2 * Copyright (C) 2007, Robin Rosenberg <robin.rosenberg@dewire.com>
3 * Copyright (C) 2008, Shawn O. Pearce <spearce@spearce.org>
7 * Redistribution and use in source and binary forms, with or
8 * without modification, are permitted provided that the following
11 * - Redistributions of source code must retain the above copyright
12 * notice, this list of conditions and the following disclaimer.
14 * - Redistributions in binary form must reproduce the above
15 * copyright notice, this list of conditions and the following
16 * disclaimer in the documentation and/or other materials provided
17 * with the distribution.
19 * - Neither the name of the Git Development Community nor the
20 * names of its contributors may be used to endorse or promote
21 * products derived from this software without specific prior
24 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
25 * CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
26 * INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
27 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
28 * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR
29 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
30 * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
31 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
32 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
33 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
34 * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
35 * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
36 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
39 package org
.spearce
.jgit
.treewalk
;
41 import java
.io
.IOException
;
42 import java
.nio
.ByteBuffer
;
43 import java
.nio
.CharBuffer
;
45 import org
.spearce
.jgit
.errors
.CorruptObjectException
;
46 import org
.spearce
.jgit
.errors
.IncorrectObjectTypeException
;
47 import org
.spearce
.jgit
.lib
.Constants
;
48 import org
.spearce
.jgit
.lib
.FileMode
;
49 import org
.spearce
.jgit
.lib
.ObjectId
;
50 import org
.spearce
.jgit
.lib
.Repository
;
51 import org
.spearce
.jgit
.treewalk
.filter
.TreeFilter
;
54 * Walks a Git tree (directory) in Git sort order.
56 * A new iterator instance should be positioned before the first entry. The data
57 * about the first entry is not available until after the first call to
58 * {@link #next()} is made.
60 * Implementors must walk a tree in the Git sort order, which has the following
68 * In the second item, <code>A</code> is the name of a subtree and
69 * <code>c</code> is a file within that subtree. The other two items are files
70 * in the root level tree.
72 * @see CanonicalTreeParser
74 public abstract class AbstractTreeIterator
{
75 private static final int DEFAULT_PATH_SIZE
= 128;
77 /** A dummy object id buffer that matches the zero ObjectId. */
78 protected static final byte[] zeroid
= new byte[Constants
.OBJECT_ID_LENGTH
];
80 /** Iterator for the parent tree; null if we are the root iterator. */
81 final AbstractTreeIterator parent
;
83 /** The iterator this current entry is path equal to. */
84 AbstractTreeIterator matches
;
87 * Mode bits for the current entry.
89 * A numerical value from FileMode is usually faster for an iterator to
90 * obtain from its data source so this is the preferred representation.
92 * @see org.spearce.jgit.lib.FileMode
97 * Path buffer for the current entry.
99 * This buffer is pre-allocated at the start of walking and is shared from
100 * parent iterators down into their subtree iterators. The sharing allows
101 * the current entry to always be a full path from the root, while each
102 * subtree only needs to populate the part that is under their control.
104 protected byte[] path
;
107 * Position within {@link #path} this iterator starts writing at.
109 * This is the first offset in {@link #path} that this iterator must
110 * populate during {@link #next}. At the root level (when {@link #parent}
111 * is null) this is 0. For a subtree iterator the index before this position
112 * should have the value '/'.
114 protected final int pathOffset
;
117 * Total length of the current entry's complete path from the root.
119 * This is the number of bytes within {@link #path} that pertain to the
120 * current entry. Values at this index through the end of the array are
121 * garbage and may be randomly populated from prior entries.
123 protected int pathLen
;
125 /** Create a new iterator with no parent. */
126 protected AbstractTreeIterator() {
128 path
= new byte[DEFAULT_PATH_SIZE
];
133 * Create a new iterator with no parent and a prefix.
135 * The prefix path supplied is inserted in front of all paths generated by
136 * this iterator. It is intended to be used when an iterator is being
137 * created for a subsection of an overall repository and needs to be
138 * combined with other iterators that are created to run over the entire
139 * repository namespace.
142 * position of this iterator in the repository tree. The value
143 * may be null or the empty string to indicate the prefix is the
144 * root of the repository. A trailing slash ('/') is
145 * automatically appended if the prefix does not end in '/'.
147 protected AbstractTreeIterator(final String prefix
) {
150 if (prefix
!= null && prefix
.length() > 0) {
153 b
= Constants
.CHARSET
.encode(CharBuffer
.wrap(prefix
));
155 path
= new byte[Math
.max(DEFAULT_PATH_SIZE
, pathLen
+ 1)];
156 b
.get(path
, 0, pathLen
);
157 if (path
[pathLen
- 1] != '/')
158 path
[pathLen
++] = '/';
159 pathOffset
= pathLen
;
161 path
= new byte[DEFAULT_PATH_SIZE
];
167 * Create an iterator for a subtree of an existing iterator.
170 * parent tree iterator.
172 protected AbstractTreeIterator(final AbstractTreeIterator p
) {
175 pathOffset
= p
.pathLen
+ 1;
177 path
[pathOffset
- 1] = '/';
178 } catch (ArrayIndexOutOfBoundsException e
) {
180 path
[pathOffset
- 1] = '/';
185 * Create an iterator for a subtree of an existing iterator.
187 * The caller is responsible for setting up the path of the child iterator.
190 * parent tree iterator.
192 * path array to be used by the child iterator. This path must
193 * contain the path from the top of the walk to the first child
194 * and must end with a '/'.
195 * @param childPathOffset
196 * position within <code>childPath</code> where the child can
197 * insert its data. The value at
198 * <code>childPath[childPathOffset-1]</code> must be '/'.
200 protected AbstractTreeIterator(final AbstractTreeIterator p
,
201 final byte[] childPath
, final int childPathOffset
) {
204 pathOffset
= childPathOffset
;
208 * Grow the path buffer larger.
211 * number of live bytes in the path buffer. This many bytes will
212 * be moved into the larger buffer.
214 protected void growPath(final int len
) {
215 final byte[] n
= new byte[path
.length
<< 1];
216 System
.arraycopy(path
, 0, n
, 0, len
);
217 for (AbstractTreeIterator p
= this; p
!= null; p
= p
.parent
)
222 * Compare the path of this current entry to another iterator's entry.
225 * the other iterator to compare the path against.
226 * @return -1 if this entry sorts first; 0 if the entries are equal; 1 if
227 * p's entry sorts first.
229 public int pathCompare(final AbstractTreeIterator p
) {
230 return pathCompare(p
, p
.mode
);
233 int pathCompare(final AbstractTreeIterator p
, final int pMode
) {
234 final byte[] a
= path
;
235 final byte[] b
= p
.path
;
236 final int aLen
= pathLen
;
237 final int bLen
= p
.pathLen
;
240 // Its common when we are a subtree for both parents to match;
241 // when this happens everything in path[0..cPos] is known to
242 // be equal and does not require evaluation again.
244 cPos
= alreadyMatch(this, p
);
246 for (; cPos
< aLen
&& cPos
< bLen
; cPos
++) {
247 final int cmp
= (a
[cPos
] & 0xff) - (b
[cPos
] & 0xff);
253 return (a
[cPos
] & 0xff) - lastPathChar(pMode
);
255 return lastPathChar(mode
) - (b
[cPos
] & 0xff);
256 return lastPathChar(mode
) - lastPathChar(pMode
);
259 private static int alreadyMatch(AbstractTreeIterator a
,
260 AbstractTreeIterator b
) {
262 final AbstractTreeIterator ap
= a
.parent
;
263 final AbstractTreeIterator bp
= b
.parent
;
264 if (ap
== null || bp
== null)
266 if (ap
.matches
== bp
.matches
)
273 private static int lastPathChar(final int mode
) {
274 return FileMode
.TREE
.equals(mode
) ?
'/' : '\0';
278 * Check if the current entry of both iterators has the same id.
280 * This method is faster than {@link #getEntryObjectId()} as it does not
281 * require copying the bytes out of the buffers. A direct {@link #idBuffer}
282 * compare operation is performed.
284 * @param otherIterator
285 * the other iterator to test against.
286 * @return true if both iterators have the same object id; false otherwise.
288 public boolean idEqual(final AbstractTreeIterator otherIterator
) {
289 return ObjectId
.equals(idBuffer(), idOffset(),
290 otherIterator
.idBuffer(), otherIterator
.idOffset());
294 * Get the object id of the current entry.
296 * @return an object id for the current entry.
298 public ObjectId
getEntryObjectId() {
299 return ObjectId
.fromRaw(idBuffer(), idOffset());
303 * Get the byte array buffer object IDs must be copied out of.
305 * The id buffer contains the bytes necessary to construct an ObjectId for
306 * the current entry of this iterator. The buffer can be the same buffer for
307 * all entries, or it can be a unique buffer per-entry. Implementations are
308 * encouraged to expose their private buffer whenever possible to reduce
309 * garbage generation and copying costs.
311 * @return byte array the implementation stores object IDs within.
312 * @see #getEntryObjectId()
314 public abstract byte[] idBuffer();
317 * Get the position within {@link #idBuffer()} of this entry's ObjectId.
319 * @return offset into the array returned by {@link #idBuffer()} where the
320 * ObjectId must be copied out of.
322 public abstract int idOffset();
325 * Create a new iterator for the current entry's subtree.
327 * The parent reference of the iterator must be <code>this</code>,
328 * otherwise the caller would not be able to exit out of the subtree
329 * iterator correctly and return to continue walking <code>this</code>.
332 * repository to load the tree data from.
333 * @return a new parser that walks over the current subtree.
334 * @throws IncorrectObjectTypeException
335 * the current entry is not actually a tree and cannot be parsed
336 * as though it were a tree.
337 * @throws IOException
338 * a loose object or pack file could not be read.
340 public abstract AbstractTreeIterator
createSubtreeIterator(Repository repo
)
341 throws IncorrectObjectTypeException
, IOException
;
344 * Is this tree iterator at its EOF point (no more entries)?
346 * An iterator is at EOF if there is no current entry.
348 * @return true if we have walked all entries and have none left.
350 public abstract boolean eof();
353 * Advance to the next tree entry, populating this iterator with its data.
355 * Implementations must populate the following members:
357 * <li>{@link #mode}</li>
358 * <li>{@link #path} (from {@link #pathOffset} to {@link #pathLen})</li>
359 * <li>{@link #pathLen}</li>
361 * as well as any implementation dependent information necessary to
362 * accurately return data from {@link #idBuffer()} and {@link #idOffset()}
365 * @throws CorruptObjectException
366 * the tree is invalid.
368 public abstract void next() throws CorruptObjectException
;
371 * Advance to the next tree entry, populating this iterator with its data.
373 * This method behaves like {@link #next()} but is called by
374 * {@link TreeWalk} only if a {@link TreeFilter} was used and ruled out the
375 * current entry from the results. In such cases this tree iterator may
376 * perform special behavior.
378 * @throws CorruptObjectException
379 * the tree is invalid.
381 public void skip() throws CorruptObjectException
{
386 * Indicates to the iterator that no more entries will be read.
388 * This is only invoked by TreeWalk when the iteration is aborted early due
389 * to a {@link org.spearce.jgit.errors.StopWalkException} being thrown from
390 * within a TreeFilter.
392 public void stopWalk() {
393 // Do nothing by default. Most iterators do not care.