조성현 / csh-parser

Go to a project
Toggle navigation Toggle navigation pinning
조성현
8f14a245 0 parents

init

Inline Side-by-side
Showing 4 changed files with 317 additions and 0 deletions
1 +
2 +Microsoft Visual Studio Solution File, Format Version 12.00
3 +# Visual Studio 14
4 +VisualStudioVersion = 14.0.25420.1
5 +MinimumVisualStudioVersion = 10.0.40219.1
6 +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "dblpJsonParser", "dblpJsonParser\dblpJsonParser.vcxproj", "{D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}"
7 +EndProject
8 +Global
9 + GlobalSection(SolutionConfigurationPlatforms) = preSolution
10 + Debug|x64 = Debug|x64
11 + Debug|x86 = Debug|x86
12 + Release|x64 = Release|x64
13 + Release|x86 = Release|x86
14 + EndGlobalSection
15 + GlobalSection(ProjectConfigurationPlatforms) = postSolution
16 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Debug|x64.ActiveCfg = Debug|x64
17 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Debug|x64.Build.0 = Debug|x64
18 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Debug|x86.ActiveCfg = Debug|Win32
19 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Debug|x86.Build.0 = Debug|Win32
20 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Release|x64.ActiveCfg = Release|x64
21 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Release|x64.Build.0 = Release|x64
22 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Release|x86.ActiveCfg = Release|Win32
23 + {D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}.Release|x86.Build.0 = Release|Win32
24 + EndGlobalSection
25 + GlobalSection(SolutionProperties) = preSolution
26 + HideSolutionNode = FALSE
27 + EndGlobalSection
28 +EndGlobal
1 +<?xml version="1.0" encoding="utf-8"?>
2 +<Project DefaultTargets="Build" ToolsVersion="14.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
3 + <ItemGroup Label="ProjectConfigurations">
4 + <ProjectConfiguration Include="Debug|Win32">
5 + <Configuration>Debug</Configuration>
6 + <Platform>Win32</Platform>
7 + </ProjectConfiguration>
8 + <ProjectConfiguration Include="Release|Win32">
9 + <Configuration>Release</Configuration>
10 + <Platform>Win32</Platform>
11 + </ProjectConfiguration>
12 + <ProjectConfiguration Include="Debug|x64">
13 + <Configuration>Debug</Configuration>
14 + <Platform>x64</Platform>
15 + </ProjectConfiguration>
16 + <ProjectConfiguration Include="Release|x64">
17 + <Configuration>Release</Configuration>
18 + <Platform>x64</Platform>
19 + </ProjectConfiguration>
20 + </ItemGroup>
21 + <PropertyGroup Label="Globals">
22 + <ProjectGuid>{D0FA588C-B5C0-422F-8CA1-4FD40054BD1C}</ProjectGuid>
23 + <Keyword>Win32Proj</Keyword>
24 + <RootNamespace>dblpJsonParser</RootNamespace>
25 + <WindowsTargetPlatformVersion>8.1</WindowsTargetPlatformVersion>
26 + </PropertyGroup>
27 + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.Default.props" />
28 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'" Label="Configuration">
29 + <ConfigurationType>Application</ConfigurationType>
30 + <UseDebugLibraries>true</UseDebugLibraries>
31 + <PlatformToolset>v140</PlatformToolset>
32 + <CharacterSet>Unicode</CharacterSet>
33 + </PropertyGroup>
34 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'" Label="Configuration">
35 + <ConfigurationType>Application</ConfigurationType>
36 + <UseDebugLibraries>false</UseDebugLibraries>
37 + <PlatformToolset>v140</PlatformToolset>
38 + <WholeProgramOptimization>true</WholeProgramOptimization>
39 + <CharacterSet>Unicode</CharacterSet>
40 + </PropertyGroup>
41 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'" Label="Configuration">
42 + <ConfigurationType>Application</ConfigurationType>
43 + <UseDebugLibraries>true</UseDebugLibraries>
44 + <PlatformToolset>v140</PlatformToolset>
45 + <CharacterSet>Unicode</CharacterSet>
46 + </PropertyGroup>
47 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'" Label="Configuration">
48 + <ConfigurationType>Application</ConfigurationType>
49 + <UseDebugLibraries>false</UseDebugLibraries>
50 + <PlatformToolset>v140</PlatformToolset>
51 + <WholeProgramOptimization>true</WholeProgramOptimization>
52 + <CharacterSet>Unicode</CharacterSet>
53 + </PropertyGroup>
54 + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.props" />
55 + <ImportGroup Label="ExtensionSettings">
56 + </ImportGroup>
57 + <ImportGroup Label="Shared">
58 + </ImportGroup>
59 + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
60 + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
61 + </ImportGroup>
62 + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
63 + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
64 + </ImportGroup>
65 + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
66 + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
67 + </ImportGroup>
68 + <ImportGroup Label="PropertySheets" Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
69 + <Import Project="$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props" Condition="exists('$(UserRootDir)\Microsoft.Cpp.$(Platform).user.props')" Label="LocalAppDataPlatform" />
70 + </ImportGroup>
71 + <PropertyGroup Label="UserMacros" />
72 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
73 + <LinkIncremental>true</LinkIncremental>
74 + <IncludePath>C:\boost\boost_1_62_0;C:\JsonCpp\jsoncpp-master\include;$(IncludePath)</IncludePath>
75 + <LibraryPath>C:\boost\boost_1_62_0\stage\lib;C:\JsonCpp\jsoncpp-master\build\vs71\debug\lib_json;$(LibraryPath)</LibraryPath>
76 + </PropertyGroup>
77 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
78 + <LinkIncremental>true</LinkIncremental>
79 + <IncludePath>C:\boost\boost_1_62_0;C:\JsonCpp\jsoncpp-master\include;$(IncludePath)</IncludePath>
80 + <LibraryPath>C:\boost\boost_1_62_0\stage\lib;C:\JsonCpp\jsoncpp-master\build\vs71\debug\lib_json;$(LibraryPath)</LibraryPath>
81 + </PropertyGroup>
82 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
83 + <LinkIncremental>false</LinkIncremental>
84 + </PropertyGroup>
85 + <PropertyGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
86 + <LinkIncremental>false</LinkIncremental>
87 + </PropertyGroup>
88 + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|Win32'">
89 + <ClCompile>
90 + <PrecompiledHeader>
91 + </PrecompiledHeader>
92 + <WarningLevel>Level3</WarningLevel>
93 + <Optimization>Disabled</Optimization>
94 + <PreprocessorDefinitions>WIN32;_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
95 + <SDLCheck>true</SDLCheck>
96 + <AdditionalIncludeDirectories>%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
97 + </ClCompile>
98 + <Link>
99 + <SubSystem>Console</SubSystem>
100 + <GenerateDebugInformation>true</GenerateDebugInformation>
101 + <AdditionalLibraryDirectories>%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
102 + <AdditionalDependencies>json_vc71_libmtd.lib;%(AdditionalDependencies)</AdditionalDependencies>
103 + </Link>
104 + </ItemDefinitionGroup>
105 + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Debug|x64'">
106 + <ClCompile>
107 + <PrecompiledHeader>
108 + </PrecompiledHeader>
109 + <WarningLevel>Level3</WarningLevel>
110 + <Optimization>Disabled</Optimization>
111 + <PreprocessorDefinitions>_DEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
112 + <SDLCheck>true</SDLCheck>
113 + <AdditionalIncludeDirectories>%(AdditionalIncludeDirectories)</AdditionalIncludeDirectories>
114 + </ClCompile>
115 + <Link>
116 + <SubSystem>Console</SubSystem>
117 + <GenerateDebugInformation>true</GenerateDebugInformation>
118 + <AdditionalLibraryDirectories>%(AdditionalLibraryDirectories)</AdditionalLibraryDirectories>
119 + <AdditionalDependencies>json_vc71_libmtd.lib;%(AdditionalDependencies)</AdditionalDependencies>
120 + </Link>
121 + </ItemDefinitionGroup>
122 + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|Win32'">
123 + <ClCompile>
124 + <WarningLevel>Level3</WarningLevel>
125 + <PrecompiledHeader>
126 + </PrecompiledHeader>
127 + <Optimization>MaxSpeed</Optimization>
128 + <FunctionLevelLinking>true</FunctionLevelLinking>
129 + <IntrinsicFunctions>true</IntrinsicFunctions>
130 + <PreprocessorDefinitions>WIN32;NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
131 + <SDLCheck>true</SDLCheck>
132 + </ClCompile>
133 + <Link>
134 + <SubSystem>Console</SubSystem>
135 + <EnableCOMDATFolding>true</EnableCOMDATFolding>
136 + <OptimizeReferences>true</OptimizeReferences>
137 + <GenerateDebugInformation>true</GenerateDebugInformation>
138 + </Link>
139 + </ItemDefinitionGroup>
140 + <ItemDefinitionGroup Condition="'$(Configuration)|$(Platform)'=='Release|x64'">
141 + <ClCompile>
142 + <WarningLevel>Level3</WarningLevel>
143 + <PrecompiledHeader>
144 + </PrecompiledHeader>
145 + <Optimization>MaxSpeed</Optimization>
146 + <FunctionLevelLinking>true</FunctionLevelLinking>
147 + <IntrinsicFunctions>true</IntrinsicFunctions>
148 + <PreprocessorDefinitions>NDEBUG;_CONSOLE;%(PreprocessorDefinitions)</PreprocessorDefinitions>
149 + <SDLCheck>true</SDLCheck>
150 + </ClCompile>
151 + <Link>
152 + <SubSystem>Console</SubSystem>
153 + <EnableCOMDATFolding>true</EnableCOMDATFolding>
154 + <OptimizeReferences>true</OptimizeReferences>
155 + <GenerateDebugInformation>true</GenerateDebugInformation>
156 + </Link>
157 + </ItemDefinitionGroup>
158 + <ItemGroup>
159 + <ClCompile Include="main.cpp" />
160 + </ItemGroup>
161 + <Import Project="$(VCTargetsPath)\Microsoft.Cpp.targets" />
162 + <ImportGroup Label="ExtensionTargets">
163 + </ImportGroup>
164 +</Project>
...\ No newline at end of file ...\ No newline at end of file
1 +<?xml version="1.0" encoding="utf-8"?>
2 +<Project ToolsVersion="4.0" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
3 + <ItemGroup>
4 + <Filter Include="소스 파일">
5 + <UniqueIdentifier>{4FC737F1-C7A5-4376-A066-2A32D752A2FF}</UniqueIdentifier>
6 + <Extensions>cpp;c;cc;cxx;def;odl;idl;hpj;bat;asm;asmx</Extensions>
7 + </Filter>
8 + <Filter Include="헤더 파일">
9 + <UniqueIdentifier>{93995380-89BD-4b04-88EB-625FBE52EBFB}</UniqueIdentifier>
10 + <Extensions>h;hh;hpp;hxx;hm;inl;inc;xsd</Extensions>
11 + </Filter>
12 + <Filter Include="리소스 파일">
13 + <UniqueIdentifier>{67DA6AB6-F800-4c08-8B7A-83BB121AAD01}</UniqueIdentifier>
14 + <Extensions>rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav;mfcribbon-ms</Extensions>
15 + </Filter>
16 + </ItemGroup>
17 + <ItemGroup>
18 + <ClCompile Include="main.cpp">
19 + <Filter>소스 파일</Filter>
20 + </ClCompile>
21 + </ItemGroup>
22 +</Project>
...\ No newline at end of file ...\ No newline at end of file
1 +#include <cstdio>
2 +#include <exception>
3 +#include <fstream>
4 +#include <iostream>
5 +#include <string>
6 +#include <vector>
7 +
8 +#include <boost/regex.hpp>
9 +
10 +#include <json/json.h>
11 +
12 +using namespace std;
13 +
14 +const char* DBLP_FILENAME = "test.json";
15 +const char* DBLP_COAUTHOR_FILENAME = "tmp_dblp_coauthorship.json";
16 +const string COLUMN_DELIMITER = "||";
17 +const string AUTHOR_DELIMITER = "&&";
18 +
19 +int main() {
20 + //init
21 + Json::Value root;
22 + Json::Reader reader;
23 + ifstream dblp_paper, dblp_coauthor;
24 + ofstream dblp_paper_out, dblp_coauthor_out;
25 +
26 + try {
27 + //1. dblp paper dataset
28 + dblp_paper.open(DBLP_FILENAME);
29 + dblp_paper_out.open((string(DBLP_FILENAME)+string(".out")).c_str());
30 + if (!dblp_paper || !dblp_paper_out) {
31 + throw exception("dblp paper file error");
32 + }
33 + printf("* complete open\n");
34 +
35 + if (!reader.parse(dblp_paper, root)) {
36 + throw exception("parse error");
37 + }
38 + printf("* complete parse\n");
39 +
40 +
41 + Json::Value row;
42 + Json::Value coauthors;
43 +
44 + std::string paper_key;
45 + std::vector<string> coauthor_list;
46 + int year;
47 +
48 + int count = 1;
49 + for (auto it=root.begin();
50 + it!=root.end();
51 + ++it) {
52 + //Àüó¸®
53 + row.clear();
54 + coauthors.clear();
55 + coauthor_list.clear();
56 +
57 + //print
58 + if (count%1000000 == 0) {
59 + printf("* [%d]\n", count);
60 + }
61 +
62 + //row ´ÜÀ§·Î read
63 + row = *it;
64 + paper_key = row[0].asString();
65 + coauthors = row[1];
66 + for (auto coit=coauthors.begin(); coit!=coauthors.end(); ++coit) {
67 + coauthor_list.push_back(coit->asString());
68 + }
69 + year = ((row[2].isNull())?-1:row[2].asInt());
70 +
71 + //write
72 + dblp_paper_out << paper_key << COLUMN_DELIMITER;
73 + for (auto auit=coauthor_list.begin(); auit!=coauthor_list.end(); ++auit) {
74 + dblp_paper_out << (*auit);
75 + if () {
76 + dblp_paper_out << AUTHOR_DELIMITER;
77 + }
78 + }
79 + dblp_paper_out << endl;
80 +
81 + //ÈÄó¸®
82 + ++count;
83 + }
84 + printf("* complete convert dblp paper\n");
85 +
86 +
87 + //2. dblp coauthorship dataset
88 + //dblp_coauthor.open(DBLP_COAUTHOR_FILENAME);
89 +
90 +
91 + }
92 + catch (const exception& e) {
93 + cerr << "Error: " << e.what() << endl;
94 + return -1;
95 + }
96 +
97 + //release
98 + dblp_paper.close();
99 + dblp_coauthor.close();
100 + dblp_paper_out.close();
101 + dblp_coauthor_out.close();
102 + return 0;
103 +}
...\ No newline at end of file ...\ No newline at end of file