src/UnionAllHandler.h
author František Kučera <franta-hg@frantovo.cz>
Tue, 19 Apr 2022 23:17:15 +0200
branchv_0
changeset 1 d53041bb781b
permissions -rw-r--r--
first version of deserializer
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
1
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     1
/**
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     2
 * Relational pipes
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     3
 * Copyright © 2022 František Kučera (Frantovo.cz, GlobalCode.info)
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     4
 *
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     5
 * This program is free software: you can redistribute it and/or modify
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     6
 * it under the terms of the GNU General Public License as published by
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     7
 * the Free Software Foundation, version 3 of the License.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     8
 *
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     9
 * This program is distributed in the hope that it will be useful,
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    10
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    11
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    12
 * GNU General Public License for more details.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    13
 *
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    14
 * You should have received a copy of the GNU General Public License
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    15
 * along with this program. If not, see <http://www.gnu.org/licenses/>.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    16
 */
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    17
#pragma once
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    18
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    19
#include <regex>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    20
#include <stdexcept>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    21
#include <sstream>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    22
#include <codecvt>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    23
#include <iomanip>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    24
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    25
#include <relpipe/common/type/typedefs.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    26
#include <relpipe/reader/TypeId.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    27
#include <relpipe/reader/handlers/RelationalReaderStringHandler.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    28
#include <relpipe/reader/handlers/AttributeMetadata.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    29
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    30
#include <relpipe/writer/Factory.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    31
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    32
#include <relpipe/cli/RelpipeCLIException.h>
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    33
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    34
#include "Configuration.h"
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    35
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    36
namespace relpipe {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    37
namespace tr {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    38
namespace deserialize {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    39
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    40
/**
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    41
 * Forwards all attributes to given writer.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    42
 * If an relation has same name and attributes as the previous one, records are just appended to it.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    43
 * If name or attributes differ, new relation is started and records are written to it.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    44
 * 
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    45
 * So it does something like UNION ALL in SQL for sequences of relations that have same name and attributes.
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    46
 */
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    47
class UnionAllHandler : public relpipe::reader::handlers::RelationalReaderValueHandler {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    48
private:
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    49
	Configuration configuration;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    50
	shared_ptr<relpipe::writer::RelationalWriter> writer;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    51
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    52
	class RelationContext {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    53
	public:
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    54
		relpipe::common::type::StringX name;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    55
		std::vector<relpipe::reader::handlers::AttributeMetadata> readerMetadata;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    56
		std::vector<relpipe::writer::AttributeMetadata> writerMetadata;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    57
	} relationContext;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    58
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    59
	void convert(const std::vector<relpipe::reader::handlers::AttributeMetadata>& r, std::vector<relpipe::writer::AttributeMetadata>& w) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    60
		w.clear();
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    61
		for (const relpipe::reader::handlers::AttributeMetadata& a : r) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    62
			w.push_back({a.getAttributeName(), writer->toTypeId(a.getTypeName())});
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    63
		}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    64
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    65
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    66
	bool equals(const std::vector<relpipe::reader::handlers::AttributeMetadata>& a, const std::vector<relpipe::reader::handlers::AttributeMetadata>& b) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    67
		if (a.size() != b.size()) return false;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    68
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    69
		for (size_t i = 0, limit = a.size(); i < limit; i++) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    70
			if (a[i].getAttributeName() != b[i].getAttributeName()) return false;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    71
			if (a[i].getTypeId() != b[i].getTypeId()) return false;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    72
		}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    73
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    74
		return true;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    75
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    76
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    77
public:
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    78
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    79
	UnionAllHandler(shared_ptr<relpipe::writer::RelationalWriter> writer, Configuration configuration) : writer(writer), configuration(configuration) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    80
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    81
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    82
	virtual ~UnionAllHandler() = default;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    83
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    84
	void startRelation(relpipe::common::type::StringX name, std::vector<relpipe::reader::handlers::AttributeMetadata> attributes) override {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    85
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    86
		bool continuation = name == relationContext.name && equals(attributes, relationContext.readerMetadata);
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    87
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    88
		relationContext = RelationContext();
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    89
		relationContext.name = name;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    90
		relationContext.readerMetadata = attributes;
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    91
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    92
		if (continuation) {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    93
			// same relation → just append records
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    94
		} else {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    95
			convert(attributes, relationContext.writerMetadata);
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    96
			writer->startRelation(name, relationContext.writerMetadata, true);
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    97
		}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    98
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    99
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   100
	void attribute(const void* value, const std::type_info& type) override {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   101
		writer->writeAttribute(value, type);
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   102
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   103
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   104
	void endOfPipe() {
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   105
	}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   106
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   107
};
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   108
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   109
}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   110
}
d53041bb781b first version of deserializer
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   111
}