123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650651652653654655656657658659660661662663 |
- /*
- * Copyright (C) 2009, Google Inc.
- * Copyright (C) 2008-2009, Johannes E. Schindelin <johannes.schindelin@gmx.de>
- * 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.diff;
-
- import static org.eclipse.jgit.lib.Constants.encode;
- import static org.eclipse.jgit.lib.Constants.encodeASCII;
- import static org.eclipse.jgit.lib.FileMode.GITLINK;
-
- import java.io.ByteArrayOutputStream;
- import java.io.IOException;
- import java.io.OutputStream;
- import java.util.Collection;
- import java.util.List;
-
- import org.eclipse.jgit.JGitText;
- import org.eclipse.jgit.errors.AmbiguousObjectException;
- import org.eclipse.jgit.errors.CorruptObjectException;
- import org.eclipse.jgit.errors.MissingObjectException;
- import org.eclipse.jgit.lib.AbbreviatedObjectId;
- import org.eclipse.jgit.lib.Constants;
- import org.eclipse.jgit.lib.CoreConfig;
- import org.eclipse.jgit.lib.FileMode;
- import org.eclipse.jgit.lib.ObjectId;
- import org.eclipse.jgit.lib.ObjectLoader;
- import org.eclipse.jgit.lib.ObjectReader;
- import org.eclipse.jgit.lib.Repository;
- import org.eclipse.jgit.patch.FileHeader;
- import org.eclipse.jgit.patch.HunkHeader;
- import org.eclipse.jgit.patch.FileHeader.PatchType;
- import org.eclipse.jgit.util.QuotedString;
- import org.eclipse.jgit.util.io.DisabledOutputStream;
-
- /**
- * Format a Git style patch script.
- */
- public class DiffFormatter {
- private static final byte[] noNewLine = encodeASCII("\\ No newline at end of file\n");
-
- private final OutputStream out;
-
- private Repository db;
-
- private int context = 3;
-
- private int abbreviationLength = 7;
-
- private RawText.Factory rawTextFactory = RawText.FACTORY;
-
- private int bigFileThreshold = 50 * 1024 * 1024;
-
- /**
- * Create a new formatter with a default level of context.
- *
- * @param out
- * the stream the formatter will write line data to. This stream
- * should have buffering arranged by the caller, as many small
- * writes are performed to it.
- */
- public DiffFormatter(OutputStream out) {
- this.out = out;
- }
-
- /** @return the stream we are outputting data to. */
- protected OutputStream getOutputStream() {
- return out;
- }
-
- /**
- * Set the repository the formatter can load object contents from.
- *
- * @param repository
- * source repository holding referenced objects.
- */
- public void setRepository(Repository repository) {
- db = repository;
-
- CoreConfig cfg = db.getConfig().get(CoreConfig.KEY);
- bigFileThreshold = cfg.getStreamFileThreshold();
- }
-
- /**
- * Change the number of lines of context to display.
- *
- * @param lineCount
- * number of lines of context to see before the first
- * modification and after the last modification within a hunk of
- * the modified file.
- */
- public void setContext(final int lineCount) {
- if (lineCount < 0)
- throw new IllegalArgumentException(
- JGitText.get().contextMustBeNonNegative);
- context = lineCount;
- }
-
- /**
- * Change the number of digits to show in an ObjectId.
- *
- * @param count
- * number of digits to show in an ObjectId.
- */
- public void setAbbreviationLength(final int count) {
- if (count < 0)
- throw new IllegalArgumentException(
- JGitText.get().abbreviationLengthMustBeNonNegative);
- abbreviationLength = count;
- }
-
- /**
- * Set the helper that constructs difference output.
- *
- * @param type
- * the factory to create different output. Different types of
- * factories can produce different whitespace behavior, for
- * example.
- * @see RawText#FACTORY
- * @see RawTextIgnoreAllWhitespace#FACTORY
- * @see RawTextIgnoreLeadingWhitespace#FACTORY
- * @see RawTextIgnoreTrailingWhitespace#FACTORY
- * @see RawTextIgnoreWhitespaceChange#FACTORY
- */
- public void setRawTextFactory(RawText.Factory type) {
- rawTextFactory = type;
- }
-
- /**
- * Set the maximum file size that should be considered for diff output.
- * <p>
- * Text files that are larger than this size will not have a difference
- * generated during output.
- *
- * @param bigFileThreshold
- * the limit, in bytes.
- */
- public void setBigFileThreshold(int bigFileThreshold) {
- this.bigFileThreshold = bigFileThreshold;
- }
-
- /**
- * Flush the underlying output stream of this formatter.
- *
- * @throws IOException
- * the stream's own flush method threw an exception.
- */
- public void flush() throws IOException {
- out.flush();
- }
-
- /**
- * Format a patch script from a list of difference entries.
- *
- * @param entries
- * entries describing the affected files.
- * @throws IOException
- * a file's content cannot be read, or the output stream cannot
- * be written to.
- */
- public void format(List<? extends DiffEntry> entries) throws IOException {
- for (DiffEntry ent : entries)
- format(ent);
- }
-
- /**
- * Format a patch script for one file entry.
- *
- * @param ent
- * the entry to be formatted.
- * @throws IOException
- * a file's content cannot be read, or the output stream cannot
- * be written to.
- */
- public void format(DiffEntry ent) throws IOException {
- writeDiffHeader(out, ent);
-
- if (ent.getOldMode() == GITLINK || ent.getNewMode() == GITLINK) {
- writeGitLinkDiffText(out, ent);
- } else {
- if (db == null)
- throw new IllegalStateException(
- JGitText.get().repositoryIsRequired);
-
- ObjectReader reader = db.newObjectReader();
- byte[] aRaw, bRaw;
- try {
- aRaw = open(reader, ent.getOldMode(), ent.getOldId());
- bRaw = open(reader, ent.getNewMode(), ent.getNewId());
- } finally {
- reader.release();
- }
-
- if (RawText.isBinary(aRaw) || RawText.isBinary(bRaw)) {
- out.write(encodeASCII("Binary files differ\n"));
-
- } else {
- RawText a = rawTextFactory.create(aRaw);
- RawText b = rawTextFactory.create(bRaw);
- formatEdits(a, b, new MyersDiff(a, b).getEdits());
- }
- }
- }
-
- private void writeGitLinkDiffText(OutputStream o, DiffEntry ent)
- throws IOException {
- if (ent.getOldMode() == GITLINK) {
- o.write(encodeASCII("-Subproject commit " + ent.getOldId().name()
- + "\n"));
- }
- if (ent.getNewMode() == GITLINK) {
- o.write(encodeASCII("+Subproject commit " + ent.getNewId().name()
- + "\n"));
- }
- }
-
- private void writeDiffHeader(OutputStream o, DiffEntry ent)
- throws IOException {
- String oldName = quotePath("a/" + ent.getOldPath());
- String newName = quotePath("b/" + ent.getNewPath());
- o.write(encode("diff --git " + oldName + " " + newName + "\n"));
-
- switch (ent.getChangeType()) {
- case ADD:
- o.write(encodeASCII("new file mode "));
- ent.getNewMode().copyTo(o);
- o.write('\n');
- break;
-
- case DELETE:
- o.write(encodeASCII("deleted file mode "));
- ent.getOldMode().copyTo(o);
- o.write('\n');
- break;
-
- case RENAME:
- o.write(encodeASCII("similarity index " + ent.getScore() + "%"));
- o.write('\n');
-
- o.write(encode("rename from " + quotePath(ent.getOldPath())));
- o.write('\n');
-
- o.write(encode("rename to " + quotePath(ent.getNewPath())));
- o.write('\n');
- break;
-
- case COPY:
- o.write(encodeASCII("similarity index " + ent.getScore() + "%"));
- o.write('\n');
-
- o.write(encode("copy from " + quotePath(ent.getOldPath())));
- o.write('\n');
-
- o.write(encode("copy to " + quotePath(ent.getNewPath())));
- o.write('\n');
-
- if (!ent.getOldMode().equals(ent.getNewMode())) {
- o.write(encodeASCII("new file mode "));
- ent.getNewMode().copyTo(o);
- o.write('\n');
- }
- break;
- case MODIFY:
- int score = ent.getScore();
- if (0 < score && score <= 100) {
- o.write(encodeASCII("dissimilarity index " + (100 - score)
- + "%"));
- o.write('\n');
- }
- break;
- }
-
- switch (ent.getChangeType()) {
- case RENAME:
- case MODIFY:
- if (!ent.getOldMode().equals(ent.getNewMode())) {
- o.write(encodeASCII("old mode "));
- ent.getOldMode().copyTo(o);
- o.write('\n');
-
- o.write(encodeASCII("new mode "));
- ent.getNewMode().copyTo(o);
- o.write('\n');
- }
- }
-
- o.write(encodeASCII("index " //
- + format(ent.getOldId()) //
- + ".." //
- + format(ent.getNewId())));
- if (ent.getOldMode().equals(ent.getNewMode())) {
- o.write(' ');
- ent.getNewMode().copyTo(o);
- }
- o.write('\n');
- o.write(encode("--- " + oldName + '\n'));
- o.write(encode("+++ " + newName + '\n'));
- }
-
- private String format(AbbreviatedObjectId id) {
- if (id.isComplete() && db != null) {
- ObjectReader reader = db.newObjectReader();
- try {
- id = reader.abbreviate(id.toObjectId(), abbreviationLength);
- } catch (IOException cannotAbbreviate) {
- // Ignore this. We'll report the full identity.
- } finally {
- reader.release();
- }
- }
- return id.name();
- }
-
- private static String quotePath(String name) {
- String q = QuotedString.GIT_PATH.quote(name);
- return ('"' + name + '"').equals(q) ? name : q;
- }
-
- private byte[] open(ObjectReader reader, FileMode mode,
- AbbreviatedObjectId id) throws IOException {
- if (mode == FileMode.MISSING)
- return new byte[] {};
-
- if (mode.getObjectType() != Constants.OBJ_BLOB)
- return new byte[] {};
-
- if (!id.isComplete()) {
- Collection<ObjectId> ids = reader.resolve(id);
- if (ids.size() == 1)
- id = AbbreviatedObjectId.fromObjectId(ids.iterator().next());
- else if (ids.size() == 0)
- throw new MissingObjectException(id, Constants.OBJ_BLOB);
- else
- throw new AmbiguousObjectException(id, ids);
- }
-
- ObjectLoader ldr = reader.open(id.toObjectId());
- return ldr.getCachedBytes(bigFileThreshold);
- }
-
- /**
- * Format a patch script, reusing a previously parsed FileHeader.
- * <p>
- * This formatter is primarily useful for editing an existing patch script
- * to increase or reduce the number of lines of context within the script.
- * All header lines are reused as-is from the supplied FileHeader.
- *
- * @param head
- * existing file header containing the header lines to copy.
- * @param a
- * text source for the pre-image version of the content. This
- * must match the content of {@link FileHeader#getOldId()}.
- * @param b
- * text source for the post-image version of the content. This
- * must match the content of {@link FileHeader#getNewId()}.
- * @throws IOException
- * writing to the supplied stream failed.
- */
- public void format(final FileHeader head, final RawText a, final RawText b)
- throws IOException {
- // Reuse the existing FileHeader as-is by blindly copying its
- // header lines, but avoiding its hunks. Instead we recreate
- // the hunks from the text instances we have been supplied.
- //
- final int start = head.getStartOffset();
- int end = head.getEndOffset();
- if (!head.getHunks().isEmpty())
- end = head.getHunks().get(0).getStartOffset();
- out.write(head.getBuffer(), start, end - start);
-
- formatEdits(a, b, head.toEditList());
- }
-
- /**
- * Formats a list of edits in unified diff format
- *
- * @param a
- * the text A which was compared
- * @param b
- * the text B which was compared
- * @param edits
- * some differences which have been calculated between A and B
- * @throws IOException
- */
- public void formatEdits(final RawText a, final RawText b,
- final EditList edits) throws IOException {
- for (int curIdx = 0; curIdx < edits.size();) {
- Edit curEdit = edits.get(curIdx);
- final int endIdx = findCombinedEnd(edits, curIdx);
- final Edit endEdit = edits.get(endIdx);
-
- int aCur = Math.max(0, curEdit.getBeginA() - context);
- int bCur = Math.max(0, curEdit.getBeginB() - context);
- final int aEnd = Math.min(a.size(), endEdit.getEndA() + context);
- final int bEnd = Math.min(b.size(), endEdit.getEndB() + context);
-
- writeHunkHeader(aCur, aEnd, bCur, bEnd);
-
- while (aCur < aEnd || bCur < bEnd) {
- if (aCur < curEdit.getBeginA() || endIdx + 1 < curIdx) {
- writeContextLine(a, aCur);
- if (isEndOfLineMissing(a, aCur))
- out.write(noNewLine);
- aCur++;
- bCur++;
- } else if (aCur < curEdit.getEndA()) {
- writeRemovedLine(a, aCur);
- if (isEndOfLineMissing(a, aCur))
- out.write(noNewLine);
- aCur++;
- } else if (bCur < curEdit.getEndB()) {
- writeAddedLine(b, bCur);
- if (isEndOfLineMissing(b, bCur))
- out.write(noNewLine);
- bCur++;
- }
-
- if (end(curEdit, aCur, bCur) && ++curIdx < edits.size())
- curEdit = edits.get(curIdx);
- }
- }
- }
-
- /**
- * Output a line of context (unmodified line).
- *
- * @param text
- * RawText for accessing raw data
- * @param line
- * the line number within text
- * @throws IOException
- */
- protected void writeContextLine(final RawText text, final int line)
- throws IOException {
- writeLine(' ', text, line);
- }
-
- private boolean isEndOfLineMissing(final RawText text, final int line) {
- return line + 1 == text.size() && text.isMissingNewlineAtEnd();
- }
-
- /**
- * Output an added line.
- *
- * @param text
- * RawText for accessing raw data
- * @param line
- * the line number within text
- * @throws IOException
- */
- protected void writeAddedLine(final RawText text, final int line)
- throws IOException {
- writeLine('+', text, line);
- }
-
- /**
- * Output a removed line
- *
- * @param text
- * RawText for accessing raw data
- * @param line
- * the line number within text
- * @throws IOException
- */
- protected void writeRemovedLine(final RawText text, final int line)
- throws IOException {
- writeLine('-', text, line);
- }
-
- /**
- * Output a hunk header
- *
- * @param aStartLine
- * within first source
- * @param aEndLine
- * within first source
- * @param bStartLine
- * within second source
- * @param bEndLine
- * within second source
- * @throws IOException
- */
- protected void writeHunkHeader(int aStartLine, int aEndLine,
- int bStartLine, int bEndLine) throws IOException {
- out.write('@');
- out.write('@');
- writeRange('-', aStartLine + 1, aEndLine - aStartLine);
- writeRange('+', bStartLine + 1, bEndLine - bStartLine);
- out.write(' ');
- out.write('@');
- out.write('@');
- out.write('\n');
- }
-
- private void writeRange(final char prefix, final int begin, final int cnt)
- throws IOException {
- out.write(' ');
- out.write(prefix);
- switch (cnt) {
- case 0:
- // If the range is empty, its beginning number must be the
- // line just before the range, or 0 if the range is at the
- // start of the file stream. Here, begin is always 1 based,
- // so an empty file would produce "0,0".
- //
- out.write(encodeASCII(begin - 1));
- out.write(',');
- out.write('0');
- break;
-
- case 1:
- // If the range is exactly one line, produce only the number.
- //
- out.write(encodeASCII(begin));
- break;
-
- default:
- out.write(encodeASCII(begin));
- out.write(',');
- out.write(encodeASCII(cnt));
- break;
- }
- }
-
- /**
- * Write a standard patch script line.
- *
- * @param prefix
- * prefix before the line, typically '-', '+', ' '.
- * @param text
- * the text object to obtain the line from.
- * @param cur
- * line number to output.
- * @throws IOException
- * the stream threw an exception while writing to it.
- */
- protected void writeLine(final char prefix, final RawText text,
- final int cur) throws IOException {
- out.write(prefix);
- text.writeLine(out, cur);
- out.write('\n');
- }
-
- /**
- * Creates a {@link FileHeader} representing the given {@link DiffEntry}
- * <p>
- * This method does not use the OutputStream associated with this
- * DiffFormatter instance. It is therefore safe to instantiate this
- * DiffFormatter instance with a {@link DisabledOutputStream} if this method
- * is the only one that will be used.
- *
- * @param ent
- * the DiffEntry to create the FileHeader for
- * @return a FileHeader representing the DiffEntry. The FileHeader's buffer
- * will contain only the header of the diff output. It will also
- * contain one {@link HunkHeader}.
- * @throws IOException
- * the stream threw an exception while writing to it, or one of
- * the blobs referenced by the DiffEntry could not be read.
- * @throws CorruptObjectException
- * one of the blobs referenced by the DiffEntry is corrupt.
- * @throws MissingObjectException
- * one of the blobs referenced by the DiffEntry is missing.
- */
- public FileHeader createFileHeader(DiffEntry ent) throws IOException,
- CorruptObjectException, MissingObjectException {
- ByteArrayOutputStream buf = new ByteArrayOutputStream();
- final EditList editList;
- final FileHeader.PatchType type;
-
- writeDiffHeader(buf, ent);
-
- if (ent.getOldMode() == GITLINK || ent.getNewMode() == GITLINK) {
- writeGitLinkDiffText(buf, ent);
- editList = new EditList();
- type = PatchType.UNIFIED;
- } else {
- if (db == null)
- throw new IllegalStateException(
- JGitText.get().repositoryIsRequired);
- ObjectReader reader = db.newObjectReader();
- byte[] aRaw, bRaw;
- try {
- aRaw = open(reader, ent.getOldMode(), ent.getOldId());
- bRaw = open(reader, ent.getNewMode(), ent.getNewId());
- } finally {
- reader.release();
- }
-
- if (RawText.isBinary(aRaw) || RawText.isBinary(bRaw)) {
- buf.write(encodeASCII("Binary files differ\n"));
- editList = new EditList();
- type = PatchType.BINARY;
- } else {
- RawText a = rawTextFactory.create(aRaw);
- RawText b = rawTextFactory.create(bRaw);
- editList = new MyersDiff(a, b).getEdits();
- type = PatchType.UNIFIED;
- }
- }
-
- return new FileHeader(buf.toByteArray(), editList, type);
- }
-
- private int findCombinedEnd(final List<Edit> edits, final int i) {
- int end = i + 1;
- while (end < edits.size()
- && (combineA(edits, end) || combineB(edits, end)))
- end++;
- return end - 1;
- }
-
- private boolean combineA(final List<Edit> e, final int i) {
- return e.get(i).getBeginA() - e.get(i - 1).getEndA() <= 2 * context;
- }
-
- private boolean combineB(final List<Edit> e, final int i) {
- return e.get(i).getBeginB() - e.get(i - 1).getEndB() <= 2 * context;
- }
-
- private static boolean end(final Edit edit, final int a, final int b) {
- return edit.getEndA() <= a && edit.getEndB() <= b;
- }
- }
|