
uno.informatics.common.io.IOUtilities Maven / Gradle / Ivy
/*******************************************************************************
* Copyright 2015 Guy Davenport
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*******************************************************************************/
package uno.informatics.common.io;
import static uno.informatics.common.io.TextFileHandler.COMMA;
import static uno.informatics.common.io.TextFileHandler.TAB;
import java.io.BufferedReader;
import java.io.BufferedWriter;
import java.io.IOException;
import java.nio.file.Path;
import java.util.ArrayList;
import java.util.List;
import uno.informatics.common.io.text.TextFileRowReader;
import uno.informatics.common.io.text.TextFileRowWriter;
import uno.informatics.data.io.FileType;
/**
* @author Guy Davenport
*
*/
public class IOUtilities {
public static final RowReader createRowReader(Path path, FileType type, int... options)
throws IOException {
RowReader reader = null;
switch (type) {
case CSV:
TextFileRowReader textFileRowStringReader = new TextFileRowReader(path);
textFileRowStringReader.setDelimiterString(COMMA);
textFileRowStringReader.setOptions(getOptions(options));
reader = textFileRowStringReader;
break;
case TXT:
textFileRowStringReader = new TextFileRowReader(path);
textFileRowStringReader.setDelimiterString(TAB);
textFileRowStringReader.setOptions(getOptions(options));
reader = textFileRowStringReader;
break;
case XLS:
throw new IOException("XLS Type not supported");
// break;
case XLSX:
throw new IOException("XLSX Type not supported");
// break;
default:
break;
}
return reader;
}
public static RowReader createRowReader(BufferedReader bufferedReader, FileType type, int... options)
throws IOException {
RowReader reader = null;
switch (type) {
case CSV:
TextFileRowReader textFileRowStringReader = new TextFileRowReader(bufferedReader);
textFileRowStringReader.setDelimiterString(COMMA);
textFileRowStringReader.setOptions(getOptions(options));
reader = textFileRowStringReader;
break;
case TXT:
textFileRowStringReader = new TextFileRowReader(bufferedReader);
textFileRowStringReader.setDelimiterString(TAB);
textFileRowStringReader.setOptions(getOptions(options));
reader = textFileRowStringReader;
break;
case XLS:
throw new IOException("XLS Type not supported");
// break;
case XLSX:
throw new IOException("XLSX Type not supported");
// break;
default:
break;
}
return reader;
}
public static final RowWriter createRowWriter(Path filePath, FileType type, int... options)
throws IOException {
RowWriter writer = null;
switch (type) {
case CSV:
TextFileRowWriter textFileRowStringWriter = new TextFileRowWriter(filePath);
textFileRowStringWriter.setDelimiterString(COMMA);
writer = textFileRowStringWriter;
break;
case TXT:
textFileRowStringWriter = new TextFileRowWriter(filePath);
textFileRowStringWriter.setDelimiterString(TAB);
writer = textFileRowStringWriter;
break;
case XLS:
throw new IOException("XLS Type not supported");
// break;
case XLSX:
throw new IOException("XLSX Type not supported");
// break;
default:
break;
}
return writer;
}
public static final RowWriter createRowWriter(BufferedWriter bufferedWriter, FileType type,
int... options) throws IOException {
RowWriter writer = null;
switch (type) {
case CSV:
TextFileRowWriter textFileRowStringWriter = new TextFileRowWriter(bufferedWriter);
textFileRowStringWriter.setDelimiterString(COMMA);
writer = textFileRowStringWriter;
break;
case TXT:
textFileRowStringWriter = new TextFileRowWriter(bufferedWriter);
textFileRowStringWriter.setDelimiterString(TAB);
writer = textFileRowStringWriter;
break;
case XLS:
throw new IOException("XLS Type not supported");
// break;
case XLSX:
throw new IOException("XLSX Type not supported");
// break;
default:
break;
}
return writer;
}
public static final List getSheets(Path filePath, FileType type, int... options)
throws IOException {
// validate arguments
if (filePath == null) {
throw new IllegalArgumentException("File path not defined.");
}
if (!filePath.toFile().exists()) {
throw new IOException("File does not exist : " + filePath + ".");
}
if (type == null) {
throw new IllegalArgumentException("File type not defined.");
}
if (type != FileType.TXT && type != FileType.CSV) {
throw new IllegalArgumentException(
String.format("Only file types TXT and CSV are supported. Got: %s.", type));
}
switch (type) {
case CSV:
return new ArrayList();
case TXT:
return new ArrayList();
case XLS:
return new ArrayList();
// JXLExcelRowStringReader jxlExcelRowStringReader = new
// JXLExcelRowStringReader(fileProperties.getFile()) ;
// return jxlExcelRowStringReader.getAllSpreadSheetNames() ;
case XLSX:
return new ArrayList();
// jxlExcelRowStringReader = new
// JXLExcelRowStringReader(fileProperties.getFile()) ;
// return jxlExcelRowStringReader.getAllSpreadSheetNames() ;
default:
return new ArrayList();
}
}
private static final int getOptions(int... options) {
int combinedOptions = 0;
for (int i = 0; i < options.length; ++i)
combinedOptions = combinedOptions | options[i];
return combinedOptions;
}
}
© 2015 - 2025 Weber Informatics LLC | Privacy Policy