
com.puppycrawl.tools.checkstyle.checks.whitespace.SeparatorWrapCheck Maven / Gradle / Ivy
Show all versions of checkstyle Show documentation
////////////////////////////////////////////////////////////////////////////////
// checkstyle: Checks Java source code for adherence to a set of rules.
// Copyright (C) 2001-2016 the original author or authors.
//
// This library is free software; you can redistribute it and/or
// modify it under the terms of the GNU Lesser General Public
// License as published by the Free Software Foundation; either
// version 2.1 of the License, or (at your option) any later version.
//
// This library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
// Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public
// License along with this library; if not, write to the Free Software
// Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
////////////////////////////////////////////////////////////////////////////////
package com.puppycrawl.tools.checkstyle.checks.whitespace;
import java.util.Locale;
import org.apache.commons.beanutils.ConversionException;
import com.puppycrawl.tools.checkstyle.api.AbstractCheck;
import com.puppycrawl.tools.checkstyle.api.DetailAST;
import com.puppycrawl.tools.checkstyle.api.TokenTypes;
import com.puppycrawl.tools.checkstyle.utils.CommonUtils;
/**
*
* Checks line wrapping with separators.
* The policy to verify is specified using the {@link WrapOption} class
* and defaults to {@link WrapOption#EOL}.
*
* By default the check will check the following separators:
* {@link TokenTypes#DOT DOT},
* {@link TokenTypes#COMMA COMMA},
* Other acceptable tokens are
* {@link TokenTypes#SEMI SEMI},
* {@link TokenTypes#ELLIPSIS ELLIPSIS},
* {@link TokenTypes#AT AT},
* {@link TokenTypes#LPAREN LPAREN},
* {@link TokenTypes#RPAREN RPAREN},
* {@link TokenTypes#ARRAY_DECLARATOR ARRAY_DECLARATOR},
* {@link TokenTypes#RBRACK RBRACK},
*
*
* Code example for comma and dot at the new line:
*
*
* s
* .isEmpty();
* foo(i
* ,s);
*
*
* An example of how to configure the check is:
*
*
* <module name="SeparatorWrap"/>
*
*
* Code example for comma and dot at the previous line:
*
*
* s.
* isEmpty();
* foo(i,
* s);
*
* An example of how to configure the check for comma at the
* new line is:
*
*
* <module name="SeparatorWrap">
* <property name="tokens" value="COMMA"/>
* <property name="option" value="nl"/>
* </module>
*
*
* @author maxvetrenko
*/
public class SeparatorWrapCheck
extends AbstractCheck {
/**
* A key is pointing to the warning message text in "messages.properties"
* file.
*/
public static final String MSG_LINE_PREVIOUS = "line.previous";
/**
* A key is pointing to the warning message text in "messages.properties"
* file.
*/
public static final String MSG_LINE_NEW = "line.new";
/** The policy to enforce. */
private WrapOption option = WrapOption.EOL;
/**
* Set the option to enforce.
* @param optionStr string to decode option from
* @throws ConversionException if unable to decode
*/
public void setOption(String optionStr) {
try {
option = WrapOption.valueOf(optionStr.trim().toUpperCase(Locale.ENGLISH));
}
catch (IllegalArgumentException iae) {
throw new ConversionException("unable to parse " + optionStr, iae);
}
}
@Override
public int[] getDefaultTokens() {
return new int[] {
TokenTypes.DOT,
TokenTypes.COMMA,
};
}
@Override
public int[] getAcceptableTokens() {
return new int[] {
TokenTypes.DOT,
TokenTypes.COMMA,
TokenTypes.SEMI,
TokenTypes.ELLIPSIS,
TokenTypes.AT,
TokenTypes.LPAREN,
TokenTypes.RPAREN,
TokenTypes.ARRAY_DECLARATOR,
TokenTypes.RBRACK,
};
}
@Override
public int[] getRequiredTokens() {
return CommonUtils.EMPTY_INT_ARRAY;
}
@Override
public void visitToken(DetailAST ast) {
final String text = ast.getText();
final int colNo = ast.getColumnNo();
final int lineNo = ast.getLineNo();
final String currentLine = getLines()[lineNo - 1];
final String substringAfterToken =
currentLine.substring(colNo + text.length()).trim();
final String substringBeforeToken =
currentLine.substring(0, colNo).trim();
if (option == WrapOption.EOL
&& substringBeforeToken.isEmpty()) {
log(lineNo, colNo, MSG_LINE_PREVIOUS, text);
}
else if (option == WrapOption.NL
&& substringAfterToken.isEmpty()) {
log(lineNo, colNo, MSG_LINE_NEW, text);
}
}
}