////////////////////////////////////////////////////////////////////////////////
// checkstyle: Checks Java source code for adherence to a set of rules.
// Copyright (C) 2001-2020 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.imports;

import java.util.ArrayList;
import java.util.HashSet;
import java.util.List;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;

import com.puppycrawl.tools.checkstyle.FileStatefulCheck;
import com.puppycrawl.tools.checkstyle.api.AbstractCheck;
import com.puppycrawl.tools.checkstyle.api.DetailAST;
import com.puppycrawl.tools.checkstyle.api.FileContents;
import com.puppycrawl.tools.checkstyle.api.FullIdent;
import com.puppycrawl.tools.checkstyle.api.TextBlock;
import com.puppycrawl.tools.checkstyle.api.TokenTypes;
import com.puppycrawl.tools.checkstyle.checks.javadoc.JavadocTag;
import com.puppycrawl.tools.checkstyle.utils.CommonUtil;
import com.puppycrawl.tools.checkstyle.utils.JavadocUtil;

/**
 * <p>
 * Checks for unused import statements. Checkstyle uses a simple but very
 * reliable algorithm to report on unused import statements. An import statement
 * is considered unused if:
 * </p>
 * <ul>
 * <li>
 * It is not referenced in the file. The algorithm does not support wild-card
 * imports like {@code import java.io.*;}. Most IDE's provide very sophisticated
 * checks for imports that handle wild-card imports.
 * </li>
 * <li>
 * It is a duplicate of another import. This is when a class is imported more
 * than once.
 * </li>
 * <li>
 * The class imported is from the {@code java.lang} package. For example
 * importing {@code java.lang.String}.
 * </li>
 * <li>
 * The class imported is from the same package.
 * </li>
 * <li>
 * <b>Optionally:</b> it is referenced in Javadoc comments. This check is on by
 * default, but it is considered bad practice to introduce a compile time
 * dependency for documentation purposes only. As an example, the import
 * {@code java.util.List} would be considered referenced with the Javadoc
 * comment {@code {@link List}}. The alternative to avoid introducing a compile
 * time dependency would be to write the Javadoc comment as {@code {&#64;link java.util.List}}.
 * </li>
 * </ul>
 * <p>
 * The main limitation of this check is handling the case where an imported type
 * has the same name as a declaration, such as a member variable.
 * </p>
 * <p>
 * For example, in the following case the import {@code java.awt.Component}
 * will not be flagged as unused:
 * </p>
 * <pre>
 * import java.awt.Component;
 * class FooBar {
 *   private Object Component; // a bad practice in my opinion
 *   ...
 * }
 * </pre>
 * <ul>
 * <li>
 * Property {@code processJavadoc} - Control whether to process Javadoc comments.
 * Type is {@code boolean}.
 * Default value is {@code true}.
 * </li>
 * </ul>
 * <p>
 * To configure the check:
 * </p>
 * <pre>
 * &lt;module name="UnusedImports"/&gt;
 * </pre>
 * <p>
 * Parent is {@code com.puppycrawl.tools.checkstyle.TreeWalker}
 * </p>
 * <p>
 * Violation Message Keys:
 * </p>
 * <ul>
 * <li>
 * {@code import.unused}
 * </li>
 * </ul>
 *
 * @since 3.0
 */
@FileStatefulCheck
public class UnusedImportsCheck extends AbstractCheck {

    /**
     * A key is pointing to the warning message text in "messages.properties"
     * file.
     */
    public static final String MSG_KEY = "import.unused";

    /** Regex to match class names. */
    private static final Pattern CLASS_NAME = CommonUtil.createPattern(
           "((:?[\\p{L}_$][\\p{L}\\p{N}_$]*\\.)*[\\p{L}_$][\\p{L}\\p{N}_$]*)");
    /** Regex to match the first class name. */
    private static final Pattern FIRST_CLASS_NAME = CommonUtil.createPattern(
           "^" + CLASS_NAME);
    /** Regex to match argument names. */
    private static final Pattern ARGUMENT_NAME = CommonUtil.createPattern(
           "[(,]\\s*" + CLASS_NAME.pattern());

    /** Regexp pattern to match java.lang package. */
    private static final Pattern JAVA_LANG_PACKAGE_PATTERN =
        CommonUtil.createPattern("^java\\.lang\\.[a-zA-Z]+$");

    /** Suffix for the star import. */
    private static final String STAR_IMPORT_SUFFIX = ".*";

    /** Set of the imports. */
    private final Set<FullIdent> imports = new HashSet<>();

    /** Set of references - possibly to imports or other things. */
    private final Set<String> referenced = new HashSet<>();

    /** Flag to indicate when time to start collecting references. */
    private boolean collect;
    /** Control whether to process Javadoc comments. */
    private boolean processJavadoc = true;

    /**
     * Setter to control whether to process Javadoc comments.
     *
     * @param value Flag for processing Javadoc comments.
     */
    public void setProcessJavadoc(boolean value) {
        processJavadoc = value;
    }

    @Override
    public void beginTree(DetailAST rootAST) {
        collect = false;
        imports.clear();
        referenced.clear();
    }

    @Override
    public void finishTree(DetailAST rootAST) {
        // loop over all the imports to see if referenced.
        imports.stream()
            .filter(imprt -> isUnusedImport(imprt.getText()))
            .forEach(imprt -> log(imprt.getDetailAst(), MSG_KEY, imprt.getText()));
    }

    @Override
    public int[] getDefaultTokens() {
        return getRequiredTokens();
    }

    @Override
    public int[] getRequiredTokens() {
        return new int[] {
            TokenTypes.IDENT,
            TokenTypes.IMPORT,
            TokenTypes.STATIC_IMPORT,
            // Definitions that may contain Javadoc...
            TokenTypes.PACKAGE_DEF,
            TokenTypes.ANNOTATION_DEF,
            TokenTypes.ANNOTATION_FIELD_DEF,
            TokenTypes.ENUM_DEF,
            TokenTypes.ENUM_CONSTANT_DEF,
            TokenTypes.CLASS_DEF,
            TokenTypes.INTERFACE_DEF,
            TokenTypes.METHOD_DEF,
            TokenTypes.CTOR_DEF,
            TokenTypes.VARIABLE_DEF,
        };
    }

    @Override
    public int[] getAcceptableTokens() {
        return getRequiredTokens();
    }

    @Override
    public void visitToken(DetailAST ast) {
        if (ast.getType() == TokenTypes.IDENT) {
            if (collect) {
                processIdent(ast);
            }
        }
        else if (ast.getType() == TokenTypes.IMPORT) {
            processImport(ast);
        }
        else if (ast.getType() == TokenTypes.STATIC_IMPORT) {
            processStaticImport(ast);
        }
        else {
            collect = true;
            if (processJavadoc) {
                collectReferencesFromJavadoc(ast);
            }
        }
    }

    /**
     * Checks whether an import is unused.
     *
     * @param imprt an import.
     * @return true if an import is unused.
     */
    private boolean isUnusedImport(String imprt) {
        final Matcher javaLangPackageMatcher = JAVA_LANG_PACKAGE_PATTERN.matcher(imprt);
        return !referenced.contains(CommonUtil.baseClassName(imprt))
            || javaLangPackageMatcher.matches();
    }

    /**
     * Collects references made by IDENT.
     *
     * @param ast the IDENT node to process
     */
    private void processIdent(DetailAST ast) {
        final DetailAST parent = ast.getParent();
        final int parentType = parent.getType();
        if (parentType != TokenTypes.DOT
            && parentType != TokenTypes.METHOD_DEF
            || parentType == TokenTypes.DOT
                && ast.getNextSibling() != null) {
            referenced.add(ast.getText());
        }
    }

    /**
     * Collects the details of imports.
     *
     * @param ast node containing the import details
     */
    private void processImport(DetailAST ast) {
        final FullIdent name = FullIdent.createFullIdentBelow(ast);
        if (!name.getText().endsWith(STAR_IMPORT_SUFFIX)) {
            imports.add(name);
        }
    }

    /**
     * Collects the details of static imports.
     *
     * @param ast node containing the static import details
     */
    private void processStaticImport(DetailAST ast) {
        final FullIdent name =
            FullIdent.createFullIdent(
                ast.getFirstChild().getNextSibling());
        if (!name.getText().endsWith(STAR_IMPORT_SUFFIX)) {
            imports.add(name);
        }
    }

    /**
     * Collects references made in Javadoc comments.
     *
     * @param ast node to inspect for Javadoc
     */
    private void collectReferencesFromJavadoc(DetailAST ast) {
        final FileContents contents = getFileContents();
        final int lineNo = ast.getLineNo();
        final TextBlock textBlock = contents.getJavadocBefore(lineNo);
        if (textBlock != null) {
            referenced.addAll(collectReferencesFromJavadoc(textBlock));
        }
    }

    /**
     * Process a javadoc {@link TextBlock} and return the set of classes
     * referenced within.
     *
     * @param textBlock The javadoc block to parse
     * @return a set of classes referenced in the javadoc block
     */
    private static Set<String> collectReferencesFromJavadoc(TextBlock textBlock) {
        final List<JavadocTag> tags = new ArrayList<>();
        // gather all the inline tags, like @link
        // INLINE tags inside BLOCKs get hidden when using ALL
        tags.addAll(getValidTags(textBlock, JavadocUtil.JavadocTagType.INLINE));
        // gather all the block-level tags, like @throws and @see
        tags.addAll(getValidTags(textBlock, JavadocUtil.JavadocTagType.BLOCK));

        final Set<String> references = new HashSet<>();

        tags.stream()
            .filter(JavadocTag::canReferenceImports)
            .forEach(tag -> references.addAll(processJavadocTag(tag)));
        return references;
    }

    /**
     * Returns the list of valid tags found in a javadoc {@link TextBlock}.
     *
     * @param cmt The javadoc block to parse
     * @param tagType The type of tags we're interested in
     * @return the list of tags
     */
    private static List<JavadocTag> getValidTags(TextBlock cmt,
            JavadocUtil.JavadocTagType tagType) {
        return JavadocUtil.getJavadocTags(cmt, tagType).getValidTags();
    }

    /**
     * Returns a list of references found in a javadoc {@link JavadocTag}.
     *
     * @param tag The javadoc tag to parse
     * @return A list of references found in this tag
     */
    private static Set<String> processJavadocTag(JavadocTag tag) {
        final Set<String> references = new HashSet<>();
        final String identifier = tag.getFirstArg().trim();
        for (Pattern pattern : new Pattern[]
        {FIRST_CLASS_NAME, ARGUMENT_NAME}) {
            references.addAll(matchPattern(identifier, pattern));
        }
        return references;
    }

    /**
     * Extracts a list of texts matching a {@link Pattern} from a
     * {@link String}.
     *
     * @param identifier The String to match the pattern against
     * @param pattern The Pattern used to extract the texts
     * @return A list of texts which matched the pattern
     */
    private static Set<String> matchPattern(String identifier, Pattern pattern) {
        final Set<String> references = new HashSet<>();
        final Matcher matcher = pattern.matcher(identifier);
        while (matcher.find()) {
            references.add(topLevelType(matcher.group(1)));
        }
        return references;
    }

    /**
     * If the given type string contains "." (e.g. "Map.Entry"), returns the
     * top level type (e.g. "Map"), as that is what must be imported for the
     * type to resolve. Otherwise, returns the type as-is.
     *
     * @param type A possibly qualified type name
     * @return The simple name of the top level type
     */
    private static String topLevelType(String type) {
        final String topLevelType;
        final int dotIndex = type.indexOf('.');
        if (dotIndex == -1) {
            topLevelType = type;
        }
        else {
            topLevelType = type.substring(0, dotIndex);
        }
        return topLevelType;
    }

}