { + rb_raise(rb_eArgError, "opts needs to be like a hash"); + } else { + VALUE tmp = ID2SYM(i_max_nesting); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE max_nesting = rb_hash_aref(opts, tmp); + if (RTEST(max_nesting)) { + Check_Type(max_nesting, T_FIXNUM); + json->max_nesting = FIX2INT(max_nesting); + } else { + json->max_nesting = 0; + } + } else { + json->max_nesting = 19; + } + tmp = ID2SYM(i_allow_nan); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE allow_nan = rb_hash_aref(opts, tmp); + json->allow_nan = RTEST(allow_nan) ? 1 : 0; + } else { + json->allow_nan = 0; + } + tmp = ID2SYM(i_create_additions); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE create_additions = rb_hash_aref(opts, tmp); + if (RTEST(create_additions)) { + json->create_id = rb_funcall(mJSON, i_create_id, 0); + } else { + json->create_id = Qnil; + } + } else { + json->create_id = rb_funcall(mJSON, i_create_id, 0); + } + tmp = ID2SYM(i_object_class); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + json->object_class = rb_hash_aref(opts, tmp); + } else { + json->object_class = Qnil; + } + tmp = ID2SYM(i_array_class); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + json->array_class = rb_hash_aref(opts, tmp); + } else { + json->array_class = Qnil; + } + } + } else { + json->max_nesting = 19; + json->allow_nan = 0; + json->create_id = rb_funcall(mJSON, i_create_id, 0); + json->object_class = Qnil; + json->array_class = Qnil; + } + json->current_nesting = 0; + json->len = len; + json->source = ptr; + json->Vsource = source; + return self; +} + +/* + * call-seq: parse() + * + * Parses the current JSON text _source_ and returns the complete data + * structure as a result. + */ +static VALUE cParser_parse(VALUE self) +{ + char *p, *pe; + int cs = EVIL; + VALUE result = Qnil; + GET_PARSER; + + +#line 1685 "parser.c" + { + cs = JSON_start; + } + +#line 682 "parser.rl" + p = json->source; + pe = p + json->len; + +#line 1694 "parser.c" + { + if ( p == pe ) + goto _test_eof; + switch ( cs ) + { +st1: + if ( ++p == pe ) + goto _test_eof1; +case 1: + switch( (*p) ) { + case 13: goto st1; + case 32: goto st1; + case 47: goto st2; + case 91: goto tr3; + case 123: goto tr4; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st1; + goto st0; +st0: +cs = 0; + goto _out; +st2: + if ( ++p == pe ) + goto _test_eof2; +case 2: + switch( (*p) ) { + case 42: goto st3; + case 47: goto st5; + } + goto st0; +st3: + if ( ++p == pe ) + goto _test_eof3; +case 3: + if ( (*p) == 42 ) + goto st4; + goto st3; +st4: + if ( ++p == pe ) + goto _test_eof4; +case 4: + switch( (*p) ) { + case 42: goto st4; + case 47: goto st1; + } + goto st3; +st5: + if ( ++p == pe ) + goto _test_eof5; +case 5: + if ( (*p) == 10 ) + goto st1; + goto st5; +tr3: +#line 500 "parser.rl" + { + char *np; + json->current_nesting = 1; + np = JSON_parse_array(json, p, pe, &result); + if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;} + } + goto st10; +tr4: +#line 493 "parser.rl" + { + char *np; + json->current_nesting = 1; + np = JSON_parse_object(json, p, pe, &result); + if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;} + } + goto st10; +st10: + if ( ++p == pe ) + goto _test_eof10; +case 10: +#line 1771 "parser.c" + switch( (*p) ) { + case 13: goto st10; + case 32: goto st10; + case 47: goto st6; + } + if ( 9 <= (*p) && (*p) <= 10 ) + goto st10; + goto st0; +st6: + if ( ++p == pe ) + goto _test_eof6; +case 6: + switch( (*p) ) { + case 42: goto st7; + case 47: goto st9; + } + goto st0; +st7: + if ( ++p == pe ) + goto _test_eof7; +case 7: + if ( (*p) == 42 ) + goto st8; + goto st7; +st8: + if ( ++p == pe ) + goto _test_eof8; +case 8: + switch( (*p) ) { + case 42: goto st8; + case 47: goto st10; + } + goto st7; +st9: + if ( ++p == pe ) + goto _test_eof9; +case 9: + if ( (*p) == 10 ) + goto st10; + goto st9; + } + _test_eof1: cs = 1; goto _test_eof; + _test_eof2: cs = 2; goto _test_eof; + _test_eof3: cs = 3; goto _test_eof; + _test_eof4: cs = 4; goto _test_eof; + _test_eof5: cs = 5; goto _test_eof; + _test_eof10: cs = 10; goto _test_eof; + _test_eof6: cs = 6; goto _test_eof; + _test_eof7: cs = 7; goto _test_eof; + _test_eof8: cs = 8; goto _test_eof; + _test_eof9: cs = 9; goto _test_eof; + + _test_eof: {} + _out: {} + } + +#line 685 "parser.rl" + + if (cs >= JSON_first_final && p == pe) { + return result; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } +} + +static JSON_Parser *JSON_allocate() +{ + JSON_Parser *json = ALLOC(JSON_Parser); + MEMZERO(json, JSON_Parser, 1); + return json; +} + +static void JSON_mark(JSON_Parser *json) +{ + rb_gc_mark_maybe(json->Vsource); + rb_gc_mark_maybe(json->create_id); + rb_gc_mark_maybe(json->object_class); + rb_gc_mark_maybe(json->array_class); +} + +static void JSON_free(JSON_Parser *json) +{ + ruby_xfree(json); +} + +static VALUE cJSON_parser_s_allocate(VALUE klass) +{ + JSON_Parser *json = JSON_allocate(); + return Data_Wrap_Struct(klass, JSON_mark, JSON_free, json); +} + +/* + * call-seq: source() + * + * Returns a copy of the current _source_ string, that was used to construct + * this Parser. + */ +static VALUE cParser_source(VALUE self) +{ + GET_PARSER; + return rb_str_dup(json->Vsource); +} + +void Init_parser() +{ + rb_require("json/common"); + mJSON = rb_define_module("JSON"); + mExt = rb_define_module_under(mJSON, "Ext"); + cParser = rb_define_class_under(mExt, "Parser", rb_cObject); + eParserError = rb_path2class("JSON::ParserError"); + eNestingError = rb_path2class("JSON::NestingError"); + rb_define_alloc_func(cParser, cJSON_parser_s_allocate); + rb_define_method(cParser, "initialize", cParser_initialize, -1); + rb_define_method(cParser, "parse", cParser_parse, 0); + rb_define_method(cParser, "source", cParser_source, 0); + + CNaN = rb_const_get(mJSON, rb_intern("NaN")); + CInfinity = rb_const_get(mJSON, rb_intern("Infinity")); + CMinusInfinity = rb_const_get(mJSON, rb_intern("MinusInfinity")); + + i_json_creatable_p = rb_intern("json_creatable?"); + i_json_create = rb_intern("json_create"); + i_create_id = rb_intern("create_id"); + i_create_additions = rb_intern("create_additions"); + i_chr = rb_intern("chr"); + i_max_nesting = rb_intern("max_nesting"); + i_allow_nan = rb_intern("allow_nan"); + i_object_class = rb_intern("object_class"); + i_array_class = rb_intern("array_class"); +#ifdef HAVE_RUBY_ENCODING_H + CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); + CEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be")); + CEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le")); + CEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be")); + CEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le")); + CEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit")); + i_encoding = rb_intern("encoding"); + i_encode = rb_intern("encode"); + i_encode_bang = rb_intern("encode!"); + i_force_encoding = rb_intern("force_encoding"); +#else + i_iconv = rb_intern("iconv"); +#endif +} diff --git a/ext/json/ext/parser.h b/ext/json/ext/parser.h new file mode 100644 index 0000000..5f2506d --- /dev/null +++ b/ext/json/ext/parser.h @@ -0,0 +1,79 @@ +#ifndef _PARSER_H_ +#define _PARSER_H_ + +#include "ruby.h" + +#if HAVE_RE_H +#include "re.h" +#endif + +#if HAVE_RUBY_ST_H +#include "ruby/st.h" +#endif + +#if HAVE_ST_H +#include "st.h" +#endif + +#ifdef HAVE_RUBY_ENCODING_H +#include "ruby/encoding.h" +#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding()) +#else +#define FORCE_UTF8(obj) +#endif + +#ifndef RHASH_TBL +#define RHASH_TBL(hsh) (RHASH(hsh)->tbl) +#endif + +/* unicode */ + +typedef unsigned long UTF32; /* at least 32 bits */ +typedef unsigned short UTF16; /* at least 16 bits */ +typedef unsigned char UTF8; /* typically 8 bits */ + +#define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD +#define UNI_SUR_HIGH_START (UTF32)0xD800 +#define UNI_SUR_HIGH_END (UTF32)0xDBFF +#define UNI_SUR_LOW_START (UTF32)0xDC00 +#define UNI_SUR_LOW_END (UTF32)0xDFFF + +typedef struct JSON_ParserStruct { + VALUE Vsource; + char *source; + long len; + char *memo; + VALUE create_id; + int max_nesting; + int current_nesting; + int allow_nan; + VALUE object_class; + VALUE array_class; +} JSON_Parser; + +#define GET_PARSER \ + JSON_Parser *json; \ + Data_Get_Struct(self, JSON_Parser, json) + +#define MinusInfinity "-Infinity" +#define EVIL 0x666 + +static UTF32 unescape_unicode(const unsigned char *p); +static int convert_UTF32_to_UTF8(char *buf, UTF32 ch); +static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result); +static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result); +static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result); +static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result); +static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result); +static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd); +static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result); +static VALUE convert_encoding(VALUE source); +static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self); +static VALUE cParser_parse(VALUE self); +static JSON_Parser *JSON_allocate(); +static void JSON_mark(JSON_Parser *json); +static void JSON_free(JSON_Parser *json); +static VALUE cJSON_parser_s_allocate(VALUE klass); +static VALUE cParser_source(VALUE self); + +#endif diff --git a/ext/json/ext/parser.rl b/ext/json/ext/parser.rl new file mode 100644 index 0000000..b91ac00 --- /dev/null +++ b/ext/json/ext/parser.rl @@ -0,0 +1,771 @@ +#include "parser.h" + +/* unicode */ + +static const char digit_values[256] = { + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, + -1, -1, -1, -1, -1, -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, + -1, -1, -1, -1, -1, -1, -1 +}; + +static UTF32 unescape_unicode(const unsigned char *p) +{ + char b; + UTF32 result = 0; + b = digit_values[p[0]]; + if (b < 0) return UNI_REPLACEMENT_CHAR; + result = (result << 4) | b; + b = digit_values[p[1]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + b = digit_values[p[2]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + b = digit_values[p[3]]; + result = (result << 4) | b; + if (b < 0) return UNI_REPLACEMENT_CHAR; + return result; +} + +static int convert_UTF32_to_UTF8(char *buf, UTF32 ch) +{ + int len = 1; + if (ch <= 0x7F) { + buf[0] = (char) ch; + } else if (ch <= 0x07FF) { + buf[0] = (char) ((ch >> 6) | 0xC0); + buf[1] = (char) ((ch & 0x3F) | 0x80); + len++; + } else if (ch <= 0xFFFF) { + buf[0] = (char) ((ch >> 12) | 0xE0); + buf[1] = (char) (((ch >> 6) & 0x3F) | 0x80); + buf[2] = (char) ((ch & 0x3F) | 0x80); + len += 2; + } else if (ch <= 0x1fffff) { + buf[0] =(char) ((ch >> 18) | 0xF0); + buf[1] =(char) (((ch >> 12) & 0x3F) | 0x80); + buf[2] =(char) (((ch >> 6) & 0x3F) | 0x80); + buf[3] =(char) ((ch & 0x3F) | 0x80); + len += 3; + } else { + buf[0] = '?'; + } + return len; +} + + +#ifdef HAVE_RUBY_ENCODING_H +static VALUE CEncoding_ASCII_8BIT, CEncoding_UTF_8, CEncoding_UTF_16BE, + CEncoding_UTF_16LE, CEncoding_UTF_32BE, CEncoding_UTF_32LE; +static ID i_encoding, i_encode, i_encode_bang, i_force_encoding; +#else +static ID i_iconv; +#endif + +static VALUE mJSON, mExt, cParser, eParserError, eNestingError; +static VALUE CNaN, CInfinity, CMinusInfinity; + +static ID i_json_creatable_p, i_json_create, i_create_id, i_create_additions, + i_chr, i_max_nesting, i_allow_nan, i_object_class, i_array_class; + +%%{ + machine JSON_common; + + cr = '\n'; + cr_neg = [^\n]; + ws = [ \t\r\n]; + c_comment = '/*' ( any* - (any* '*/' any* ) ) '*/'; + cpp_comment = '//' cr_neg* cr; + comment = c_comment | cpp_comment; + ignore = ws | comment; + name_separator = ':'; + value_separator = ','; + Vnull = 'null'; + Vfalse = 'false'; + Vtrue = 'true'; + VNaN = 'NaN'; + VInfinity = 'Infinity'; + VMinusInfinity = '-Infinity'; + begin_value = [nft"\-[{NI] | digit; + begin_object = '{'; + end_object = '}'; + begin_array = '['; + end_array = ']'; + begin_string = '"'; + begin_name = begin_string; + begin_number = digit | '-'; +}%% + +%%{ + machine JSON_object; + include JSON_common; + + write data; + + action parse_value { + VALUE v = Qnil; + char *np = JSON_parse_value(json, fpc, pe, &v); + if (np == NULL) { + fhold; fbreak; + } else { + rb_hash_aset(*result, last_name, v); + fexec np; + } + } + + action parse_name { + char *np = JSON_parse_string(json, fpc, pe, &last_name); + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + action exit { fhold; fbreak; } + + a_pair = ignore* begin_name >parse_name + ignore* name_separator ignore* + begin_value >parse_value; + + main := begin_object + (a_pair (ignore* value_separator a_pair)*)? + ignore* end_object @exit; +}%% + +static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + VALUE last_name = Qnil; + VALUE object_class = json->object_class; + + if (json->max_nesting && json->current_nesting > json->max_nesting) { + rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); + } + + *result = NIL_P(object_class) ? rb_hash_new() : rb_class_new_instance(0, 0, object_class); + + %% write init; + %% write exec; + + if (cs >= JSON_object_first_final) { + if (RTEST(json->create_id)) { + VALUE klassname = rb_hash_aref(*result, json->create_id); + if (!NIL_P(klassname)) { + VALUE klass = rb_path2class(StringValueCStr(klassname)); + if RTEST(rb_funcall(klass, i_json_creatable_p, 0)) { + *result = rb_funcall(klass, i_json_create, 1, *result); + } + } + } + return p + 1; + } else { + return NULL; + } +} + +%%{ + machine JSON_value; + include JSON_common; + + write data; + + action parse_null { + *result = Qnil; + } + action parse_false { + *result = Qfalse; + } + action parse_true { + *result = Qtrue; + } + action parse_nan { + if (json->allow_nan) { + *result = CNaN; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 2); + } + } + action parse_infinity { + if (json->allow_nan) { + *result = CInfinity; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 8); + } + } + action parse_string { + char *np = JSON_parse_string(json, fpc, pe, result); + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + action parse_number { + char *np; + if(pe > fpc + 9 && !strncmp(MinusInfinity, fpc, 9)) { + if (json->allow_nan) { + *result = CMinusInfinity; + fexec p + 10; + fhold; fbreak; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } + } + np = JSON_parse_float(json, fpc, pe, result); + if (np != NULL) fexec np; + np = JSON_parse_integer(json, fpc, pe, result); + if (np != NULL) fexec np; + fhold; fbreak; + } + + action parse_array { + char *np; + json->current_nesting++; + np = JSON_parse_array(json, fpc, pe, result); + json->current_nesting--; + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + action parse_object { + char *np; + json->current_nesting++; + np = JSON_parse_object(json, fpc, pe, result); + json->current_nesting--; + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + action exit { fhold; fbreak; } + +main := ( + Vnull @parse_null | + Vfalse @parse_false | + Vtrue @parse_true | + VNaN @parse_nan | + VInfinity @parse_infinity | + begin_number >parse_number | + begin_string >parse_string | + begin_array >parse_array | + begin_object >parse_object + ) %*exit; +}%% + +static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + %% write init; + %% write exec; + + if (cs >= JSON_value_first_final) { + return p; + } else { + return NULL; + } +} + +%%{ + machine JSON_integer; + + write data; + + action exit { fhold; fbreak; } + + main := '-'? ('0' | [1-9][0-9]*) (^[0-9] @exit); +}%% + +static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + %% write init; + json->memo = p; + %% write exec; + + if (cs >= JSON_integer_first_final) { + long len = p - json->memo; + *result = rb_Integer(rb_str_new(json->memo, len)); + return p + 1; + } else { + return NULL; + } +} + +%%{ + machine JSON_float; + include JSON_common; + + write data; + + action exit { fhold; fbreak; } + + main := '-'? ( + (('0' | [1-9][0-9]*) '.' [0-9]+ ([Ee] [+\-]?[0-9]+)?) + | (('0' | [1-9][0-9]*) ([Ee] [+\-]?[0-9]+)) + ) (^[0-9Ee.\-] @exit ); +}%% + +static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + %% write init; + json->memo = p; + %% write exec; + + if (cs >= JSON_float_first_final) { + long len = p - json->memo; + *result = rb_Float(rb_str_new(json->memo, len)); + return p + 1; + } else { + return NULL; + } +} + + +%%{ + machine JSON_array; + include JSON_common; + + write data; + + action parse_value { + VALUE v = Qnil; + char *np = JSON_parse_value(json, fpc, pe, &v); + if (np == NULL) { + fhold; fbreak; + } else { + rb_ary_push(*result, v); + fexec np; + } + } + + action exit { fhold; fbreak; } + + next_element = value_separator ignore* begin_value >parse_value; + + main := begin_array ignore* + ((begin_value >parse_value ignore*) + (ignore* next_element ignore*)*)? + end_array @exit; +}%% + +static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + VALUE array_class = json->array_class; + + if (json->max_nesting && json->current_nesting > json->max_nesting) { + rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); + } + *result = NIL_P(array_class) ? rb_ary_new() : rb_class_new_instance(0, 0, array_class); + + %% write init; + %% write exec; + + if(cs >= JSON_array_first_final) { + return p + 1; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } +} + +static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd) +{ + char *p = string, *pe = string, *unescape; + int unescape_len; + + while (pe < stringEnd) { + if (*pe == '\\') { + unescape = (char *) "?"; + unescape_len = 1; + if (pe > p) rb_str_buf_cat(result, p, pe - p); + switch (*++pe) { + case 'n': + unescape = (char *) "\n"; + break; + case 'r': + unescape = (char *) "\r"; + break; + case 't': + unescape = (char *) "\t"; + break; + case '"': + unescape = (char *) "\""; + break; + case '\\': + unescape = (char *) "\\"; + break; + case 'b': + unescape = (char *) "\b"; + break; + case 'f': + unescape = (char *) "\f"; + break; + case 'u': + if (pe > stringEnd - 4) { + return Qnil; + } else { + char buf[4]; + UTF32 ch = unescape_unicode((unsigned char *) ++pe); + pe += 3; + if (UNI_SUR_HIGH_START == (ch & 0xFC00)) { + pe++; + if (pe > stringEnd - 6) return Qnil; + if (pe[0] == '\\' && pe[1] == 'u') { + UTF32 sur = unescape_unicode((unsigned char *) pe + 2); + ch = (((ch & 0x3F) << 10) | ((((ch >> 6) & 0xF) + 1) << 16) + | (sur & 0x3FF)); + pe += 5; + } else { + unescape = (char *) "?"; + break; + } + } + unescape_len = convert_UTF32_to_UTF8(buf, ch); + unescape = buf; + } + break; + default: + p = pe; + continue; + } + rb_str_buf_cat(result, unescape, unescape_len); + p = ++pe; + } else { + pe++; + } + } + rb_str_buf_cat(result, p, pe - p); + return result; +} + +%%{ + machine JSON_string; + include JSON_common; + + write data; + + action parse_string { + *result = json_string_unescape(*result, json->memo + 1, p); + if (NIL_P(*result)) { + fhold; + fbreak; + } else { + FORCE_UTF8(*result); + fexec p + 1; + } + } + + action exit { fhold; fbreak; } + + main := '"' ((^(["\\] | 0..0x1f) | '\\'["\\/bfnrt] | '\\u'[0-9a-fA-F]{4} | '\\'^(["\\/bfnrtu]|0..0x1f))* %parse_string) '"' @exit; +}%% + +static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result) +{ + int cs = EVIL; + + *result = rb_str_buf_new(0); + %% write init; + json->memo = p; + %% write exec; + + if (cs >= JSON_string_first_final) { + return p + 1; + } else { + return NULL; + } +} + + +%%{ + machine JSON; + + write data; + + include JSON_common; + + action parse_object { + char *np; + json->current_nesting = 1; + np = JSON_parse_object(json, fpc, pe, &result); + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + action parse_array { + char *np; + json->current_nesting = 1; + np = JSON_parse_array(json, fpc, pe, &result); + if (np == NULL) { fhold; fbreak; } else fexec np; + } + + main := ignore* ( + begin_object >parse_object | + begin_array >parse_array + ) ignore*; +}%% + +/* + * Document-class: JSON::Ext::Parser + * + * This is the JSON parser implemented as a C extension. It can be configured + * to be used by setting + * + * JSON.parser = JSON::Ext::Parser + * + * with the method parser= in JSON. + * + */ + +static VALUE convert_encoding(VALUE source) +{ + char *ptr = RSTRING_PTR(source); + long len = RSTRING_LEN(source); + if (len < 2) { + rb_raise(eParserError, "A JSON text must at least contain two octets!"); + } +#ifdef HAVE_RUBY_ENCODING_H + { + VALUE encoding = rb_funcall(source, i_encoding, 0); + if (encoding == CEncoding_ASCII_8BIT) { + if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32BE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16BE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32LE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { + source = rb_str_dup(source); + rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16LE); + source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); + } else { + FORCE_UTF8(source); + } + } else { + source = rb_funcall(source, i_encode, 1, CEncoding_UTF_8); + } + } +#else + if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32be"), source); + } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16be"), source); + } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32le"), source); + } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { + source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16le"), source); + } +#endif + return source; +} + +/* + * call-seq: new(source, opts => {}) + * + * Creates a new JSON::Ext::Parser instance for the string _source_. + * + * Creates a new JSON::Ext::Parser instance for the string _source_. + * + * It will be configured by the _opts_ hash. _opts_ can have the following + * keys: + * + * _opts_ can have the following keys: + * * *max_nesting*: The maximum depth of nesting allowed in the parsed data + * structures. Disable depth checking with :max_nesting => false|nil|0, it + * defaults to 19. + * * *allow_nan*: If set to true, allow NaN, Infinity and -Infinity in + * defiance of RFC 4627 to be parsed by the Parser. This option defaults to + * false. + * * *create_additions*: If set to false, the Parser doesn't create + * additions even if a matchin class and create_id was found. This option + * defaults to true. + * * *object_class*: Defaults to Hash + * * *array_class*: Defaults to Array + */ +static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self) +{ + char *ptr; + long len; + VALUE source, opts; + GET_PARSER; + rb_scan_args(argc, argv, "11", &source, &opts); + source = convert_encoding(StringValue(source)); + ptr = RSTRING_PTR(source); + len = RSTRING_LEN(source); + if (!NIL_P(opts)) { + opts = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); + if (NIL_P(opts)) { + rb_raise(rb_eArgError, "opts needs to be like a hash"); + } else { + VALUE tmp = ID2SYM(i_max_nesting); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE max_nesting = rb_hash_aref(opts, tmp); + if (RTEST(max_nesting)) { + Check_Type(max_nesting, T_FIXNUM); + json->max_nesting = FIX2INT(max_nesting); + } else { + json->max_nesting = 0; + } + } else { + json->max_nesting = 19; + } + tmp = ID2SYM(i_allow_nan); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE allow_nan = rb_hash_aref(opts, tmp); + json->allow_nan = RTEST(allow_nan) ? 1 : 0; + } else { + json->allow_nan = 0; + } + tmp = ID2SYM(i_create_additions); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + VALUE create_additions = rb_hash_aref(opts, tmp); + if (RTEST(create_additions)) { + json->create_id = rb_funcall(mJSON, i_create_id, 0); + } else { + json->create_id = Qnil; + } + } else { + json->create_id = rb_funcall(mJSON, i_create_id, 0); + } + tmp = ID2SYM(i_object_class); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + json->object_class = rb_hash_aref(opts, tmp); + } else { + json->object_class = Qnil; + } + tmp = ID2SYM(i_array_class); + if (st_lookup(RHASH_TBL(opts), tmp, 0)) { + json->array_class = rb_hash_aref(opts, tmp); + } else { + json->array_class = Qnil; + } + } + } else { + json->max_nesting = 19; + json->allow_nan = 0; + json->create_id = rb_funcall(mJSON, i_create_id, 0); + json->object_class = Qnil; + json->array_class = Qnil; + } + json->current_nesting = 0; + json->len = len; + json->source = ptr; + json->Vsource = source; + return self; +} + +/* + * call-seq: parse() + * + * Parses the current JSON text _source_ and returns the complete data + * structure as a result. + */ +static VALUE cParser_parse(VALUE self) +{ + char *p, *pe; + int cs = EVIL; + VALUE result = Qnil; + GET_PARSER; + + %% write init; + p = json->source; + pe = p + json->len; + %% write exec; + + if (cs >= JSON_first_final && p == pe) { + return result; + } else { + rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); + } +} + +static JSON_Parser *JSON_allocate() +{ + JSON_Parser *json = ALLOC(JSON_Parser); + MEMZERO(json, JSON_Parser, 1); + return json; +} + +static void JSON_mark(JSON_Parser *json) +{ + rb_gc_mark_maybe(json->Vsource); + rb_gc_mark_maybe(json->create_id); + rb_gc_mark_maybe(json->object_class); + rb_gc_mark_maybe(json->array_class); +} + +static void JSON_free(JSON_Parser *json) +{ + ruby_xfree(json); +} + +static VALUE cJSON_parser_s_allocate(VALUE klass) +{ + JSON_Parser *json = JSON_allocate(); + return Data_Wrap_Struct(klass, JSON_mark, JSON_free, json); +} + +/* + * call-seq: source() + * + * Returns a copy of the current _source_ string, that was used to construct + * this Parser. + */ +static VALUE cParser_source(VALUE self) +{ + GET_PARSER; + return rb_str_dup(json->Vsource); +} + +void Init_parser() +{ + rb_require("json/common"); + mJSON = rb_define_module("JSON"); + mExt = rb_define_module_under(mJSON, "Ext"); + cParser = rb_define_class_under(mExt, "Parser", rb_cObject); + eParserError = rb_path2class("JSON::ParserError"); + eNestingError = rb_path2class("JSON::NestingError"); + rb_define_alloc_func(cParser, cJSON_parser_s_allocate); + rb_define_method(cParser, "initialize", cParser_initialize, -1); + rb_define_method(cParser, "parse", cParser_parse, 0); + rb_define_method(cParser, "source", cParser_source, 0); + + CNaN = rb_const_get(mJSON, rb_intern("NaN")); + CInfinity = rb_const_get(mJSON, rb_intern("Infinity")); + CMinusInfinity = rb_const_get(mJSON, rb_intern("MinusInfinity")); + + i_json_creatable_p = rb_intern("json_creatable?"); + i_json_create = rb_intern("json_create"); + i_create_id = rb_intern("create_id"); + i_create_additions = rb_intern("create_additions"); + i_chr = rb_intern("chr"); + i_max_nesting = rb_intern("max_nesting"); + i_allow_nan = rb_intern("allow_nan"); + i_object_class = rb_intern("object_class"); + i_array_class = rb_intern("array_class"); +#ifdef HAVE_RUBY_ENCODING_H + CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); + CEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be")); + CEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le")); + CEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be")); + CEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le")); + CEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit")); + i_encoding = rb_intern("encoding"); + i_encode = rb_intern("encode"); + i_encode_bang = rb_intern("encode!"); + i_force_encoding = rb_intern("force_encoding"); +#else + i_iconv = rb_intern("iconv"); +#endif +} diff --git a/ext/json/ext/parser/extconf.rb b/ext/json/ext/parser/extconf.rb deleted file mode 100644 index 9662e9a..0000000 --- a/ext/json/ext/parser/extconf.rb +++ /dev/null @@ -1,14 +0,0 @@ -require 'mkmf' -require 'rbconfig' - -unless $CFLAGS.gsub!(/ -O[\dsz]?/, ' -O3') - $CFLAGS << ' -O3' -end -if CONFIG['CC'] =~ /gcc/ - $CFLAGS << ' -Wall' - #$CFLAGS.gsub!(/ -O[\dsz]?/, ' -O0 -ggdb') -end - -have_header("ruby/st.h") || have_header("st.h") -have_header("re.h") -create_makefile 'parser' diff --git a/ext/json/ext/parser/parser.c b/ext/json/ext/parser/parser.c deleted file mode 100644 index 0122571..0000000 --- a/ext/json/ext/parser/parser.c +++ /dev/null @@ -1,1914 +0,0 @@ - -#line 1 "parser.rl" -#include "parser.h" - -/* unicode */ - -static const char digit_values[256] = { - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, - -1, -1, -1, -1, -1, -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1 -}; - -static UTF32 unescape_unicode(const unsigned char *p) -{ - char b; - UTF32 result = 0; - b = digit_values[p[0]]; - if (b < 0) return UNI_REPLACEMENT_CHAR; - result = (result << 4) | b; - b = digit_values[p[1]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - b = digit_values[p[2]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - b = digit_values[p[3]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - return result; -} - -static int convert_UTF32_to_UTF8(char *buf, UTF32 ch) -{ - int len = 1; - if (ch <= 0x7F) { - buf[0] = (char) ch; - } else if (ch <= 0x07FF) { - buf[0] = (char) ((ch >> 6) | 0xC0); - buf[1] = (char) ((ch & 0x3F) | 0x80); - len++; - } else if (ch <= 0xFFFF) { - buf[0] = (char) ((ch >> 12) | 0xE0); - buf[1] = (char) (((ch >> 6) & 0x3F) | 0x80); - buf[2] = (char) ((ch & 0x3F) | 0x80); - len += 2; - } else if (ch <= 0x1fffff) { - buf[0] =(char) ((ch >> 18) | 0xF0); - buf[1] =(char) (((ch >> 12) & 0x3F) | 0x80); - buf[2] =(char) (((ch >> 6) & 0x3F) | 0x80); - buf[3] =(char) ((ch & 0x3F) | 0x80); - len += 3; - } else { - buf[0] = '?'; - } - return len; -} - - -#ifdef HAVE_RUBY_ENCODING_H -static VALUE CEncoding_ASCII_8BIT, CEncoding_UTF_8, CEncoding_UTF_16BE, - CEncoding_UTF_16LE, CEncoding_UTF_32BE, CEncoding_UTF_32LE; -static ID i_encoding, i_encode, i_encode_bang, i_force_encoding; -#else -static ID i_iconv; -#endif - -static VALUE mJSON, mExt, cParser, eParserError, eNestingError; -static VALUE CNaN, CInfinity, CMinusInfinity; - -static ID i_json_creatable_p, i_json_create, i_create_id, i_create_additions, - i_chr, i_max_nesting, i_allow_nan, i_object_class, i_array_class; - - -#line 108 "parser.rl" - - - -#line 90 "parser.c" -static const int JSON_object_start = 1; -static const int JSON_object_first_final = 27; -static const int JSON_object_error = 0; - -static const int JSON_object_en_main = 1; - - -#line 141 "parser.rl" - - -static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - VALUE last_name = Qnil; - VALUE object_class = json->object_class; - - if (json->max_nesting && json->current_nesting > json->max_nesting) { - rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); - } - - *result = NIL_P(object_class) ? rb_hash_new() : rb_class_new_instance(0, 0, object_class); - - -#line 114 "parser.c" - { - cs = JSON_object_start; - } - -#line 156 "parser.rl" - -#line 121 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - if ( (*p) == 123 ) - goto st2; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - switch( (*p) ) { - case 13: goto st2; - case 32: goto st2; - case 34: goto tr2; - case 47: goto st23; - case 125: goto tr4; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st2; - goto st0; -tr2: -#line 127 "parser.rl" - { - char *np = JSON_parse_string(json, p, pe, &last_name); - if (np == NULL) { p--; {p++; cs = 3; goto _out;} } else {p = (( np))-1;} - } - goto st3; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: -#line 159 "parser.c" - switch( (*p) ) { - case 13: goto st3; - case 32: goto st3; - case 47: goto st4; - case 58: goto st8; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st3; - goto st0; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - switch( (*p) ) { - case 42: goto st5; - case 47: goto st7; - } - goto st0; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - if ( (*p) == 42 ) - goto st6; - goto st5; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - switch( (*p) ) { - case 42: goto st6; - case 47: goto st3; - } - goto st5; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( (*p) == 10 ) - goto st3; - goto st7; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - switch( (*p) ) { - case 13: goto st8; - case 32: goto st8; - case 34: goto tr11; - case 45: goto tr11; - case 47: goto st19; - case 73: goto tr11; - case 78: goto tr11; - case 91: goto tr11; - case 102: goto tr11; - case 110: goto tr11; - case 116: goto tr11; - case 123: goto tr11; - } - if ( (*p) > 10 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr11; - } else if ( (*p) >= 9 ) - goto st8; - goto st0; -tr11: -#line 116 "parser.rl" - { - VALUE v = Qnil; - char *np = JSON_parse_value(json, p, pe, &v); - if (np == NULL) { - p--; {p++; cs = 9; goto _out;} - } else { - rb_hash_aset(*result, last_name, v); - {p = (( np))-1;} - } - } - goto st9; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: -#line 242 "parser.c" - switch( (*p) ) { - case 13: goto st9; - case 32: goto st9; - case 44: goto st10; - case 47: goto st15; - case 125: goto tr4; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st9; - goto st0; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: - switch( (*p) ) { - case 13: goto st10; - case 32: goto st10; - case 34: goto tr2; - case 47: goto st11; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st10; - goto st0; -st11: - if ( ++p == pe ) - goto _test_eof11; -case 11: - switch( (*p) ) { - case 42: goto st12; - case 47: goto st14; - } - goto st0; -st12: - if ( ++p == pe ) - goto _test_eof12; -case 12: - if ( (*p) == 42 ) - goto st13; - goto st12; -st13: - if ( ++p == pe ) - goto _test_eof13; -case 13: - switch( (*p) ) { - case 42: goto st13; - case 47: goto st10; - } - goto st12; -st14: - if ( ++p == pe ) - goto _test_eof14; -case 14: - if ( (*p) == 10 ) - goto st10; - goto st14; -st15: - if ( ++p == pe ) - goto _test_eof15; -case 15: - switch( (*p) ) { - case 42: goto st16; - case 47: goto st18; - } - goto st0; -st16: - if ( ++p == pe ) - goto _test_eof16; -case 16: - if ( (*p) == 42 ) - goto st17; - goto st16; -st17: - if ( ++p == pe ) - goto _test_eof17; -case 17: - switch( (*p) ) { - case 42: goto st17; - case 47: goto st9; - } - goto st16; -st18: - if ( ++p == pe ) - goto _test_eof18; -case 18: - if ( (*p) == 10 ) - goto st9; - goto st18; -tr4: -#line 132 "parser.rl" - { p--; {p++; cs = 27; goto _out;} } - goto st27; -st27: - if ( ++p == pe ) - goto _test_eof27; -case 27: -#line 338 "parser.c" - goto st0; -st19: - if ( ++p == pe ) - goto _test_eof19; -case 19: - switch( (*p) ) { - case 42: goto st20; - case 47: goto st22; - } - goto st0; -st20: - if ( ++p == pe ) - goto _test_eof20; -case 20: - if ( (*p) == 42 ) - goto st21; - goto st20; -st21: - if ( ++p == pe ) - goto _test_eof21; -case 21: - switch( (*p) ) { - case 42: goto st21; - case 47: goto st8; - } - goto st20; -st22: - if ( ++p == pe ) - goto _test_eof22; -case 22: - if ( (*p) == 10 ) - goto st8; - goto st22; -st23: - if ( ++p == pe ) - goto _test_eof23; -case 23: - switch( (*p) ) { - case 42: goto st24; - case 47: goto st26; - } - goto st0; -st24: - if ( ++p == pe ) - goto _test_eof24; -case 24: - if ( (*p) == 42 ) - goto st25; - goto st24; -st25: - if ( ++p == pe ) - goto _test_eof25; -case 25: - switch( (*p) ) { - case 42: goto st25; - case 47: goto st2; - } - goto st24; -st26: - if ( ++p == pe ) - goto _test_eof26; -case 26: - if ( (*p) == 10 ) - goto st2; - goto st26; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof11: cs = 11; goto _test_eof; - _test_eof12: cs = 12; goto _test_eof; - _test_eof13: cs = 13; goto _test_eof; - _test_eof14: cs = 14; goto _test_eof; - _test_eof15: cs = 15; goto _test_eof; - _test_eof16: cs = 16; goto _test_eof; - _test_eof17: cs = 17; goto _test_eof; - _test_eof18: cs = 18; goto _test_eof; - _test_eof27: cs = 27; goto _test_eof; - _test_eof19: cs = 19; goto _test_eof; - _test_eof20: cs = 20; goto _test_eof; - _test_eof21: cs = 21; goto _test_eof; - _test_eof22: cs = 22; goto _test_eof; - _test_eof23: cs = 23; goto _test_eof; - _test_eof24: cs = 24; goto _test_eof; - _test_eof25: cs = 25; goto _test_eof; - _test_eof26: cs = 26; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 157 "parser.rl" - - if (cs >= JSON_object_first_final) { - if (RTEST(json->create_id)) { - VALUE klassname = rb_hash_aref(*result, json->create_id); - if (!NIL_P(klassname)) { - VALUE klass = rb_path2class(StringValueCStr(klassname)); - if RTEST(rb_funcall(klass, i_json_creatable_p, 0)) { - *result = rb_funcall(klass, i_json_create, 1, *result); - } - } - } - return p + 1; - } else { - return NULL; - } -} - - -#line 455 "parser.c" -static const int JSON_value_start = 1; -static const int JSON_value_first_final = 21; -static const int JSON_value_error = 0; - -static const int JSON_value_en_main = 1; - - -#line 255 "parser.rl" - - -static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - -#line 471 "parser.c" - { - cs = JSON_value_start; - } - -#line 262 "parser.rl" - -#line 478 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - switch( (*p) ) { - case 34: goto tr0; - case 45: goto tr2; - case 73: goto st2; - case 78: goto st9; - case 91: goto tr5; - case 102: goto st11; - case 110: goto st15; - case 116: goto st18; - case 123: goto tr9; - } - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr2; - goto st0; -st0: -cs = 0; - goto _out; -tr0: -#line 203 "parser.rl" - { - char *np = JSON_parse_string(json, p, pe, result); - if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} - } - goto st21; -tr2: -#line 208 "parser.rl" - { - char *np; - if(pe > p + 9 && !strncmp(MinusInfinity, p, 9)) { - if (json->allow_nan) { - *result = CMinusInfinity; - {p = (( p + 10))-1;} - p--; {p++; cs = 21; goto _out;} - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } - } - np = JSON_parse_float(json, p, pe, result); - if (np != NULL) {p = (( np))-1;} - np = JSON_parse_integer(json, p, pe, result); - if (np != NULL) {p = (( np))-1;} - p--; {p++; cs = 21; goto _out;} - } - goto st21; -tr5: -#line 226 "parser.rl" - { - char *np; - json->current_nesting++; - np = JSON_parse_array(json, p, pe, result); - json->current_nesting--; - if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} - } - goto st21; -tr9: -#line 234 "parser.rl" - { - char *np; - json->current_nesting++; - np = JSON_parse_object(json, p, pe, result); - json->current_nesting--; - if (np == NULL) { p--; {p++; cs = 21; goto _out;} } else {p = (( np))-1;} - } - goto st21; -tr16: -#line 196 "parser.rl" - { - if (json->allow_nan) { - *result = CInfinity; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 8); - } - } - goto st21; -tr18: -#line 189 "parser.rl" - { - if (json->allow_nan) { - *result = CNaN; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 2); - } - } - goto st21; -tr22: -#line 183 "parser.rl" - { - *result = Qfalse; - } - goto st21; -tr25: -#line 180 "parser.rl" - { - *result = Qnil; - } - goto st21; -tr28: -#line 186 "parser.rl" - { - *result = Qtrue; - } - goto st21; -st21: - if ( ++p == pe ) - goto _test_eof21; -case 21: -#line 242 "parser.rl" - { p--; {p++; cs = 21; goto _out;} } -#line 593 "parser.c" - goto st0; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - if ( (*p) == 110 ) - goto st3; - goto st0; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: - if ( (*p) == 102 ) - goto st4; - goto st0; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - if ( (*p) == 105 ) - goto st5; - goto st0; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - if ( (*p) == 110 ) - goto st6; - goto st0; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - if ( (*p) == 105 ) - goto st7; - goto st0; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( (*p) == 116 ) - goto st8; - goto st0; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - if ( (*p) == 121 ) - goto tr16; - goto st0; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: - if ( (*p) == 97 ) - goto st10; - goto st0; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: - if ( (*p) == 78 ) - goto tr18; - goto st0; -st11: - if ( ++p == pe ) - goto _test_eof11; -case 11: - if ( (*p) == 97 ) - goto st12; - goto st0; -st12: - if ( ++p == pe ) - goto _test_eof12; -case 12: - if ( (*p) == 108 ) - goto st13; - goto st0; -st13: - if ( ++p == pe ) - goto _test_eof13; -case 13: - if ( (*p) == 115 ) - goto st14; - goto st0; -st14: - if ( ++p == pe ) - goto _test_eof14; -case 14: - if ( (*p) == 101 ) - goto tr22; - goto st0; -st15: - if ( ++p == pe ) - goto _test_eof15; -case 15: - if ( (*p) == 117 ) - goto st16; - goto st0; -st16: - if ( ++p == pe ) - goto _test_eof16; -case 16: - if ( (*p) == 108 ) - goto st17; - goto st0; -st17: - if ( ++p == pe ) - goto _test_eof17; -case 17: - if ( (*p) == 108 ) - goto tr25; - goto st0; -st18: - if ( ++p == pe ) - goto _test_eof18; -case 18: - if ( (*p) == 114 ) - goto st19; - goto st0; -st19: - if ( ++p == pe ) - goto _test_eof19; -case 19: - if ( (*p) == 117 ) - goto st20; - goto st0; -st20: - if ( ++p == pe ) - goto _test_eof20; -case 20: - if ( (*p) == 101 ) - goto tr28; - goto st0; - } - _test_eof21: cs = 21; goto _test_eof; - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof11: cs = 11; goto _test_eof; - _test_eof12: cs = 12; goto _test_eof; - _test_eof13: cs = 13; goto _test_eof; - _test_eof14: cs = 14; goto _test_eof; - _test_eof15: cs = 15; goto _test_eof; - _test_eof16: cs = 16; goto _test_eof; - _test_eof17: cs = 17; goto _test_eof; - _test_eof18: cs = 18; goto _test_eof; - _test_eof19: cs = 19; goto _test_eof; - _test_eof20: cs = 20; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 263 "parser.rl" - - if (cs >= JSON_value_first_final) { - return p; - } else { - return NULL; - } -} - - -#line 764 "parser.c" -static const int JSON_integer_start = 1; -static const int JSON_integer_first_final = 5; -static const int JSON_integer_error = 0; - -static const int JSON_integer_en_main = 1; - - -#line 279 "parser.rl" - - -static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - -#line 780 "parser.c" - { - cs = JSON_integer_start; - } - -#line 286 "parser.rl" - json->memo = p; - -#line 788 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - switch( (*p) ) { - case 45: goto st2; - case 48: goto st3; - } - if ( 49 <= (*p) && (*p) <= 57 ) - goto st4; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - if ( (*p) == 48 ) - goto st3; - if ( 49 <= (*p) && (*p) <= 57 ) - goto st4; - goto st0; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st0; - goto tr4; -tr4: -#line 276 "parser.rl" - { p--; {p++; cs = 5; goto _out;} } - goto st5; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: -#line 829 "parser.c" - goto st0; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st4; - goto tr4; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 288 "parser.rl" - - if (cs >= JSON_integer_first_final) { - long len = p - json->memo; - *result = rb_Integer(rb_str_new(json->memo, len)); - return p + 1; - } else { - return NULL; - } -} - - -#line 860 "parser.c" -static const int JSON_float_start = 1; -static const int JSON_float_first_final = 10; -static const int JSON_float_error = 0; - -static const int JSON_float_en_main = 1; - - -#line 310 "parser.rl" - - -static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - -#line 876 "parser.c" - { - cs = JSON_float_start; - } - -#line 317 "parser.rl" - json->memo = p; - -#line 884 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - switch( (*p) ) { - case 45: goto st2; - case 48: goto st3; - } - if ( 49 <= (*p) && (*p) <= 57 ) - goto st9; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - if ( (*p) == 48 ) - goto st3; - if ( 49 <= (*p) && (*p) <= 57 ) - goto st9; - goto st0; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: - switch( (*p) ) { - case 46: goto st4; - case 69: goto st6; - case 101: goto st6; - } - goto st0; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st5; - goto st0; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - switch( (*p) ) { - case 69: goto st6; - case 101: goto st6; - } - if ( (*p) > 46 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st5; - } else if ( (*p) >= 45 ) - goto st0; - goto tr7; -tr7: -#line 304 "parser.rl" - { p--; {p++; cs = 10; goto _out;} } - goto st10; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: -#line 949 "parser.c" - goto st0; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - switch( (*p) ) { - case 43: goto st7; - case 45: goto st7; - } - if ( 48 <= (*p) && (*p) <= 57 ) - goto st8; - goto st0; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( 48 <= (*p) && (*p) <= 57 ) - goto st8; - goto st0; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - switch( (*p) ) { - case 69: goto st0; - case 101: goto st0; - } - if ( (*p) > 46 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st8; - } else if ( (*p) >= 45 ) - goto st0; - goto tr7; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: - switch( (*p) ) { - case 46: goto st4; - case 69: goto st6; - case 101: goto st6; - } - if ( 48 <= (*p) && (*p) <= 57 ) - goto st9; - goto st0; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 319 "parser.rl" - - if (cs >= JSON_float_first_final) { - long len = p - json->memo; - *result = rb_Float(rb_str_new(json->memo, len)); - return p + 1; - } else { - return NULL; - } -} - - - -#line 1023 "parser.c" -static const int JSON_array_start = 1; -static const int JSON_array_first_final = 17; -static const int JSON_array_error = 0; - -static const int JSON_array_en_main = 1; - - -#line 355 "parser.rl" - - -static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - VALUE array_class = json->array_class; - - if (json->max_nesting && json->current_nesting > json->max_nesting) { - rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); - } - *result = NIL_P(array_class) ? rb_ary_new() : rb_class_new_instance(0, 0, array_class); - - -#line 1045 "parser.c" - { - cs = JSON_array_start; - } - -#line 368 "parser.rl" - -#line 1052 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - if ( (*p) == 91 ) - goto st2; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - switch( (*p) ) { - case 13: goto st2; - case 32: goto st2; - case 34: goto tr2; - case 45: goto tr2; - case 47: goto st13; - case 73: goto tr2; - case 78: goto tr2; - case 91: goto tr2; - case 93: goto tr4; - case 102: goto tr2; - case 110: goto tr2; - case 116: goto tr2; - case 123: goto tr2; - } - if ( (*p) > 10 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr2; - } else if ( (*p) >= 9 ) - goto st2; - goto st0; -tr2: -#line 336 "parser.rl" - { - VALUE v = Qnil; - char *np = JSON_parse_value(json, p, pe, &v); - if (np == NULL) { - p--; {p++; cs = 3; goto _out;} - } else { - rb_ary_push(*result, v); - {p = (( np))-1;} - } - } - goto st3; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: -#line 1107 "parser.c" - switch( (*p) ) { - case 13: goto st3; - case 32: goto st3; - case 44: goto st4; - case 47: goto st9; - case 93: goto tr4; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st3; - goto st0; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - switch( (*p) ) { - case 13: goto st4; - case 32: goto st4; - case 34: goto tr2; - case 45: goto tr2; - case 47: goto st5; - case 73: goto tr2; - case 78: goto tr2; - case 91: goto tr2; - case 102: goto tr2; - case 110: goto tr2; - case 116: goto tr2; - case 123: goto tr2; - } - if ( (*p) > 10 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto tr2; - } else if ( (*p) >= 9 ) - goto st4; - goto st0; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - switch( (*p) ) { - case 42: goto st6; - case 47: goto st8; - } - goto st0; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - if ( (*p) == 42 ) - goto st7; - goto st6; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - switch( (*p) ) { - case 42: goto st7; - case 47: goto st4; - } - goto st6; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - if ( (*p) == 10 ) - goto st4; - goto st8; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: - switch( (*p) ) { - case 42: goto st10; - case 47: goto st12; - } - goto st0; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: - if ( (*p) == 42 ) - goto st11; - goto st10; -st11: - if ( ++p == pe ) - goto _test_eof11; -case 11: - switch( (*p) ) { - case 42: goto st11; - case 47: goto st3; - } - goto st10; -st12: - if ( ++p == pe ) - goto _test_eof12; -case 12: - if ( (*p) == 10 ) - goto st3; - goto st12; -tr4: -#line 347 "parser.rl" - { p--; {p++; cs = 17; goto _out;} } - goto st17; -st17: - if ( ++p == pe ) - goto _test_eof17; -case 17: -#line 1214 "parser.c" - goto st0; -st13: - if ( ++p == pe ) - goto _test_eof13; -case 13: - switch( (*p) ) { - case 42: goto st14; - case 47: goto st16; - } - goto st0; -st14: - if ( ++p == pe ) - goto _test_eof14; -case 14: - if ( (*p) == 42 ) - goto st15; - goto st14; -st15: - if ( ++p == pe ) - goto _test_eof15; -case 15: - switch( (*p) ) { - case 42: goto st15; - case 47: goto st2; - } - goto st14; -st16: - if ( ++p == pe ) - goto _test_eof16; -case 16: - if ( (*p) == 10 ) - goto st2; - goto st16; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof11: cs = 11; goto _test_eof; - _test_eof12: cs = 12; goto _test_eof; - _test_eof17: cs = 17; goto _test_eof; - _test_eof13: cs = 13; goto _test_eof; - _test_eof14: cs = 14; goto _test_eof; - _test_eof15: cs = 15; goto _test_eof; - _test_eof16: cs = 16; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 369 "parser.rl" - - if(cs >= JSON_array_first_final) { - return p + 1; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } -} - -static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd) -{ - char *p = string, *pe = string, *unescape; - int unescape_len; - - while (pe < stringEnd) { - if (*pe == '\\') { - unescape = (char *) "?"; - unescape_len = 1; - if (pe > p) rb_str_buf_cat(result, p, pe - p); - switch (*++pe) { - case 'n': - unescape = (char *) "\n"; - break; - case 'r': - unescape = (char *) "\r"; - break; - case 't': - unescape = (char *) "\t"; - break; - case '"': - unescape = (char *) "\""; - break; - case '\\': - unescape = (char *) "\\"; - break; - case 'b': - unescape = (char *) "\b"; - break; - case 'f': - unescape = (char *) "\f"; - break; - case 'u': - if (pe > stringEnd - 4) { - return Qnil; - } else { - char buf[4]; - UTF32 ch = unescape_unicode((unsigned char *) ++pe); - pe += 3; - if (UNI_SUR_HIGH_START == (ch & 0xFC00)) { - pe++; - if (pe > stringEnd - 6) return Qnil; - if (pe[0] == '\\' && pe[1] == 'u') { - UTF32 sur = unescape_unicode((unsigned char *) pe + 2); - ch = (((ch & 0x3F) << 10) | ((((ch >> 6) & 0xF) + 1) << 16) - | (sur & 0x3FF)); - pe += 5; - } else { - unescape = (char *) "?"; - break; - } - } - unescape_len = convert_UTF32_to_UTF8(buf, ch); - unescape = buf; - } - break; - default: - p = pe; - continue; - } - rb_str_buf_cat(result, unescape, unescape_len); - p = ++pe; - } else { - pe++; - } - } - rb_str_buf_cat(result, p, pe - p); - return result; -} - - -#line 1350 "parser.c" -static const int JSON_string_start = 1; -static const int JSON_string_first_final = 8; -static const int JSON_string_error = 0; - -static const int JSON_string_en_main = 1; - - -#line 467 "parser.rl" - - -static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - *result = rb_str_buf_new(0); - -#line 1367 "parser.c" - { - cs = JSON_string_start; - } - -#line 475 "parser.rl" - json->memo = p; - -#line 1375 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -case 1: - if ( (*p) == 34 ) - goto st2; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - switch( (*p) ) { - case 34: goto tr2; - case 92: goto st3; - } - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st2; -tr2: -#line 453 "parser.rl" - { - *result = json_string_unescape(*result, json->memo + 1, p); - if (NIL_P(*result)) { - p--; - {p++; cs = 8; goto _out;} - } else { - FORCE_UTF8(*result); - {p = (( p + 1))-1;} - } - } -#line 464 "parser.rl" - { p--; {p++; cs = 8; goto _out;} } - goto st8; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: -#line 1418 "parser.c" - goto st0; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: - if ( (*p) == 117 ) - goto st4; - if ( 0 <= (*p) && (*p) <= 31 ) - goto st0; - goto st2; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st5; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st5; - } else - goto st5; - goto st0; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st6; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st6; - } else - goto st6; - goto st0; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st7; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st7; - } else - goto st7; - goto st0; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( (*p) < 65 ) { - if ( 48 <= (*p) && (*p) <= 57 ) - goto st2; - } else if ( (*p) > 70 ) { - if ( 97 <= (*p) && (*p) <= 102 ) - goto st2; - } else - goto st2; - goto st0; - } - _test_eof2: cs = 2; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 477 "parser.rl" - - if (cs >= JSON_string_first_final) { - return p + 1; - } else { - return NULL; - } -} - - - -#line 1505 "parser.c" -static const int JSON_start = 1; -static const int JSON_first_final = 10; -static const int JSON_error = 0; - -static const int JSON_en_main = 1; - - -#line 511 "parser.rl" - - -/* - * Document-class: JSON::Ext::Parser - * - * This is the JSON parser implemented as a C extension. It can be configured - * to be used by setting - * - * JSON.parser = JSON::Ext::Parser - * - * with the method parser= in JSON. - * - */ - -static VALUE convert_encoding(VALUE source) -{ - char *ptr = RSTRING_PTR(source); - long len = RSTRING_LEN(source); - if (len < 2) { - rb_raise(eParserError, "A JSON text must at least contain two octets!"); - } -#ifdef HAVE_RUBY_ENCODING_H - { - VALUE encoding = rb_funcall(source, i_encoding, 0); - if (encoding == CEncoding_ASCII_8BIT) { - if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32BE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16BE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32LE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16LE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else { - FORCE_UTF8(source); - } - } else { - source = rb_funcall(source, i_encode, 1, CEncoding_UTF_8); - } - } -#else - if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32be"), source); - } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16be"), source); - } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32le"), source); - } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16le"), source); - } -#endif - return source; -} - -/* - * call-seq: new(source, opts => {}) - * - * Creates a new JSON::Ext::Parser instance for the string _source_. - * - * Creates a new JSON::Ext::Parser instance for the string _source_. - * - * It will be configured by the _opts_ hash. _opts_ can have the following - * keys: - * - * _opts_ can have the following keys: - * * *max_nesting*: The maximum depth of nesting allowed in the parsed data - * structures. Disable depth checking with :max_nesting => false|nil|0, it - * defaults to 19. - * * *allow_nan*: If set to true, allow NaN, Infinity and -Infinity in - * defiance of RFC 4627 to be parsed by the Parser. This option defaults to - * false. - * * *create_additions*: If set to false, the Parser doesn't create - * additions even if a matchin class and create_id was found. This option - * defaults to true. - * * *object_class*: Defaults to Hash - * * *array_class*: Defaults to Array - */ -static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self) -{ - char *ptr; - long len; - VALUE source, opts; - GET_PARSER; - rb_scan_args(argc, argv, "11", &source, &opts); - source = convert_encoding(StringValue(source)); - ptr = RSTRING_PTR(source); - len = RSTRING_LEN(source); - if (!NIL_P(opts)) { - opts = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); - if (NIL_P(opts)) { - rb_raise(rb_eArgError, "opts needs to be like a hash"); - } else { - VALUE tmp = ID2SYM(i_max_nesting); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE max_nesting = rb_hash_aref(opts, tmp); - if (RTEST(max_nesting)) { - Check_Type(max_nesting, T_FIXNUM); - json->max_nesting = FIX2INT(max_nesting); - } else { - json->max_nesting = 0; - } - } else { - json->max_nesting = 19; - } - tmp = ID2SYM(i_allow_nan); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE allow_nan = rb_hash_aref(opts, tmp); - json->allow_nan = RTEST(allow_nan) ? 1 : 0; - } else { - json->allow_nan = 0; - } - tmp = ID2SYM(i_create_additions); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE create_additions = rb_hash_aref(opts, tmp); - if (RTEST(create_additions)) { - json->create_id = rb_funcall(mJSON, i_create_id, 0); - } else { - json->create_id = Qnil; - } - } else { - json->create_id = rb_funcall(mJSON, i_create_id, 0); - } - tmp = ID2SYM(i_object_class); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - json->object_class = rb_hash_aref(opts, tmp); - } else { - json->object_class = Qnil; - } - tmp = ID2SYM(i_array_class); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - json->array_class = rb_hash_aref(opts, tmp); - } else { - json->array_class = Qnil; - } - } - } else { - json->max_nesting = 19; - json->allow_nan = 0; - json->create_id = rb_funcall(mJSON, i_create_id, 0); - json->object_class = Qnil; - json->array_class = Qnil; - } - json->current_nesting = 0; - json->len = len; - json->source = ptr; - json->Vsource = source; - return self; -} - -/* - * call-seq: parse() - * - * Parses the current JSON text _source_ and returns the complete data - * structure as a result. - */ -static VALUE cParser_parse(VALUE self) -{ - char *p, *pe; - int cs = EVIL; - VALUE result = Qnil; - GET_PARSER; - - -#line 1685 "parser.c" - { - cs = JSON_start; - } - -#line 682 "parser.rl" - p = json->source; - pe = p + json->len; - -#line 1694 "parser.c" - { - if ( p == pe ) - goto _test_eof; - switch ( cs ) - { -st1: - if ( ++p == pe ) - goto _test_eof1; -case 1: - switch( (*p) ) { - case 13: goto st1; - case 32: goto st1; - case 47: goto st2; - case 91: goto tr3; - case 123: goto tr4; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st1; - goto st0; -st0: -cs = 0; - goto _out; -st2: - if ( ++p == pe ) - goto _test_eof2; -case 2: - switch( (*p) ) { - case 42: goto st3; - case 47: goto st5; - } - goto st0; -st3: - if ( ++p == pe ) - goto _test_eof3; -case 3: - if ( (*p) == 42 ) - goto st4; - goto st3; -st4: - if ( ++p == pe ) - goto _test_eof4; -case 4: - switch( (*p) ) { - case 42: goto st4; - case 47: goto st1; - } - goto st3; -st5: - if ( ++p == pe ) - goto _test_eof5; -case 5: - if ( (*p) == 10 ) - goto st1; - goto st5; -tr3: -#line 500 "parser.rl" - { - char *np; - json->current_nesting = 1; - np = JSON_parse_array(json, p, pe, &result); - if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;} - } - goto st10; -tr4: -#line 493 "parser.rl" - { - char *np; - json->current_nesting = 1; - np = JSON_parse_object(json, p, pe, &result); - if (np == NULL) { p--; {p++; cs = 10; goto _out;} } else {p = (( np))-1;} - } - goto st10; -st10: - if ( ++p == pe ) - goto _test_eof10; -case 10: -#line 1771 "parser.c" - switch( (*p) ) { - case 13: goto st10; - case 32: goto st10; - case 47: goto st6; - } - if ( 9 <= (*p) && (*p) <= 10 ) - goto st10; - goto st0; -st6: - if ( ++p == pe ) - goto _test_eof6; -case 6: - switch( (*p) ) { - case 42: goto st7; - case 47: goto st9; - } - goto st0; -st7: - if ( ++p == pe ) - goto _test_eof7; -case 7: - if ( (*p) == 42 ) - goto st8; - goto st7; -st8: - if ( ++p == pe ) - goto _test_eof8; -case 8: - switch( (*p) ) { - case 42: goto st8; - case 47: goto st10; - } - goto st7; -st9: - if ( ++p == pe ) - goto _test_eof9; -case 9: - if ( (*p) == 10 ) - goto st10; - goto st9; - } - _test_eof1: cs = 1; goto _test_eof; - _test_eof2: cs = 2; goto _test_eof; - _test_eof3: cs = 3; goto _test_eof; - _test_eof4: cs = 4; goto _test_eof; - _test_eof5: cs = 5; goto _test_eof; - _test_eof10: cs = 10; goto _test_eof; - _test_eof6: cs = 6; goto _test_eof; - _test_eof7: cs = 7; goto _test_eof; - _test_eof8: cs = 8; goto _test_eof; - _test_eof9: cs = 9; goto _test_eof; - - _test_eof: {} - _out: {} - } - -#line 685 "parser.rl" - - if (cs >= JSON_first_final && p == pe) { - return result; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } -} - -static JSON_Parser *JSON_allocate() -{ - JSON_Parser *json = ALLOC(JSON_Parser); - MEMZERO(json, JSON_Parser, 1); - return json; -} - -static void JSON_mark(JSON_Parser *json) -{ - rb_gc_mark_maybe(json->Vsource); - rb_gc_mark_maybe(json->create_id); - rb_gc_mark_maybe(json->object_class); - rb_gc_mark_maybe(json->array_class); -} - -static void JSON_free(JSON_Parser *json) -{ - ruby_xfree(json); -} - -static VALUE cJSON_parser_s_allocate(VALUE klass) -{ - JSON_Parser *json = JSON_allocate(); - return Data_Wrap_Struct(klass, JSON_mark, JSON_free, json); -} - -/* - * call-seq: source() - * - * Returns a copy of the current _source_ string, that was used to construct - * this Parser. - */ -static VALUE cParser_source(VALUE self) -{ - GET_PARSER; - return rb_str_dup(json->Vsource); -} - -void Init_parser() -{ - rb_require("json/common"); - mJSON = rb_define_module("JSON"); - mExt = rb_define_module_under(mJSON, "Ext"); - cParser = rb_define_class_under(mExt, "Parser", rb_cObject); - eParserError = rb_path2class("JSON::ParserError"); - eNestingError = rb_path2class("JSON::NestingError"); - rb_define_alloc_func(cParser, cJSON_parser_s_allocate); - rb_define_method(cParser, "initialize", cParser_initialize, -1); - rb_define_method(cParser, "parse", cParser_parse, 0); - rb_define_method(cParser, "source", cParser_source, 0); - - CNaN = rb_const_get(mJSON, rb_intern("NaN")); - CInfinity = rb_const_get(mJSON, rb_intern("Infinity")); - CMinusInfinity = rb_const_get(mJSON, rb_intern("MinusInfinity")); - - i_json_creatable_p = rb_intern("json_creatable?"); - i_json_create = rb_intern("json_create"); - i_create_id = rb_intern("create_id"); - i_create_additions = rb_intern("create_additions"); - i_chr = rb_intern("chr"); - i_max_nesting = rb_intern("max_nesting"); - i_allow_nan = rb_intern("allow_nan"); - i_object_class = rb_intern("object_class"); - i_array_class = rb_intern("array_class"); -#ifdef HAVE_RUBY_ENCODING_H - CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); - CEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be")); - CEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le")); - CEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be")); - CEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le")); - CEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit")); - i_encoding = rb_intern("encoding"); - i_encode = rb_intern("encode"); - i_encode_bang = rb_intern("encode!"); - i_force_encoding = rb_intern("force_encoding"); -#else - i_iconv = rb_intern("iconv"); -#endif -} diff --git a/ext/json/ext/parser/parser.h b/ext/json/ext/parser/parser.h deleted file mode 100644 index 5f2506d..0000000 --- a/ext/json/ext/parser/parser.h +++ /dev/null @@ -1,79 +0,0 @@ -#ifndef _PARSER_H_ -#define _PARSER_H_ - -#include "ruby.h" - -#if HAVE_RE_H -#include "re.h" -#endif - -#if HAVE_RUBY_ST_H -#include "ruby/st.h" -#endif - -#if HAVE_ST_H -#include "st.h" -#endif - -#ifdef HAVE_RUBY_ENCODING_H -#include "ruby/encoding.h" -#define FORCE_UTF8(obj) rb_enc_associate((obj), rb_utf8_encoding()) -#else -#define FORCE_UTF8(obj) -#endif - -#ifndef RHASH_TBL -#define RHASH_TBL(hsh) (RHASH(hsh)->tbl) -#endif - -/* unicode */ - -typedef unsigned long UTF32; /* at least 32 bits */ -typedef unsigned short UTF16; /* at least 16 bits */ -typedef unsigned char UTF8; /* typically 8 bits */ - -#define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD -#define UNI_SUR_HIGH_START (UTF32)0xD800 -#define UNI_SUR_HIGH_END (UTF32)0xDBFF -#define UNI_SUR_LOW_START (UTF32)0xDC00 -#define UNI_SUR_LOW_END (UTF32)0xDFFF - -typedef struct JSON_ParserStruct { - VALUE Vsource; - char *source; - long len; - char *memo; - VALUE create_id; - int max_nesting; - int current_nesting; - int allow_nan; - VALUE object_class; - VALUE array_class; -} JSON_Parser; - -#define GET_PARSER \ - JSON_Parser *json; \ - Data_Get_Struct(self, JSON_Parser, json) - -#define MinusInfinity "-Infinity" -#define EVIL 0x666 - -static UTF32 unescape_unicode(const unsigned char *p); -static int convert_UTF32_to_UTF8(char *buf, UTF32 ch); -static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result); -static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result); -static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result); -static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result); -static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result); -static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd); -static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result); -static VALUE convert_encoding(VALUE source); -static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self); -static VALUE cParser_parse(VALUE self); -static JSON_Parser *JSON_allocate(); -static void JSON_mark(JSON_Parser *json); -static void JSON_free(JSON_Parser *json); -static VALUE cJSON_parser_s_allocate(VALUE klass); -static VALUE cParser_source(VALUE self); - -#endif diff --git a/ext/json/ext/parser/parser.rl b/ext/json/ext/parser/parser.rl deleted file mode 100644 index b91ac00..0000000 --- a/ext/json/ext/parser/parser.rl +++ /dev/null @@ -1,771 +0,0 @@ -#include "parser.h" - -/* unicode */ - -static const char digit_values[256] = { - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, -1, - -1, -1, -1, -1, -1, -1, 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - 10, 11, 12, 13, 14, 15, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, -1, - -1, -1, -1, -1, -1, -1, -1 -}; - -static UTF32 unescape_unicode(const unsigned char *p) -{ - char b; - UTF32 result = 0; - b = digit_values[p[0]]; - if (b < 0) return UNI_REPLACEMENT_CHAR; - result = (result << 4) | b; - b = digit_values[p[1]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - b = digit_values[p[2]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - b = digit_values[p[3]]; - result = (result << 4) | b; - if (b < 0) return UNI_REPLACEMENT_CHAR; - return result; -} - -static int convert_UTF32_to_UTF8(char *buf, UTF32 ch) -{ - int len = 1; - if (ch <= 0x7F) { - buf[0] = (char) ch; - } else if (ch <= 0x07FF) { - buf[0] = (char) ((ch >> 6) | 0xC0); - buf[1] = (char) ((ch & 0x3F) | 0x80); - len++; - } else if (ch <= 0xFFFF) { - buf[0] = (char) ((ch >> 12) | 0xE0); - buf[1] = (char) (((ch >> 6) & 0x3F) | 0x80); - buf[2] = (char) ((ch & 0x3F) | 0x80); - len += 2; - } else if (ch <= 0x1fffff) { - buf[0] =(char) ((ch >> 18) | 0xF0); - buf[1] =(char) (((ch >> 12) & 0x3F) | 0x80); - buf[2] =(char) (((ch >> 6) & 0x3F) | 0x80); - buf[3] =(char) ((ch & 0x3F) | 0x80); - len += 3; - } else { - buf[0] = '?'; - } - return len; -} - - -#ifdef HAVE_RUBY_ENCODING_H -static VALUE CEncoding_ASCII_8BIT, CEncoding_UTF_8, CEncoding_UTF_16BE, - CEncoding_UTF_16LE, CEncoding_UTF_32BE, CEncoding_UTF_32LE; -static ID i_encoding, i_encode, i_encode_bang, i_force_encoding; -#else -static ID i_iconv; -#endif - -static VALUE mJSON, mExt, cParser, eParserError, eNestingError; -static VALUE CNaN, CInfinity, CMinusInfinity; - -static ID i_json_creatable_p, i_json_create, i_create_id, i_create_additions, - i_chr, i_max_nesting, i_allow_nan, i_object_class, i_array_class; - -%%{ - machine JSON_common; - - cr = '\n'; - cr_neg = [^\n]; - ws = [ \t\r\n]; - c_comment = '/*' ( any* - (any* '*/' any* ) ) '*/'; - cpp_comment = '//' cr_neg* cr; - comment = c_comment | cpp_comment; - ignore = ws | comment; - name_separator = ':'; - value_separator = ','; - Vnull = 'null'; - Vfalse = 'false'; - Vtrue = 'true'; - VNaN = 'NaN'; - VInfinity = 'Infinity'; - VMinusInfinity = '-Infinity'; - begin_value = [nft"\-[{NI] | digit; - begin_object = '{'; - end_object = '}'; - begin_array = '['; - end_array = ']'; - begin_string = '"'; - begin_name = begin_string; - begin_number = digit | '-'; -}%% - -%%{ - machine JSON_object; - include JSON_common; - - write data; - - action parse_value { - VALUE v = Qnil; - char *np = JSON_parse_value(json, fpc, pe, &v); - if (np == NULL) { - fhold; fbreak; - } else { - rb_hash_aset(*result, last_name, v); - fexec np; - } - } - - action parse_name { - char *np = JSON_parse_string(json, fpc, pe, &last_name); - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - action exit { fhold; fbreak; } - - a_pair = ignore* begin_name >parse_name - ignore* name_separator ignore* - begin_value >parse_value; - - main := begin_object - (a_pair (ignore* value_separator a_pair)*)? - ignore* end_object @exit; -}%% - -static char *JSON_parse_object(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - VALUE last_name = Qnil; - VALUE object_class = json->object_class; - - if (json->max_nesting && json->current_nesting > json->max_nesting) { - rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); - } - - *result = NIL_P(object_class) ? rb_hash_new() : rb_class_new_instance(0, 0, object_class); - - %% write init; - %% write exec; - - if (cs >= JSON_object_first_final) { - if (RTEST(json->create_id)) { - VALUE klassname = rb_hash_aref(*result, json->create_id); - if (!NIL_P(klassname)) { - VALUE klass = rb_path2class(StringValueCStr(klassname)); - if RTEST(rb_funcall(klass, i_json_creatable_p, 0)) { - *result = rb_funcall(klass, i_json_create, 1, *result); - } - } - } - return p + 1; - } else { - return NULL; - } -} - -%%{ - machine JSON_value; - include JSON_common; - - write data; - - action parse_null { - *result = Qnil; - } - action parse_false { - *result = Qfalse; - } - action parse_true { - *result = Qtrue; - } - action parse_nan { - if (json->allow_nan) { - *result = CNaN; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 2); - } - } - action parse_infinity { - if (json->allow_nan) { - *result = CInfinity; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p - 8); - } - } - action parse_string { - char *np = JSON_parse_string(json, fpc, pe, result); - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - action parse_number { - char *np; - if(pe > fpc + 9 && !strncmp(MinusInfinity, fpc, 9)) { - if (json->allow_nan) { - *result = CMinusInfinity; - fexec p + 10; - fhold; fbreak; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } - } - np = JSON_parse_float(json, fpc, pe, result); - if (np != NULL) fexec np; - np = JSON_parse_integer(json, fpc, pe, result); - if (np != NULL) fexec np; - fhold; fbreak; - } - - action parse_array { - char *np; - json->current_nesting++; - np = JSON_parse_array(json, fpc, pe, result); - json->current_nesting--; - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - action parse_object { - char *np; - json->current_nesting++; - np = JSON_parse_object(json, fpc, pe, result); - json->current_nesting--; - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - action exit { fhold; fbreak; } - -main := ( - Vnull @parse_null | - Vfalse @parse_false | - Vtrue @parse_true | - VNaN @parse_nan | - VInfinity @parse_infinity | - begin_number >parse_number | - begin_string >parse_string | - begin_array >parse_array | - begin_object >parse_object - ) %*exit; -}%% - -static char *JSON_parse_value(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - %% write init; - %% write exec; - - if (cs >= JSON_value_first_final) { - return p; - } else { - return NULL; - } -} - -%%{ - machine JSON_integer; - - write data; - - action exit { fhold; fbreak; } - - main := '-'? ('0' | [1-9][0-9]*) (^[0-9] @exit); -}%% - -static char *JSON_parse_integer(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - %% write init; - json->memo = p; - %% write exec; - - if (cs >= JSON_integer_first_final) { - long len = p - json->memo; - *result = rb_Integer(rb_str_new(json->memo, len)); - return p + 1; - } else { - return NULL; - } -} - -%%{ - machine JSON_float; - include JSON_common; - - write data; - - action exit { fhold; fbreak; } - - main := '-'? ( - (('0' | [1-9][0-9]*) '.' [0-9]+ ([Ee] [+\-]?[0-9]+)?) - | (('0' | [1-9][0-9]*) ([Ee] [+\-]?[0-9]+)) - ) (^[0-9Ee.\-] @exit ); -}%% - -static char *JSON_parse_float(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - %% write init; - json->memo = p; - %% write exec; - - if (cs >= JSON_float_first_final) { - long len = p - json->memo; - *result = rb_Float(rb_str_new(json->memo, len)); - return p + 1; - } else { - return NULL; - } -} - - -%%{ - machine JSON_array; - include JSON_common; - - write data; - - action parse_value { - VALUE v = Qnil; - char *np = JSON_parse_value(json, fpc, pe, &v); - if (np == NULL) { - fhold; fbreak; - } else { - rb_ary_push(*result, v); - fexec np; - } - } - - action exit { fhold; fbreak; } - - next_element = value_separator ignore* begin_value >parse_value; - - main := begin_array ignore* - ((begin_value >parse_value ignore*) - (ignore* next_element ignore*)*)? - end_array @exit; -}%% - -static char *JSON_parse_array(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - VALUE array_class = json->array_class; - - if (json->max_nesting && json->current_nesting > json->max_nesting) { - rb_raise(eNestingError, "nesting of %d is too deep", json->current_nesting); - } - *result = NIL_P(array_class) ? rb_ary_new() : rb_class_new_instance(0, 0, array_class); - - %% write init; - %% write exec; - - if(cs >= JSON_array_first_final) { - return p + 1; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } -} - -static VALUE json_string_unescape(VALUE result, char *string, char *stringEnd) -{ - char *p = string, *pe = string, *unescape; - int unescape_len; - - while (pe < stringEnd) { - if (*pe == '\\') { - unescape = (char *) "?"; - unescape_len = 1; - if (pe > p) rb_str_buf_cat(result, p, pe - p); - switch (*++pe) { - case 'n': - unescape = (char *) "\n"; - break; - case 'r': - unescape = (char *) "\r"; - break; - case 't': - unescape = (char *) "\t"; - break; - case '"': - unescape = (char *) "\""; - break; - case '\\': - unescape = (char *) "\\"; - break; - case 'b': - unescape = (char *) "\b"; - break; - case 'f': - unescape = (char *) "\f"; - break; - case 'u': - if (pe > stringEnd - 4) { - return Qnil; - } else { - char buf[4]; - UTF32 ch = unescape_unicode((unsigned char *) ++pe); - pe += 3; - if (UNI_SUR_HIGH_START == (ch & 0xFC00)) { - pe++; - if (pe > stringEnd - 6) return Qnil; - if (pe[0] == '\\' && pe[1] == 'u') { - UTF32 sur = unescape_unicode((unsigned char *) pe + 2); - ch = (((ch & 0x3F) << 10) | ((((ch >> 6) & 0xF) + 1) << 16) - | (sur & 0x3FF)); - pe += 5; - } else { - unescape = (char *) "?"; - break; - } - } - unescape_len = convert_UTF32_to_UTF8(buf, ch); - unescape = buf; - } - break; - default: - p = pe; - continue; - } - rb_str_buf_cat(result, unescape, unescape_len); - p = ++pe; - } else { - pe++; - } - } - rb_str_buf_cat(result, p, pe - p); - return result; -} - -%%{ - machine JSON_string; - include JSON_common; - - write data; - - action parse_string { - *result = json_string_unescape(*result, json->memo + 1, p); - if (NIL_P(*result)) { - fhold; - fbreak; - } else { - FORCE_UTF8(*result); - fexec p + 1; - } - } - - action exit { fhold; fbreak; } - - main := '"' ((^(["\\] | 0..0x1f) | '\\'["\\/bfnrt] | '\\u'[0-9a-fA-F]{4} | '\\'^(["\\/bfnrtu]|0..0x1f))* %parse_string) '"' @exit; -}%% - -static char *JSON_parse_string(JSON_Parser *json, char *p, char *pe, VALUE *result) -{ - int cs = EVIL; - - *result = rb_str_buf_new(0); - %% write init; - json->memo = p; - %% write exec; - - if (cs >= JSON_string_first_final) { - return p + 1; - } else { - return NULL; - } -} - - -%%{ - machine JSON; - - write data; - - include JSON_common; - - action parse_object { - char *np; - json->current_nesting = 1; - np = JSON_parse_object(json, fpc, pe, &result); - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - action parse_array { - char *np; - json->current_nesting = 1; - np = JSON_parse_array(json, fpc, pe, &result); - if (np == NULL) { fhold; fbreak; } else fexec np; - } - - main := ignore* ( - begin_object >parse_object | - begin_array >parse_array - ) ignore*; -}%% - -/* - * Document-class: JSON::Ext::Parser - * - * This is the JSON parser implemented as a C extension. It can be configured - * to be used by setting - * - * JSON.parser = JSON::Ext::Parser - * - * with the method parser= in JSON. - * - */ - -static VALUE convert_encoding(VALUE source) -{ - char *ptr = RSTRING_PTR(source); - long len = RSTRING_LEN(source); - if (len < 2) { - rb_raise(eParserError, "A JSON text must at least contain two octets!"); - } -#ifdef HAVE_RUBY_ENCODING_H - { - VALUE encoding = rb_funcall(source, i_encoding, 0); - if (encoding == CEncoding_ASCII_8BIT) { - if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32BE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16BE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_32LE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { - source = rb_str_dup(source); - rb_funcall(source, i_force_encoding, 1, CEncoding_UTF_16LE); - source = rb_funcall(source, i_encode_bang, 1, CEncoding_UTF_8); - } else { - FORCE_UTF8(source); - } - } else { - source = rb_funcall(source, i_encode, 1, CEncoding_UTF_8); - } - } -#else - if (len >= 4 && ptr[0] == 0 && ptr[1] == 0 && ptr[2] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32be"), source); - } else if (len >= 4 && ptr[0] == 0 && ptr[2] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16be"), source); - } else if (len >= 4 && ptr[1] == 0 && ptr[2] == 0 && ptr[3] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-32le"), source); - } else if (len >= 4 && ptr[1] == 0 && ptr[3] == 0) { - source = rb_funcall(mJSON, i_iconv, 3, rb_str_new2("utf-8"), rb_str_new2("utf-16le"), source); - } -#endif - return source; -} - -/* - * call-seq: new(source, opts => {}) - * - * Creates a new JSON::Ext::Parser instance for the string _source_. - * - * Creates a new JSON::Ext::Parser instance for the string _source_. - * - * It will be configured by the _opts_ hash. _opts_ can have the following - * keys: - * - * _opts_ can have the following keys: - * * *max_nesting*: The maximum depth of nesting allowed in the parsed data - * structures. Disable depth checking with :max_nesting => false|nil|0, it - * defaults to 19. - * * *allow_nan*: If set to true, allow NaN, Infinity and -Infinity in - * defiance of RFC 4627 to be parsed by the Parser. This option defaults to - * false. - * * *create_additions*: If set to false, the Parser doesn't create - * additions even if a matchin class and create_id was found. This option - * defaults to true. - * * *object_class*: Defaults to Hash - * * *array_class*: Defaults to Array - */ -static VALUE cParser_initialize(int argc, VALUE *argv, VALUE self) -{ - char *ptr; - long len; - VALUE source, opts; - GET_PARSER; - rb_scan_args(argc, argv, "11", &source, &opts); - source = convert_encoding(StringValue(source)); - ptr = RSTRING_PTR(source); - len = RSTRING_LEN(source); - if (!NIL_P(opts)) { - opts = rb_convert_type(opts, T_HASH, "Hash", "to_hash"); - if (NIL_P(opts)) { - rb_raise(rb_eArgError, "opts needs to be like a hash"); - } else { - VALUE tmp = ID2SYM(i_max_nesting); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE max_nesting = rb_hash_aref(opts, tmp); - if (RTEST(max_nesting)) { - Check_Type(max_nesting, T_FIXNUM); - json->max_nesting = FIX2INT(max_nesting); - } else { - json->max_nesting = 0; - } - } else { - json->max_nesting = 19; - } - tmp = ID2SYM(i_allow_nan); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE allow_nan = rb_hash_aref(opts, tmp); - json->allow_nan = RTEST(allow_nan) ? 1 : 0; - } else { - json->allow_nan = 0; - } - tmp = ID2SYM(i_create_additions); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - VALUE create_additions = rb_hash_aref(opts, tmp); - if (RTEST(create_additions)) { - json->create_id = rb_funcall(mJSON, i_create_id, 0); - } else { - json->create_id = Qnil; - } - } else { - json->create_id = rb_funcall(mJSON, i_create_id, 0); - } - tmp = ID2SYM(i_object_class); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - json->object_class = rb_hash_aref(opts, tmp); - } else { - json->object_class = Qnil; - } - tmp = ID2SYM(i_array_class); - if (st_lookup(RHASH_TBL(opts), tmp, 0)) { - json->array_class = rb_hash_aref(opts, tmp); - } else { - json->array_class = Qnil; - } - } - } else { - json->max_nesting = 19; - json->allow_nan = 0; - json->create_id = rb_funcall(mJSON, i_create_id, 0); - json->object_class = Qnil; - json->array_class = Qnil; - } - json->current_nesting = 0; - json->len = len; - json->source = ptr; - json->Vsource = source; - return self; -} - -/* - * call-seq: parse() - * - * Parses the current JSON text _source_ and returns the complete data - * structure as a result. - */ -static VALUE cParser_parse(VALUE self) -{ - char *p, *pe; - int cs = EVIL; - VALUE result = Qnil; - GET_PARSER; - - %% write init; - p = json->source; - pe = p + json->len; - %% write exec; - - if (cs >= JSON_first_final && p == pe) { - return result; - } else { - rb_raise(eParserError, "%u: unexpected token at '%s'", __LINE__, p); - } -} - -static JSON_Parser *JSON_allocate() -{ - JSON_Parser *json = ALLOC(JSON_Parser); - MEMZERO(json, JSON_Parser, 1); - return json; -} - -static void JSON_mark(JSON_Parser *json) -{ - rb_gc_mark_maybe(json->Vsource); - rb_gc_mark_maybe(json->create_id); - rb_gc_mark_maybe(json->object_class); - rb_gc_mark_maybe(json->array_class); -} - -static void JSON_free(JSON_Parser *json) -{ - ruby_xfree(json); -} - -static VALUE cJSON_parser_s_allocate(VALUE klass) -{ - JSON_Parser *json = JSON_allocate(); - return Data_Wrap_Struct(klass, JSON_mark, JSON_free, json); -} - -/* - * call-seq: source() - * - * Returns a copy of the current _source_ string, that was used to construct - * this Parser. - */ -static VALUE cParser_source(VALUE self) -{ - GET_PARSER; - return rb_str_dup(json->Vsource); -} - -void Init_parser() -{ - rb_require("json/common"); - mJSON = rb_define_module("JSON"); - mExt = rb_define_module_under(mJSON, "Ext"); - cParser = rb_define_class_under(mExt, "Parser", rb_cObject); - eParserError = rb_path2class("JSON::ParserError"); - eNestingError = rb_path2class("JSON::NestingError"); - rb_define_alloc_func(cParser, cJSON_parser_s_allocate); - rb_define_method(cParser, "initialize", cParser_initialize, -1); - rb_define_method(cParser, "parse", cParser_parse, 0); - rb_define_method(cParser, "source", cParser_source, 0); - - CNaN = rb_const_get(mJSON, rb_intern("NaN")); - CInfinity = rb_const_get(mJSON, rb_intern("Infinity")); - CMinusInfinity = rb_const_get(mJSON, rb_intern("MinusInfinity")); - - i_json_creatable_p = rb_intern("json_creatable?"); - i_json_create = rb_intern("json_create"); - i_create_id = rb_intern("create_id"); - i_create_additions = rb_intern("create_additions"); - i_chr = rb_intern("chr"); - i_max_nesting = rb_intern("max_nesting"); - i_allow_nan = rb_intern("allow_nan"); - i_object_class = rb_intern("object_class"); - i_array_class = rb_intern("array_class"); -#ifdef HAVE_RUBY_ENCODING_H - CEncoding_UTF_8 = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-8")); - CEncoding_UTF_16BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16be")); - CEncoding_UTF_16LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-16le")); - CEncoding_UTF_32BE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32be")); - CEncoding_UTF_32LE = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("utf-32le")); - CEncoding_ASCII_8BIT = rb_funcall(rb_path2class("Encoding"), rb_intern("find"), 1, rb_str_new2("ascii-8bit")); - i_encoding = rb_intern("encoding"); - i_encode = rb_intern("encode"); - i_encode_bang = rb_intern("encode!"); - i_force_encoding = rb_intern("force_encoding"); -#else - i_iconv = rb_intern("iconv"); -#endif -} -- cgit v1.2.1