src/SedHandler.h
author František Kučera <franta-hg@frantovo.cz>
Sat, 30 Mar 2019 16:14:42 +0100
branchv_0
changeset 15 09981db6ad87
parent 9 f5c9a6daf4ea
child 19 9bac174d11b6
permissions -rw-r--r--
fix typo: Hadler → Handler
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     1
/**
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     2
 * Relational pipes
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     3
 * Copyright © 2018 František Kučera (Frantovo.cz, GlobalCode.info)
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     4
 *
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     5
 * This program is free software: you can redistribute it and/or modify
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     6
 * it under the terms of the GNU General Public License as published by
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     7
 * the Free Software Foundation, either version 3 of the License, or
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     8
 * (at your option) any later version.
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
     9
 *
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    10
 * This program is distributed in the hope that it will be useful,
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    11
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    12
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    13
 * GNU General Public License for more details.
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    14
 *
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    15
 * You should have received a copy of the GNU General Public License
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    16
 * along with this program. If not, see <http://www.gnu.org/licenses/>.
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    17
 */
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    18
#pragma once
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    19
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    20
#include <memory>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    21
#include <string>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    22
#include <vector>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    23
#include <iostream>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    24
#include <sstream>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    25
#include <locale>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    26
#include <codecvt>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    27
#include <regex>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    28
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    29
#include <relpipe/reader/typedefs.h>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    30
#include <relpipe/reader/TypeId.h>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    31
#include <relpipe/reader/handlers/RelationalReaderStringHandler.h>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    32
#include <relpipe/reader/handlers/AttributeMetadata.h>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    33
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    34
#include <relpipe/writer/Factory.h>
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    35
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    36
#include <relpipe/cli/RelpipeCLIException.h>
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    37
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    38
namespace relpipe {
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    39
namespace tr {
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    40
namespace sed {
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    41
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    42
using namespace std;
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    43
using namespace relpipe;
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    44
using namespace relpipe::reader;
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    45
using namespace relpipe::reader::handlers;
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    46
15
09981db6ad87 fix typo: Hadler → Handler
František Kučera <franta-hg@frantovo.cz>
parents: 9
diff changeset
    47
class SedHandler : public RelationalReaderStringHandler {
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    48
private:
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    49
	shared_ptr<writer::RelationalWriter> relationalWriter;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    50
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    51
	wregex relationNameRegEx;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    52
	wregex attributeNameRegEx;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    53
	wregex searchRegEx;
8
de3def6505a9 working regex_replace
František Kučera <franta-hg@frantovo.cz>
parents: 7
diff changeset
    54
	string_t replacement;
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    55
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    56
	vector<boolean_t> currentReplacableAttributes;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    57
	integer_t currentAttributeIndex = 0;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    58
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    59
public:
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    60
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    61
	SedHandler(ostream& output, const vector<string_t>& arguments) {
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    62
		relationalWriter.reset(writer::Factory::create(output));
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    63
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    64
		if (arguments.size() == 4) {
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    65
			relationNameRegEx = wregex(arguments[0]);
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    66
			attributeNameRegEx = wregex(arguments[1]);
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    67
			searchRegEx = wregex(arguments[2]);
8
de3def6505a9 working regex_replace
František Kučera <franta-hg@frantovo.cz>
parents: 7
diff changeset
    68
			replacement = arguments[3];
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    69
		} else {
9
f5c9a6daf4ea small text update
František Kučera <franta-hg@frantovo.cz>
parents: 8
diff changeset
    70
			throw cli::RelpipeCLIException(L"Usage: relpipe-tr-sed <relationNameRegExp> <attributeNameRegExp> <searchRegExp> <replacement>", cli::CLI::EXIT_CODE_UNKNOWN_COMMAND);
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    71
		}
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    72
	}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    73
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    74
	void startRelation(string_t name, vector<AttributeMetadata> attributes) override {
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    75
		// TODO: move to a reusable method (or use same metadata on both reader and writer side?)
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    76
		vector<writer::AttributeMetadata> writerMetadata;
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    77
		for (AttributeMetadata readerMetadata : attributes) {
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    78
			writerMetadata.push_back({readerMetadata.getAttributeName(), relationalWriter->toTypeId(readerMetadata.getTypeName())});
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    79
		}
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    80
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    81
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    82
		currentReplacableAttributes.resize(attributes.size(), false);
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    83
		if (regex_match(name, relationNameRegEx)) {
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    84
			for (int i = 0; i < currentReplacableAttributes.size(); i++) {
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    85
				currentReplacableAttributes[i] = regex_match(attributes[i].getAttributeName(), attributeNameRegEx);
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    86
			}
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    87
		}
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    88
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    89
		relationalWriter->startRelation(name, writerMetadata, true);
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    90
	}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    91
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
    92
	void attribute(const string_t& value) override {
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    93
		if (currentReplacableAttributes[currentAttributeIndex]) {
8
de3def6505a9 working regex_replace
František Kučera <franta-hg@frantovo.cz>
parents: 7
diff changeset
    94
			relationalWriter->writeAttribute(regex_replace(value, searchRegEx, replacement));
7
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    95
		} else {
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    96
			relationalWriter->writeAttribute(value);
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    97
		}
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    98
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
    99
		currentAttributeIndex++;
92d85e02b276 regex match relation and attribute + replace with a fixed value
František Kučera <franta-hg@frantovo.cz>
parents: 3
diff changeset
   100
		currentAttributeIndex = currentAttributeIndex % currentReplacableAttributes.size();
3
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   101
	}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   102
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   103
	void endOfPipe() {
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   104
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   105
	}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   106
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   107
};
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   108
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   109
}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   110
}
8731263d44f1 PassthroughHandler (string-based version)
František Kučera <franta-hg@frantovo.cz>
parents:
diff changeset
   111
}