Parse signature of GPG-signed commits

In order to support GPG-signed commits, add some methods which will
allow GPG signatures to be parsed out of RevCommit objects.

Later, we can add code to verify the signatures.

Change-Id: Ifcf6b3ac79115c15d3ec4b4eaed07315534d09ac
Signed-off-by: David Turner <dturner@twosigma.com>
Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>

org.eclipse.jgit.test/tst/org/eclipse/jgit/revwalk/RevCommitParseTest.java

@@ -486,4 +486,36 @@ public class RevCommitParseTest extends RepositoryTestCase {
 	private static ObjectId id(String str) {
 		return ObjectId.fromString(str);
 	}
+
+	@Test
+	public void testParse_gpgSig() throws Exception {
+		String commit = "tree e3a1035abd2b319bb01e57d69b0ba6cab289297e\n" +
+		"parent 54e895b87c0768d2317a2b17062e3ad9f76a8105\n" +
+		"committer A U Thor <author@xample.com 1528968566 +0200\n" +
+		"gpgsig -----BEGIN PGP SIGNATURE-----\n" +
+		" \n" +
+		" wsBcBAABCAAQBQJbGB4pCRBK7hj4Ov3rIwAAdHIIAENrvz23867ZgqrmyPemBEZP\n" +
+		" U24B1Tlq/DWvce2buaxmbNQngKZ0pv2s8VMc11916WfTIC9EKvioatmpjduWvhqj\n" +
+		" znQTFyiMor30pyYsfrqFuQZvqBW01o8GEWqLg8zjf9Rf0R3LlOEw86aT8CdHRlm6\n" +
+		" wlb22xb8qoX4RB+LYfz7MhK5F+yLOPXZdJnAVbuyoMGRnDpwdzjL5Hj671+XJxN5\n" +
+		" SasRdhxkkfw/ZnHxaKEc4juMz8Nziz27elRwhOQqlTYoXNJnsV//wy5Losd7aKi1\n" +
+		" xXXyUpndEOmT0CIcKHrN/kbYoVL28OJaxoBuva3WYQaRrzEe3X02NMxZe9gkSqA=\n" +
+		" =TClh\n" +
+		" -----END PGP SIGNATURE-----\n" +
+		"some other header\n\n" +
+		"commit message";
+
+		final RevCommit c;
+		c = new RevCommit(id("9473095c4cb2f12aefe1db8a355fe3fafba42f67"));
+		c.parseCanonical(new RevWalk(db), commit.getBytes(UTF_8));
+		String gpgSig = new String(c.getRawGpgSignature(), UTF_8);
+		assertTrue(gpgSig.startsWith("-----BEGIN"));
+		assertTrue(gpgSig.endsWith("END PGP SIGNATURE-----"));
+	}
+
+	@Test
+	public void testParse_NoGpgSig() throws Exception {
+		final RevCommit c = create("a message");
+		assertNull(c.getRawGpgSignature());
+	}
 }

org.eclipse.jgit.test/tst/org/eclipse/jgit/util/RawParseUtilsTest.java

@@ -52,7 +52,24 @@ import java.nio.charset.UnsupportedCharsetException;
 import org.eclipse.jgit.lib.Constants;
 import org.junit.Test;
 
+import static java.nio.charset.StandardCharsets.UTF_8;
+
 public class RawParseUtilsTest {
+	String commit = "tree e3a1035abd2b319bb01e57d69b0ba6cab289297e\n" +
+		"parent 54e895b87c0768d2317a2b17062e3ad9f76a8105\n" +
+		"committer A U Thor <author@xample.com 1528968566 +0200\n" +
+		"gpgsig -----BEGIN PGP SIGNATURE-----\n" +
+		" \n" +
+		" wsBcBAABCAAQBQJbGB4pCRBK7hj4Ov3rIwAAdHIIAENrvz23867ZgqrmyPemBEZP\n" +
+		" U24B1Tlq/DWvce2buaxmbNQngKZ0pv2s8VMc11916WfTIC9EKvioatmpjduWvhqj\n" +
+		" znQTFyiMor30pyYsfrqFuQZvqBW01o8GEWqLg8zjf9Rf0R3LlOEw86aT8CdHRlm6\n" +
+		" wlb22xb8qoX4RB+LYfz7MhK5F+yLOPXZdJnAVbuyoMGRnDpwdzjL5Hj671+XJxN5\n" +
+		" SasRdhxkkfw/ZnHxaKEc4juMz8Nziz27elRwhOQqlTYoXNJnsV//wy5Losd7aKi1\n" +
+		" xXXyUpndEOmT0CIcKHrN/kbYoVL28OJaxoBuva3WYQaRrzEe3X02NMxZe9gkSqA=\n" +
+		" =TClh\n" +
+		" -----END PGP SIGNATURE-----\n" +
+		"some other header\n\n" +
+		"commit message";
 
 	@Test
 	public void testParseEncoding_ISO8859_1_encoding() {
@@ -79,4 +96,30 @@ public class RawParseUtilsTest {
 		}
 	}
 
+	@Test
+	public void testHeaderStart() {
+		byte[] headerName = "some".getBytes(UTF_8);
+		byte[] commitBytes = commit.getBytes(UTF_8);
+		assertEquals(625, RawParseUtils.headerStart(headerName, commitBytes, 0));
+		assertEquals(625, RawParseUtils.headerStart(headerName, commitBytes, 4));
+
+		byte[] missingHeaderName = "missing".getBytes(UTF_8);
+		assertEquals(-1, RawParseUtils.headerStart(missingHeaderName,
+							   commitBytes, 0));
+
+		byte[] fauxHeaderName = "other".getBytes(UTF_8);
+		assertEquals(-1, RawParseUtils.headerStart(fauxHeaderName, commitBytes, 625 + 4));
+	}
+
+	@Test
+	public void testHeaderEnd() {
+		byte[] commitBytes = commit.getBytes(UTF_8);
+		int[] expected = new int[] {45, 93, 148, 619, 637};
+		int start = 0;
+		for (int i = 0; i < expected.length; i++) {
+			start = RawParseUtils.headerEnd(commitBytes, start);
+			assertEquals(expected[i], start);
+			start += 1;
+		}
+	}
 }

org.eclipse.jgit/src/org/eclipse/jgit/revwalk/RevCommit.java

@@ -51,6 +51,7 @@ import java.nio.charset.Charset;
 import java.nio.charset.IllegalCharsetNameException;
 import java.nio.charset.UnsupportedCharsetException;
 import java.util.ArrayList;
+import java.util.Arrays;
 import java.util.Collections;
 import java.util.List;
 
@@ -389,6 +390,34 @@ public class RevCommit extends RevObject {
 		return buffer;
 	}
 
+	/**
+	 * Parse the gpg signature from the raw buffer.
+	 * <p>
+	 * This method parses and returns the raw content of the gpgsig lines. This
+	 * method is fairly expensive and produces a new byte[] instance on each
+	 * invocation. Callers should invoke this method only if they are certain
+	 * they will need, and should cache the return value for as long as
+	 * necessary to use all information from it.
+	 * <p>
+	 * RevFilter implementations should try to use
+	 * {@link org.eclipse.jgit.util.RawParseUtils} to scan the
+	 * {@link #getRawBuffer()} instead, as this will allow faster evaluation of
+	 * commits.
+	 *
+	 * @return contents of the gpg signature; null if the commit was not signed.
+	 * @since 5.1
+	 */
+	public final @Nullable byte[] getRawGpgSignature() {
+		final byte[] raw = buffer;
+		final byte[] header = {'g', 'p', 'g', 's', 'i', 'g'};
+		final int start = RawParseUtils.headerStart(header, raw, 0);
+		if (start < 0) {
+			return null;
+		}
+		final int end = RawParseUtils.headerEnd(raw, start);
+		return Arrays.copyOfRange(raw, start, end);
+	}
+
 	/**
 	 * Parse the author identity from the raw buffer.
 	 * <p>

org.eclipse.jgit/src/org/eclipse/jgit/util/RawParseUtils.java

@@ -548,6 +548,62 @@ public final class RawParseUtils {
 		return ptr;
 	}
 
+	/**
+	 * Locate the end of the header.  Note that headers may be
+	 * more than one line long.
+	 * @param b
+	 *            buffer to scan.
+	 * @param ptr
+	 *            position within buffer to start looking for the end-of-header.
+	 * @return new position just after the header.  This is either
+	 * b.length, or the index of the header's terminating newline.
+	 * @since 5.1
+	 */
+	public static final int headerEnd(final byte[] b, int ptr) {
+		final int sz = b.length;
+		while (ptr < sz) {
+			final byte c = b[ptr++];
+			if (c == '\n' && (ptr == sz || b[ptr] != ' ')) {
+				return ptr - 1;
+			}
+		}
+		return ptr - 1;
+	}
+
+	/**
+	 * Find the start of the contents of a given header.
+	 *
+	 * @param b
+	 *            buffer to scan.
+	 * @param headerName
+	 *            header to search for
+	 * @param ptr
+	 *            position within buffer to start looking for header at.
+	 * @return new position at the start of the header's contents, -1 for
+	 *         not found
+	 * @since 5.1
+	 */
+	public static final int headerStart(byte[] headerName, byte[] b, int ptr) {
+		// Start by advancing to just past a LF or buffer start
+		if (ptr != 0) {
+			ptr = nextLF(b, ptr - 1);
+		}
+		while (ptr < b.length - (headerName.length + 1)) {
+			boolean found = true;
+			for (int i = 0; i < headerName.length; i++) {
+				if (headerName[i] != b[ptr++]) {
+					found = false;
+					break;
+				}
+			}
+			if (found && b[ptr++] == ' ') {
+				return ptr;
+			}
+			ptr = nextLF(b, ptr);
+		}
+		return -1;
+	}
+
 	/**
 	 * Locate the first position before a given character.
 	 *