8

我正在研究 Google 的免费(开源)Java diff 库结果,其中似乎有很多(其中一些甚至使用通用对象,而不仅仅是字符串)。

在我挖掘大量搜索结果但找不到我要搜索的内容之前,我先在这里问一下:

这些差异库中的任何一个是否支持诸如 cvs annotate 或 svn blame 之类的功能。我想要

  • 将电流传递String[]给函数
  • 继续将旧版本String[]String[]很贵,所以我想尽早停止)
  • 调用一个函数,它会ìnt[]告诉我当前版本的每一行,最后一次更改的是哪个版本,或者它是否根本没有更改(即最后一次更改是在第一个版本中)。

支持不是Strings 的对象很好,但不是必须的。如果 API 不完全是这样,我想我可以接受它。

如果没有,任何人都可以建议一个可以轻松添加该功能的可扩展差异库,最好是希望接收该功能作为贡献的一个(并且在他们接受贡献之前不需要填写大量文书工作,例如GNU 项目)?那么,我会自愿(至少尝试)将其添加到那里。

4

3 回答 3

2

我决定自己为 Dmitry Naumenko 的java-diff-utils库实现它:

/*
   Copyright 2010 Michael Schierl (schierlm@gmx.de)

   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 difflib.annotate;

import java.util.*;

import difflib.*;

/**
 * Generates an annotated version of a revision based on a list of older
 * revisions, like <tt>cvs annotate</tt> or <tt>svn blame</tt>.
 * 
 * @author <a href="schierlm@gmx.de">Michael Schierl</a>
 * 
 * @param <R>
 *            Type of the revision metadata
 */
public class Annotate<R> {

    private final List<R> revisions;
    private final int[] lineNumbers;
    private R currentRevision;
    private final List<Object> currentLines;
    private final List<Integer> currentLineMap;

    /**
     * Creates a new annotation generator.
     * 
     * @param revision
     *            Revision metadata for the revision to be annotated
     * @param targetLines
     *            Lines of the revision to be annotated
     */
    public Annotate(R revision, List<?> targetLines) {
        revisions = new ArrayList<R>();
        lineNumbers = new int[targetLines.size()];
        currentRevision = revision;
        currentLines = new ArrayList<Object>(targetLines);
        currentLineMap = new ArrayList<Integer>();
        for (int i = 0; i < lineNumbers.length; i++) {
            lineNumbers[i] = -1;
            revisions.add(null);
            currentLineMap.add(i);
        }
    }

    /**
     * Check whether there are still lines that are unannotated. In that case,
     * more older revisions should be retrieved and passed to the function. Note
     * that as soon as you pass an empty revision, all lines will be annotated
     * (with a later revision), therefore if you do not have any more revisions,
     * pass an empty revision to annotate the rest of the lines.
     */
    public boolean areLinesUnannotated() {
        for (int i = 0; i < lineNumbers.length; i++) {
            if (lineNumbers[i] == -1 || revisions.get(i) == null)
                return true;
        }
        return false;
    }

    /**
     * Add the previous revision and update annotation info.
     * 
     * @param revision
     *            Revision metadata for this revision
     * @param lines
     *            Lines of this revision
     */
    public void addRevision(R revision, List<?> lines) {
        Patch patch = DiffUtils.diff(currentLines, lines);
        int lineOffset = 0; // remember number of already deleted/added lines
        for (Delta d : patch.getDeltas()) {
            Chunk original = d.getOriginal();
            Chunk revised = d.getRevised();
            int pos = original.getPosition() + lineOffset;
            // delete lines
            for (int i = 0; i < original.getSize(); i++) {
                int origLine = currentLineMap.remove(pos);
                currentLines.remove(pos);
                if (origLine != -1) {
                    lineNumbers[origLine] = original.getPosition() + i;
                    revisions.set(origLine, currentRevision);
                }
            }
            for (int i = 0; i < revised.getSize(); i++) {
                currentLines.add(pos + i, revised.getLines().get(i));
                currentLineMap.add(pos + i, -1);
            }
            lineOffset += revised.getSize() - original.getSize();
        }

        currentRevision = revision;
        if (!currentLines.equals(lines))
            throw new RuntimeException("Patch application failed");
    }

    /**
     * Return the result of the annotation. It will be a List of the same length
     * as the target revision, for which every entry states the revision where
     * the line appeared last.
     */
    public List<R> getAnnotatedRevisions() {
        return Collections.unmodifiableList(revisions);
    }

    /**
     * Return the result of the annotation. It will be a List of the same length
     * as the target revision, for which every entry states the line number in
     * the revision where the line appeared last.
     */
    public int[] getAnnotatedLineNumbers() {
        return (int[]) lineNumbers.clone();
    }
}

我还将它发送给 Dmitry Naumenko(带有一些测试用例),以防他想要包含它。

于 2010-12-27T15:22:33.347 回答
1

您可以使用 xwiki-commons-blame-api。它实际上使用了该线程已接受答案中的代码感谢 Michael Schierl 在 StackOverflow 上共享此代码

你可以在它的单元测试中看到如何在 Java 中使用它。

或者在 Scala 中,例如:

import java.util
import org.xwiki.blame.AnnotatedContent
import org.xwiki.blame.internal.DefaultBlameManager

case class Revision(id: Int,
                    parentId: Option[Int] = None,
                    content: Option[String] = None)

def createAnnotation(revisions: Seq[Revision]): Option[AnnotatedContent[Revision, String]] = {
    val blameManager = new DefaultBlameManager()

    val annotatedContent = revisions.foldLeft(null.asInstanceOf[AnnotatedContent[Revision, String]]){
      (annotation, revision) =>
        blameManager.blame(annotation, revision, splitByWords(revision.content))
    }
    Option(annotatedContent)
}

def splitByWords(content: Option[String]): util.List[String] = {
    val array = content.fold(Array[String]())(_.split("[^\\pL_\\pN]+"))
    util.Arrays.asList(array:_*)
}
于 2015-02-21T17:32:14.157 回答
1

我可能错了,但我认为 annotate/blame 需要一个版本控制系统才能工作,因为它需要访问文件的历史记录。通用差异库无法做到这一点。因此,如果这是您的目标,请查看与这些 VCS 一起使用的库,例如svnkit。如果没有,这样的库可能是一个很好的起点,了解如何完成注释/责备,这通常涉及区分文件所有版本的链。

于 2010-12-23T14:10:05.897 回答