Package org.eclipse.jgit.treewalk

Source Code of org.eclipse.jgit.treewalk.WorkingTreeIterator$IteratorState

/*
* Copyright (C) 2008, Shawn O. Pearce <spearce@spearce.org>
* Copyright (C) 2010, Christian Halstrick <christian.halstrick@sap.com>
* Copyright (C) 2010, Matthias Sohn <matthias.sohn@sap.com>
* and other copyright owners as documented in the project's IP log.
*
* This program and the accompanying materials are made available
* under the terms of the Eclipse Distribution License v1.0 which
* accompanies this distribution, is reproduced below, and is
* available at http://www.eclipse.org/org/documents/edl-v10.php
*
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or
* without modification, are permitted provided that the following
* conditions are met:
*
* - Redistributions of source code must retain the above copyright
*   notice, this list of conditions and the following disclaimer.
*
* - Redistributions in binary form must reproduce the above
*   copyright notice, this list of conditions and the following
*   disclaimer in the documentation and/or other materials provided
*   with the distribution.
*
* - Neither the name of the Eclipse Foundation, Inc. nor the
*   names of its contributors may be used to endorse or promote
*   products derived from this software without specific prior
*   written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
* CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
* INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
* OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
* CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
* LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
* CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
* STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
* ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/

package org.eclipse.jgit.treewalk;

import java.io.ByteArrayInputStream;
import java.io.File;
import java.io.FileInputStream;
import java.io.FileNotFoundException;
import java.io.IOException;
import java.io.InputStream;
import java.nio.ByteBuffer;
import java.nio.CharBuffer;
import java.nio.charset.CharacterCodingException;
import java.nio.charset.CharsetEncoder;
import java.security.MessageDigest;
import java.text.MessageFormat;
import java.util.Arrays;
import java.util.Collections;
import java.util.Comparator;

import org.eclipse.jgit.JGitText;
import org.eclipse.jgit.diff.RawText;
import org.eclipse.jgit.dircache.DirCache;
import org.eclipse.jgit.dircache.DirCacheEntry;
import org.eclipse.jgit.dircache.DirCacheIterator;
import org.eclipse.jgit.errors.CorruptObjectException;
import org.eclipse.jgit.ignore.IgnoreNode;
import org.eclipse.jgit.ignore.IgnoreRule;
import org.eclipse.jgit.lib.Constants;
import org.eclipse.jgit.lib.CoreConfig;
import org.eclipse.jgit.lib.FileMode;
import org.eclipse.jgit.lib.Repository;
import org.eclipse.jgit.util.FS;
import org.eclipse.jgit.util.IO;
import org.eclipse.jgit.util.io.EolCanonicalizingInputStream;

/**
* Walks a working directory tree as part of a {@link TreeWalk}.
* <p>
* Most applications will want to use the standard implementation of this
* iterator, {@link FileTreeIterator}, as that does all IO through the standard
* <code>java.io</code> package. Plugins for a Java based IDE may however wish
* to create their own implementations of this class to allow traversal of the
* IDE's project space, as well as benefit from any caching the IDE may have.
*
* @see FileTreeIterator
*/
public abstract class WorkingTreeIterator extends AbstractTreeIterator {
  /** An empty entry array, suitable for {@link #init(Entry[])}. */
  protected static final Entry[] EOF = {};

  /** Size we perform file IO in if we have to read and hash a file. */
  static final int BUFFER_SIZE = 2048;

  /**
   * Maximum size of files which may be read fully into memory for performance
   * reasons.
   */
  private static final long MAXIMUM_FILE_SIZE_TO_READ_FULLY = 65536;

  /** Inherited state of this iterator, describing working tree, etc. */
  private final IteratorState state;

  /** The {@link #idBuffer()} for the current entry. */
  private byte[] contentId;

  /** Index within {@link #entries} that {@link #contentId} came from. */
  private int contentIdFromPtr;

  /** List of entries obtained from the subclass. */
  private Entry[] entries;

  /** Total number of entries in {@link #entries} that are valid. */
  private int entryCnt;

  /** Current position within {@link #entries}. */
  private int ptr;

  /** If there is a .gitignore file present, the parsed rules from it. */
  private IgnoreNode ignoreNode;

  /**
   * Create a new iterator with no parent.
   *
   * @param options
   *            working tree options to be used
   */
  protected WorkingTreeIterator(WorkingTreeOptions options) {
    super();
    state = new IteratorState(options);
  }

  /**
   * Create a new iterator with no parent and a prefix.
   * <p>
   * The prefix path supplied is inserted in front of all paths generated by
   * this iterator. It is intended to be used when an iterator is being
   * created for a subsection of an overall repository and needs to be
   * combined with other iterators that are created to run over the entire
   * repository namespace.
   *
   * @param prefix
   *            position of this iterator in the repository tree. The value
   *            may be null or the empty string to indicate the prefix is the
   *            root of the repository. A trailing slash ('/') is
   *            automatically appended if the prefix does not end in '/'.
   * @param options
   *            working tree options to be used
   */
  protected WorkingTreeIterator(final String prefix,
      WorkingTreeOptions options) {
    super(prefix);
    state = new IteratorState(options);
  }

  /**
   * Create an iterator for a subtree of an existing iterator.
   *
   * @param p
   *            parent tree iterator.
   */
  protected WorkingTreeIterator(final WorkingTreeIterator p) {
    super(p);
    state = p.state;
  }

  /**
   * Initialize this iterator for the root level of a repository.
   * <p>
   * This method should only be invoked after calling {@link #init(Entry[])},
   * and only for the root iterator.
   *
   * @param repo
   *            the repository.
   */
  protected void initRootIterator(Repository repo) {
    Entry entry;
    if (ignoreNode instanceof PerDirectoryIgnoreNode)
      entry = ((PerDirectoryIgnoreNode) ignoreNode).entry;
    else
      entry = null;
    ignoreNode = new RootIgnoreNode(entry, repo);
  }

  /**
   * Define the matching {@link DirCacheIterator}, to optimize ObjectIds.
   *
   * Once the DirCacheIterator has been set this iterator must only be
   * advanced by the TreeWalk that is supplied, as it assumes that itself and
   * the corresponding DirCacheIterator are positioned on the same file path
   * whenever {@link #idBuffer()} is invoked.
   *
   * @param walk
   *            the walk that will be advancing this iterator.
   * @param treeId
   *            index of the matching {@link DirCacheIterator}.
   */
  public void setDirCacheIterator(TreeWalk walk, int treeId) {
    state.walk = walk;
    state.dirCacheTree = treeId;
  }

  @Override
  public boolean hasId() {
    if (contentIdFromPtr == ptr)
      return true;
    return (mode & FileMode.TYPE_MASK) == FileMode.TYPE_FILE;
  }

  @Override
  public byte[] idBuffer() {
    if (contentIdFromPtr == ptr)
      return contentId;

    if (state.walk != null) {
      // If there is a matching DirCacheIterator, we can reuse
      // its idBuffer, but only if we appear to be clean against
      // the cached index information for the path.
      //
      DirCacheIterator i = state.walk.getTree(state.dirCacheTree,
          DirCacheIterator.class);
      if (i != null) {
        DirCacheEntry ent = i.getDirCacheEntry();
        if (ent != null && compareMetadata(ent) == MetadataDiff.EQUAL)
          return i.idBuffer();
      }
    }

    switch (mode & FileMode.TYPE_MASK) {
    case FileMode.TYPE_FILE:
      contentIdFromPtr = ptr;
      return contentId = idBufferBlob(entries[ptr]);
    case FileMode.TYPE_SYMLINK:
      // Java does not support symbolic links, so we should not
      // have reached this particular part of the walk code.
      //
      return zeroid;
    case FileMode.TYPE_GITLINK:
      // TODO: Support obtaining current HEAD SHA-1 from nested repository
      //
      return zeroid;
    }
    return zeroid;
  }

  private static final byte[] digits = { '0', '1', '2', '3', '4', '5', '6',
      '7', '8', '9' };

  private static final byte[] hblob = Constants
      .encodedTypeString(Constants.OBJ_BLOB);

  private byte[] idBufferBlob(final Entry e) {
    try {
      final InputStream is = e.openInputStream();
      if (is == null)
        return zeroid;
      try {
        state.initializeDigestAndReadBuffer();

        final long len = e.getLength();
        if (!mightNeedCleaning())
          return computeHash(is, len);

        if (len <= MAXIMUM_FILE_SIZE_TO_READ_FULLY) {
          ByteBuffer rawbuf = IO.readWholeStream(is, (int) len);
          byte[] raw = rawbuf.array();
          int n = rawbuf.limit();
          if (!isBinary(raw, n)) {
            rawbuf = filterClean(raw, n);
            raw = rawbuf.array();
            n = rawbuf.limit();
          }
          return computeHash(new ByteArrayInputStream(raw, 0, n), n);
        }

        if (isBinary(e))
          return computeHash(is, len);

        final long canonLen;
        final InputStream lenIs = filterClean(e.openInputStream());
        try {
          canonLen = computeLength(lenIs);
        } finally {
          safeClose(lenIs);
        }

        return computeHash(filterClean(is), canonLen);
      } finally {
        safeClose(is);
      }
    } catch (IOException err) {
      // Can't read the file? Don't report the failure either.
      return zeroid;
    }
  }

  private static void safeClose(final InputStream in) {
    try {
      in.close();
    } catch (IOException err2) {
      // Suppress any error related to closing an input
      // stream. We don't care, we should not have any
      // outstanding data to flush or anything like that.
    }
  }

  private boolean mightNeedCleaning() {
    switch (getOptions().getAutoCRLF()) {
    case FALSE:
    default:
      return false;

    case TRUE:
    case INPUT:
      return true;
    }
  }

  private boolean isBinary(byte[] content, int sz) {
    return RawText.isBinary(content, sz);
  }

  private boolean isBinary(Entry entry) throws IOException {
    InputStream in = entry.openInputStream();
    try {
      return RawText.isBinary(in);
    } finally {
      safeClose(in);
    }
  }

  private ByteBuffer filterClean(byte[] src, int n)
      throws IOException {
    InputStream in = new ByteArrayInputStream(src);
    return IO.readWholeStream(filterClean(in), n);
  }

  private InputStream filterClean(InputStream in) {
    return new EolCanonicalizingInputStream(in);
  }

  /**
   * Returns the working tree options used by this iterator.
   *
   * @return working tree options
   */
  public WorkingTreeOptions getOptions() {
    return state.options;
  }

  @Override
  public int idOffset() {
    return 0;
  }

  @Override
  public void reset() {
    if (!first()) {
      ptr = 0;
      if (!eof())
        parseEntry();
    }
  }

  @Override
  public boolean first() {
    return ptr == 0;
  }

  @Override
  public boolean eof() {
    return ptr == entryCnt;
  }

  @Override
  public void next(final int delta) throws CorruptObjectException {
    ptr += delta;
    if (!eof())
      parseEntry();
  }

  @Override
  public void back(final int delta) throws CorruptObjectException {
    ptr -= delta;
    parseEntry();
  }

  private void parseEntry() {
    final Entry e = entries[ptr];
    mode = e.getMode().getBits();

    final int nameLen = e.encodedNameLen;
    ensurePathCapacity(pathOffset + nameLen, pathOffset);
    System.arraycopy(e.encodedName, 0, path, pathOffset, nameLen);
    pathLen = pathOffset + nameLen;
  }

  /**
   * Get the byte length of this entry.
   *
   * @return size of this file, in bytes.
   */
  public long getEntryLength() {
    return current().getLength();
  }

  /**
   * Get the last modified time of this entry.
   *
   * @return last modified time of this file, in milliseconds since the epoch
   *         (Jan 1, 1970 UTC).
   */
  public long getEntryLastModified() {
    return current().getLastModified();
  }

  /**
   * Obtain an input stream to read the file content.
   * <p>
   * Efficient implementations are not required. The caller will usually
   * obtain the stream only once per entry, if at all.
   * <p>
   * The input stream should not use buffering if the implementation can avoid
   * it. The caller will buffer as necessary to perform efficient block IO
   * operations.
   * <p>
   * The caller will close the stream once complete.
   *
   * @return a stream to read from the file.
   * @throws IOException
   *             the file could not be opened for reading.
   */
  public InputStream openEntryStream() throws IOException {
    return current().openInputStream();
  }

  /**
   * Determine if the current entry path is ignored by an ignore rule.
   *
   * @return true if the entry was ignored by an ignore rule file.
   * @throws IOException
   *             a relevant ignore rule file exists but cannot be read.
   */
  public boolean isEntryIgnored() throws IOException {
    return isEntryIgnored(pathLen);
  }

  /**
   * Determine if the entry path is ignored by an ignore rule.
   *
   * @param pLen
   *            the length of the path in the path buffer.
   * @return true if the entry is ignored by an ignore rule.
   * @throws IOException
   *             a relevant ignore rule file exists but cannot be read.
   */
  protected boolean isEntryIgnored(final int pLen) throws IOException {
    IgnoreNode rules = getIgnoreNode();
    if (rules != null) {
      // The ignore code wants path to start with a '/' if possible.
      // If we have the '/' in our path buffer because we are inside
      // a subdirectory include it in the range we convert to string.
      //
      int pOff = pathOffset;
      if (0 < pOff)
        pOff--;
      String p = TreeWalk.pathOf(path, pOff, pLen);
      switch (rules.isIgnored(p, FileMode.TREE.equals(mode))) {
      case IGNORED:
        return true;
      case NOT_IGNORED:
        return false;
      case CHECK_PARENT:
        break;
      }
    }
    if (parent instanceof WorkingTreeIterator)
      return ((WorkingTreeIterator) parent).isEntryIgnored(pLen);
    return false;
  }

  private IgnoreNode getIgnoreNode() throws IOException {
    if (ignoreNode instanceof PerDirectoryIgnoreNode)
      ignoreNode = ((PerDirectoryIgnoreNode) ignoreNode).load();
    return ignoreNode;
  }

  private static final Comparator<Entry> ENTRY_CMP = new Comparator<Entry>() {
    public int compare(final Entry o1, final Entry o2) {
      final byte[] a = o1.encodedName;
      final byte[] b = o2.encodedName;
      final int aLen = o1.encodedNameLen;
      final int bLen = o2.encodedNameLen;
      int cPos;

      for (cPos = 0; cPos < aLen && cPos < bLen; cPos++) {
        final int cmp = (a[cPos] & 0xff) - (b[cPos] & 0xff);
        if (cmp != 0)
          return cmp;
      }

      if (cPos < aLen)
        return (a[cPos] & 0xff) - lastPathChar(o2);
      if (cPos < bLen)
        return lastPathChar(o1) - (b[cPos] & 0xff);
      return lastPathChar(o1) - lastPathChar(o2);
    }
  };

  static int lastPathChar(final Entry e) {
    return e.getMode() == FileMode.TREE ? '/' : '\0';
  }

  /**
   * Constructor helper.
   *
   * @param list
   *            files in the subtree of the work tree this iterator operates
   *            on
   */
  protected void init(final Entry[] list) {
    // Filter out nulls, . and .. as these are not valid tree entries,
    // also cache the encoded forms of the path names for efficient use
    // later on during sorting and iteration.
    //
    entries = list;
    int i, o;

    final CharsetEncoder nameEncoder = state.nameEncoder;
    for (i = 0, o = 0; i < entries.length; i++) {
      final Entry e = entries[i];
      if (e == null)
        continue;
      final String name = e.getName();
      if (".".equals(name) || "..".equals(name))
        continue;
      if (Constants.DOT_GIT.equals(name))
        continue;
      if (Constants.DOT_GIT_IGNORE.equals(name))
        ignoreNode = new PerDirectoryIgnoreNode(e);
      if (i != o)
        entries[o] = e;
      e.encodeName(nameEncoder);
      o++;
    }
    entryCnt = o;
    Arrays.sort(entries, 0, entryCnt, ENTRY_CMP);

    contentIdFromPtr = -1;
    ptr = 0;
    if (!eof())
      parseEntry();
  }

  /**
   * Obtain the current entry from this iterator.
   *
   * @return the currently selected entry.
   */
  protected Entry current() {
    return entries[ptr];
  }

  /**
   * The result of a metadata-comparison between the current entry and a
   * {@link DirCacheEntry}
   */
  public enum MetadataDiff {
    /**
     * The entries are equal by metaData (mode, length,
     * modification-timestamp) or the <code>assumeValid</code> attribute of
     * the index entry is set
     */
    EQUAL,

    /**
     * The entries are not equal by metaData (mode, length) or the
     * <code>isUpdateNeeded</code> attribute of the index entry is set
     */
    DIFFER_BY_METADATA,

    /** index entry is smudged - can't use that entry for comparison */
    SMUDGED,

    /**
     * The entries are equal by metaData (mode, length) but differ by
     * modification-timestamp.
     */
    DIFFER_BY_TIMESTAMP
  }

  /**
   * Compare the metadata (mode, length, modification-timestamp) of the
   * current entry and a {@link DirCacheEntry}
   *
   * @param entry
   *            the {@link DirCacheEntry} to compare with
   * @return a {@link MetadataDiff} which tells whether and how the entries
   *         metadata differ
   */
  public MetadataDiff compareMetadata(DirCacheEntry entry) {
    if (entry.isAssumeValid())
      return MetadataDiff.EQUAL;

    if (entry.isUpdateNeeded())
      return MetadataDiff.DIFFER_BY_METADATA;

    if (!entry.isSmudged() && (getEntryLength() != entry.getLength()))
      return MetadataDiff.DIFFER_BY_METADATA;

    // Determine difference in mode-bits of file and index-entry. In the
    // bitwise presentation of modeDiff we'll have a '1' when the two modes
    // differ at this position.
    int modeDiff = getEntryRawMode() ^ entry.getRawMode();

    // Do not rely on filemode differences in case of symbolic links
    if (modeDiff != 0 && !FileMode.SYMLINK.equals(entry.getRawMode())) {
      // Ignore the executable file bits if WorkingTreeOptions tell me to
      // do so. Ignoring is done by setting the bits representing a
      // EXECUTABLE_FILE to '0' in modeDiff
      if (!state.options.isFileMode())
        modeDiff &= ~FileMode.EXECUTABLE_FILE.getBits();
      if (modeDiff != 0)
        // Report a modification if the modes still (after potentially
        // ignoring EXECUTABLE_FILE bits) differ
        return MetadataDiff.DIFFER_BY_METADATA;
    }

    // Git under windows only stores seconds so we round the timestamp
    // Java gives us if it looks like the timestamp in index is seconds
    // only. Otherwise we compare the timestamp at millisecond precision.
    long cacheLastModified = entry.getLastModified();
    long fileLastModified = getEntryLastModified();
    if (cacheLastModified % 1000 == 0)
      fileLastModified = fileLastModified - fileLastModified % 1000;

    if (fileLastModified != cacheLastModified)
      return MetadataDiff.DIFFER_BY_TIMESTAMP;
    else if (!entry.isSmudged())
      // The file is clean when you look at timestamps.
      return MetadataDiff.EQUAL;
    else
      return MetadataDiff.SMUDGED;
  }

  /**
   * Checks whether this entry differs from a given entry from the
   * {@link DirCache}.
   *
   * File status information is used and if status is same we consider the
   * file identical to the state in the working directory. Native git uses
   * more stat fields than we have accessible in Java.
   *
   * @param entry
   *            the entry from the dircache we want to compare against
   * @param forceContentCheck
   *            True if the actual file content should be checked if
   *            modification time differs.
   * @return true if content is most likely different.
   */
  public boolean isModified(DirCacheEntry entry, boolean forceContentCheck) {
    MetadataDiff diff = compareMetadata(entry);
    switch (diff) {
    case DIFFER_BY_TIMESTAMP:
      if (forceContentCheck)
        // But we are told to look at content even though timestamps
        // tell us about modification
        return contentCheck(entry);
      else
        // We are told to assume a modification if timestamps differs
        return true;
    case SMUDGED:
      // The file is clean by timestamps but the entry was smudged.
      // Lets do a content check
      return contentCheck(entry);
    case EQUAL:
      return false;
    case DIFFER_BY_METADATA:
      return true;
    default:
      throw new IllegalStateException(MessageFormat.format(
          JGitText.get().unexpectedCompareResult, diff.name()));
    }
  }

  /**
   * Compares the entries content with the content in the filesystem.
   * Unsmudges the entry when it is detected that it is clean.
   *
   * @param entry
   *            the entry to be checked
   * @return <code>true</code> if the content matches, <code>false</code>
   *         otherwise
   */
  private boolean contentCheck(DirCacheEntry entry) {
    if (getEntryObjectId().equals(entry.getObjectId())) {
      // Content has not changed

      // We know the entry can't be racily clean because it's still clean.
      // Therefore we unsmudge the entry!
      // If by any chance we now unsmudge although we are still in the
      // same time-slot as the last modification to the index file the
      // next index write operation will smudge again.
      // Caution: we are unsmudging just by setting the length of the
      // in-memory entry object. It's the callers task to detect that we
      // have modified the entry and to persist the modified index.
      entry.setLength((int) getEntryLength());

      return false;
    } else {
      // Content differs: that's a real change!
      return true;
    }
  }

  private long computeLength(InputStream in) throws IOException {
    // Since we only care about the length, use skip. The stream
    // may be able to more efficiently wade through its data.
    //
    long length = 0;
    for (;;) {
      long n = in.skip(1 << 20);
      if (n <= 0)
        break;
      length += n;
    }
    return length;
  }

  private byte[] computeHash(InputStream in, long length) throws IOException {
    final MessageDigest contentDigest = state.contentDigest;
    final byte[] contentReadBuffer = state.contentReadBuffer;

    contentDigest.reset();
    contentDigest.update(hblob);
    contentDigest.update((byte) ' ');

    long sz = length;
    if (sz == 0) {
      contentDigest.update((byte) '0');
    } else {
      final int bufn = contentReadBuffer.length;
      int p = bufn;
      do {
        contentReadBuffer[--p] = digits[(int) (sz % 10)];
        sz /= 10;
      } while (sz > 0);
      contentDigest.update(contentReadBuffer, p, bufn - p);
    }
    contentDigest.update((byte) 0);

    for (;;) {
      final int r = in.read(contentReadBuffer);
      if (r <= 0)
        break;
      contentDigest.update(contentReadBuffer, 0, r);
      sz += r;
    }
    if (sz != length)
      return zeroid;
    return contentDigest.digest();
  }

  /** A single entry within a working directory tree. */
  protected static abstract class Entry {
    byte[] encodedName;

    int encodedNameLen;

    void encodeName(final CharsetEncoder enc) {
      final ByteBuffer b;
      try {
        b = enc.encode(CharBuffer.wrap(getName()));
      } catch (CharacterCodingException e) {
        // This should so never happen.
        throw new RuntimeException(MessageFormat.format(
            JGitText.get().unencodeableFile, getName()));
      }

      encodedNameLen = b.limit();
      if (b.hasArray() && b.arrayOffset() == 0)
        encodedName = b.array();
      else
        b.get(encodedName = new byte[encodedNameLen]);
    }

    public String toString() {
      return getMode().toString() + " " + getName();
    }

    /**
     * Get the type of this entry.
     * <p>
     * <b>Note: Efficient implementation required.</b>
     * <p>
     * The implementation of this method must be efficient. If a subclass
     * needs to compute the value they should cache the reference within an
     * instance member instead.
     *
     * @return a file mode constant from {@link FileMode}.
     */
    public abstract FileMode getMode();

    /**
     * Get the byte length of this entry.
     * <p>
     * <b>Note: Efficient implementation required.</b>
     * <p>
     * The implementation of this method must be efficient. If a subclass
     * needs to compute the value they should cache the reference within an
     * instance member instead.
     *
     * @return size of this file, in bytes.
     */
    public abstract long getLength();

    /**
     * Get the last modified time of this entry.
     * <p>
     * <b>Note: Efficient implementation required.</b>
     * <p>
     * The implementation of this method must be efficient. If a subclass
     * needs to compute the value they should cache the reference within an
     * instance member instead.
     *
     * @return time since the epoch (in ms) of the last change.
     */
    public abstract long getLastModified();

    /**
     * Get the name of this entry within its directory.
     * <p>
     * Efficient implementations are not required. The caller will obtain
     * the name only once and cache it once obtained.
     *
     * @return name of the entry.
     */
    public abstract String getName();

    /**
     * Obtain an input stream to read the file content.
     * <p>
     * Efficient implementations are not required. The caller will usually
     * obtain the stream only once per entry, if at all.
     * <p>
     * The input stream should not use buffering if the implementation can
     * avoid it. The caller will buffer as necessary to perform efficient
     * block IO operations.
     * <p>
     * The caller will close the stream once complete.
     *
     * @return a stream to read from the file.
     * @throws IOException
     *             the file could not be opened for reading.
     */
    public abstract InputStream openInputStream() throws IOException;
  }

  /** Magic type indicating we know rules exist, but they aren't loaded. */
  private static class PerDirectoryIgnoreNode extends IgnoreNode {
    final Entry entry;

    PerDirectoryIgnoreNode(Entry entry) {
      super(Collections.<IgnoreRule> emptyList());
      this.entry = entry;
    }

    IgnoreNode load() throws IOException {
      IgnoreNode r = new IgnoreNode();
      InputStream in = entry.openInputStream();
      try {
        r.parse(in);
      } finally {
        in.close();
      }
      return r.getRules().isEmpty() ? null : r;
    }
  }

  /** Magic type indicating there may be rules for the top level. */
  private static class RootIgnoreNode extends PerDirectoryIgnoreNode {
    final Repository repository;

    RootIgnoreNode(Entry entry, Repository repository) {
      super(entry);
      this.repository = repository;
    }

    @Override
    IgnoreNode load() throws IOException {
      IgnoreNode r;
      if (entry != null) {
        r = super.load();
        if (r == null)
          r = new IgnoreNode();
      } else {
        r = new IgnoreNode();
      }

      FS fs = repository.getFS();
      String path = repository.getConfig().get(CoreConfig.KEY)
          .getExcludesFile();
      if (path != null) {
        File excludesfile;
        if (path.startsWith("~/"))
          excludesfile = fs.resolve(fs.userHome(), path.substring(2));
        else
          excludesfile = fs.resolve(null, path);
        loadRulesFromFile(r, excludesfile);
      }

      File exclude = fs
          .resolve(repository.getDirectory(), "info/exclude");
      loadRulesFromFile(r, exclude);

      return r.getRules().isEmpty() ? null : r;
    }

    private void loadRulesFromFile(IgnoreNode r, File exclude)
        throws FileNotFoundException, IOException {
      if (exclude.exists()) {
        FileInputStream in = new FileInputStream(exclude);
        try {
          r.parse(in);
        } finally {
          in.close();
        }
      }
    }
  }

  private static final class IteratorState {
    /** Options used to process the working tree. */
    final WorkingTreeOptions options;

    /** File name character encoder. */
    final CharsetEncoder nameEncoder;

    /** Digest computer for {@link #contentId} computations. */
    MessageDigest contentDigest;

    /** Buffer used to perform {@link #contentId} computations. */
    byte[] contentReadBuffer;

    /** TreeWalk with a (supposedly) matching DirCacheIterator. */
    TreeWalk walk;

    /** Position of the matching {@link DirCacheIterator}. */
    int dirCacheTree;

    IteratorState(WorkingTreeOptions options) {
      this.options = options;
      this.nameEncoder = Constants.CHARSET.newEncoder();
    }

    void initializeDigestAndReadBuffer() {
      if (contentDigest == null) {
        contentDigest = Constants.newMessageDigest();
        contentReadBuffer = new byte[BUFFER_SIZE];
      }
    }
  }
}
TOP

Related Classes of org.eclipse.jgit.treewalk.WorkingTreeIterator$IteratorState

TOP
Copyright © 2018 www.massapi.com. All rights reserved.
All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.