Adds the file content merge alorithm and tests for merge to jgit.
The merge algorithm:

- Gets as input parameters the common base, the two new contents
  called "ours" and "theirs".

- Computes the Edits from base to ours and from base to theirs with
  the help of MyersDiff.

- Iterates over the edits.

- Independent edits from ours or from theirs will just be applied
  to the result.

- For conflicting edits we first harmonize the ranges of the edits
  so that in the end we have exactly two edits starting and ending
  at the same points in the common base. Then we write the two
  conclicting contents into the result stream.

Change-Id: I411862393e7bf416b6f33ca55ec5af608ff4663
Signed-off-by: Christian Halstrick <christian.halstrick@sap.com>
[sp: Fixed up two awkard comments in documentation.]
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Add some tests which make sure that the diff algorithm really behaves in the
promised O(N*D) manner. This tests compute diffs between multiple big chunks
of data, measure time for computing the diffs and fail if the measured times
are off O(N*D) by more than a factor 10

Signed-off-by: Christian Halstrick <christian.halstrick@sap.com>
Change-Id: I8e1e0be60299472828718371b231f1d8a9dc21a7
Signed-off-by: Robin Rosenberg <robin.rosenberg@dewire.com>

This change only fixes warnings of the eclipse build regarding missing javadocs.
Some comments where just missing, so they have been added. Other comments where
at the wrong (from eclipse point of view) place, so eclipse was complaining.
Also two method which existed for debugging purposes have been removed to get
rid of Eclipse warngins about unused code.

* changes:
  Fixed MyersDiff to be able to handle more than 100k

MyersDiff was crashing with ArrayIndexOutOfBoundsException when
diffing huge files. This was because the snake data (begin and
end position while processing chunks of the text) was sequeezed
into a single int. But with longer texts to diff these values
exceeded the range of java int values. It is fixed by squeezing
the two ints into a long and by adding the LongList helper
class.

Change-Id: Iacb9082e1b076e994d1486aa8e512342ad7432b3
Signed-off-by: Christian Halstrick <christian.halstrick@sap.com>

My Galileo eclipse was complaining about unneeded import statements and
how static methods have been accessed in Diff.java. There was also one method
call which could be removed because he had no sideeffects and the return value
was not used. I fixed this so that there are no warnings anymore.

Signed-off-by: Christian Halstrick <christian.halstrick@sap.com>
Change-Id: I48d7e01536aab1524140d72af574e3fd7149cd23

This commit contains fixes provided by Christian Halstrick.

Bug: 291083
Eclipse-CQ: 3559
Change-Id: If8d187e70a51f31040f99098489e513791dc6415
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Bug: 291083
Eclipse-CQ: 3559
Change-Id: Ia02f346a96b5f1e24f8bc9676bd428b968a41222
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Bug: 291083
Eclipse-CQ: 3559
Change-Id: I5a10946637438052e7596489b9f9de3a0c6b2066
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Myers' algorithm is the standard way to generate diff scripts in
an efficient manner (especially memory-wise).

The source contains extensive documentation about the principal
ideas of the algorithm.

Bug: 291083
Eclipse-CQ: 3559
Change-Id: Id873e7c2c204692cb4a1a05d24b733154b9ed49c
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Some applications may wish to modify an int list.

Bug: 291083
Eclipse-CQ: 3559
Change-Id: Iea871443ec661230aec92397229f1eda6c74216f
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

* changes:
  Move HttpSupport's configureHttpProxy to jgit-pgm
  Move AmazonS3 command line utility to jgit-pgm
  Move setupReflog test function to only test that cares
  Remove useless test005_todopack test
  Standardize the source code formatter for Eclipse

Our project plan calls for us to build 0.6 as the next version,
but I forgot to also update the MANIFEST.MF for the plugin when
I edited the Maven pom.

Change-Id: Ic1a6c64374a4384a65a3dd0306adddfc73adac52
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

[sp: Externalized all strings for translation]

Change-Id: Ib3f6e58e9e1cbcbc41705b6e1be37a89ca38f1a9
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

This is the last chunk of code in jgit-core which references the awtui
package.  Moving it to the only consumer in jgit-pgm allows us to move
the awtui package over to the jgit-awtui module.

Change-Id: I2fd81be2076117b2f2c5f8ed45de7f29272af6cf
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

This removes one of the few remaining dependencies on AWTAuthenticator
from the core library.  For the most part the interface is identical to
the prior main method.  The jgit-pgm Main class already sets up the
HTTP proxy and authenticator for us, so we don't need to do that in
our new run method.

Change-Id: Ia2cce34d34c97b88214a8bd8f2cc542845e19032
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Only one test class actually needs this function, so instead of
us inheriting it down into every test, move it to that one class.

Change-Id: I5700ca48df4177153f2b3861dec7c538c621e775
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

This test doesn't work because it requires a pack file which we have
lost to the ages.  We couldn't include it because the pack was actually
a copy of the GPL'd C git.git project, and was there to test some sort
of corner case that the test never documented properly.

Change-Id: I282ee1c6a637a8654df93a3847507a6c60e4cfab
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

We now supply an exported format description for anyone to import
into their own workbench, and all projects reference this style in
a consistent way.

Change-Id: Ic243544a761ef2db29025a89ba6bb932a3a3ce34
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

* changes:
  Mark the next version as 0.6

Our project plan calls for us to build 0.6 as the next version.

Change-Id: I11dc620009d83e5715f7f2c17bbc658cc6e49f20
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

* changes:
  Add support for logAllRefUpdates configuration parameter

Honor the configuration parameter core.logAllRefUpdates when writing
reflogs.  Instead of writing reflog entries always only write
reflogs if this parameter is set to true or if the corresponding
file in the <git-dir>/logs directory already exists. In other words:
if you are updating a ref and this parameter is set to false and
there is no file corresponding to your ref in the <git-dir>/logs
folder then no reflog will be written.

This is a fix for the issue http://code.google.com/p/egit/issues/detail?id=4



Change-Id: I908e4c77e3630dc3223b2d2a47cb4534dbe4ed42
Signed-off-by: Matthias Sohn <matthias.sohn@sap.com>
Signed-off-by: Christian Halstrick <christian.halstrick@sap.com>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

There is no longer any value in keeping track of things we need to
implement in a file in the top level of the source code.  Time has
proven we don't keep it current, as some of these features are
already implemented, and some are still being worked on.

Worse, many of these items were for the EGit plugin, which has no
relationship to JGit.

Change-Id: Iaa83491819518a1293810aafb566354bdc4b5e28
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

I somehow missed that SUBMITTING_PATCHES referenced the older "send
to git mailing list by email" patch process we used for the first
few years of the JGit project's life.  Now that we are hosted under
the Eclipse Foundation umbrella we need to follow their IP process
for all contributions.

Change-Id: Ia29b31746826ca5b914e0a1d108c2166ff64d268
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Change-Id: I1cf5b4355a147c39107c8b97e56108f482c96d8d
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

The process is likely to change as we experiment with Gerrit Code
Review and learn the Eclipse IP process.  Lets point the potential
contributor at a wiki that we can more easily keep up-to-date.

Change-Id: I99f7cb9edadd0f23024f59ec139dd1e35f5fbd3f
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Recently Eclipse started complaining about undocumented protected
members here, so lets document them to make the errors go away.

Change-Id: Ia3bc9fa69854953db5dc9a982821a2f3abe8d05a
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

This allows SwingGraphPanes to be reused by simply clearing and
re-filling. Requires RevObjectList initialization to not call clear()
from its constructor, because this will lead PlotCommitList.clear()
to be called before all variables have been initialized.

Change-Id: I14a07124441b58cd88c67da088ba52ef9c30b043
Signed-off-by: Jonas Fonseca <fonseca@diku.dk>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

When throwing MissingBundlePrerequisiteException we
also include the short description, if available, of
each missing object.

This is the fix for the following issue:
http://code.google.com/p/egit/issues/detail?id=25



Change-Id: I5d45aec7873af76a12170d9a500626a7264f2c42
Signed-off-by: Sasa Zivkov <sasa.zivkov@sap.com>
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

- We shouldn't be re-exporting bundles, JSch in this case.
  Instead used a Import-Package clause.

- Some packages weren't exported.  In Eclipse we have a policy to
  export all the packages in a bundle

- Exporting version numbers on the export packages.

Bug: 291108
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

This script can be executed by a developer to download and graft
on the old JGit history, from before we moved the project to the
eclipse.org namespace and the Eclipse Foundation servers.

Executing this script is only necessary if you need to run log or
blame past the migration boundary, and isn't always recommended when
it comes to pushing objects to a remote server.  As mentioned in the
script, it is best to use a specialized repository with this graft,
not your main work repository.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>

Per CQ 3448 this is the initial contribution of the JGit project
to eclipse.org.  It is derived from the historical JGit repository
at commit 3a2dd9921c8a08740a9e02c421469e5b1a9e47cb.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>