/*
 * Copyright (C) 2010, Google Inc. and others
 *
 * This program and the accompanying materials are made available under the
 * terms of the Eclipse Distribution License v. 1.0 which is available at
 * https://www.eclipse.org/org/documents/edl-v10.php.
 *
 * SPDX-License-Identifier: BSD-3-Clause
 */

package org.eclipse.jgit.notes;

import static org.eclipse.jgit.lib.Constants.OBJECT_ID_STRING_LENGTH;
import static org.eclipse.jgit.lib.Constants.encodeASCII;
import static org.eclipse.jgit.lib.FileMode.TREE;
import static org.eclipse.jgit.util.RawParseUtils.parseHexInt4;

import java.io.IOException;

import org.eclipse.jgit.errors.IncorrectObjectTypeException;
import org.eclipse.jgit.lib.AbbreviatedObjectId;
import org.eclipse.jgit.lib.FileMode;
import org.eclipse.jgit.lib.MutableObjectId;
import org.eclipse.jgit.lib.ObjectId;
import org.eclipse.jgit.lib.ObjectReader;
import org.eclipse.jgit.treewalk.CanonicalTreeParser;

Custom tree parser to select note bucket type and load it.
/** Custom tree parser to select note bucket type and load it. */
final class NoteParser extends CanonicalTreeParser {
Parse a tree object into a NoteBucket instance. The type of note tree is automatically detected by examining the items within the tree, and allocating the proper storage type based on the first note-like entry encountered. Since the method parses by guessing the type on the first element, malformed note trees can be read as the wrong type of tree. This method is not recursive, it parses the one tree given to it and returns the bucket. If there are subtrees for note storage, they are setup as lazy pointers that will be resolved at a later time.
Params:
  • prefix – common hex digits that all notes within this tree share. The root tree has prefix.length() == 0, the first-level subtrees should be prefix.length()==2, etc.
  • treeId – the tree to read from the repository.
  • reader – reader to access the tree object.
Throws:
Returns:bucket to holding the notes of the specified tree.
/** * Parse a tree object into a {@link NoteBucket} instance. * * The type of note tree is automatically detected by examining the items * within the tree, and allocating the proper storage type based on the * first note-like entry encountered. Since the method parses by guessing * the type on the first element, malformed note trees can be read as the * wrong type of tree. * * This method is not recursive, it parses the one tree given to it and * returns the bucket. If there are subtrees for note storage, they are * setup as lazy pointers that will be resolved at a later time. * * @param prefix * common hex digits that all notes within this tree share. The * root tree has {@code prefix.length() == 0}, the first-level * subtrees should be {@code prefix.length()==2}, etc. * @param treeId * the tree to read from the repository. * @param reader * reader to access the tree object. * @return bucket to holding the notes of the specified tree. * @throws IOException * {@code treeId} cannot be accessed. */
static InMemoryNoteBucket parse(AbbreviatedObjectId prefix, final ObjectId treeId, final ObjectReader reader) throws IOException { return new NoteParser(prefix, reader, treeId).parse(); } private final int prefixLen; private final int pathPadding; private NonNoteEntry firstNonNote; private NonNoteEntry lastNonNote; private NoteParser(AbbreviatedObjectId prefix, ObjectReader r, ObjectId t) throws IncorrectObjectTypeException, IOException { super(encodeASCII(prefix.name()), r, t); prefixLen = prefix.length(); // Our path buffer has a '/' that we don't want after the prefix. // Drop it by shifting the path down one position. pathPadding = 0 < prefixLen ? 1 : 0; if (0 < pathPadding) System.arraycopy(path, 0, path, pathPadding, prefixLen); } private InMemoryNoteBucket parse() { InMemoryNoteBucket r = parseTree(); r.nonNotes = firstNonNote; return r; } private InMemoryNoteBucket parseTree() { for (; !eof(); next(1)) { if (pathLen == pathPadding + OBJECT_ID_STRING_LENGTH && isHex()) return parseLeafTree(); else if (getNameLength() == 2 && isHex() && isTree()) return parseFanoutTree(); else storeNonNote(); } // If we cannot determine the style used, assume its a leaf. return new LeafBucket(prefixLen); } private LeafBucket parseLeafTree() { final LeafBucket leaf = new LeafBucket(prefixLen); final MutableObjectId idBuf = new MutableObjectId(); for (; !eof(); next(1)) { if (parseObjectId(idBuf)) leaf.parseOneEntry(idBuf, getEntryObjectId()); else storeNonNote(); } return leaf; } private boolean parseObjectId(MutableObjectId id) { if (pathLen == pathPadding + OBJECT_ID_STRING_LENGTH) { try { id.fromString(path, pathPadding); return true; } catch (ArrayIndexOutOfBoundsException notHex) { return false; } } return false; } private FanoutBucket parseFanoutTree() { final FanoutBucket fanout = new FanoutBucket(prefixLen); for (; !eof(); next(1)) { final int cell = parseFanoutCell(); if (0 <= cell) fanout.setBucket(cell, getEntryObjectId()); else storeNonNote(); } return fanout; } private int parseFanoutCell() { if (getNameLength() == 2 && isTree()) { try { return (parseHexInt4(path[pathOffset + 0]) << 4) | parseHexInt4(path[pathOffset + 1]); } catch (ArrayIndexOutOfBoundsException notHex) { return -1; } } return -1; } private void storeNonNote() { ObjectId id = getEntryObjectId(); FileMode fileMode = getEntryFileMode(); byte[] name = new byte[getNameLength()]; getName(name, 0); NonNoteEntry ent = new NonNoteEntry(name, fileMode, id); if (firstNonNote == null) firstNonNote = ent; if (lastNonNote != null) lastNonNote.next = ent; lastNonNote = ent; } private boolean isTree() { return TREE.equals(mode); } private boolean isHex() { try { for (int i = pathOffset; i < pathLen; i++) parseHexInt4(path[i]); return true; } catch (ArrayIndexOutOfBoundsException fail) { return false; } } }