Merge: Add diff3 style merge conflict formatter.

Add base section to the merge conflict hunks.

Bug: 442284
Change-Id: I977b43e7dd8119d6b72d11f09c4e8ec241750383
This commit is contained in:
Haamed Gheibi 2023-07-24 17:50:34 -07:00
parent 0f4af2bc36
commit 462c57ec8d
5 changed files with 258 additions and 48 deletions

View File

@ -47,7 +47,10 @@ public MergeAlgorithmTest(boolean newlineAtEnd) {
@Test
public void testTwoConflictingModifications() throws IOException {
assertEquals(t("a<b=Z>Zdefghij"),
merge("abcdefghij", "abZdefghij", "aZZdefghij"));
merge("abcdefghij", "abZdefghij", "aZZdefghij", false));
assertEquals(t("a<b|b=Z>Zdefghij"),
merge("abcdefghij", "abZdefghij", "aZZdefghij", true));
}
/**
@ -60,7 +63,10 @@ public void testTwoConflictingModifications() throws IOException {
@Test
public void testOneAgainstTwoConflictingModifications() throws IOException {
assertEquals(t("aZ<Z=c>Zefghij"),
merge("abcdefghij", "aZZZefghij", "aZcZefghij"));
merge("abcdefghij", "aZZZefghij", "aZcZefghij", false));
assertEquals(t("aZ<Z|c=c>Zefghij"),
merge("abcdefghij", "aZZZefghij", "aZcZefghij", true));
}
/**
@ -72,7 +78,10 @@ public void testOneAgainstTwoConflictingModifications() throws IOException {
@Test
public void testNoAgainstOneModification() throws IOException {
assertEquals(t("aZcZefghij"),
merge("abcdefghij", "abcdefghij", "aZcZefghij"));
merge("abcdefghij", "abcdefghij", "aZcZefghij", false));
assertEquals(t("aZcZefghij"),
merge("abcdefghij", "abcdefghij", "aZcZefghij", true));
}
/**
@ -84,7 +93,10 @@ public void testNoAgainstOneModification() throws IOException {
@Test
public void testTwoNonConflictingModifications() throws IOException {
assertEquals(t("YbZdefghij"),
merge("abcdefghij", "abZdefghij", "Ybcdefghij"));
merge("abcdefghij", "abZdefghij", "Ybcdefghij", false));
assertEquals(t("YbZdefghij"),
merge("abcdefghij", "abZdefghij", "Ybcdefghij", true));
}
/**
@ -96,7 +108,10 @@ public void testTwoNonConflictingModifications() throws IOException {
@Test
public void testTwoComplicatedModifications() throws IOException {
assertEquals(t("a<ZZZZfZhZj=bYdYYYYiY>"),
merge("abcdefghij", "aZZZZfZhZj", "abYdYYYYiY"));
merge("abcdefghij", "aZZZZfZhZj", "abYdYYYYiY", false));
assertEquals(t("a<ZZZZfZhZj|bcdefghij=bYdYYYYiY>"),
merge("abcdefghij", "aZZZZfZhZj", "abYdYYYYiY", true));
}
/**
@ -109,7 +124,9 @@ public void testTwoComplicatedModifications() throws IOException {
@Test
public void testTwoModificationsWithSharedDelete() throws IOException {
assertEquals(t("Cb}n}"),
merge("ab}n}n}", "ab}n}", "Cb}n}"));
merge("ab}n}n}", "ab}n}", "Cb}n}", false));
assertEquals(t("Cb}n}"), merge("ab}n}n}", "ab}n}", "Cb}n}", true));
}
/**
@ -122,7 +139,11 @@ public void testTwoModificationsWithSharedDelete() throws IOException {
@Test
public void testModificationsWithMiddleInsert() throws IOException {
assertEquals(t("aBcd123123uvwxPq"),
merge("abcd123uvwxpq", "aBcd123123uvwxPq", "abcd123123uvwxpq"));
merge("abcd123uvwxpq", "aBcd123123uvwxPq", "abcd123123uvwxpq",
false));
assertEquals(t("aBcd123123uvwxPq"), merge("abcd123uvwxpq",
"aBcd123123uvwxPq", "abcd123123uvwxpq", true));
}
/**
@ -135,7 +156,10 @@ public void testModificationsWithMiddleInsert() throws IOException {
@Test
public void testModificationsWithMiddleDelete() throws IOException {
assertEquals(t("Abz}z123Q"),
merge("abz}z}z123q", "Abz}z123Q", "abz}z123q"));
merge("abz}z}z123q", "Abz}z123Q", "abz}z123q", false));
assertEquals(t("Abz}z123Q"),
merge("abz}z}z123q", "Abz}z123Q", "abz}z123q", true));
}
/**
@ -146,7 +170,10 @@ public void testModificationsWithMiddleDelete() throws IOException {
@Test
public void testConflictAtStart() throws IOException {
assertEquals(t("<Z=Y>bcdefghij"),
merge("abcdefghij", "Zbcdefghij", "Ybcdefghij"));
merge("abcdefghij", "Zbcdefghij", "Ybcdefghij", false));
assertEquals(t("<Z|a=Y>bcdefghij"),
merge("abcdefghij", "Zbcdefghij", "Ybcdefghij", true));
}
/**
@ -157,7 +184,10 @@ public void testConflictAtStart() throws IOException {
@Test
public void testConflictAtEnd() throws IOException {
assertEquals(t("abcdefghi<Z=Y>"),
merge("abcdefghij", "abcdefghiZ", "abcdefghiY"));
merge("abcdefghij", "abcdefghiZ", "abcdefghiY", false));
assertEquals(t("abcdefghi<Z|j=Y>"),
merge("abcdefghij", "abcdefghiZ", "abcdefghiY", true));
}
/**
@ -169,7 +199,10 @@ public void testConflictAtEnd() throws IOException {
@Test
public void testSameModification() throws IOException {
assertEquals(t("abZdefghij"),
merge("abcdefghij", "abZdefghij", "abZdefghij"));
merge("abcdefghij", "abZdefghij", "abZdefghij", false));
assertEquals(t("abZdefghij"),
merge("abcdefghij", "abZdefghij", "abZdefghij", true));
}
/**
@ -181,27 +214,36 @@ public void testSameModification() throws IOException {
@Test
public void testDeleteVsModify() throws IOException {
assertEquals(t("ab<=Z>defghij"),
merge("abcdefghij", "abdefghij", "abZdefghij"));
merge("abcdefghij", "abdefghij", "abZdefghij", false));
assertEquals(t("ab<|c=Z>defghij"),
merge("abcdefghij", "abdefghij", "abZdefghij", true));
}
@Test
public void testInsertVsModify() throws IOException {
assertEquals(t("a<bZ=XY>"), merge("ab", "abZ", "aXY"));
assertEquals(t("a<bZ=XY>"), merge("ab", "abZ", "aXY", false));
assertEquals(t("a<bZ|b=XY>"), merge("ab", "abZ", "aXY", true));
}
@Test
public void testAdjacentModifications() throws IOException {
assertEquals(t("a<Zc=bY>d"), merge("abcd", "aZcd", "abYd"));
assertEquals(t("a<Zc=bY>d"), merge("abcd", "aZcd", "abYd", false));
assertEquals(t("a<Zc|bc=bY>d"), merge("abcd", "aZcd", "abYd", true));
}
@Test
public void testSeparateModifications() throws IOException {
assertEquals(t("aZcYe"), merge("abcde", "aZcde", "abcYe"));
assertEquals(t("aZcYe"), merge("abcde", "aZcde", "abcYe", false));
assertEquals(t("aZcYe"), merge("abcde", "aZcde", "abcYe", true));
}
@Test
public void testBlankLines() throws IOException {
assertEquals(t("aZc\nYe"), merge("abc\nde", "aZc\nde", "abc\nYe"));
assertEquals(t("aZc\nYe"),
merge("abc\nde", "aZc\nde", "abc\nYe", false));
assertEquals(t("aZc\nYe"),
merge("abc\nde", "aZc\nde", "abc\nYe", true));
}
/**
@ -214,11 +256,22 @@ public void testBlankLines() throws IOException {
*/
@Test
public void testTwoSimilarModsAndOneInsert() throws IOException {
assertEquals(t("aBcDde"), merge("abcde", "aBcde", "aBcDde"));
assertEquals(t("IAAAJCAB"), merge("iACAB", "IACAB", "IAAAJCAB"));
assertEquals(t("HIAAAJCAB"), merge("HiACAB", "HIACAB", "HIAAAJCAB"));
assertEquals(t("aBcDde"), merge("abcde", "aBcde", "aBcDde", false));
assertEquals(t("aBcDde"), merge("abcde", "aBcde", "aBcDde", true));
assertEquals(t("IAAAJCAB"), merge("iACAB", "IACAB", "IAAAJCAB", false));
assertEquals(t("IAAAJCAB"), merge("iACAB", "IACAB", "IAAAJCAB", true));
assertEquals(t("HIAAAJCAB"),
merge("HiACAB", "HIACAB", "HIAAAJCAB", false));
assertEquals(t("HIAAAJCAB"),
merge("HiACAB", "HIACAB", "HIAAAJCAB", true));
assertEquals(t("AGADEFHIAAAJCAB"),
merge("AGADEFHiACAB", "AGADEFHIACAB", "AGADEFHIAAAJCAB"));
merge("AGADEFHiACAB", "AGADEFHIACAB", "AGADEFHIAAAJCAB",
false));
assertEquals(t("AGADEFHIAAAJCAB"),
merge("AGADEFHiACAB", "AGADEFHIACAB", "AGADEFHIAAAJCAB", true));
}
/**
@ -232,18 +285,28 @@ public void testTwoSimilarModsAndOneInsert() throws IOException {
@Test
public void testTwoSimilarModsAndOneInsertAtEnd() throws IOException {
Assume.assumeTrue(newlineAtEnd);
assertEquals(t("IAAJ"), merge("iA", "IA", "IAAJ"));
assertEquals(t("IAJ"), merge("iA", "IA", "IAJ"));
assertEquals(t("IAAAJ"), merge("iA", "IA", "IAAAJ"));
assertEquals(t("IAAJ"), merge("iA", "IA", "IAAJ", false));
assertEquals(t("IAAJ"), merge("iA", "IA", "IAAJ", true));
assertEquals(t("IAJ"), merge("iA", "IA", "IAJ", false));
assertEquals(t("IAJ"), merge("iA", "IA", "IAJ", true));
assertEquals(t("IAAAJ"), merge("iA", "IA", "IAAAJ", false));
assertEquals(t("IAAAJ"), merge("iA", "IA", "IAAAJ", true));
}
@Test
public void testTwoSimilarModsAndOneInsertAtEndNoNewlineAtEnd()
throws IOException {
Assume.assumeFalse(newlineAtEnd);
assertEquals(t("I<A=AAJ>"), merge("iA", "IA", "IAAJ"));
assertEquals(t("I<A=AJ>"), merge("iA", "IA", "IAJ"));
assertEquals(t("I<A=AAAJ>"), merge("iA", "IA", "IAAAJ"));
assertEquals(t("I<A=AAJ>"), merge("iA", "IA", "IAAJ", false));
assertEquals(t("I<A|A=AAJ>"), merge("iA", "IA", "IAAJ", true));
assertEquals(t("I<A=AJ>"), merge("iA", "IA", "IAJ", false));
assertEquals(t("I<A|A=AJ>"), merge("iA", "IA", "IAJ", true));
assertEquals(t("I<A=AAAJ>"), merge("iA", "IA", "IAAAJ", false));
assertEquals(t("I<A|A=AAAJ>"), merge("iA", "IA", "IAAAJ", true));
}
/**
@ -254,22 +317,34 @@ public void testTwoSimilarModsAndOneInsertAtEndNoNewlineAtEnd()
@Test
public void testEmptyTexts() throws IOException {
// test modification against deletion
assertEquals(t("<AB=>"), merge("A", "AB", ""));
assertEquals(t("<=AB>"), merge("A", "", "AB"));
assertEquals(t("<AB=>"), merge("A", "AB", "", false));
assertEquals(t("<AB|A=>"), merge("A", "AB", "", true));
assertEquals(t("<=AB>"), merge("A", "", "AB", false));
assertEquals(t("<|A=AB>"), merge("A", "", "AB", true));
// test unmodified against deletion
assertEquals(t(""), merge("AB", "AB", ""));
assertEquals(t(""), merge("AB", "", "AB"));
assertEquals(t(""), merge("AB", "AB", "", false));
assertEquals(t(""), merge("AB", "AB", "", true));
assertEquals(t(""), merge("AB", "", "AB", false));
assertEquals(t(""), merge("AB", "", "AB", true));
// test deletion against deletion
assertEquals(t(""), merge("AB", "", ""));
assertEquals(t(""), merge("AB", "", "", false));
assertEquals(t(""), merge("AB", "", "", true));
}
private String merge(String commonBase, String ours, String theirs) throws IOException {
private String merge(String commonBase, String ours, String theirs,
boolean diff3) throws IOException {
MergeResult r = new MergeAlgorithm().merge(RawTextComparator.DEFAULT,
T(commonBase), T(ours), T(theirs));
ByteArrayOutputStream bo=new ByteArrayOutputStream(50);
fmt.formatMerge(bo, r, "B", "O", "T", UTF_8);
if (diff3) {
fmt.formatMergeDiff3(bo, r, "B", "O", "T", UTF_8);
} else {
fmt.formatMerge(bo, r, "B", "O", "T", UTF_8);
}
return new String(bo.toByteArray(), UTF_8);
}
@ -284,6 +359,9 @@ public String t(String text) {
case '=':
r.append("=======\n");
break;
case '|':
r.append("||||||| B\n");
break;
case '>':
r.append(">>>>>>> T\n");
break;

View File

@ -127,6 +127,8 @@ public <S extends Sequence> MergeResult<S> merge(
// Let their complete content conflict with empty text
result.add(1, 0, 0,
ConflictState.FIRST_CONFLICTING_RANGE);
result.add(0, 0, base.size(),
ConflictState.BASE_CONFLICTING_RANGE);
result.add(2, 0, theirs.size(),
ConflictState.NEXT_CONFLICTING_RANGE);
break;
@ -155,6 +157,8 @@ public <S extends Sequence> MergeResult<S> merge(
// Let our complete content conflict with empty text
result.add(1, 0, ours.size(),
ConflictState.FIRST_CONFLICTING_RANGE);
result.add(0, 0, base.size(),
ConflictState.BASE_CONFLICTING_RANGE);
result.add(2, 0, 0, ConflictState.NEXT_CONFLICTING_RANGE);
break;
}
@ -324,6 +328,14 @@ public <S extends Sequence> MergeResult<S> merge(
result.add(1, oursBeginB + commonPrefix,
oursEndB - commonSuffix,
ConflictState.FIRST_CONFLICTING_RANGE);
int baseBegin = Math.min(oursBeginB, theirsBeginB)
+ commonPrefix;
int baseEnd = Math.min(base.size(),
Math.max(oursEndB, theirsEndB)) - commonSuffix;
result.add(0, baseBegin, baseEnd,
ConflictState.BASE_CONFLICTING_RANGE);
result.add(2, theirsBeginB + commonPrefix,
theirsEndB - commonSuffix,
ConflictState.NEXT_CONFLICTING_RANGE);

View File

@ -29,14 +29,22 @@ public enum ConflictState {
NO_CONFLICT,
/**
* This chunk does belong to a conflict and is the first one of the
* This chunk does belong to a conflict and is the ours section of the
* conflicting chunks
*/
FIRST_CONFLICTING_RANGE,
/**
* This chunk does belong to a conflict but is not the first one of the
* conflicting chunks. It's a subsequent one.
* This chunk does belong to a conflict and is the base section of the
* conflicting chunks
*
* @since 6.7
*/
BASE_CONFLICTING_RANGE,
/**
* This chunk does belong to a conflict and is the theirs section of
* the conflicting chunks. It's a subsequent one.
*/
NEXT_CONFLICTING_RANGE
}

View File

@ -82,6 +82,35 @@ public void formatMerge(OutputStream out, MergeResult<RawText> res,
new MergeFormatterPass(out, res, seqName, charset).formatMerge();
}
/**
* Formats the results of a merge of {@link org.eclipse.jgit.diff.RawText}
* objects in a Git conformant way using diff3 style. This method also
* assumes that the {@link org.eclipse.jgit.diff.RawText} objects being
* merged are line oriented files which use LF as delimiter. This method
* will also use LF to separate chunks and conflict metadata, therefore it
* fits only to texts that are LF-separated lines.
*
* @param out
* the output stream where to write the textual presentation
* @param res
* the merge result which should be presented
* @param seqName
* When a conflict is reported each conflicting range will get a
* name. This name is following the "&lt;&lt;&lt;&lt;&lt;&lt;&lt;
* ", "|||||||" or "&gt;&gt;&gt;&gt;&gt;&gt;&gt; " conflict
* markers. The names for the sequences are given in this list
* @param charset
* the character set used when writing conflict metadata
* @throws java.io.IOException
* if an IO error occurred
* @since 6.7
*/
public void formatMergeDiff3(OutputStream out,
MergeResult<RawText> res, List<String> seqName, Charset charset)
throws IOException {
new MergeFormatterPass(out, res, seqName, charset, true).formatMerge();
}
/**
* Formats the results of a merge of exactly two
* {@link org.eclipse.jgit.diff.RawText} objects in a Git conformant way.
@ -150,4 +179,39 @@ public void formatMerge(OutputStream out, MergeResult res, String baseName,
names.add(theirsName);
formatMerge(out, res, names, charset);
}
/**
* Formats the results of a merge of three
* {@link org.eclipse.jgit.diff.RawText} objects in a Git conformant way,
* using diff-3 style. This convenience method accepts the names for the
* three sequences (base and the two merged sequences) as explicit
* parameters and doesn't require the caller to specify a List
*
* @param out
* the {@link java.io.OutputStream} where to write the textual
* presentation
* @param res
* the merge result which should be presented
* @param baseName
* the name ranges from the base should get
* @param oursName
* the name ranges from ours should get
* @param theirsName
* the name ranges from theirs should get
* @param charset
* the character set used when writing conflict metadata
* @throws java.io.IOException
* if an IO error occurred
* @since 6.7
*/
@SuppressWarnings("unchecked")
public void formatMergeDiff3(OutputStream out,
MergeResult res, String baseName, String oursName,
String theirsName, Charset charset) throws IOException {
List<String> names = new ArrayList<>(3);
names.add(baseName);
names.add(oursName);
names.add(theirsName);
formatMergeDiff3(out, res, names, charset);
}
}

View File

@ -31,6 +31,8 @@ class MergeFormatterPass {
private final boolean threeWayMerge;
private final boolean writeBase; // diff3-style requested
private String lastConflictingName; // is set to non-null whenever we are in
// a conflict
@ -50,22 +52,47 @@ class MergeFormatterPass {
*/
MergeFormatterPass(OutputStream out, MergeResult<RawText> res,
List<String> seqName, Charset charset) {
this(out, res, seqName, charset, false);
}
/**
* @param out
* the {@link java.io.OutputStream} where to write the textual
* presentation
* @param res
* the merge result which should be presented
* @param seqName
* When a conflict is reported each conflicting range will get a
* name. This name is following the "&lt;&lt;&lt;&lt;&lt;&lt;&lt;
* ", "|||||||" or "&gt;&gt;&gt;&gt;&gt;&gt;&gt; " conflict
* markers. The names for the sequences are given in this list
* @param charset
* the character set used when writing conflict metadata
* @param writeBase
* base's contribution should be written in conflicts
*/
MergeFormatterPass(OutputStream out, MergeResult<RawText> res,
List<String> seqName, Charset charset, boolean writeBase) {
this.out = new EolAwareOutputStream(out);
this.res = res;
this.seqName = seqName;
this.charset = charset;
this.threeWayMerge = (res.getSequences().size() == 3);
this.writeBase = writeBase;
}
void formatMerge() throws IOException {
boolean missingNewlineAtEnd = false;
for (MergeChunk chunk : res) {
RawText seq = res.getSequences().get(chunk.getSequenceIndex());
writeConflictMetadata(chunk);
// the lines with conflict-metadata are written. Now write the chunk
for (int i = chunk.getBegin(); i < chunk.getEnd(); i++)
writeLine(seq, i);
missingNewlineAtEnd = seq.isMissingNewlineAtEnd();
if (!isBase(chunk) || writeBase) {
RawText seq = res.getSequences().get(chunk.getSequenceIndex());
writeConflictMetadata(chunk);
// the lines with conflict-metadata are written. Now write the
// chunk
for (int i = chunk.getBegin(); i < chunk.getEnd(); i++)
writeLine(seq, i);
missingNewlineAtEnd = seq.isMissingNewlineAtEnd();
}
}
// one possible leftover: if the merge result ended with a conflict we
// have to close the last conflict here
@ -77,16 +104,19 @@ void formatMerge() throws IOException {
private void writeConflictMetadata(MergeChunk chunk) throws IOException {
if (lastConflictingName != null
&& chunk.getConflictState() != ConflictState.NEXT_CONFLICTING_RANGE) {
// found the end of an conflict
&& !isTheirs(chunk) && !isBase(chunk)) {
// found the end of a conflict
writeConflictEnd();
}
if (chunk.getConflictState() == ConflictState.FIRST_CONFLICTING_RANGE) {
// found the start of an conflict
if (isOurs(chunk)) {
// found the start of a conflict
writeConflictStart(chunk);
} else if (chunk.getConflictState() == ConflictState.NEXT_CONFLICTING_RANGE) {
// found another conflicting chunk
} else if (isTheirs(chunk)) {
// found the theirs conflicting chunk
writeConflictChange(chunk);
} else if (isBase(chunk)) {
// found the base conflicting chunk
writeConflictBase(chunk);
}
}
@ -113,6 +143,11 @@ private void writeConflictChange(MergeChunk chunk) throws IOException {
+ lastConflictingName);
}
private void writeConflictBase(MergeChunk chunk) throws IOException {
lastConflictingName = seqName.get(chunk.getSequenceIndex());
writeln("||||||| " + lastConflictingName); //$NON-NLS-1$
}
private void writeln(String s) throws IOException {
out.beginln();
out.write((s + "\n").getBytes(charset)); //$NON-NLS-1$
@ -125,4 +160,17 @@ private void writeLine(RawText seq, int i) throws IOException {
if (out.isBeginln())
out.write('\n');
}
private boolean isBase(MergeChunk chunk) {
return chunk.getConflictState() == ConflictState.BASE_CONFLICTING_RANGE;
}
private boolean isOurs(MergeChunk chunk) {
return chunk
.getConflictState() == ConflictState.FIRST_CONFLICTING_RANGE;
}
private boolean isTheirs(MergeChunk chunk) {
return chunk.getConflictState() == ConflictState.NEXT_CONFLICTING_RANGE;
}
}