jelmer/python-fastimport.git
12 years agoImplement 'done' feature.
Jelmer Vernooij [Tue, 28 Feb 2012 11:49:37 +0000 (12:49 +0100)]
Implement 'done' feature.

12 years agoMerge support for squash_empty_commit argument to FilterProcessor.
Jelmer Vernooij [Sat, 22 Oct 2011 21:47:22 +0000 (14:47 -0700)]
Merge support for squash_empty_commit argument to FilterProcessor.

12 years agoOnly fix references to commits that have been squashed rather than removing everythin...
Jelmer Vernooij [Sat, 22 Oct 2011 21:39:03 +0000 (14:39 -0700)]
Only fix references to commits that have been squashed rather than removing everything that we haven't touched yet

12 years agoRename flag to squash_empty_commits and add few more tests with includes/excludes
Oleksandr Usov [Fri, 21 Oct 2011 09:35:03 +0000 (10:35 +0100)]
Rename flag to squash_empty_commits and add few more tests with includes/excludes

12 years agoAdd preserve_all_history flag to filter_processor and tests for it
Oleksandr Usov [Thu, 20 Oct 2011 18:14:27 +0000 (19:14 +0100)]
Add preserve_all_history flag to filter_processor and tests for it

12 years agoRemove old FSF address and point at FSF website instead.
Jelmer Vernooij [Thu, 6 Oct 2011 00:24:50 +0000 (02:24 +0200)]
Remove old FSF address and point at FSF website instead.

Thanks Dan Callaghan.

13 years agoStart on 0.9.1.
Jelmer Vernooij [Mon, 31 Jan 2011 00:44:49 +0000 (01:44 +0100)]
Start on 0.9.1.

13 years agorelease 0.9.0
Jelmer Vernooij [Mon, 31 Jan 2011 00:40:02 +0000 (01:40 +0100)]
release 0.9.0

13 years agoFix typo in setup.py
Jonathan Nieder [Sun, 30 Jan 2011 09:41:28 +0000 (10:41 +0100)]
Fix typo in setup.py

13 years agoAvoid attempting to utf-8 decode/encode committer and author information.
Jelmer Vernooij [Sun, 12 Dec 2010 03:52:43 +0000 (04:52 +0100)]
Avoid attempting to utf-8 decode/encode committer and author information.

13 years agoCheck for leading slashes in paths, invalid path types.
Jelmer Vernooij [Sun, 12 Dec 2010 03:41:09 +0000 (04:41 +0100)]
Check for leading slashes in paths, invalid path types.

13 years agoStop utf-8 decoding several strings as they are not guaranteed to be unicode.
Jelmer Vernooij [Sun, 12 Dec 2010 03:39:24 +0000 (04:39 +0100)]
Stop utf-8 decoding several strings as they are not guaranteed to be unicode.

13 years agoIf there is no mark then don't track it.
Jelmer Vernooij [Sat, 11 Dec 2010 19:11:53 +0000 (20:11 +0100)]
If there is no mark then don't track it.

13 years agoAvoid file_iter, use iter_files instead.
Jelmer Vernooij [Mon, 8 Nov 2010 09:59:33 +0000 (10:59 +0100)]
Avoid file_iter, use iter_files instead.

file_iter should ideally be made private.

13 years agoFix tests.
Jelmer Vernooij [Sat, 6 Nov 2010 17:06:28 +0000 (18:06 +0100)]
Fix tests.

13 years agoSupport out of band processing of commands.
Jelmer Vernooij [Thu, 28 Oct 2010 18:22:16 +0000 (11:22 -0700)]
Support out of band processing of commands.

13 years agoCope with FileModifyCommand now having mode rather than is_executable and kind.
Jelmer Vernooij [Sat, 16 Oct 2010 19:00:07 +0000 (21:00 +0200)]
Cope with FileModifyCommand now having mode rather than is_executable and kind.

13 years agoInstall fastimport.processors.
Jelmer Vernooij [Sat, 16 Oct 2010 18:53:47 +0000 (20:53 +0200)]
Install fastimport.processors.

13 years agoremove dependency on bzrlib.
Jelmer Vernooij [Sun, 12 Sep 2010 00:46:46 +0000 (02:46 +0200)]
remove dependency on bzrlib.

13 years agoDon't rely on bzrlib.trace.
Jelmer Vernooij [Sun, 12 Sep 2010 00:39:11 +0000 (02:39 +0200)]
Don't rely on bzrlib.trace.

13 years agoUse modes directly rather than bzr file kinds and executable flags.
Jelmer Vernooij [Sat, 11 Sep 2010 22:34:22 +0000 (00:34 +0200)]
Use modes directly rather than bzr file kinds and executable flags.

13 years agoImport processors from bzr-fastimport.
Jelmer Vernooij [Sun, 5 Sep 2010 23:41:23 +0000 (01:41 +0200)]
Import processors from bzr-fastimport.

13 years agoSplit out reftracker.
Jelmer Vernooij [Sun, 5 Sep 2010 23:16:15 +0000 (01:16 +0200)]
Split out reftracker.

13 years agoFix typo, handle bzr-specific locking in GenericProcessor.
Jelmer Vernooij [Sun, 5 Sep 2010 21:42:04 +0000 (23:42 +0200)]
Fix typo, handle bzr-specific locking in GenericProcessor.

13 years agoremove bzrisms from Processor.
Jelmer Vernooij [Sun, 5 Sep 2010 21:30:48 +0000 (23:30 +0200)]
remove bzrisms from Processor.

13 years agoRemove bzr-specific bits from processor.
Jelmer Vernooij [Sat, 4 Sep 2010 21:59:19 +0000 (23:59 +0200)]
Remove bzr-specific bits from processor.

13 years agoOnly require passing BzrDir to bzr-specific processors.
Jelmer Vernooij [Sat, 4 Sep 2010 21:59:08 +0000 (23:59 +0200)]
Only require passing BzrDir to bzr-specific processors.

13 years agoIgnore testrepository directories.
Jelmer Vernooij [Sat, 4 Sep 2010 21:53:47 +0000 (23:53 +0200)]
Ignore testrepository directories.

13 years agoSupport the 'done' command.
Sverre Rabbelier [Sat, 4 Sep 2010 21:50:57 +0000 (23:50 +0200)]
Support the 'done' command.

13 years agoFix some imports.
Jelmer Vernooij [Sat, 4 Sep 2010 17:46:30 +0000 (19:46 +0200)]
Fix some imports.

13 years agoAdd testrepository configuration, test_suite() function.
Jelmer Vernooij [Sat, 4 Sep 2010 14:33:18 +0000 (16:33 +0200)]
Add testrepository configuration, test_suite() function.

13 years agoRemove bzr-specific code, whitespace.
Jelmer Vernooij [Sat, 4 Sep 2010 14:07:12 +0000 (16:07 +0200)]
Remove bzr-specific code, whitespace.

13 years agoSplit out fastimport, import it from the system.
Jelmer Vernooij [Fri, 3 Sep 2010 21:51:42 +0000 (23:51 +0200)]
Split out fastimport, import it from the system.

13 years agoSplit python-fastimport into its own separate package.
Jelmer Vernooij [Fri, 3 Sep 2010 21:28:11 +0000 (23:28 +0200)]
Split python-fastimport into its own separate package.

13 years agoMove pure-fastimport code into its own directory, in preparation of splitting it...
Jelmer Vernooij [Fri, 3 Sep 2010 21:09:05 +0000 (23:09 +0200)]
Move pure-fastimport code into its own directory, in preparation of splitting it into a separate package.

13 years agoMerge fixes from Miklos for darcs fast import.
Jelmer Vernooij [Tue, 27 Jul 2010 23:55:20 +0000 (01:55 +0200)]
Merge fixes from Miklos for darcs fast import.

13 years agodarcs-fast-import: don't fail for first empty commit
Miklos Vajna [Sun, 27 Jun 2010 23:21:13 +0000 (01:21 +0200)]
darcs-fast-import: don't fail for first empty commit

13 years agodarcs-fast-import: add --progress switch
Miklos Vajna [Sun, 27 Jun 2010 21:28:23 +0000 (23:28 +0200)]
darcs-fast-import: add --progress switch

13 years agox2d: pass options to d-f-i
Miklos Vajna [Sun, 27 Jun 2010 21:11:37 +0000 (23:11 +0200)]
x2d: pass options to d-f-i

13 years agodarcs-fast-import: Create symlinks after darcs record
Miklos Vajna [Sun, 27 Jun 2010 21:03:53 +0000 (23:03 +0200)]
darcs-fast-import: Create symlinks after darcs record

Otherwise darcs diff will show the file removals but even darcs record
-a won't record them.

13 years agodarcs-fast-import: Allow more liberal renaming
Miklos Vajna [Sun, 27 Jun 2010 18:40:01 +0000 (20:40 +0200)]
darcs-fast-import: Allow more liberal renaming

Allow renames like BUG -> bug. Darcs does not allow it by default.

13 years agodarcs-fast-import: add --symhack option
Miklos Vajna [Sun, 27 Jun 2010 17:37:50 +0000 (19:37 +0200)]
darcs-fast-import: add --symhack option

13 years agodarcs-fast-import: always invoke darcs add using --boring
Miklos Vajna [Sun, 27 Jun 2010 17:02:07 +0000 (19:02 +0200)]
darcs-fast-import: always invoke darcs add using --boring

13 years agodarcs-fast-import: do not silently ignore darcs failures
Miklos Vajna [Sun, 27 Jun 2010 17:00:08 +0000 (19:00 +0200)]
darcs-fast-import: do not silently ignore darcs failures

13 years agodarcs-fast-import: better error handling for symlinks
Miklos Vajna [Sun, 27 Jun 2010 16:51:11 +0000 (18:51 +0200)]
darcs-fast-import: better error handling for symlinks

The importer can't handle them till Darcs does not so, but at least we
can give an easy-to-understand error message and a separate exit code.

13 years agodarcs-fast-import: No longer throw away the blob after writing it
Miklos Vajna [Sun, 27 Jun 2010 15:59:55 +0000 (17:59 +0200)]
darcs-fast-import: No longer throw away the blob after writing it

It turns out that a mark can be used multiple times, and given that
reading it from darcs may be slow, let's just keep it in the memory for
now.

Later - if it turns out the memory usage is too heavy - we can still use
darcs show contents instead.

14 years agogit-darcs: Fix exit/return confusion.
Anders Kaseorg [Wed, 21 Apr 2010 21:23:07 +0000 (23:23 +0200)]
git-darcs: Fix exit/return confusion.

Shell functions should return instead of exiting the whole script;
they should return a nonzero status on error and a zero status on
success.  Notably, “Nothing fetched” is not an error, and pull should
go ahead and do a merge or rebase even if fetch didn’t fetch anything.

14 years agogit-darcs: Import to a real ref (refs/remotes/darcs/$remote).
Anders Kaseorg [Wed, 21 Apr 2010 21:23:07 +0000 (23:23 +0200)]
git-darcs: Import to a real ref (refs/remotes/darcs/$remote).

This allows the imported darcs branch to be seen by tools like ‘git
branch -r’ and ‘gitk --all’, and prevents the imported commits from
being garbage-collected.

14 years agofix timezones in darcs-fast-export, take 2
D Herring [Wed, 21 Apr 2010 21:23:07 +0000 (23:23 +0200)]
fix timezones in darcs-fast-export, take 2

The previous patch had a flaw; it assumed that darcs was storing the committer's timezone.
Instead, darcs always stores UTC timestamps in an ISO 8601 format.
Tools like "darcs changes" convert this into the user's local time as a convenience.

I couldn't find an authoritative spec, but here are some relevant references.
http://wiki.darcs.net/NamedPatch
http://search.cpan.org/~david/Darcs-Inventory-1.4/lib/Darcs/Inventory/Patch.pm
http://bugs.darcs.net/issue140

To resolve the issue, this patch always reports that the timezone is UTC.

14 years agoMerge patches from Miklos for darcs-fast-export
Ian Clatworthy [Wed, 17 Mar 2010 01:33:49 +0000 (11:33 +1000)]
Merge patches from Miklos for darcs-fast-export

14 years agofix timezones in darcs-fast-export
D Herring [Wed, 17 Mar 2010 00:14:48 +0000 (01:14 +0100)]
fix timezones in darcs-fast-export

Previously, the export was mangling timezones.  There were conversion errors when the darcs TZ
did not match the conversion TZ.  Also, the conversion timezone was always reported (this is
bad since two conversions may now differ).

This patch fixes both problems on my system, but it has not been extensively tested.
Can 'local_date' be used reliably?  What about my TZ manipulations?
I am no expert in darcs or python.  Developed with darcs 2.3.1 and Python 2.6.4.

To check for errors, I compared a darcs repo against a "git darcs fetch" into a new repo.
The following two commands were helpful.
# darcs changes | grep for dates
# git log --pretty=format:%ad REF | cat

Example -- EDT=UTC-4 and EST=UTC-5; I am processing in EDT.
darcs date:
"Fri Mar 31 10:48:00 EST 2006"
Before patch, git reported:
"Fri Mar 31 12:48:00 2006 -0400" (should be 11:48 -0400 or 10:48 -0500)
After patch, git reported:
"Fri Mar 31 10:48:00 2006 -0500"

14 years agogit-darcs: add a see-also section to manpage
Miklos Vajna [Wed, 17 Mar 2010 00:09:30 +0000 (01:09 +0100)]
git-darcs: add a see-also section to manpage

14 years agogit-darcs: add support for {pre,post}-{fetch,push} hooks
Miklos Vajna [Wed, 17 Mar 2010 00:09:30 +0000 (01:09 +0100)]
git-darcs: add support for {pre,post}-{fetch,push} hooks

14 years agogit-darcs: don't add empty darcs-fast-export-options config entry
Miklos Vajna [Wed, 17 Mar 2010 00:09:30 +0000 (01:09 +0100)]
git-darcs: don't add empty darcs-fast-export-options config entry

14 years agoRelease 0.9
Ian Clatworthy [Sat, 27 Feb 2010 14:47:56 +0000 (00:47 +1000)]
Release 0.9

14 years agoDon't bail for http://path/to/repo darcs remotes
Ian Clatworthy [Fri, 12 Feb 2010 20:42:47 +0000 (06:42 +1000)]
Don't bail for http://path/to/repo darcs remotes

14 years agoDon't bail for http://path/to/repo darcs remotes
Justin Bogner [Thu, 11 Feb 2010 09:11:22 +0000 (10:11 +0100)]
Don't bail for http://path/to/repo darcs remotes

14 years agoFallback to system configobj if none in bzrib.util
Ian Clatworthy [Tue, 9 Feb 2010 00:05:42 +0000 (10:05 +1000)]
Fallback to system configobj if none in bzrib.util

14 years agoCope with Debian's bzr using the system configobj, not shipping with its own.
Jelmer Vernooij [Mon, 8 Feb 2010 13:26:12 +0000 (14:26 +0100)]
Cope with Debian's bzr using the system configobj, not shipping with its own.

14 years agoAdd explorer integration
Ian Clatworthy [Sun, 31 Jan 2010 06:42:32 +0000 (16:42 +1000)]
Add explorer integration

14 years agoIgnore progress messages unless in verbose mode
Ian Clatworthy [Thu, 28 Jan 2010 06:00:31 +0000 (16:00 +1000)]
Ignore progress messages unless in verbose mode

14 years agoFix Monotone prefix and command name: mnt -> mtn
Ian Clatworthy [Thu, 28 Jan 2010 05:27:08 +0000 (15:27 +1000)]
Fix Monotone prefix and command name: mnt -> mtn

14 years agoMerge Python 2.5 fix for darcs-fast-export
Ian Clatworthy [Mon, 25 Jan 2010 01:59:10 +0000 (11:59 +1000)]
Merge Python 2.5 fix for darcs-fast-export

14 years agodarcs-fast-export: don't use urllib.urlopen().getcode()
Miklos Vajna [Sun, 24 Jan 2010 23:45:31 +0000 (00:45 +0100)]
darcs-fast-export: don't use urllib.urlopen().getcode()

As it's missing from python2.5.

14 years agoInclude the known_graph code.
John Arbash Meinel [Tue, 5 Jan 2010 07:24:14 +0000 (01:24 -0600)]
Include the known_graph code.

14 years agoSwitch to using CommitBuilder
John Arbash Meinel [Fri, 11 Dec 2009 14:55:04 +0000 (08:55 -0600)]
Switch to using CommitBuilder
Rather than holding most of the logic ourselves.

14 years agoAdd a bunch of direct tests for the _TreeShim interface.
John Arbash Meinel [Wed, 9 Dec 2009 20:14:46 +0000 (14:14 -0600)]
Add a bunch of direct tests for the _TreeShim interface.

14 years agoMerge John's test refactoring - uses multiplication now
Ian Clatworthy [Tue, 8 Dec 2009 06:31:53 +0000 (16:31 +1000)]
Merge John's test refactoring - uses multiplication now

14 years agoMerge John's improvements to info_processor.py
Ian Clatworthy [Tue, 8 Dec 2009 06:29:47 +0000 (16:29 +1000)]
Merge John's improvements to info_processor.py

14 years agoMerge John's smarter caching of blobs to improve memory footprint
Ian Clatworthy [Tue, 8 Dec 2009 06:26:34 +0000 (16:26 +1000)]
Merge John's smarter caching of blobs to improve memory footprint

14 years agoFaster post-processing of the inventory delta
Ian Clatworthy [Fri, 4 Dec 2009 06:52:54 +0000 (16:52 +1000)]
Faster post-processing of the inventory delta

14 years agoSome debugging code. It looks like the main bugs involve files that are deleted and...
John Arbash Meinel [Tue, 1 Dec 2009 15:15:19 +0000 (09:15 -0600)]
Some debugging code. It looks like the main bugs involve files that are deleted and restored.

14 years agoWhen post-processing the delta stream, don't ask to generate a full inventory to...
John Arbash Meinel [Mon, 30 Nov 2009 21:31:31 +0000 (15:31 -0600)]
When post-processing the delta stream, don't ask to generate a full inventory to check for deletions.
Instead, just use 'create_by_apply_delta'.

14 years agoUse the new KnownGraph.add_node() functionality.
John Arbash Meinel [Mon, 30 Nov 2009 19:00:44 +0000 (13:00 -0600)]
Use the new KnownGraph.add_node() functionality.

This changes the file_id heads lookups to use a KnownGraph over all revision-ids
which should be a lot better than using a regular Graph instance.
Initial testing w/ xserver doesn't show much, since it seems to be purely
linear for quite a bit of x history.

Oddly enough the heads result seems to give a different number of file texts
which is worrying.

14 years agoThe fix that landed in bzr did not set builder.new_inventory. :(
John Arbash Meinel [Mon, 30 Nov 2009 17:57:02 +0000 (11:57 -0600)]
The fix that landed in bzr did not set builder.new_inventory. :(

This still works, but it means that we don't keep the in-memory deserialized bits
from the previous inventory. Instead we start from-scratch each time.

14 years agoMerge Jelmer's fix for exporters installation
Ian Clatworthy [Sun, 29 Nov 2009 10:37:32 +0000 (20:37 +1000)]
Merge Jelmer's fix for exporters installation

14 years agoFix installation of the exporters module.
Jelmer Vernooij [Thu, 26 Nov 2009 19:03:48 +0000 (20:03 +0100)]
Fix installation of the exporters module.

14 years agoDefault branch nick to mapped git ref name (Max Bowsher)
Ian Clatworthy [Mon, 23 Nov 2009 07:06:23 +0000 (17:06 +1000)]
Default branch nick to mapped git ref name (Max Bowsher)

14 years agobzr *does* run atexit functions when exiting, but doesn't run deconstructors.
John Arbash Meinel [Sun, 15 Nov 2009 20:42:36 +0000 (14:42 -0600)]
bzr *does* run atexit functions when exiting, but doesn't run deconstructors.

Also, shrink the 'small blob' size a bit to allow data to be reclaimed. Though it did
show up as *lots* of small files in the qt import. Something like 1-2k files in the
first 2 dumps.

14 years agoSwitch to closing the large-content blobs that we store to disk.
John Arbash Meinel [Fri, 13 Nov 2009 19:16:20 +0000 (13:16 -0600)]
Switch to closing the large-content blobs that we store to disk.

During the qt import, we end up with >2000 large blobs, and a few hundred MB of
small blobs. We don't want to create more small blobs, because their disk space
is not reclaimed, but >2000 is too many open file handles (at least on Windows).
Using filenames works, but we aren't as guaranteed that things will clean up
nicely.

14 years agoDump sticky blobs to disk when memory pressure gets high.
John Arbash Meinel [Fri, 13 Nov 2009 08:00:16 +0000 (02:00 -0600)]
Dump sticky blobs to disk when memory pressure gets high.

14 years agoFall back to the repository for cases where the content is not present in the stream...
John Arbash Meinel [Thu, 12 Nov 2009 23:25:04 +0000 (17:25 -0600)]
Fall back to the repository for cases where the content is not present in the stream yet.

14 years agoWe need to handle when the object has been deleted.
John Arbash Meinel [Thu, 12 Nov 2009 22:29:42 +0000 (16:29 -0600)]
We need to handle when the object has been deleted.

14 years agoFound a bug in CommitBuilder.finish_inventory().
John Arbash Meinel [Thu, 12 Nov 2009 22:23:23 +0000 (16:23 -0600)]
Found a bug in CommitBuilder.finish_inventory().

It turns out that the add_revision_by_delta code was not properly setting
inv_sha1, but it didn't matter because apparently the other code wasn't really
using it....

Anyway, this gives us a proper new inventory, and seems to generally be working.
The next step is that the 'merge' file text is not present because the
delta stream does not contain it...

14 years ago(broken) Start working towards using CommitBuilder rather than using a custom impleme...
John Arbash Meinel [Thu, 12 Nov 2009 14:07:24 +0000 (08:07 -0600)]
(broken) Start working towards using CommitBuilder rather than using a custom implementation.

14 years agoadd the failing test that we preserve the last-modified revision
John Arbash Meinel [Thu, 12 Nov 2009 08:02:42 +0000 (02:02 -0600)]
add the failing test that we preserve the last-modified revision

14 years agoChange to multiplying tests rather than manually.
John Arbash Meinel [Thu, 12 Nov 2009 07:51:21 +0000 (01:51 -0600)]
Change to multiplying tests rather than manually.

14 years agoChange (.keys(), .values()) to using (.iteritems())
John Arbash Meinel [Thu, 12 Nov 2009 05:37:43 +0000 (23:37 -0600)]
Change (.keys(), .values()) to using (.iteritems())

14 years agoMerge Max's fixes for tags created via commit commands
Ian Clatworthy [Fri, 6 Nov 2009 08:06:27 +0000 (18:06 +1000)]
Merge Max's fixes for tags created via commit commands

14 years agoAlso catch tagging via commit when resuming a crashed import.
Max Bowsher [Wed, 4 Nov 2009 01:58:36 +0000 (01:58 +0000)]
Also catch tagging via commit when resuming a crashed import.

14 years agoDefault branch-nick to mapped git ref name.
Max Bowsher [Wed, 4 Nov 2009 01:52:34 +0000 (01:52 +0000)]
Default branch-nick to mapped git ref name.

14 years agoStore the BranchMapper in the CacheManager so it can be got from other places.
Max Bowsher [Wed, 4 Nov 2009 01:27:59 +0000 (01:27 +0000)]
Store the BranchMapper in the CacheManager so it can be got from other places.

14 years agoMake BranchMapper just map one name per call.
Max Bowsher [Wed, 4 Nov 2009 01:04:34 +0000 (01:04 +0000)]
Make BranchMapper just map one name per call.
Move building a dict to the one callsite which actually wants that.

14 years agoRemove unused and unimplemented BranchMapper.bzr_to_git, and update docstring.
Max Bowsher [Wed, 4 Nov 2009 00:52:27 +0000 (00:52 +0000)]
Remove unused and unimplemented BranchMapper.bzr_to_git, and update docstring.

14 years agoAvoid spurious 'git-' being prefixed on branches whose names happen to end with ...
Max Bowsher [Wed, 4 Nov 2009 00:49:10 +0000 (00:49 +0000)]
Avoid spurious 'git-' being prefixed on branches whose names happen to end with 'trunk',
by tightening logic in BranchMapper._git_to_bzr_name.  Also document.

14 years agoSet a tag when touching a refs/tags/ ref with a commit command.
Max Bowsher [Tue, 3 Nov 2009 23:52:58 +0000 (23:52 +0000)]
Set a tag when touching a refs/tags/ ref with a commit command.

14 years agohttp read support for darcs-fast-export
Ian Clatworthy [Tue, 27 Oct 2009 05:49:55 +0000 (15:49 +1000)]
http read support for darcs-fast-export

14 years agoHandle multi-level branches
Ian Clatworthy [Mon, 26 Oct 2009 00:15:57 +0000 (10:15 +1000)]
Handle multi-level branches

14 years agoGet fastimport working on non-chk repositories again for bzr versions after 2.0.0
Ian Clatworthy [Sun, 25 Oct 2009 22:05:48 +0000 (08:05 +1000)]
Get fastimport working on non-chk repositories again for bzr versions after 2.0.0

14 years agodarcs-fast-export: refactor to a python class
Miklos Vajna [Thu, 22 Oct 2009 10:26:32 +0000 (12:26 +0200)]
darcs-fast-export: refactor to a python class

The code should be a lot more readable now and we can avoid function
names like open_() as well.

All testcases still pass.