src/CSVHandler.h
author František Kučera <franta-hg@frantovo.cz>
Thu, 10 Jan 2019 00:21:38 +0100
branchv_0
changeset 3 b5a5bf32f7ff
parent 1 82f86dc48339
child 6 de4c706edf41
permissions -rw-r--r--
generate CRLF line ends according to RFC 4180
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
0
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     1
/**
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     2
 * Relational pipes
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     3
 * Copyright © 2018 František Kučera (Frantovo.cz, GlobalCode.info)
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     4
 *
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     5
 * This program is free software: you can redistribute it and/or modify
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     6
 * it under the terms of the GNU General Public License as published by
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     7
 * the Free Software Foundation, either version 3 of the License, or
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     8
 * (at your option) any later version.
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     9
 *
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    10
 * This program is distributed in the hope that it will be useful,
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    11
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    12
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    13
 * GNU General Public License for more details.
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    14
 *
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    15
 * You should have received a copy of the GNU General Public License
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    16
 * along with this program. If not, see <http://www.gnu.org/licenses/>.
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    17
 */
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    18
#pragma once
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    19
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    20
#include <memory>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    21
#include <string>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    22
#include <vector>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    23
#include <iostream>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    24
#include <sstream>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    25
#include <locale>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    26
#include <codecvt>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    27
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    28
#include <relpipe/reader/typedefs.h>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    29
#include <relpipe/reader/TypeId.h>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    30
#include <relpipe/reader/handlers/RelationalReaderStringHandler.h>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    31
#include <relpipe/reader/handlers/AttributeMetadata.h>
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    32
1
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    33
#include "RelpipeCSVWriterException.h"
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    34
0
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    35
namespace relpipe {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    36
namespace out {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    37
namespace csv {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    38
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    39
using namespace relpipe;
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    40
using namespace relpipe::reader;
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    41
using namespace relpipe::reader::handlers;
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    42
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    43
class CSVHandler : public RelationalReaderStringHadler {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    44
private:
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    45
	std::ostream& output;
1
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    46
	const char QUOTE = '"';
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    47
	std::wstring_convert<std::codecvt_utf8<wchar_t>> convertor; // TODO: local system encoding or generate CSV always in UTF-8 like XML?
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    48
	std::vector<AttributeMetadata> firstAttributes;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    49
	integer_t valueCount = 0;
0
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    50
public:
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    51
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    52
	CSVHandler(std::ostream& output) : output(output) {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    53
	}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    54
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    55
	void startRelation(string_t name, std::vector<AttributeMetadata> attributes) override {
1
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    56
		if (firstAttributes.empty()) {
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    57
			firstAttributes = attributes;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    58
			for (auto attr : attributes) attribute(attr.getAttributeName());
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    59
		} else {
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    60
			// TODO: UNION ALL if data types and attribute count matches
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    61
			throw RelpipeCSVWriterException(L"Only a single relation can be converted to the CSV format.");
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    62
		}
0
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    63
	}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    64
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    65
	void attribute(const string_t& value) override {
1
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    66
		valueCount++;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    67
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    68
		if (value.size() > 0) {
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    69
			output << QUOTE;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    70
			for (auto ch : convertor.to_bytes(value)) {
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    71
				if (ch == QUOTE) output << QUOTE << QUOTE;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    72
				else output << ch;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    73
			}
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    74
			output << QUOTE;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    75
		}
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    76
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    77
		if (valueCount % firstAttributes.size()) {
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    78
			output << ",";
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    79
		} else {
3
b5a5bf32f7ff generate CRLF line ends according to RFC 4180
František Kučera <franta-hg@frantovo.cz>
parents: 1
diff changeset
    80
			output << "\r\n";
1
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    81
			valueCount = 0;
82f86dc48339 first working version
František Kučera <franta-hg@frantovo.cz>
parents: 0
diff changeset
    82
		}
0
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    83
	}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    84
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    85
	void endOfPipe() {
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    86
		output.flush();
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    87
	}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    88
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    89
};
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    90
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    91
}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    92
}
97967db4b95b project skeleton
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    93
}