Matthias Sohn [Fri, 27 Jul 2018 12:06:14 +0000 (14:06 +0200)]
Merge branch 'stable-4.11' into stable-5.0
* stable-4.11:
Prepare 4.7.3-SNAPSHOT builds
JGit v4.7.2.201807261330-r
Delete all loose refs empty directories
Use java.nio to delete path to get detailed errors
GC: Remove empty references folders
Do not ignore path deletion errors
Change-Id: I780d4a1624fbb97a2acdf5321b9c3244b0626444 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Matthias Sohn [Fri, 27 Jul 2018 11:38:21 +0000 (13:38 +0200)]
Merge branch 'stable-4.10' into stable-4.11
* stable-4.10:
Prepare 4.7.3-SNAPSHOT builds
JGit v4.7.2.201807261330-r
Delete all loose refs empty directories
Use java.nio to delete path to get detailed errors
GC: Remove empty references folders
Do not ignore path deletion errors
Change-Id: I2b44d862869d4453c57db668fc7c925da591f671 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Matthias Sohn [Fri, 27 Jul 2018 09:03:11 +0000 (11:03 +0200)]
Merge branch 'stable-4.9' into stable-4.10
* stable-4.9:
Prepare 4.7.3-SNAPSHOT builds
JGit v4.7.2.201807261330-r
Delete all loose refs empty directories
Use java.nio to delete path to get detailed errors
GC: Remove empty references folders
Do not ignore path deletion errors
Change-Id: Ie7029bc91621af32e7bfd2e0d76a424b991b1995 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Matthias Sohn [Fri, 27 Jul 2018 08:51:58 +0000 (10:51 +0200)]
Merge branch 'stable-4.8' into stable-4.9
* stable-4.8:
Prepare 4.7.3-SNAPSHOT builds
JGit v4.7.2.201807261330-r
Delete all loose refs empty directories
Use java.nio to delete path to get detailed errors
GC: Remove empty references folders
Do not ignore path deletion errors
Change-Id: I6ab2b951dd94a9fc1c4f5283847a3e2ec37d0895 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
David Pursehouse [Fri, 27 Jul 2018 07:22:51 +0000 (08:22 +0100)]
Merge branch 'stable-4.7' into stable-4.8
* stable-4.7:
Prepare 4.7.3-SNAPSHOT builds
JGit v4.7.2.201807261330-r
Delete all loose refs empty directories
Use java.nio to delete path to get detailed errors
GC: Remove empty references folders
Do not ignore path deletion errors
Change-Id: Iadc8275fbaa3d6f7d08a96ab66d49f392f6aab78 Signed-off-by: David Pursehouse <david.pursehouse@gmail.com>
After packaging references, the folders containing these references are
not deleted. In a busy repository, this causes operations to slow down
as traversing the references tree becomes longer.
Delete empty reference folders after the loose references have been
packed.
To avoid deleting a folder that was just created by another concurrent
operation, only delete folders that were not modified in the last 30
seconds.
Marco Miller [Thu, 21 Jun 2018 18:18:48 +0000 (14:18 -0400)]
ResolveMerger: Fix encoding with string; use bytes
This change fixes the issue [1]. Before this fix, a merge involving
the caching of consecutive yet similar filenames with Norwegian
characters [2] used to throw an IllegalStateException: Duplicate
stages not allowed. This was caused by inaccurate decoding of the
filenames, using string values assuming default encoding. In the
toString method of DirCacheEntry, used before through getPathString,
UTF-8 encoding is used, but the end result becomes default encoding,
through Object's default toString usage. The special characters in
those two consecutive (particular) filenames [2] were becoming the
very same decoded /single character, lending consecutive -but then
identical- filenames. Thus the perceived duplicate 0-staging of the
file(s).
Replace getPathString usage with getRawPath for this specific case,
or use byte array representations of cached entries instead of string.
Adding a test for this change is not possible, as there is no known
way to change the default encoding for filenames such as [2] (e.g.).
JGitTestUtil does write file contents through UTF-8, but encoding like
so does not apply to the actual file name. Hence there is no way to
create files with names properly made of special characters such as
[2]'s. And the test that is necessary for this case assumes such
Norwegian (or similar characters) filenames. Changing the default
locale programmatically in a test has no effect either. And changing
the LANG value passed to the JVM is only possible upon starting it.
then I expect to have a full history, just as though I had fetched
without --depth in the first place. Instead, it reports success
but does not fetch enough objects:
The false success indicates problems in the client and the server.
Git 2.18-rc2 (the client) ought to have been more defensive, noticing
the incomplete history. The greater error is in JGit (the server),
which neglects to send the objects requested.
When serving protocol v0 requests, JGit sends the correct objects by
taking unshallowCommits into account when generating the pack to send
to the client. Do the same in the protocol v2 code path. I forgot to
do this in v5.0.0.201806050710-rc3~6 (Teach UploadPack shallow fetch
in protocol v2, 2018-03-15).
Reported-by: Russ Cox <rsc@golang.org>
Change-Id: I282b45f47616a641b9e8d6210b4a070d3efdbb9b Signed-off-by: Jonathan Nieder <jrn@google.com>
Matthias Sohn [Wed, 13 Jun 2018 07:03:20 +0000 (09:03 +0200)]
Downgrade Apache httpclient to 4.5.2.v20170210-0925
Eclipse platform uses this version from the Oxygen Orbit release for
Photon. In order to avoid that we end up with two slightly different
versions in the same p2 repository of the simultaneous release we
downgrade temporarily from the version 4.5.2.v20180410-1551 in the
Photon Orbit release.
See
https://dev.eclipse.org/mhonarc/lists/cross-project-issues-dev/msg15659.html
Change-Id: Id46a840aa4b1010af7fe311498f17f1f2e5b81e0 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Matthias Sohn [Tue, 12 Jun 2018 12:09:39 +0000 (14:09 +0200)]
Propagate failure of ssh command to caller of SshSupport
When SshSupport.runSshCommand fails since the executed external ssh
command failed throw a CommandFailedException.
If discovery of LFS server fails due to failure of the
git-lfs-authenticate command chain the CommandFailureException to the
LfsConfigInvalidException in order to allow root cause analysis in the
application using that.
Change-Id: I2f9ea2be11274549f6d845937164c248b3d840b2 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Marcel Trautwein [Fri, 23 Feb 2018 06:27:52 +0000 (07:27 +0100)]
Make JGit describe behaves same as c-git for lightweight tags
JGit now considers lightweight tags only if the --tags option is set
i.e. `git.describe().setAllTags(true)` has to be set, else the default
is now as in c git:
Only annotated tags are evaluated unless you pass true
equivalent to --tags (or --all) by the option setAllTags.
Hint: This (still) doesn't address any difference between c-git
`--all` and `!--all --tags` behavior;
perhaps this might be a follow up request
Markus Duft [Mon, 11 Jun 2018 15:12:00 +0000 (17:12 +0200)]
Fix issues with LFS on GitHub (SSH)
* URIish seems to have a tiny feature (bug?). The path of the URI
starts with a '/' only if the URI has a port set (it seems).
* GitHub does not return SSH authorization on a single line as Gerrit
does - need to account for that.
* Increase the SSH git-lfs-authenticate timeout, as GitHub sometimes
responds slower than expected.
* Guard against NPE in case the download action does not contain any
additional headers.
Change-Id: Icd1ead3d015479fd4b8bbd42ed42129b0abfb95c Signed-off-by: Markus Duft <markus.duft@ssi-schaefer.com>
Michael Keppler [Sun, 10 Jun 2018 12:15:45 +0000 (14:15 +0200)]
Fix Javadoc typo
Change-Id: Ib4ebc57236bdea663f27295764886413e2550580 Signed-off-by: Michael Keppler <Michael.Keppler@gmx.de> Signed-off-by: David Pursehouse <david.pursehouse@gmail.com>
Thomas Wolf [Fri, 8 Jun 2018 13:47:05 +0000 (15:47 +0200)]
Ensure Jsch checks all configured algorithms
Jsch checks only for the availability of the algorithms given by
Jsch-internal config keys "CheckCiphers", "CheckKexes", and
"CheckSignatures". If the ssh config defines any algorithms
unknown to Jsch not listed in those keys, it'll still propose them
during the negotiation phase, and run into an NPE later on if the
server happens to propose such an algorithm and it gets chosen.
Jsch reads those "CheckCiphers" and the other values from either a
session-local config, or the global static Jsch config. It bypasses
~/.ssh/config for these values.
Therefore, copy these values from the config as read from
~/.ssh/config into the session-specific config. That makes Jsch
check _all_ configured algorithms up front, discarding any for
which it has no implementation. Thus it proposes only algorithms
it actually can handle.
Bug: 535672
Change-Id: I6a68e54f4d9a3267e895c536bcf3c58099826ad5 Signed-off-by: Thomas Wolf <thomas.wolf@paranor.ch>
Dave Borowitz [Fri, 5 Jan 2018 18:02:47 +0000 (13:02 -0500)]
Ensure DirectoryStream is closed promptly
From the javadoc for Files.list:
"The returned stream encapsulates a DirectoryStream. If timely disposal
of file system resources is required, the try-with-resources construct
should be used to ensure that the stream's close method is invoked
after the stream operations are completed."
This is the only call to Files#newDirectoryStream that is not already in
a try-with-resources.
Matthias Sohn [Fri, 8 Jun 2018 15:45:00 +0000 (17:45 +0200)]
Validate branch names on branch creation
Since v2.16.0-rc0~89^2~1 (branch: correctly reject
refs/heads/{-dash,HEAD}, 2017-11-14),
native git does not allow branch names
- refs/heads/HEAD
- starting with '-'
Bug: 535655
Change-Id: Ib1c4ec9ea844073901a4ebe6a29ff6cc8ae58e93 Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Update maven plugins to fix Zip Slip vulnerability
Zip Slip [1] is an arbitrary file write generic vulnerability, that can
be achieved using a specially crafted zip (or bzip2, gzip, tar, xz, war)
archive, that holds path traversal filenames.
According to Maven's announcement [2] several plugins use plexus-archiver to
unpack dependencies to disk and have been identified as potential triggers
for exposing the vulnerability.
Of those, JGit uses the maven-dependency-plugin and the maven-javadoc-plugin.
Update them to the fixed versions reported in [2].
See the corresponding issues for the maven-dependency-plugin [3] and the
maven-javadoc-plugin [4] for details.
Jonathan Tan [Mon, 30 Apr 2018 20:21:43 +0000 (13:21 -0700)]
Teach UploadPack "filter" in protocol v2 fetch
If the configuration variable uploadpack.allowfilter is true, advertise
that "filter" is supported, and support it if the client sends such an
argument.
Change-Id: I7de66c0a0ada46ff71c5ba124d4ffa7c47254c3b Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Jonathan Tan [Wed, 2 May 2018 23:35:48 +0000 (16:35 -0700)]
Refactor test of capabilities output
A subsequent patch will dynamically generate the capability
advertisement, so the capability advertisements produced are not always
the same. Separate the checking of the advertisements into its own test
method.
Change-Id: I768d14b9d1a244d5d886c42ffd62ef3957b518fb Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Jonathan Tan [Wed, 2 May 2018 22:43:50 +0000 (15:43 -0700)]
Refactor v2 advertisement into own function
A subsequent patch needs dynamic generation of this advertisement
depending on a configuration variable in the underlying repository, so
refactor it into a function instead of using a constant list.
Change-Id: Ie00584add1fb56c9e88c7b57f75703981ea5bb85 Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Jonathan Tan [Thu, 3 May 2018 00:17:04 +0000 (17:17 -0700)]
Disallow unknown args to "fetch" in protocol v2
JGit's implementation of the fetch command of protocol v2, unlike its
implementation of ls-refs, currently tolerates unknown arguments.
Tighten fetch to not allow unrecognized arguments and add tests to
verify this behavior for both ls-refs and fetch.
Change-Id: I321161d568bd638252fab1a47b06b924d472a669 Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Jonathan Tan [Thu, 15 Mar 2018 22:56:50 +0000 (15:56 -0700)]
Teach UploadPack shallow fetch in protocol v2
Add support for the "shallow" and "deepen" parameters in the "fetch"
command in the fetch-pack/upload-pack protocol v2. Advertise support for
this in the capability advertisement.
Change-Id: I7ffd80d6c38872f9d713ac7d6e0412106b3766d7 Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrn@google.com>
Jonathan Tan [Tue, 22 May 2018 22:19:04 +0000 (15:19 -0700)]
Refactor unshallowCommits to local variable
This reduces the amount of state held as instance variables in
UploadPack, and makes it easier for a future patch to contain a clearer
version of UploadPack#processShallow.
Change-Id: I6df80b42f9e5118fda1420692e02e417670cced3 Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrn@google.com>
Jonathan Tan [Wed, 28 Feb 2018 22:36:44 +0000 (14:36 -0800)]
Add protocol v2 support in http
Teach UploadPack to support protocol v2 with non-bidirectional pipes,
and add support to the HTTP protocol for v2. This is only activated if
the repository's config has "protocol.version" equal to 2.
Change-Id: I093a14acd2c3850b8b98e14936a716958f35a848 Helped-by: Matthias Sohn <matthias.sohn@sap.com> Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrn@google.com>
Jonathan Tan [Wed, 25 Apr 2018 20:59:52 +0000 (13:59 -0700)]
Give info/refs services more control over response
Currently, SmartServiceInfoRefs always prints "# service=serviceName"
followed by a flush packet in response to an info/refs request, and then
hands it off to the specific service class. Printing of "#
service=serviceName" is mandated for protocol v0, but not v2.
Therefore, the existing code works for protocol v0, but whenever a
service that supports protocol v2 receives an info/refs request, it must
first determine which protocol version is to be used (depending on, for
example, the request and any relevant configuration variables), and then
decide if "# service=serviceName" needs to be printed.
Create a new method that v2-supporting service classes can override,
covering the printing of both "# service=serviceName" and everything
that the #advertise method prints. This will be used in a subsequent
commit in which UploadPackServlet (and the other classes it uses) is
updated to support protocol v2.
Change-Id: Ia026b06e96a6b15937514096babd024ef77df1ea Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrn@google.com>
Thomas Wolf [Thu, 24 May 2018 14:26:36 +0000 (16:26 +0200)]
Don't prune symbolic refs when fetch.prune = true
The canonical implementation also doesn't. Compare current
code in remote.c, function get_stale_heads_cb.[1] Not handling
symrefs in this case was introduced in canonical git in [2]
in 2008.
Matthias Sohn [Tue, 22 May 2018 21:00:37 +0000 (23:00 +0200)]
Merge branch 'master' into stable-5.0
* master:
DescribeCommand: Refactor to not use deprecated Repository#peel
Repository: Deprecate #peel method
Repository: Make #exactRef and #findRef final
Skip ignored directories in FileTreeIterator
Repository: Deprecate getTags method
InfoRefsServlet: Refactor to not use deprecated methods
RefAdvertiser: Add send(Collection<Ref>) and deprecate send(Map<String, Ref>)
Remove deprecated Repository#notifyIndexChanged
Implementors should override Repository#notifyIndexChanged(boolean)
Revive Repository#notifyIndexChanged()
Remove further unnecessary 'final' keywords
Execute AdvertiseRefsHook only for protocol v0 and v1
Add protocol v2 support in "jgit daemon"
Teach UploadPack "ofs-delta" in "fetch"
Teach UploadPack "include-tag" in "fetch"
Avoid using #refs in UploadPack#sendPack
FileRepository: Don't use deprecated RefDatabase#getRefs(String)
BatchRefUpdate: Don't use deprecated RefDatabase#getRefs(String)
Change-Id: I16c5da62d09262d3f4070aa0f466dd6c8352b5ea Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
David Pursehouse [Tue, 22 May 2018 02:27:30 +0000 (11:27 +0900)]
Repository: Deprecate #peel method
Callers should use getRefDatabase().peel(ref) instead since it
doesn't swallow the IOException.
Adapt all trivial callers to user the alternative.
DescribeCommand still uses the deprecated method and is not adapted in
this change since it will require more refactoring to add handling of
the IOException.
Change-Id: I14d4a95a5e0570548753b9fc5c03d024dc3ff832 Signed-off-by: David Pursehouse <david.pursehouse@gmail.com>
David Pursehouse [Tue, 22 May 2018 00:47:16 +0000 (09:47 +0900)]
Repository: Make #exactRef and #findRef final
This means less cognitive overhead for both implementors and callers,
since this way we can guarantee that they are always synonyms for
RefDatabase#exactRef and RefDatabase#findRef, respectively.
Change-Id: Ic8aeb52fc7ed65672f3f6cd1da39a66908d88baa Signed-off-by: David Pursehouse <david.pursehouse@gmail.com>
Thomas Wolf [Tue, 27 Mar 2018 20:22:09 +0000 (22:22 +0200)]
Skip ignored directories in FileTreeIterator
Make FileTreeIterator not enter ignored directories by default. We
only need to enter ignored directories if we do some operation against
git, and there is at least one tracked file underneath an ignored
directory.
Walking ignored directories should be avoided as much as possible as
it is a potential performance bottleneck. Some projects have a lot of
files or very deep hierarchies in ignored directories; walking those
may be costly (especially so on Windows). See for instance also bug
500106.
Provide a FileTreeIterator.setWalkIgnoredDirectories() operation to
force the iterator to iterate also through otherwise ignored
directories. Useful for tests (IgnoreNodeTest, CGitIgnoreTest), or
to implement things like "git ls-files --ignored".
Add tests in DirCacheCheckoutTest, and amend IndexDiffTest to test a
little bit more.
Bug: 388582
Change-Id: I6ff584a42c55a07120a4369fd308409431bdb94a Signed-off-by: Thomas Wolf <thomas.wolf@paranor.ch>
Jonathan Nieder [Fri, 18 May 2018 16:16:16 +0000 (09:16 -0700)]
Implementors should override Repository#notifyIndexChanged(boolean)
Declare Repository#notifyIndexChanged() final and
Repository#notifyIndexChanged(boolean) abstract to force implementors
to switch to overriding the latter method. This makes Repository less
error-prone to extend since implementors no longer need to remember to
override one of those two methods.
Change-Id: I721db0f4a4865db3b35212ee0a2045d5b31c96af Signed-off-by: Jonathan Nieder <jrn@google.com>
Jonathan Nieder [Fri, 18 May 2018 16:15:30 +0000 (09:15 -0700)]
Revive Repository#notifyIndexChanged()
e9e150fdd24d (Store in IndexChangedEvent if it was caused by JGit
itself, 2018-05-13) modified Repository#notifyIndexChanged to take a
boolean argument to indicate whether the index change happened under
the current process's control or externally, for use by EGit. In
other words, the function signature changed from
public abstract void notifyIndexChanged();
to
public abstract void notifyIndexChanged(boolean internal);
Callers outside JGit itself notifying a Repository about index changes
are expected to be rare, so this is not very disruptive to them. In
most cases they would be notifying about changes that they made
themselves, so treating their notifyIndexChanged() calls as
notifyIndexChanged(true) should be relatively safe.
Implementors have the opposite problem: adding the new "abstract void
notifyIndexChanged(boolean)" method means they are obligated to
override it. Add a default implementation that calls their existing
override of notifyIndexChanged() to make their migration easier.
The main downside is that authors of new Repository subclasses that
do not realize they need to override notifyIndexChanged would end up
with a default implementation which calls notifyIndexChanged(true),
in turn calling notifyIndexChanged() again and so on, resulting in
StackOverflowException. Add an implementors' note to the class
Javadoc to avoid this issue. A followup commit will force
implementors to adapt to the new API by changing the methods to
@Deprecated
public final void notifyIndexChanged() {
notifyIndexChanged(true);
}
public abstract void notifyIndexChanged(boolean internal);
Change-Id: I7d014890ee19abf283ea824d9baa9044bfdde130 Signed-off-by: Jonathan Nieder <jrn@google.com>
Masaya Suzuki [Mon, 14 May 2018 21:13:11 +0000 (14:13 -0700)]
Execute AdvertiseRefsHook only for protocol v0 and v1
Refs are not advertised as part of the protocol v2 capability
advertisement. Don't call AdvertiseRefsHook.
Noticed because many implementations of AdvertiseRefsHook read all
refs in order to call UploadPack#setAdvertisedRefs, causing the
capability advertisement to be as slow as a v0 ref advertisement with
some RefDatabase implementations.
Such an AdvertiseRefsHook is of dubious utility (a better place to
determine which refs are advertised is in the RefDatabase
implementation itself, as in Gerrit), but at any rate since it's not
bringing about any benefit here, we can skip the hook call.
Jonathan Tan [Sat, 24 Feb 2018 00:55:03 +0000 (16:55 -0800)]
Add protocol v2 support in "jgit daemon"
With this patch, a server spawned by "jgit daemon" can be accessed using
protocol v2 from a Git client that supports it (for example, "git" with
the appropriate patches). This is only activated if the repository's
config has "protocol.version" be 2.
This required a change to the package-private #execute methods in
DaemonService to allow passing of extra parameters.
This has been tested with a patched Git.
Change-Id: Icf043efec7ce956d72b075fc6dc7a87d5a2da82a Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Jonathan Tan [Tue, 13 Mar 2018 18:07:36 +0000 (11:07 -0700)]
Teach UploadPack "include-tag" in "fetch"
Add support for the "include-tag" parameter in the "fetch" command in
the fetch-pack/upload-pack protocol v2.
In order to determine which tags to include, only objects pointed to by
refs starting with "refs/tags/" are checked. This restriction is for
performance reasons and to match the behavior of Git (see add_ref_tag()
in builtin/pack-objects.c).
Change-Id: I7d70aa09bcc8a525218ff1559e286c2a610258ca Signed-off-by: Jonathan Tan <jonathantanmy@google.com> Signed-off-by: Jonathan Nieder <jrn@google.com>
projects / jgit.git / log