5
This document describes the Bazaar internals and the development process.
6
It's meant for people interested in developing Bazaar, and some parts will
7
also be useful to people developing Bazaar plugins.
9
If you have any questions or something seems to be incorrect, unclear or
10
missing, please talk to us in ``irc://irc.freenode.net/#bzr``, or write to
11
the Bazaar mailing list. To propose a correction or addition to this
12
document, send a merge request or new text to the mailing list.
14
The current version of this document is available in the file
15
``doc/developers/HACKING.txt`` in the source tree, or at
16
http://doc.bazaar-vcs.org/bzr.dev/en/developer-guide/HACKING.html
19
`Bazaar Developer Documentation Catalog <../../developers/index.html>`_.
27
Exploring the Bazaar Platform
28
=============================
30
Before making changes, it's a good idea to explore the work already
31
done by others. Perhaps the new feature or improvement you're looking
32
for is available in another plug-in already? If you find a bug,
33
perhaps someone else has already fixed it?
35
To answer these questions and more, take a moment to explore the
36
overall Bazaar Platform. Here are some links to browse:
38
* The Plugins page on the Wiki - http://bazaar-vcs.org/BzrPlugins
40
* The Bazaar product family on Launchpad - https://launchpad.net/bazaar
42
* Bug Tracker for the core product - https://bugs.launchpad.net/bzr/
44
* Blueprint Tracker for the core product - https://blueprints.launchpad.net/bzr/
46
If nothing else, perhaps you'll find inspiration in how other developers
47
have solved their challenges.
50
Planning and Discussing Changes
51
===============================
53
There is a very active community around Bazaar. Mostly we meet on IRC
54
(#bzr on irc.freenode.net) and on the mailing list. To join the Bazaar
55
community, see http://bazaar-vcs.org/BzrSupport.
57
If you are planning to make a change, it's a very good idea to mention it
58
on the IRC channel and/or on the mailing list. There are many advantages
59
to involving the community before you spend much time on a change.
62
* you get to build on the wisdom on others, saving time
64
* if others can direct you to similar code, it minimises the work to be done
66
* it assists everyone in coordinating direction, priorities and effort.
68
In summary, maximising the input from others typically minimises the
69
total effort required to get your changes merged. The community is
70
friendly, helpful and always keen to welcome newcomers.
73
Bazaar Development in a Nutshell
74
================================
76
Looking for a 10 minute introduction to submitting a change?
77
See http://bazaar-vcs.org/BzrGivingBack.
79
TODO: Merge that Wiki page into this document.
82
Understanding the Development Process
83
=====================================
85
The development team follows many practices including:
87
* a public roadmap and planning process in which anyone can participate
89
* time based milestones everyone can work towards and plan around
91
* extensive code review and feedback to contributors
93
* complete and rigorous test coverage on any code contributed
95
* automated validation that all tests still pass before code is merged
96
into the main code branch.
98
The key tools we use to enable these practices are:
100
* Launchpad - https://launchpad.net/
102
* Bazaar - http://bazaar-vcs.org/
104
* Bundle Buggy - http://bundlebuggy.aaronbentley.com/
106
* Patch Queue Manager - https://launchpad.net/pqm/
108
For further information, see http://bazaar-vcs.org/BzrDevelopment.
113
Preparing a Sandbox for Making Changes to Bazaar
114
================================================
116
Bazaar supports many ways of organising your work. See
117
http://bazaar-vcs.org/SharedRepositoryLayouts for a summary of the
118
popular alternatives.
120
Of course, the best choice for you will depend on numerous factors:
121
the number of changes you may be making, the complexity of the changes, etc.
122
As a starting suggestion though:
124
* create a local copy of the main development branch (bzr.dev) by using
127
bzr branch http://bazaar-vcs.org/bzr/bzr.dev/ bzr.dev
129
* keep your copy of bzr.dev prestine (by not developing in it) and keep
130
it up to date (by using bzr pull)
132
* create a new branch off your local bzr.dev copy for each issue
133
(bug or feature) you are working on.
135
This approach makes it easy to go back and make any required changes
136
after a code review. Resubmitting the change is then simple with no
137
risk of accidentially including edits related to other issues you may
138
be working on. After the changes for an issue are accepted and merged,
139
the associated branch can be deleted or archived as you wish.
142
Navigating the Code Base
143
========================
145
.. Was at <http://bazaar-vcs.org/NewDeveloperIntroduction>
147
Some of the key files in this directory are:
150
The command you run to start Bazaar itself. This script is pretty
151
short and just does some checks then jumps into bzrlib.
154
This file covers a brief introduction to Bazaar and lists some of its
158
Summary of changes in each Bazaar release that can affect users or
162
Installs Bazaar system-wide or to your home directory. To perform
163
development work on Bazaar it is not required to run this file - you
164
can simply run the bzr command from the top level directory of your
165
development copy. Note: That if you run setup.py this will create a
166
'build' directory in your development branch. There's nothing wrong
167
with this but don't be confused by it. The build process puts a copy
168
of the main code base into this build directory, along with some other
169
files. You don't need to go in here for anything discussed in this
173
Possibly the most exciting folder of all, bzrlib holds the main code
174
base. This is where you will go to edit python files and contribute to
178
Holds documentation on a whole range of things on Bazaar from the
179
origination of ideas within the project to information on Bazaar
180
features and use cases. Within this directory there is a subdirectory
181
for each translation into a human language. All the documentation
182
is in the ReStructuredText markup language.
185
Documentation specifically targetted at Bazaar and plugin developers.
186
(Including this document.)
190
Automatically-generated API reference information is available at
191
<http://starship.python.net/crew/mwh/bzrlibapi/>.
193
See also the `Bazaar Architectural Overview <../../developers/overview.html>`_.
196
The Code Review Process
197
#######################
199
All code changes coming in to Bazaar are reviewed by someone else.
200
Normally changes by core contributors are reviewed by one other core
201
developer, and changes from other people are reviewed by two core
202
developers. Use intelligent discretion if the patch is trivial.
204
Good reviews do take time. They also regularly require a solid
205
understanding of the overall code base. In practice, this means a small
206
number of people often have a large review burden - with knowledge comes
207
responsibility. No one like their merge requests sitting in a queue going
208
nowhere, so reviewing sooner rather than later is strongly encouraged.
212
Sending patches for review
213
==========================
215
If you'd like to propose a change, please post to the
216
bazaar@lists.canonical.com list with a bundle, patch, or link to a
217
branch. Put ``[PATCH]`` or ``[MERGE]`` in the subject so Bundle Buggy
218
can pick it out, and explain the change in the email message text.
219
Remember to update the NEWS file as part of your change if it makes any
220
changes visible to users or plugin developers. Please include a diff
221
against mainline if you're giving a link to a branch.
223
You can generate a merge request like this::
225
bzr send -o bug-1234.patch
227
A ``.patch`` extension is recommended instead of .bundle as many mail clients
228
will send the latter as a binary file.
230
``bzr send`` can also send mail directly if you prefer; see the help.
232
Please do **NOT** put [PATCH] or [MERGE] in the subject line if you don't
233
want it to be merged. If you want comments from developers rather than
234
to be merged, you can put ``[RFC]`` in the subject line.
236
If this change addresses a bug, please put the bug number in the subject
237
line too, in the form ``[#1]`` so that Bundle Buggy can recognize it.
239
If the change is intended for a particular release mark that in the
240
subject too, e.g. ``[1.6]``.
246
Please put a "cover letter" on your merge request explaining:
248
* the reason **why** you're making this change
250
* **how** this change achieves this purpose
252
* anything else you may have fixed in passing
254
* anything significant that you thought of doing, such as a more
255
extensive fix or a different approach, but didn't or couldn't do now
257
A good cover letter makes reviewers' lives easier because they can decide
258
from the letter whether they agree with the purpose and approach, and then
259
assess whether the patch actually does what the cover letter says.
260
Explaining any "drive-by fixes" or roads not taken may also avoid queries
261
from the reviewer. All in all this should give faster and better reviews.
262
Sometimes writing the cover letter helps the submitter realize something
263
else they need to do. The size of the cover letter should be proportional
264
to the size and complexity of the patch.
267
Reviewing proposed changes
268
==========================
270
Anyone is welcome to review code, and reply to the thread with their
273
The simplest way to review a proposed change is to just read the patch on
274
the list or in Bundle Buggy. For more complex changes it may be useful
275
to make a new working tree or branch from trunk, and merge the proposed
276
change into it, so you can experiment with the code or look at a wider
279
There are three main requirements for code to get in:
281
* Doesn't reduce test coverage: if it adds new methods or commands,
282
there should be tests for them. There is a good test framework
283
and plenty of examples to crib from, but if you are having trouble
284
working out how to test something feel free to post a draft patch
287
* Doesn't reduce design clarity, such as by entangling objects
288
we're trying to separate. This is mostly something the more
289
experienced reviewers need to help check.
291
* Improves bugs, features, speed, or code simplicity.
293
Code that goes in should not degrade any of these aspects. Patches are
294
welcome that only cleanup the code without changing the external
295
behaviour. The core developers take care to keep the code quality high
296
and understandable while recognising that perfect is sometimes the enemy
299
It is easy for reviews to make people notice other things which should be
300
fixed but those things should not hold up the original fix being accepted.
301
New things can easily be recorded in the Bug Tracker instead.
303
It's normally much easier to review several smaller patches than one large
304
one. You might want to use ``bzr-loom`` to maintain threads of related
305
work, or submit a preparatory patch that will make your "real" change
309
Checklist for reviewers
310
=======================
312
* Do you understand what the code's doing and why?
314
* Will it perform reasonably for large inputs, both in memory size and
315
run time? Are there some scenarios where performance should be
318
* Is it tested, and are the tests at the right level? Are there both
319
blackbox (command-line level) and API-oriented tests?
321
* If this change will be visible to end users or API users, is it
322
appropriately documented in NEWS?
324
* Does it meet the coding standards below?
326
* If it changes the user-visible behaviour, does it update the help
327
strings and user documentation?
329
* If it adds a new major concept or standard practice, does it update the
330
developer documentation?
332
* (your ideas here...)
335
Bundle Buggy and review outcomes
336
================================
338
Anyone can "vote" on the mailing list by expressing an opinion. Core
339
developers can also vote using Bundle Buggy. Here are the voting codes and
342
:approve: Reviewer wants this submission merged.
343
:tweak: Reviewer wants this submission merged with small changes. (No
345
:abstain: Reviewer does not intend to vote on this patch.
346
:resubmit: Please make changes and resubmit for review.
347
:reject: Reviewer doesn't want this kind of change merged.
348
:comment: Not really a vote. Reviewer just wants to comment, for now.
350
If a change gets two approvals from core reviewers, and no rejections,
351
then it's OK to come in. Any of the core developers can bring it into the
352
bzr.dev trunk and backport it to maintenance branches if required. The
353
Release Manager will merge the change into the branch for a pending
354
release, if any. As a guideline, core developers usually merge their own
355
changes and volunteer to merge other contributions if they were the second
356
reviewer to agree to a change.
358
To track the progress of proposed changes, use Bundle Buggy. See
359
http://bundlebuggy.aaronbentley.com/help for a link to all the
360
outstanding merge requests together with an explanation of the columns.
361
Bundle Buggy will also mail you a link to track just your change.
363
Coding Style Guidelines
364
#######################
369
``hasattr`` should not be used because it swallows exceptions including
370
``KeyboardInterrupt``. Instead, say something like ::
372
if getattr(thing, 'name', None) is None
378
Please write PEP-8__ compliant code.
380
__ http://www.python.org/peps/pep-0008.html
382
One often-missed requirement is that the first line of docstrings
383
should be a self-contained one-sentence summary.
385
We use 4 space indents for blocks, and never use tab characters. (In vim,
388
Lines should be no more than 79 characters if at all possible.
389
Lines that continue a long statement may be indented in either of
392
within the parenthesis or other character that opens the block, e.g.::
398
or indented by four spaces::
404
The first is considered clearer by some people; however it can be a bit
405
harder to maintain (e.g. when the method name changes), and it does not
406
work well if the relevant parenthesis is already far to the right. Avoid
409
self.legbone.kneebone.shinbone.toebone.shake_it(one,
415
self.legbone.kneebone.shinbone.toebone.shake_it(one,
421
self.legbone.kneebone.shinbone.toebone.shake_it(
424
For long lists, we like to add a trailing comma and put the closing
425
character on the following line. This makes it easier to add new items in
428
from bzrlib.goo import (
434
There should be spaces between function paramaters, but not between the
435
keyword name and the value::
437
call(1, 3, cheese=quark)
441
;(defface my-invalid-face
442
; '((t (:background "Red" :underline t)))
443
; "Face used to highlight invalid constructs or other uglyties"
446
(defun my-python-mode-hook ()
447
;; setup preferred indentation style.
448
(setq fill-column 79)
449
(setq indent-tabs-mode nil) ; no tabs, never, I will not repeat
450
; (font-lock-add-keywords 'python-mode
451
; '(("^\\s *\t" . 'my-invalid-face) ; Leading tabs
452
; ("[ \t]+$" . 'my-invalid-face) ; Trailing spaces
453
; ("^[ \t]+$" . 'my-invalid-face)); Spaces only
457
(add-hook 'python-mode-hook 'my-python-mode-hook)
459
The lines beginning with ';' are comments. They can be activated
460
if one want to have a strong notice of some tab/space usage
467
* Imports should be done at the top-level of the file, unless there is
468
a strong reason to have them lazily loaded when a particular
469
function runs. Import statements have a cost, so try to make sure
470
they don't run inside hot functions.
472
* Module names should always be given fully-qualified,
473
i.e. ``bzrlib.hashcache`` not just ``hashcache``.
479
Functions, methods or members that are "private" to bzrlib are given
480
a leading underscore prefix. Names without a leading underscore are
481
public not just across modules but to programmers using bzrlib as an
482
API. As a consequence, a leading underscore is appropriate for names
483
exposed across modules but that are not to be exposed to bzrlib API
486
We prefer class names to be concatenated capital words (``TestCase``)
487
and variables, methods and functions to be lowercase words joined by
488
underscores (``revision_id``, ``get_revision``).
490
For the purposes of naming some names are treated as single compound
491
words: "filename", "revno".
493
Consider naming classes as nouns and functions/methods as verbs.
495
Try to avoid using abbreviations in names, because there can be
496
inconsistency if other people use the full name.
502
``revision_id`` not ``rev_id`` or ``revid``
504
Functions that transform one thing to another should be named ``x_to_y``
505
(not ``x2y`` as occurs in some old code.)
511
Python destructors (``__del__``) work differently to those of other
512
languages. In particular, bear in mind that destructors may be called
513
immediately when the object apparently becomes unreferenced, or at some
514
later time, or possibly never at all. Therefore we have restrictions on
515
what can be done inside them.
517
0. If you think you need to use a ``__del__`` method ask another
518
developer for alternatives. If you do need to use one, explain
521
1. Never rely on a ``__del__`` method running. If there is code that
522
must run, do it from a ``finally`` block instead.
524
2. Never ``import`` from inside a ``__del__`` method, or you may crash the
527
3. In some places we raise a warning from the destructor if the object
528
has not been cleaned up or closed. This is considered OK: the warning
529
may not catch every case but it's still useful sometimes.
535
In some places we have variables which point to callables that construct
536
new instances. That is to say, they can be used a lot like class objects,
537
but they shouldn't be *named* like classes:
539
> I think that things named FooBar should create instances of FooBar when
540
> called. Its plain confusing for them to do otherwise. When we have
541
> something that is going to be used as a class - that is, checked for via
542
> isinstance or other such idioms, them I would call it foo_class, so that
543
> it is clear that a callable is not sufficient. If it is only used as a
544
> factory, then yes, foo_factory is what I would use.
550
Several places in Bazaar use (or will use) a registry, which is a
551
mapping from names to objects or classes. The registry allows for
552
loading in registered code only when it's needed, and keeping
553
associated information such as a help string or description.
556
InterObject and multiple dispatch
557
=================================
559
The ``InterObject`` provides for two-way `multiple dispatch`__: matching
560
up for example a source and destination repository to find the right way
561
to transfer data between them.
563
.. __: http://en.wikipedia.org/wiki/Multiple_dispatch
565
There is a subclass ``InterObject`` classes for each type of object that is
566
dispatched this way, e.g. ``InterRepository``. Calling ``.get()`` on this
567
class will return an ``InterObject`` instance providing the best match for
568
those parameters, and this instance then has methods for operations
571
inter = InterRepository.get(source_repo, target_repo)
572
inter.fetch(revision_id)
574
``InterRepository`` also acts as a registry-like object for its
575
subclasses, and they can be added through ``.register_optimizer``. The
576
right one to run is selected by asking each class, in reverse order of
577
registration, whether it ``.is_compatible`` with the relevant objects.
582
To make startup time faster, we use the ``bzrlib.lazy_import`` module to
583
delay importing modules until they are actually used. ``lazy_import`` uses
584
the same syntax as regular python imports. So to import a few modules in a
587
from bzrlib.lazy_import import lazy_import
588
lazy_import(globals(), """
597
revision as _mod_revision,
599
import bzrlib.transport
603
At this point, all of these exist as a ``ImportReplacer`` object, ready to
604
be imported once a member is accessed. Also, when importing a module into
605
the local namespace, which is likely to clash with variable names, it is
606
recommended to prefix it as ``_mod_<module>``. This makes it clearer that
607
the variable is a module, and these object should be hidden anyway, since
608
they shouldn't be imported into other namespaces.
610
While it is possible for ``lazy_import()`` to import members of a module
611
when using the ``from module import member`` syntax, it is recommended to
612
only use that syntax to load sub modules ``from module import submodule``.
613
This is because variables and classes can frequently be used without
614
needing a sub-member for example::
616
lazy_import(globals(), """
617
from module import MyClass
621
return isinstance(x, MyClass)
623
This will incorrectly fail, because ``MyClass`` is a ``ImportReplacer``
624
object, rather than the real class.
626
It also is incorrect to assign ``ImportReplacer`` objects to other variables.
627
Because the replacer only knows about the original name, it is unable to
628
replace other variables. The ``ImportReplacer`` class will raise an
629
``IllegalUseOfScopeReplacer`` exception if it can figure out that this
630
happened. But it requires accessing a member more than once from the new
631
variable, so some bugs are not detected right away.
637
The null revision is the ancestor of all revisions. Its revno is 0, its
638
revision-id is ``null:``, and its tree is the empty tree. When referring
639
to the null revision, please use ``bzrlib.revision.NULL_REVISION``. Old
640
code sometimes uses ``None`` for the null revision, but this practice is
644
Object string representations
645
=============================
647
Python prints objects using their ``__repr__`` method when they are
648
written to logs, exception tracebacks, or the debugger. We want
649
objects to have useful representations to help in determining what went
652
If you add a new class you should generally add a ``__repr__`` method
653
unless there is an adequate method in a parent class. There should be a
656
Representations should typically look like Python constructor syntax, but
657
they don't need to include every value in the object and they don't need
658
to be able to actually execute. They're to be read by humans, not
659
machines. Don't hardcode the classname in the format, so that we get the
660
correct value if the method is inherited by a subclass. If you're
661
printing attributes of the object, including strings, you should normally
662
use ``%r`` syntax (to call their repr in turn).
664
Try to avoid the representation becoming more than one or two lines long.
665
(But balance this against including useful information, and simplicity of
668
Because repr methods are often called when something has already gone
669
wrong, they should be written somewhat more defensively than most code.
670
The object may be half-initialized or in some other way in an illegal
671
state. The repr method shouldn't raise an exception, or it may hide the
672
(probably more useful) underlying exception.
677
return '%s(%r)' % (self.__class__.__name__,
684
A bare ``except`` statement will catch all exceptions, including ones that
685
really should terminate the program such as ``MemoryError`` and
686
``KeyboardInterrupt``. They should rarely be used unless the exception is
687
later re-raised. Even then, think about whether catching just
688
``Exception`` (which excludes system errors in Python2.5 and later) would
695
All code should be exercised by the test suite. See `Guide to Testing
696
Bazaar <testing.html>`_ for detailed information about writing tests.
705
We have a commitment to 6 months API stability - any supported symbol in a
706
release of bzr MUST NOT be altered in any way that would result in
707
breaking existing code that uses it. That means that method names,
708
parameter ordering, parameter names, variable and attribute names etc must
709
not be changed without leaving a 'deprecated forwarder' behind. This even
710
applies to modules and classes.
712
If you wish to change the behaviour of a supported API in an incompatible
713
way, you need to change its name as well. For instance, if I add an optional keyword
714
parameter to branch.commit - that's fine. On the other hand, if I add a
715
keyword parameter to branch.commit which is a *required* transaction
716
object, I should rename the API - i.e. to 'branch.commit_transaction'.
718
When renaming such supported API's, be sure to leave a deprecated_method (or
719
_function or ...) behind which forwards to the new API. See the
720
bzrlib.symbol_versioning module for decorators that take care of the
721
details for you - such as updating the docstring, and issuing a warning
722
when the old api is used.
724
For unsupported API's, it does not hurt to follow this discipline, but it's
725
not required. Minimally though, please try to rename things so that
726
callers will at least get an AttributeError rather than weird results.
729
Deprecation decorators
730
----------------------
732
``bzrlib.symbol_versioning`` provides decorators that can be attached to
733
methods, functions, and other interfaces to indicate that they should no
734
longer be used. For example::
736
@deprecated_method(deprecated_in((0, 1, 4)))
738
return self._new_foo()
740
To deprecate a static method you must call ``deprecated_function``
741
(**not** method), after the staticmethod call::
744
@deprecated_function(deprecated_in((0, 1, 4)))
745
def create_repository(base, shared=False, format=None):
747
When you deprecate an API, you should not just delete its tests, because
748
then we might introduce bugs in them. If the API is still present at all,
749
it should still work. The basic approach is to use
750
``TestCase.applyDeprecated`` which in one step checks that the API gives
751
the expected deprecation message, and also returns the real result from
752
the method, so that tests can keep running.
754
Deprecation warnings will be suppressed for final releases, but not for
755
development versions or release candidates, or when running ``bzr
756
selftest``. This gives developers information about whether their code is
757
using deprecated functions, but avoids confusing users about things they
764
Processing Command Lines
765
------------------------
767
bzrlib has a standard framework for parsing command lines and calling
768
processing routines associated with various commands. See builtins.py
769
for numerous examples.
772
Standard Parameter Types
773
------------------------
775
There are some common requirements in the library: some parameters need to be
776
unicode safe, some need byte strings, and so on. At the moment we have
777
only codified one specific pattern: Parameters that need to be unicode
778
should be checked via ``bzrlib.osutils.safe_unicode``. This will coerce the
779
input into unicode in a consistent fashion, allowing trivial strings to be
780
used for programmer convenience, but not performing unpredictably in the
781
presence of different locales.
787
(The strategy described here is what we want to get to, but it's not
788
consistently followed in the code at the moment.)
790
bzrlib is intended to be a generically reusable library. It shouldn't
791
write messages to stdout or stderr, because some programs that use it
792
might want to display that information through a GUI or some other
795
We can distinguish two types of output from the library:
797
1. Structured data representing the progress or result of an
798
operation. For example, for a commit command this will be a list
799
of the modified files and the finally committed revision number
802
These should be exposed either through the return code or by calls
803
to a callback parameter.
805
A special case of this is progress indicators for long-lived
806
operations, where the caller should pass a ProgressBar object.
808
2. Unstructured log/debug messages, mostly for the benefit of the
809
developers or users trying to debug problems. This should always
810
be sent through ``bzrlib.trace`` and Python ``logging``, so that
811
it can be redirected by the client.
813
The distinction between the two is a bit subjective, but in general if
814
there is any chance that a library would want to see something as
815
structured data, we should make it so.
817
The policy about how output is presented in the text-mode client
818
should be only in the command-line tool.
825
Bazaar has online help for various topics through ``bzr help COMMAND`` or
826
equivalently ``bzr command -h``. We also have help on command options,
827
and on other help topics. (See ``help_topics.py``.)
829
As for python docstrings, the first paragraph should be a single-sentence
830
synopsis of the command.
832
The help for options should be one or more proper sentences, starting with
833
a capital letter and finishing with a full stop (period).
835
All help messages and documentation should have two spaces between
839
Handling Errors and Exceptions
840
==============================
842
Commands should return non-zero when they encounter circumstances that
843
the user should really pay attention to - which includes trivial shell
846
Recommended values are:
849
1. Conflicts in merge-like operations, or changes are present in
850
diff-like operations.
851
2. Unrepresentable diff changes (i.e. binary files that we cannot show
853
3. An error or exception has occurred.
854
4. An internal error occurred (one that shows a traceback.)
856
Errors are handled through Python exceptions. Exceptions should be defined
857
inside bzrlib.errors, so that we can see the whole tree at a glance.
859
We broadly classify errors as either being either internal or not,
860
depending on whether ``internal_error`` is set or not. If we think it's our
861
fault, we show a backtrace, an invitation to report the bug, and possibly
862
other details. This is the default for errors that aren't specifically
863
recognized as being caused by a user error. Otherwise we show a briefer
864
message, unless -Derror was given.
866
Many errors originate as "environmental errors" which are raised by Python
867
or builtin libraries -- for example IOError. These are treated as being
868
our fault, unless they're caught in a particular tight scope where we know
869
that they indicate a user errors. For example if the repository format
870
is not found, the user probably gave the wrong path or URL. But if one of
871
the files inside the repository is not found, then it's our fault --
872
either there's a bug in bzr, or something complicated has gone wrong in
873
the environment that means one internal file was deleted.
875
Many errors are defined in ``bzrlib/errors.py`` but it's OK for new errors
876
to be added near the place where they are used.
878
Exceptions are formatted for the user by conversion to a string
879
(eventually calling their ``__str__`` method.) As a convenience the
880
``._fmt`` member can be used as a template which will be mapped to the
881
error's instance dict.
883
New exception classes should be defined when callers might want to catch
884
that exception specifically, or when it needs a substantially different
887
#. If it is something that a caller can recover from, a custom exception
890
#. If it is a data consistency issue, using a builtin like
891
``ValueError``/``TypeError`` is reasonable.
893
#. If it is a programmer error (using an api incorrectly)
894
``AssertionError`` is reasonable.
896
#. Otherwise, use ``BzrError`` or ``InternalBzrError``.
898
Exception strings should start with a capital letter and should not have a
899
final fullstop. If long, they may contain newlines to break the text.
905
Do not use the Python ``assert`` statement, either in tests or elsewhere.
906
A source test checks that it is not used. It is ok to explicitly raise
911
* It makes the behaviour vary depending on whether bzr is run with -O
912
or not, therefore giving a chance for bugs that occur in one case or
913
the other, several of which have already occurred: assertions with
914
side effects, code which can't continue unless the assertion passes,
915
cases where we should give the user a proper message rather than an
917
* It's not that much shorter than an explicit if/raise.
918
* It tends to lead to fuzzy thinking about whether the check is
919
actually needed or not, and whether it's an internal error or not
920
* It tends to cause look-before-you-leap patterns.
921
* It's unsafe if the check is needed to protect the integrity of the
923
* It tends to give poor messages since the developer can get by with
924
no explanatory text at all.
925
* We can't rely on people always running with -O in normal use, so we
926
can't use it for tests that are actually expensive.
927
* Expensive checks that help developers are better turned on from the
928
test suite or a -D flag.
929
* If used instead of ``self.assert*()`` in tests it makes them falsely pass with -O.
935
When you change bzrlib, please update the relevant documentation for the
936
change you made: Changes to commands should update their help, and
937
possibly end user tutorials; changes to the core library should be
938
reflected in API documentation.
943
If you make a user-visible change, please add a note to the NEWS file.
944
The description should be written to make sense to someone who's just
945
a user of bzr, not a developer: new functions or classes shouldn't be
946
mentioned, but new commands, changes in behaviour or fixed nontrivial
947
bugs should be listed. See the existing entries for an idea of what
950
Within each release, entries in the news file should have the most
951
user-visible changes first. So the order should be approximately:
953
* changes to existing behaviour - the highest priority because the
954
user's existing knowledge is incorrect
955
* new features - should be brought to their attention
956
* bug fixes - may be of interest if the bug was affecting them, and
957
should include the bug number if any
958
* major documentation changes
959
* changes to internal interfaces
961
People who made significant contributions to each change are listed in
962
parenthesis. This can include reporting bugs (particularly with good
963
details or reproduction recipes), submitting patches, etc.
968
The docstring of a command is used by ``bzr help`` to generate help output
969
for the command. The list 'takes_options' attribute on a command is used by
970
``bzr help`` to document the options for the command - the command
971
docstring does not need to document them. Finally, the '_see_also'
972
attribute on a command can be used to reference other related help topics.
977
Functions, methods, classes and modules should have docstrings
978
describing how they are used.
980
The first line of the docstring should be a self-contained sentence.
982
For the special case of Command classes, this acts as the user-visible
983
documentation shown by the help command.
985
The docstrings should be formatted as reStructuredText_ (like this
986
document), suitable for processing using the epydoc_ tool into HTML
989
.. _reStructuredText: http://docutils.sourceforge.net/rst.html
990
.. _epydoc: http://epydoc.sourceforge.net/
999
The copyright policy for bzr was recently made clear in this email (edited
1000
for grammatical correctness)::
1002
The attached patch cleans up the copyright and license statements in
1003
the bzr source. It also adds tests to help us remember to add them
1004
with the correct text.
1006
We had the problem that lots of our files were "Copyright Canonical
1007
Development Ltd" which is not a real company, and some other variations
1008
on this theme. Also, some files were missing the GPL statements.
1010
I want to be clear about the intent of this patch, since copyright can
1011
be a little controversial.
1013
1) The big motivation for this is not to shut out the community, but
1014
just to clean up all of the invalid copyright statements.
1016
2) It has been the general policy for bzr that we want a single
1017
copyright holder for all of the core code. This is following the model
1018
set by the FSF, which makes it easier to update the code to a new
1019
license in case problems are encountered. (For example, if we want to
1020
upgrade the project universally to GPL v3 it is much simpler if there is
1021
a single copyright holder). It also makes it clearer if copyright is
1022
ever debated, there is a single holder, which makes it easier to defend
1023
in court, etc. (I think the FSF position is that if you assign them
1024
copyright, they can defend it in court rather than you needing to, and
1025
I'm sure Canonical would do the same).
1026
As such, Canonical has requested copyright assignments from all of the
1029
3) If someone wants to add code and not attribute it to Canonical, there
1030
is a specific list of files that are excluded from this check. And the
1031
test failure indicates where that is, and how to update it.
1033
4) If anyone feels that I changed a copyright statement incorrectly, just
1034
let me know, and I'll be happy to correct it. Whenever you have large
1035
mechanical changes like this, it is possible to make some mistakes.
1037
Just to reiterate, this is a community project, and it is meant to stay
1038
that way. Core bzr code is copyright Canonical for legal reasons, and
1039
the tests are just there to help us maintain that.
1042
Miscellaneous Topics
1043
####################
1048
Bazaar has a few facilities to help debug problems by going into pdb_, the
1051
.. _pdb: http://docs.python.org/lib/debugger-commands.html
1053
If the ``BZR_PDB`` environment variable is set
1054
then bzr will go into pdb post-mortem mode when an unhandled exception
1057
If you send a SIGQUIT signal to bzr, which can be done by pressing
1058
Ctrl-\\ on Unix, bzr will go into the debugger immediately. You can
1059
continue execution by typing ``c``. This can be disabled if necessary
1060
by setting the environment variable ``BZR_SIGQUIT_PDB=0``.
1066
Bazaar accepts some global options starting with ``-D`` such as
1067
``-Dhpss``. These set a value in `bzrlib.debug.debug_flags`, and
1068
typically cause more information to be written to the trace file. Most
1069
`mutter` calls should be guarded by a check of those flags so that we
1070
don't write out too much information if it's not needed.
1072
Debug flags may have effects other than just emitting trace messages.
1074
Run ``bzr help global-options`` to see them all.
1081
Integer identifier for a revision on the main line of a branch.
1082
Revision 0 is always the null revision; others are 1-based
1083
indexes into the branch's revision history.
1086
Unicode and Encoding Support
1087
============================
1089
This section discusses various techniques that Bazaar uses to handle
1090
characters that are outside the ASCII set.
1095
When a ``Command`` object is created, it is given a member variable
1096
accessible by ``self.outf``. This is a file-like object, which is bound to
1097
``sys.stdout``, and should be used to write information to the screen,
1098
rather than directly writing to ``sys.stdout`` or calling ``print``.
1099
This file has the ability to translate Unicode objects into the correct
1100
representation, based on the console encoding. Also, the class attribute
1101
``encoding_type`` will effect how unprintable characters will be
1102
handled. This parameter can take one of 3 values:
1105
Unprintable characters will be represented with a suitable replacement
1106
marker (typically '?'), and no exception will be raised. This is for
1107
any command which generates text for the user to review, rather than
1108
for automated processing.
1109
For example: ``bzr log`` should not fail if one of the entries has text
1110
that cannot be displayed.
1113
Attempting to print an unprintable character will cause a UnicodeError.
1114
This is for commands that are intended more as scripting support, rather
1115
than plain user review.
1116
For exampl: ``bzr ls`` is designed to be used with shell scripting. One
1117
use would be ``bzr ls --null --unknows | xargs -0 rm``. If ``bzr``
1118
printed a filename with a '?', the wrong file could be deleted. (At the
1119
very least, the correct file would not be deleted). An error is used to
1120
indicate that the requested action could not be performed.
1123
Do not attempt to automatically convert Unicode strings. This is used
1124
for commands that must handle conversion themselves.
1125
For example: ``bzr diff`` needs to translate Unicode paths, but should
1126
not change the exact text of the contents of the files.
1129
``bzrlib.urlutils.unescape_for_display``
1130
----------------------------------------
1132
Because Transports work in URLs (as defined earlier), printing the raw URL
1133
to the user is usually less than optimal. Characters outside the standard
1134
set are printed as escapes, rather than the real character, and local
1135
paths would be printed as ``file://`` urls. The function
1136
``unescape_for_display`` attempts to unescape a URL, such that anything
1137
that cannot be printed in the current encoding stays an escaped URL, but
1138
valid characters are generated where possible.
1144
The ``bzrlib.osutils`` module has many useful helper functions, including
1145
some more portable variants of functions in the standard library.
1147
In particular, don't use ``shutil.rmtree`` unless it's acceptable for it
1148
to fail on Windows if some files are readonly or still open elsewhere.
1149
Use ``bzrlib.osutils.rmtree`` instead.
1155
We write some extensions in C using pyrex. We design these to work in
1158
* User with no C compiler
1159
* User with C compiler
1162
The recommended way to install bzr is to have a C compiler so that the
1163
extensions can be built, but if no C compiler is present, the pure python
1164
versions we supply will work, though more slowly.
1166
For developers we recommend that pyrex be installed, so that the C
1167
extensions can be changed if needed.
1169
For the C extensions, the extension module should always match the
1170
original python one in all respects (modulo speed). This should be
1171
maintained over time.
1173
To create an extension, add rules to setup.py for building it with pyrex,
1174
and with distutils. Now start with an empty .pyx file. At the top add
1175
"include 'yourmodule.py'". This will import the contents of foo.py into this
1176
file at build time - remember that only one module will be loaded at
1177
runtime. Now you can subclass classes, or replace functions, and only your
1178
changes need to be present in the .pyx file.
1180
Note that pyrex does not support all 2.4 programming idioms, so some
1181
syntax changes may be required. I.e.
1183
- 'from foo import (bar, gam)' needs to change to not use the brackets.
1184
- 'import foo.bar as bar' needs to be 'import foo.bar; bar = foo.bar'
1186
If the changes are too dramatic, consider
1187
maintaining the python code twice - once in the .pyx, and once in the .py,
1188
and no longer including the .py file.
1191
Making Installers for OS Windows
1192
================================
1193
To build a win32 installer, see the instructions on the wiki page:
1194
http://bazaar-vcs.org/BzrWin32Installer
1197
Core Developer Tasks
1198
####################
1203
What is a Core Developer?
1204
-------------------------
1206
While everyone in the Bazaar community is welcome and encouraged to
1207
propose and submit changes, a smaller team is reponsible for pulling those
1208
changes together into a cohesive whole. In addition to the general developer
1209
stuff covered above, "core" developers have responsibility for:
1212
* reviewing blueprints
1214
* managing releases (see the `Releasing Bazaar <../../developers/releasing.html>`_)
1217
Removing barriers to community participation is a key reason for adopting
1218
distributed VCS technology. While DVCS removes many technical barriers,
1219
a small number of social barriers are often necessary instead.
1220
By documenting how the above things are done, we hope to
1221
encourage more people to participate in these activities, keeping the
1222
differences between core and non-core contributors to a minimum.
1225
Communicating and Coordinating
1226
------------------------------
1228
While it has many advantages, one of the challenges of distributed
1229
development is keeping everyone else aware of what you're working on.
1230
There are numerous ways to do this:
1232
#. Assign bugs to yourself in Launchpad
1233
#. Mention it on the mailing list
1234
#. Mention it on IRC
1236
As well as the email notifcations that occur when merge requests are sent
1237
and reviewed, you can keep others informed of where you're spending your
1238
energy by emailing the **bazaar-commits** list implicitly. To do this,
1239
install and configure the Email plugin. One way to do this is add these
1240
configuration settings to your central configuration file (e.g.
1241
``~/.bazaar/bazaar.conf`` on Linux)::
1244
email = Joe Smith <joe.smith@internode.on.net>
1245
smtp_server = mail.internode.on.net:25
1247
Then add these lines for the relevant branches in ``locations.conf``::
1249
post_commit_to = bazaar-commits@lists.canonical.com
1250
post_commit_mailer = smtplib
1252
While attending a sprint, RobertCollins' Dbus plugin is useful for the
1253
same reason. See the documentation within the plugin for information on
1254
how to set it up and configure it.
1263
Of the many workflows supported by Bazaar, the one adopted for Bazaar
1264
development itself is known as "Decentralized with automatic gatekeeper".
1265
To repeat the explanation of this given on
1266
http://bazaar-vcs.org/Workflows:
1269
In this workflow, each developer has their own branch or
1270
branches, plus read-only access to the mainline. A software gatekeeper
1271
(e.g. PQM) has commit rights to the main branch. When a developer wants
1272
their work merged, they request the gatekeeper to merge it. The gatekeeper
1273
does a merge, a compile, and runs the test suite. If the code passes, it
1274
is merged into the mainline.
1276
In a nutshell, here's the overall submission process:
1278
#. get your work ready (including review except for trivial changes)
1279
#. push to a public location
1280
#. ask PQM to merge from that location
1283
At present, PQM always takes the changes to merge from a branch
1284
at a URL that can be read by it. For Bazaar, that means a public,
1285
typically http, URL.
1287
As a result, the following things are needed to use PQM for submissions:
1289
#. A publicly available web server
1290
#. Your OpenPGP key registered with PQM (contact RobertCollins for this)
1291
#. The PQM plugin installed and configured (not strictly required but
1292
highly recommended).
1295
Selecting a Public Branch Location
1296
----------------------------------
1298
If you don't have your own web server running, branches can always be
1299
pushed to Launchpad. Here's the process for doing that:
1301
Depending on your location throughout the world and the size of your
1302
repository though, it is often quicker to use an alternative public
1303
location to Launchpad, particularly if you can set up your own repo and
1304
push into that. By using an existing repo, push only needs to send the
1305
changes, instead of the complete repository every time. Note that it is
1306
easy to register branches in other locations with Launchpad so no benefits
1307
are lost by going this way.
1310
For Canonical staff, http://people.ubuntu.com/~<user>/ is one
1311
suggestion for public http branches. Contact your manager for information
1312
on accessing this system if required.
1314
It should also be noted that best practice in this area is subject to
1315
change as things evolve. For example, once the Bazaar smart server on
1316
Launchpad supports server-side branching, the performance situation will
1317
be very different to what it is now (Jun 2007).
1320
Configuring the PQM Plug-In
1321
---------------------------
1323
While not strictly required, the PQM plugin automates a few things and
1324
reduces the chance of error. Before looking at the plugin, it helps to
1325
understand a little more how PQM operates. Basically, PQM requires an
1326
email indicating what you want it to do. The email typically looks like
1329
star-merge source-branch target-branch
1333
star-merge http://bzr.arbash-meinel.com/branches/bzr/jam-integration http://bazaar-vcs.org/bzr/bzr.dev
1335
Note that the command needs to be on one line. The subject of the email
1336
will be used for the commit message. The email also needs to be ``gpg``
1337
signed with a key that PQM accepts.
1339
The advantages of using the PQM plugin are:
1341
#. You can use the config policies to make it easy to set up public
1342
branches, so you don't have to ever type the full paths you want to merge
1345
#. It checks to make sure the public branch last revision matches the
1346
local last revision so you are submitting what you think you are.
1348
#. It uses the same public_branch and smtp sending settings as bzr-email,
1349
so if you have one set up, you have the other mostly set up.
1351
#. Thunderbird refuses to not wrap lines, and request lines are usually
1352
pretty long (you have 2 long URLs in there).
1354
Here are sample configuration settings for the PQM plugin. Here are the
1355
lines in bazaar.conf::
1358
email = Joe Smith <joe.smith@internode.on.net>
1359
smtp_server=mail.internode.on.net:25
1361
And here are the lines in ``locations.conf`` (or ``branch.conf`` for
1362
dirstate-tags branches)::
1364
[/home/joe/bzr/my-integration]
1365
push_location = sftp://joe-smith@bazaar.launchpad.net/%7Ejoe-smith/bzr/my-integration/
1366
push_location:policy = norecurse
1367
public_branch = http://bazaar.launchpad.net/~joe-smith/bzr/my-integration/
1368
public_branch:policy = appendpath
1369
pqm_email = Bazaar PQM <pqm@bazaar-vcs.org>
1370
pqm_branch = http://bazaar-vcs.org/bzr/bzr.dev
1372
Note that the push settings will be added by the first ``push`` on
1373
a branch. Indeed the preferred way to generate the lines above is to use
1374
``push`` with an argument, then copy-and-paste the other lines into
1381
Here is one possible recipe once the above environment is set up:
1383
#. pull bzr.dev => my-integration
1384
#. merge patch => my-integration
1385
#. fix up any final merge conflicts (NEWS being the big killer here).
1391
The ``push`` step is not required if ``my-integration`` is a checkout of
1394
Because of defaults, you can type a single message into commit and
1395
pqm-commit will reuse that.
1398
Tracking Change Acceptance
1399
--------------------------
1401
The web interface to PQM is https://pqm.bazaar-vcs.org/. After submitting
1402
a change, you can visit this URL to confirm it was received and placed in
1405
When PQM completes processing a change, an email is sent to you with the
1409
Reviewing Blueprints
1410
====================
1412
Blueprint Tracking Using Launchpad
1413
----------------------------------
1415
New features typically require a fair amount of discussion, design and
1416
debate. For Bazaar, that information is often captured in a so-called
1417
"blueprint" on our Wiki. Overall tracking of blueprints and their status
1418
is done using Launchpad's relevant tracker,
1419
https://blueprints.launchpad.net/bzr/. Once a blueprint for ready for
1420
review, please announce it on the mailing list.
1422
Alternatively, send an email begining with [RFC] with the proposal to the
1423
list. In some cases, you may wish to attach proposed code or a proposed
1424
developer document if that best communicates the idea. Debate can then
1425
proceed using the normal merge review processes.
1428
Recording Blueprint Review Feedback
1429
-----------------------------------
1431
Unlike its Bug Tracker, Launchpad's Blueprint Tracker doesn't currently
1432
(Jun 2007) support a chronological list of comment responses. Review
1433
feedback can either be recorded on the Wiki hosting the blueprints or by
1434
using Launchpad's whiteboard feature.
1443
As the two senior developers, Martin Pool and Robert Collins coordinate
1444
the overall Bazaar product development roadmap. Core developers provide
1445
input and review into this, particularly during sprints. It's totally
1446
expected that community members ought to be working on things that
1447
interest them the most. The roadmap is valuable though because it provides
1448
context for understanding where the product is going as a whole and why.
1451
Using Releases and Milestones in Launchpad
1452
------------------------------------------
1454
TODO ... (Exact policies still under discussion)
1460
Keeping on top of bugs reported is an important part of ongoing release
1461
planning. Everyone in the community is welcome and encouraged to raise
1462
bugs, confirm bugs raised by others, and nominate a priority. Practically
1463
though, a good percentage of bug triage is often done by the core
1464
developers, partially because of their depth of product knowledge.
1466
With respect to bug triage, core developers are encouraged to play an
1467
active role with particular attention to the following tasks:
1469
* keeping the number of unconfirmed bugs low
1470
* ensuring the priorities are generally right (everything as critical - or
1471
medium - is meaningless)
1472
* looking out for regressions and turning those around sooner rather than later.
1475
As well as prioritizing bugs and nominating them against a
1476
target milestone, Launchpad lets core developers offer to mentor others in
1481
vim: ft=rst tw=74 ai
added
removed
doc/developers/HACKING.txt
