misc/mysql_splitter/splitter.cpp

#include <iostream>
#include <fstream>
#include <sstream>
#include <string>
#include <vector>


#include <boost/algorithm/string.hpp>
#include <boost/filesystem.hpp>
#include <boost/program_options.hpp>

namespace po = boost::program_options;

using namespace std;

const int BUFSIZE = 1024*1024;

string remove_comments(string in) {
        bool isComment = false;
        ostringstream out;
        for (int i=0; i<in.length() -1; i++) {
                if (in.at(i) == '/' && in.at(i+1) == '*')  
                        isComment = true;
                
                if (in.at(i) == '*' && in.at(i+1) == '/') {
                        i++;
                        isComment = false;
                        continue;
                }

                if (isComment == false)
                        out << in.at(i) ;
        }

        return out.str();
}

string getLastWord(string input) {
        input = remove_comments(input);

        boost::erase_all(input,"`");
        boost::trim(input);
        vector<string> words;
        words = boost::split(words, input, boost::is_any_of(" ") );

        string last = words.back();

        boost::erase_all(last, ";");
        boost::trim(last);


        return last;
}


int main(int argc, char** argv) {

        po::options_description desc("Mysql dump file splitter:\nAllowed options");
        desc.add_options()
        ("help", "produce help message")
        ("data", "ignore ceate tablestatements")
                ("input-file", po::value< string >(), "input file")
        ;

        po::positional_options_description p;
        p.add("input-file", -1);


        po::variables_map vm;
        po::store( po::command_line_parser(argc, argv).options(desc).positional(p).run(), vm);
        po::notify(vm);

        if (vm.count("help")) {
        cout << desc << "\n";
        return 1;
        }


        if ( vm.count("input-file") == 0) {
                cout << desc << endl;
                return 1;
        }


        char linebuf[BUFSIZE];

        string inputfile = vm["input-file"].as< string >();
        if (! boost::filesystem::exists(inputfile) ) {
                cout << "No file named " << inputfile << endl;
                return 1;
        }
        

        ifstream in( inputfile.c_str() );
        ofstream out;

        ostringstream header;

        if (!in.is_open() ) {
                cout << "Could not open " << argv[1] << endl;
                return 2;
        }

        while ( in.good() ) {
                in.getline(linebuf, BUFSIZE);

                string line(linebuf);


                //if (line.substr(0, 15) == "CREATE DATABASE" ) {
                if ( boost::starts_with(line, "CREATE DATABASE") ) {

                        if (out.is_open() ) {
                                out.close();
                        }

                        boost::trim(line);
                        string dbname = getLastWord(line);
                        cout << ">" << dbname << endl;


                        ostringstream oss;
                        oss << dbname << "_dump.sql";

            string filename = oss.str();

                        bool did_exist = boost::filesystem::exists(filename);

                        out.open( filename.c_str(), ios::app );


                        if (!out.is_open() ) {
                                cout << "could not create outfile " << filename << endl;
                                return 3;
                        }

                        if (!did_exist) {
                                out << header.str() << endl; //write preamble in new file
                        }

                }


                if (out.is_open() ) {
                        out << line << endl;
                } else {
                        header << line << endl; //collect preamble for later use
                }

        }


        if(out.is_open())
                out.close();


        in.close();
}
1	torben	1231	#include <iostream>
2			#include <fstream>
3			#include <sstream>
4			#include <string>
5			#include <vector>
6
7	torben	1282
8	torben	1231	#include <boost/algorithm/string.hpp>
9			#include <boost/filesystem.hpp>
10	torben	1282	#include <boost/program_options.hpp>
11	torben	1231
12	torben	1282	namespace po = boost::program_options;
13
14	torben	1231	using namespace std;
15
16			const int BUFSIZE = 1024*1024;
17
18	torben	1283	string remove_comments(string in) {
19			bool isComment = false;
20			ostringstream out;
21			for (int i=0; i<in.length() -1; i++) {
22			if (in.at(i) == '/' && in.at(i+1) == '*')
23			isComment = true;
24
25			if (in.at(i) == '*' && in.at(i+1) == '/') {
26			i++;
27			isComment = false;
28			continue;
29			}
30
31			if (isComment == false)
32			out << in.at(i) ;
33			}
34
35			return out.str();
36			}
37
38	torben	1231	string getLastWord(string input) {
39	torben	1283	input = remove_comments(input);
40
41			boost::erase_all(input,"`");
42			boost::trim(input);
43	torben	1231	vector<string> words;
44			words = boost::split(words, input, boost::is_any_of(" ") );
45
46			string last = words.back();
47
48			boost::erase_all(last, ";");
49			boost::trim(last);
50
51
52			return last;
53			}
54
55
56			int main(int argc, char** argv) {
57	torben	1282
58			po::options_description desc("Mysql dump file splitter:\nAllowed options");
59			desc.add_options()
60			("help", "produce help message")
61			("data", "ignore ceate tablestatements")
62			("input-file", po::value< string >(), "input file")
63			;
64
65			po::positional_options_description p;
66			p.add("input-file", -1);
67
68
69			po::variables_map vm;
70			po::store( po::command_line_parser(argc, argv).options(desc).positional(p).run(), vm);
71			po::notify(vm);
72
73			if (vm.count("help")) {
74			cout << desc << "\n";
75			return 1;
76			}
77
78
79
80			if ( vm.count("input-file") == 0) {
81			cout << desc << endl;
82	torben	1231	return 1;
83			}
84
85	torben	1282
86	torben	1231	char linebuf[BUFSIZE];
87
88	torben	1282	string inputfile = vm["input-file"].as< string >();
89			if (! boost::filesystem::exists(inputfile) ) {
90			cout << "No file named " << inputfile << endl;
91			return 1;
92			}
93
94
95			ifstream in( inputfile.c_str() );
96	torben	1231	ofstream out;
97
98			ostringstream header;
99
100			if (!in.is_open() ) {
101			cout << "Could not open " << argv[1] << endl;
102			return 2;
103			}
104
105			while ( in.good() ) {
106			in.getline(linebuf, BUFSIZE);
107
108			string line(linebuf);
109
110
111			//if (line.substr(0, 15) == "CREATE DATABASE" ) {
112			if ( boost::starts_with(line, "CREATE DATABASE") ) {
113
114			if (out.is_open() ) {
115			out.close();
116			}
117
118			boost::trim(line);
119			string dbname = getLastWord(line);
120			cout << ">" << dbname << endl;
121
122
123			ostringstream oss;
124			oss << dbname << "_dump.sql";
125
126			string filename = oss.str();
127
128			bool did_exist = boost::filesystem::exists(filename);
129
130			out.open( filename.c_str(), ios::app );
131
132
133
134			if (!out.is_open() ) {
135			cout << "could not create outfile " << filename << endl;
136			return 3;
137			}
138
139			if (!did_exist) {
140			out << header.str() << endl; //write preamble in new file
141			}
142
143			}
144
145
146			if (out.is_open() ) {
147			out << line << endl;
148			} else {
149			header << line << endl; //collect preamble for later use
150			}
151
152			}
153
154
155			if(out.is_open())
156			out.close();
157
158
159			in.close();
160			}