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 best-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.
111
A Closer Look at the Merge & Review Process
112
===========================================
114
If you'd like to propose a change, please post to the
115
bazaar@lists.canonical.com list with a bundle, patch, or link to a
116
branch. Put '[PATCH]' or '[MERGE]' in the subject so Bundle Buggy
117
can pick it out, and explain the change in the email message text.
118
Remember to update the NEWS file as part of your change if it makes any
119
changes visible to users or plugin developers. Please include a diff
120
against mainline if you're giving a link to a branch.
122
You can generate a bundle like this::
124
bzr bundle > mybundle.patch
126
A .patch extension is recommended instead of .bundle as many mail clients
127
will send the latter as a binary file. If a bundle would be too long or your
128
mailer mangles whitespace (e.g. implicitly converts Unix newlines to DOS
129
newlines), use the merge-directive command instead like this::
131
bzr merge-directive http://bazaar-vcs.org http://example.org/my_branch > my_directive.patch
133
See the help for details on the arguments to merge-directive.
135
Please do **NOT** put [PATCH] or [MERGE] in the subject line if you don't
136
want it to be merged. If you want comments from developers rather than
137
to be merged, you can put '[RFC]' in the subject line.
139
Anyone is welcome to review code. There are broadly three gates for
142
* Doesn't reduce test coverage: if it adds new methods or commands,
143
there should be tests for them. There is a good test framework
144
and plenty of examples to crib from, but if you are having trouble
145
working out how to test something feel free to post a draft patch
148
* Doesn't reduce design clarity, such as by entangling objects
149
we're trying to separate. This is mostly something the more
150
experienced reviewers need to help check.
152
* Improves bugs, features, speed, or code simplicity.
154
Code that goes in should pass all three. The core developers take care
155
to keep the code quality high and understandable while recognising that
156
perfect is sometimes the enemy of good. (It is easy for reviews to make
157
people notice other things which should be fixed but those things should
158
not hold up the original fix being accepted. New things can easily be
159
recorded in the Bug Tracker instead.)
161
Anyone can "vote" on the mailing list. Core developers can also vote using
162
Bundle Buggy. Here are the voting codes and their explanations.
164
:approve: Reviewer wants this submission merged.
165
:tweak: Reviewer wants this submission merged with small changes. (No
167
:abstain: Reviewer does not intend to vote on this patch.
168
:resubmit: Please make changes and resubmit for review.
169
:reject: Reviewer doesn't want this kind of change merged.
170
:comment: Not really a vote. Reviewer just wants to comment, for now.
172
If a change gets two approvals from core reviewers, and no rejections,
173
then it's OK to come in. Any of the core developers can bring it into the
174
bzr.dev trunk and backport it to maintenance branches if required. The
175
Release Manager will merge the change into the branch for a pending
176
release, if any. As a guideline, core developers usually merge their own
177
changes and volunteer to merge other contributions if they were the second
178
reviewer to agree to a change.
180
To track the progress of proposed changes, use Bundle Buggy. See
181
http://bundlebuggy.aaronbentley.com/help for a link to all the
182
outstanding merge requests together with an explanation of the columns.
183
Bundle Buggy will also mail you a link to track just your change.
186
Preparing a Sandbox for Making Changes to Bazaar
187
================================================
189
Bazaar supports many ways of organising your work. See
190
http://bazaar-vcs.org/SharedRepositoryLayouts for a summary of the
191
popular alternatives.
193
Of course, the best choice for you will depend on numerous factors:
194
the number of changes you may be making, the complexity of the changes, etc.
195
As a starting suggestion though:
197
* create a local copy of the main development branch (bzr.dev) by using
200
bzr branch http://bazaar-vcs.org/bzr/bzr.dev/ bzr.dev
202
* keep your copy of bzr.dev prestine (by not developing in it) and keep
203
it up to date (by using bzr pull)
205
* create a new branch off your local bzr.dev copy for each issue
206
(bug or feature) you are working on.
208
This approach makes it easy to go back and make any required changes
209
after a code review. Resubmitting the change is then simple with no
210
risk of accidentially including edits related to other issues you may
211
be working on. After the changes for an issue are accepted and merged,
212
the associated branch can be deleted or archived as you wish.
215
Navigating the Code Base
216
========================
218
.. Was at <http://bazaar-vcs.org/NewDeveloperIntroduction>
220
Some of the key files in this directory are:
223
The command you run to start Bazaar itself. This script is pretty
224
short and just does some checks then jumps into bzrlib.
227
This file covers a brief introduction to Bazaar and lists some of its
231
Summary of changes in each Bazaar release that can affect users or
235
Installs Bazaar system-wide or to your home directory. To perform
236
development work on Bazaar it is not required to run this file - you
237
can simply run the bzr command from the top level directory of your
238
development copy. Note: That if you run setup.py this will create a
239
'build' directory in your development branch. There's nothing wrong
240
with this but don't be confused by it. The build process puts a copy
241
of the main code base into this build directory, along with some other
242
files. You don't need to go in here for anything discussed in this
246
Possibly the most exciting folder of all, bzrlib holds the main code
247
base. This is where you will go to edit python files and contribute to
251
Holds documentation on a whole range of things on Bazaar from the
252
origination of ideas within the project to information on Bazaar
253
features and use cases. Within this directory there is a subdirectory
254
for each translation into a human language. All the documentation
255
is in the ReStructuredText markup language.
258
Documentation specifically targetted at Bazaar and plugin developers.
259
(Including this document.)
263
Automatically-generated API reference information is available at
264
<http://starship.python.net/crew/mwh/bzrlibapi/>.
265
(There is an experimental editable version at
266
<http://starship.python.net/crew/mwh/bzrlibapi-oe/>.)
267
See also the `Essential Domain Classes`_
268
section of this guide.
271
Essential Domain Classes
272
########################
274
Introducing the Object Model
275
============================
277
The core domain objects within the bazaar model are:
287
Transports are explained below. See http://bazaar-vcs.org/Classes/
288
for an introduction to the other key classes.
293
The ``Transport`` layer handles access to local or remote directories.
294
Each Transport object acts like a logical connection to a particular
295
directory, and it allows various operations on files within it. You can
296
*clone* a transport to get a new Transport connected to a subdirectory or
299
Transports are not used for access to the working tree. At present
300
working trees are always local and they are accessed through the regular
301
Python file io mechanisms.
306
Transports work in URLs. Take note that URLs are by definition only
307
ASCII - the decision of how to encode a Unicode string into a URL must be
308
taken at a higher level, typically in the Store. (Note that Stores also
309
escape filenames which cannot be safely stored on all filesystems, but
310
this is a different level.)
312
The main reason for this is that it's not possible to safely roundtrip a
313
URL into Unicode and then back into the same URL. The URL standard
314
gives a way to represent non-ASCII bytes in ASCII (as %-escapes), but
315
doesn't say how those bytes represent non-ASCII characters. (They're not
316
guaranteed to be UTF-8 -- that is common but doesn't happen everywhere.)
318
For example if the user enters the url ``http://example/%e0`` there's no
319
way to tell whether that character represents "latin small letter a with
320
grave" in iso-8859-1, or "latin small letter r with acute" in iso-8859-2
321
or malformed UTF-8. So we can't convert their URL to Unicode reliably.
323
Equally problematic if we're given a url-like string containing non-ascii
324
characters (such as the accented a) we can't be sure how to convert that
325
to the correct URL, because we don't know what encoding the server expects
326
for those characters. (Although this is not totally reliable we might still
327
accept these and assume they should be put into UTF-8.)
329
A similar edge case is that the url ``http://foo/sweet%2Fsour`` contains
330
one directory component whose name is "sweet/sour". The escaped slash is
331
not a directory separator. If we try to convert URLs to regular Unicode
332
paths this information will be lost.
334
This implies that Transports must natively deal with URLs; for simplicity
335
they *only* deal with URLs and conversion of other strings to URLs is done
336
elsewhere. Information they return, such as from ``list_dir``, is also in
337
the form of URL components.
340
Coding Style Guidelines
341
#######################
346
``hasattr`` should not be used because it swallows exceptions including
347
``KeyboardInterrupt``. Instead, say something like ::
349
if getattr(thing, 'name', None) is None
355
Please write PEP-8__ compliant code.
357
__ http://www.python.org/peps/pep-0008.html
359
One often-missed requirement is that the first line of docstrings
360
should be a self-contained one-sentence summary.
362
We use 4 space indents for blocks, and never use tab characters. (In vim,
365
Lines should be no more than 79 characters if at all possible.
366
Lines that continue a long statement may be indented in either of
369
within the parenthesis or other character that opens the block, e.g.::
375
or indented by four spaces::
381
The first is considered clearer by some people; however it can be a bit
382
harder to maintain (e.g. when the method name changes), and it does not
383
work well if the relevant parenthesis is already far to the right. Avoid
386
self.legbone.kneebone.shinbone.toebone.shake_it(one,
392
self.legbone.kneebone.shinbone.toebone.shake_it(one,
398
self.legbone.kneebone.shinbone.toebone.shake_it(
401
For long lists, we like to add a trailing comma and put the closing
402
character on the following line. This makes it easier to add new items in
405
from bzrlib.goo import (
411
There should be spaces between function paramaters, but not between the
412
keyword name and the value::
414
call(1, 3, cheese=quark)
418
;(defface my-invalid-face
419
; '((t (:background "Red" :underline t)))
420
; "Face used to highlight invalid constructs or other uglyties"
423
(defun my-python-mode-hook ()
424
;; setup preferred indentation style.
425
(setq fill-column 79)
426
(setq indent-tabs-mode nil) ; no tabs, never, I will not repeat
427
; (font-lock-add-keywords 'python-mode
428
; '(("^\\s *\t" . 'my-invalid-face) ; Leading tabs
429
; ("[ \t]+$" . 'my-invalid-face) ; Trailing spaces
430
; ("^[ \t]+$" . 'my-invalid-face)); Spaces only
434
(add-hook 'python-mode-hook 'my-python-mode-hook)
436
The lines beginning with ';' are comments. They can be activated
437
if one want to have a strong notice of some tab/space usage
444
* Imports should be done at the top-level of the file, unless there is
445
a strong reason to have them lazily loaded when a particular
446
function runs. Import statements have a cost, so try to make sure
447
they don't run inside hot functions.
449
* Module names should always be given fully-qualified,
450
i.e. ``bzrlib.hashcache`` not just ``hashcache``.
456
Functions, methods or members that are "private" to bzrlib are given
457
a leading underscore prefix. Names without a leading underscore are
458
public not just across modules but to programmers using bzrlib as an
459
API. As a consequence, a leading underscore is appropriate for names
460
exposed across modules but that are not to be exposed to bzrlib API
463
We prefer class names to be concatenated capital words (``TestCase``)
464
and variables, methods and functions to be lowercase words joined by
465
underscores (``revision_id``, ``get_revision``).
467
For the purposes of naming some names are treated as single compound
468
words: "filename", "revno".
470
Consider naming classes as nouns and functions/methods as verbs.
472
Try to avoid using abbreviations in names, because there can be
473
inconsistency if other people use the full name.
479
``revision_id`` not ``rev_id`` or ``revid``
481
Functions that transform one thing to another should be named ``x_to_y``
482
(not ``x2y`` as occurs in some old code.)
488
Python destructors (``__del__``) work differently to those of other
489
languages. In particular, bear in mind that destructors may be called
490
immediately when the object apparently becomes unreferenced, or at some
491
later time, or possibly never at all. Therefore we have restrictions on
492
what can be done inside them.
494
0. If you think you need to use a ``__del__`` method ask another
495
developer for alternatives. If you do need to use one, explain
498
1. Never rely on a ``__del__`` method running. If there is code that
499
must run, do it from a ``finally`` block instead.
501
2. Never ``import`` from inside a ``__del__`` method, or you may crash the
504
3. In some places we raise a warning from the destructor if the object
505
has not been cleaned up or closed. This is considered OK: the warning
506
may not catch every case but it's still useful sometimes.
512
In some places we have variables which point to callables that construct
513
new instances. That is to say, they can be used a lot like class objects,
514
but they shouldn't be *named* like classes:
516
> I think that things named FooBar should create instances of FooBar when
517
> called. Its plain confusing for them to do otherwise. When we have
518
> something that is going to be used as a class - that is, checked for via
519
> isinstance or other such idioms, them I would call it foo_class, so that
520
> it is clear that a callable is not sufficient. If it is only used as a
521
> factory, then yes, foo_factory is what I would use.
527
Several places in Bazaar use (or will use) a registry, which is a
528
mapping from names to objects or classes. The registry allows for
529
loading in registered code only when it's needed, and keeping
530
associated information such as a help string or description.
533
InterObject and multiple dispatch
534
=================================
536
The ``InterObject`` provides for two-way `multiple dispatch`__: matching
537
up for example a source and destination repository to find the right way
538
to transfer data between them.
540
.. __: http://en.wikipedia.org/wiki/Multiple_dispatch
542
There is a subclass ``InterObject`` classes for each type of object that is
543
dispatched this way, e.g. ``InterRepository``. Calling ``.get()`` on this
544
class will return an ``InterObject`` instance providing the best match for
545
those parameters, and this instance then has methods for operations
548
inter = InterRepository.get(source_repo, target_repo)
549
inter.fetch(revision_id)
551
``InterRepository`` also acts as a registry-like object for its
552
subclasses, and they can be added through ``.register_optimizer``. The
553
right one to run is selected by asking each class, in reverse order of
554
registration, whether it ``.is_compatible`` with the relevant objects.
559
To make startup time faster, we use the ``bzrlib.lazy_import`` module to
560
delay importing modules until they are actually used. ``lazy_import`` uses
561
the same syntax as regular python imports. So to import a few modules in a
564
from bzrlib.lazy_import import lazy_import
565
lazy_import(globals(), """
574
revision as _mod_revision,
576
import bzrlib.transport
580
At this point, all of these exist as a ``ImportReplacer`` object, ready to
581
be imported once a member is accessed. Also, when importing a module into
582
the local namespace, which is likely to clash with variable names, it is
583
recommended to prefix it as ``_mod_<module>``. This makes it clearer that
584
the variable is a module, and these object should be hidden anyway, since
585
they shouldn't be imported into other namespaces.
587
While it is possible for ``lazy_import()`` to import members of a module
588
when using the ``from module import member`` syntax, it is recommended to
589
only use that syntax to load sub modules ``from module import submodule``.
590
This is because variables and classes can frequently be used without
591
needing a sub-member for example::
593
lazy_import(globals(), """
594
from module import MyClass
598
return isinstance(x, MyClass)
600
This will incorrectly fail, because ``MyClass`` is a ``ImportReplacer``
601
object, rather than the real class.
603
It also is incorrect to assign ``ImportReplacer`` objects to other variables.
604
Because the replacer only knows about the original name, it is unable to
605
replace other variables. The ``ImportReplacer`` class will raise an
606
``IllegalUseOfScopeReplacer`` exception if it can figure out that this
607
happened. But it requires accessing a member more than once from the new
608
variable, so some bugs are not detected right away.
614
The null revision is the ancestor of all revisions. Its revno is 0, its
615
revision-id is ``null:``, and its tree is the empty tree. When referring
616
to the null revision, please use ``bzrlib.revision.NULL_REVISION``. Old
617
code sometimes uses ``None`` for the null revision, but this practice is
621
Object string representations
622
=============================
624
Python prints objects using their ``__repr__`` method when they are
625
written to logs, exception tracebacks, or the debugger. We want
626
objects to have useful representations to help in determining what went
629
If you add a new class you should generally add a ``__repr__`` method
630
unless there is an adequate method in a parent class. There should be a
633
Representations should typically look like Python constructor syntax, but
634
they don't need to include every value in the object and they don't need
635
to be able to actually execute. They're to be read by humans, not
636
machines. Don't hardcode the classname in the format, so that we get the
637
correct value if the method is inherited by a subclass. If you're
638
printing attributes of the object, including strings, you should normally
639
use ``%r`` syntax (to call their repr in turn).
641
Try to avoid the representation becoming more than one or two lines long.
642
(But balance this against including useful information, and simplicity of
645
Because repr methods are often called when something has already gone
646
wrong, they should be written somewhat more defensively than most code.
647
The object may be half-initialized or in some other way in an illegal
648
state. The repr method shouldn't raise an exception, or it may hide the
649
(probably more useful) underlying exception.
654
return '%s(%r)' % (self.__class__.__name__,
661
A bare ``except`` statement will catch all exceptions, including ones that
662
really should terminate the program such as ``MemoryError`` and
663
``KeyboardInterrupt``. They should rarely be used unless the exception is
664
later re-raised. Even then, think about whether catching just
665
``Exception`` (which excludes system errors in Python2.5 and later) would
672
All code should be exercised by the test suite. See `Guide to Testing
673
Bazaar <testing.html>`_ for detailed information about writing tests.
682
We have a commitment to 6 months API stability - any supported symbol in a
683
release of bzr MUST NOT be altered in any way that would result in
684
breaking existing code that uses it. That means that method names,
685
parameter ordering, parameter names, variable and attribute names etc must
686
not be changed without leaving a 'deprecated forwarder' behind. This even
687
applies to modules and classes.
689
If you wish to change the behaviour of a supported API in an incompatible
690
way, you need to change its name as well. For instance, if I add an optional keyword
691
parameter to branch.commit - that's fine. On the other hand, if I add a
692
keyword parameter to branch.commit which is a *required* transaction
693
object, I should rename the API - i.e. to 'branch.commit_transaction'.
695
When renaming such supported API's, be sure to leave a deprecated_method (or
696
_function or ...) behind which forwards to the new API. See the
697
bzrlib.symbol_versioning module for decorators that take care of the
698
details for you - such as updating the docstring, and issuing a warning
699
when the old api is used.
701
For unsupported API's, it does not hurt to follow this discipline, but it's
702
not required. Minimally though, please try to rename things so that
703
callers will at least get an AttributeError rather than weird results.
706
Deprecation decorators
707
----------------------
709
``bzrlib.symbol_versioning`` provides decorators that can be attached to
710
methods, functions, and other interfaces to indicate that they should no
711
longer be used. For example::
713
@deprecated_method(deprecated_in((0, 1, 4)))
715
return self._new_foo()
717
To deprecate a static method you must call ``deprecated_function``
718
(**not** method), after the staticmethod call::
721
@deprecated_function(deprecated_in((0, 1, 4)))
722
def create_repository(base, shared=False, format=None):
724
When you deprecate an API, you should not just delete its tests, because
725
then we might introduce bugs in them. If the API is still present at all,
726
it should still work. The basic approach is to use
727
``TestCase.applyDeprecated`` which in one step checks that the API gives
728
the expected deprecation message, and also returns the real result from
729
the method, so that tests can keep running.
731
Deprecation warnings will be suppressed for final releases, but not for
732
development versions or release candidates, or when running ``bzr
733
selftest``. This gives developers information about whether their code is
734
using deprecated functions, but avoids confusing users about things they
741
Processing Command Lines
742
------------------------
744
bzrlib has a standard framework for parsing command lines and calling
745
processing routines associated with various commands. See builtins.py
746
for numerous examples.
749
Standard Parameter Types
750
------------------------
752
There are some common requirements in the library: some parameters need to be
753
unicode safe, some need byte strings, and so on. At the moment we have
754
only codified one specific pattern: Parameters that need to be unicode
755
should be checked via ``bzrlib.osutils.safe_unicode``. This will coerce the
756
input into unicode in a consistent fashion, allowing trivial strings to be
757
used for programmer convenience, but not performing unpredictably in the
758
presence of different locales.
764
(The strategy described here is what we want to get to, but it's not
765
consistently followed in the code at the moment.)
767
bzrlib is intended to be a generically reusable library. It shouldn't
768
write messages to stdout or stderr, because some programs that use it
769
might want to display that information through a GUI or some other
772
We can distinguish two types of output from the library:
774
1. Structured data representing the progress or result of an
775
operation. For example, for a commit command this will be a list
776
of the modified files and the finally committed revision number
779
These should be exposed either through the return code or by calls
780
to a callback parameter.
782
A special case of this is progress indicators for long-lived
783
operations, where the caller should pass a ProgressBar object.
785
2. Unstructured log/debug messages, mostly for the benefit of the
786
developers or users trying to debug problems. This should always
787
be sent through ``bzrlib.trace`` and Python ``logging``, so that
788
it can be redirected by the client.
790
The distinction between the two is a bit subjective, but in general if
791
there is any chance that a library would want to see something as
792
structured data, we should make it so.
794
The policy about how output is presented in the text-mode client
795
should be only in the command-line tool.
802
Bazaar has online help for various topics through ``bzr help COMMAND`` or
803
equivalently ``bzr command -h``. We also have help on command options,
804
and on other help topics. (See ``help_topics.py``.)
806
As for python docstrings, the first paragraph should be a single-sentence
807
synopsis of the command.
809
The help for options should be one or more proper sentences, starting with
810
a capital letter and finishing with a full stop (period).
812
All help messages and documentation should have two spaces between
816
Handling Errors and Exceptions
817
==============================
819
Commands should return non-zero when they encounter circumstances that
820
the user should really pay attention to - which includes trivial shell
823
Recommended values are:
826
1. Conflicts in merge-like operations, or changes are present in
827
diff-like operations.
828
2. Unrepresentable diff changes (i.e. binary files that we cannot show
830
3. An error or exception has occurred.
831
4. An internal error occurred (one that shows a traceback.)
833
Errors are handled through Python exceptions. Exceptions should be defined
834
inside bzrlib.errors, so that we can see the whole tree at a glance.
836
We broadly classify errors as either being either internal or not,
837
depending on whether ``internal_error`` is set or not. If we think it's our
838
fault, we show a backtrace, an invitation to report the bug, and possibly
839
other details. This is the default for errors that aren't specifically
840
recognized as being caused by a user error. Otherwise we show a briefer
841
message, unless -Derror was given.
843
Many errors originate as "environmental errors" which are raised by Python
844
or builtin libraries -- for example IOError. These are treated as being
845
our fault, unless they're caught in a particular tight scope where we know
846
that they indicate a user errors. For example if the repository format
847
is not found, the user probably gave the wrong path or URL. But if one of
848
the files inside the repository is not found, then it's our fault --
849
either there's a bug in bzr, or something complicated has gone wrong in
850
the environment that means one internal file was deleted.
852
Many errors are defined in ``bzrlib/errors.py`` but it's OK for new errors
853
to be added near the place where they are used.
855
Exceptions are formatted for the user by conversion to a string
856
(eventually calling their ``__str__`` method.) As a convenience the
857
``._fmt`` member can be used as a template which will be mapped to the
858
error's instance dict.
860
New exception classes should be defined when callers might want to catch
861
that exception specifically, or when it needs a substantially different
864
Exception strings should start with a capital letter and should not have a
865
final fullstop. If long, they may contain newlines to break the text.
871
Do not use the Python ``assert`` statement, either in tests or elsewhere.
872
A source test checks that it is not used. It is ok to explicitly raise
877
* It makes the behaviour vary depending on whether bzr is run with -O
878
or not, therefore giving a chance for bugs that occur in one case or
879
the other, several of which have already occurred: assertions with
880
side effects, code which can't continue unless the assertion passes,
881
cases where we should give the user a proper message rather than an
883
* It's not that much shorter than an explicit if/raise.
884
* It tends to lead to fuzzy thinking about whether the check is
885
actually needed or not, and whether it's an internal error or not
886
* It tends to cause look-before-you-leap patterns.
887
* It's unsafe if the check is needed to protect the integrity of the
889
* It tends to give poor messages since the developer can get by with
890
no explanatory text at all.
891
* We can't rely on people always running with -O in normal use, so we
892
can't use it for tests that are actually expensive.
893
* Expensive checks that help developers are better turned on from the
894
test suite or a -D flag.
895
* If used instead of ``self.assert*()`` in tests it makes them falsely pass with -O.
901
When you change bzrlib, please update the relevant documentation for the
902
change you made: Changes to commands should update their help, and
903
possibly end user tutorials; changes to the core library should be
904
reflected in API documentation.
909
If you make a user-visible change, please add a note to the NEWS file.
910
The description should be written to make sense to someone who's just
911
a user of bzr, not a developer: new functions or classes shouldn't be
912
mentioned, but new commands, changes in behaviour or fixed nontrivial
913
bugs should be listed. See the existing entries for an idea of what
916
Within each release, entries in the news file should have the most
917
user-visible changes first. So the order should be approximately:
919
* changes to existing behaviour - the highest priority because the
920
user's existing knowledge is incorrect
921
* new features - should be brought to their attention
922
* bug fixes - may be of interest if the bug was affecting them, and
923
should include the bug number if any
924
* major documentation changes
925
* changes to internal interfaces
927
People who made significant contributions to each change are listed in
928
parenthesis. This can include reporting bugs (particularly with good
929
details or reproduction recipes), submitting patches, etc.
934
The docstring of a command is used by ``bzr help`` to generate help output
935
for the command. The list 'takes_options' attribute on a command is used by
936
``bzr help`` to document the options for the command - the command
937
docstring does not need to document them. Finally, the '_see_also'
938
attribute on a command can be used to reference other related help topics.
943
Functions, methods, classes and modules should have docstrings
944
describing how they are used.
946
The first line of the docstring should be a self-contained sentence.
948
For the special case of Command classes, this acts as the user-visible
949
documentation shown by the help command.
951
The docstrings should be formatted as reStructuredText_ (like this
952
document), suitable for processing using the epydoc_ tool into HTML
955
.. _reStructuredText: http://docutils.sourceforge.net/rst.html
956
.. _epydoc: http://epydoc.sourceforge.net/
965
The copyright policy for bzr was recently made clear in this email (edited
966
for grammatical correctness)::
968
The attached patch cleans up the copyright and license statements in
969
the bzr source. It also adds tests to help us remember to add them
970
with the correct text.
972
We had the problem that lots of our files were "Copyright Canonical
973
Development Ltd" which is not a real company, and some other variations
974
on this theme. Also, some files were missing the GPL statements.
976
I want to be clear about the intent of this patch, since copyright can
977
be a little controversial.
979
1) The big motivation for this is not to shut out the community, but
980
just to clean up all of the invalid copyright statements.
982
2) It has been the general policy for bzr that we want a single
983
copyright holder for all of the core code. This is following the model
984
set by the FSF, which makes it easier to update the code to a new
985
license in case problems are encountered. (For example, if we want to
986
upgrade the project universally to GPL v3 it is much simpler if there is
987
a single copyright holder). It also makes it clearer if copyright is
988
ever debated, there is a single holder, which makes it easier to defend
989
in court, etc. (I think the FSF position is that if you assign them
990
copyright, they can defend it in court rather than you needing to, and
991
I'm sure Canonical would do the same).
992
As such, Canonical has requested copyright assignments from all of the
995
3) If someone wants to add code and not attribute it to Canonical, there
996
is a specific list of files that are excluded from this check. And the
997
test failure indicates where that is, and how to update it.
999
4) If anyone feels that I changed a copyright statement incorrectly, just
1000
let me know, and I'll be happy to correct it. Whenever you have large
1001
mechanical changes like this, it is possible to make some mistakes.
1003
Just to reiterate, this is a community project, and it is meant to stay
1004
that way. Core bzr code is copyright Canonical for legal reasons, and
1005
the tests are just there to help us maintain that.
1008
Miscellaneous Topics
1009
####################
1014
Bazaar has a few facilities to help debug problems by going into pdb_, the
1017
.. _pdb: http://docs.python.org/lib/debugger-commands.html
1019
If the ``BZR_PDB`` environment variable is set
1020
then bzr will go into pdb post-mortem mode when an unhandled exception
1023
If you send a SIGQUIT signal to bzr, which can be done by pressing
1024
Ctrl-\\ on Unix, bzr will go into the debugger immediately. You can
1025
continue execution by typing ``c``. This can be disabled if necessary
1026
by setting the environment variable ``BZR_SIGQUIT_PDB=0``.
1033
Integer identifier for a revision on the main line of a branch.
1034
Revision 0 is always the null revision; others are 1-based
1035
indexes into the branch's revision history.
1038
Unicode and Encoding Support
1039
============================
1041
This section discusses various techniques that Bazaar uses to handle
1042
characters that are outside the ASCII set.
1047
When a ``Command`` object is created, it is given a member variable
1048
accessible by ``self.outf``. This is a file-like object, which is bound to
1049
``sys.stdout``, and should be used to write information to the screen,
1050
rather than directly writing to ``sys.stdout`` or calling ``print``.
1051
This file has the ability to translate Unicode objects into the correct
1052
representation, based on the console encoding. Also, the class attribute
1053
``encoding_type`` will effect how unprintable characters will be
1054
handled. This parameter can take one of 3 values:
1057
Unprintable characters will be represented with a suitable replacement
1058
marker (typically '?'), and no exception will be raised. This is for
1059
any command which generates text for the user to review, rather than
1060
for automated processing.
1061
For example: ``bzr log`` should not fail if one of the entries has text
1062
that cannot be displayed.
1065
Attempting to print an unprintable character will cause a UnicodeError.
1066
This is for commands that are intended more as scripting support, rather
1067
than plain user review.
1068
For exampl: ``bzr ls`` is designed to be used with shell scripting. One
1069
use would be ``bzr ls --null --unknows | xargs -0 rm``. If ``bzr``
1070
printed a filename with a '?', the wrong file could be deleted. (At the
1071
very least, the correct file would not be deleted). An error is used to
1072
indicate that the requested action could not be performed.
1075
Do not attempt to automatically convert Unicode strings. This is used
1076
for commands that must handle conversion themselves.
1077
For example: ``bzr diff`` needs to translate Unicode paths, but should
1078
not change the exact text of the contents of the files.
1081
``bzrlib.urlutils.unescape_for_display``
1082
----------------------------------------
1084
Because Transports work in URLs (as defined earlier), printing the raw URL
1085
to the user is usually less than optimal. Characters outside the standard
1086
set are printed as escapes, rather than the real character, and local
1087
paths would be printed as ``file://`` urls. The function
1088
``unescape_for_display`` attempts to unescape a URL, such that anything
1089
that cannot be printed in the current encoding stays an escaped URL, but
1090
valid characters are generated where possible.
1096
The ``bzrlib.osutils`` module has many useful helper functions, including
1097
some more portable variants of functions in the standard library.
1099
In particular, don't use ``shutil.rmtree`` unless it's acceptable for it
1100
to fail on Windows if some files are readonly or still open elsewhere.
1101
Use ``bzrlib.osutils.rmtree`` instead.
1107
We write some extensions in C using pyrex. We design these to work in
1110
* User with no C compiler
1111
* User with C compiler
1114
The recommended way to install bzr is to have a C compiler so that the
1115
extensions can be built, but if no C compiler is present, the pure python
1116
versions we supply will work, though more slowly.
1118
For developers we recommend that pyrex be installed, so that the C
1119
extensions can be changed if needed.
1121
For the C extensions, the extension module should always match the
1122
original python one in all respects (modulo speed). This should be
1123
maintained over time.
1125
To create an extension, add rules to setup.py for building it with pyrex,
1126
and with distutils. Now start with an empty .pyx file. At the top add
1127
"include 'yourmodule.py'". This will import the contents of foo.py into this
1128
file at build time - remember that only one module will be loaded at
1129
runtime. Now you can subclass classes, or replace functions, and only your
1130
changes need to be present in the .pyx file.
1132
Note that pyrex does not support all 2.4 programming idioms, so some
1133
syntax changes may be required. I.e.
1135
- 'from foo import (bar, gam)' needs to change to not use the brackets.
1136
- 'import foo.bar as bar' needs to be 'import foo.bar; bar = foo.bar'
1138
If the changes are too dramatic, consider
1139
maintaining the python code twice - once in the .pyx, and once in the .py,
1140
and no longer including the .py file.
1143
Making Installers for OS Windows
1144
================================
1145
To build a win32 installer, see the instructions on the wiki page:
1146
http://bazaar-vcs.org/BzrWin32Installer
1149
Core Developer Tasks
1150
####################
1155
What is a Core Developer?
1156
-------------------------
1158
While everyone in the Bazaar community is welcome and encouraged to
1159
propose and submit changes, a smaller team is reponsible for pulling those
1160
changes together into a cohesive whole. In addition to the general developer
1161
stuff covered above, "core" developers have responsibility for:
1164
* reviewing blueprints
1166
* managing releases (see the `Releasing Bazaar <../../developers/releasing.html>`_)
1169
Removing barriers to community participation is a key reason for adopting
1170
distributed VCS technology. While DVCS removes many technical barriers,
1171
a small number of social barriers are often necessary instead.
1172
By documenting how the above things are done, we hope to
1173
encourage more people to participate in these activities, keeping the
1174
differences between core and non-core contributors to a minimum.
1177
Communicating and Coordinating
1178
------------------------------
1180
While it has many advantages, one of the challenges of distributed
1181
development is keeping everyone else aware of what you're working on.
1182
There are numerous ways to do this:
1184
#. Assign bugs to yourself in Launchpad
1185
#. Mention it on the mailing list
1186
#. Mention it on IRC
1188
As well as the email notifcations that occur when merge requests are sent
1189
and reviewed, you can keep others informed of where you're spending your
1190
energy by emailing the **bazaar-commits** list implicitly. To do this,
1191
install and configure the Email plugin. One way to do this is add these
1192
configuration settings to your central configuration file (e.g.
1193
``~/.bazaar/bazaar.conf`` on Linux)::
1196
email = Joe Smith <joe.smith@internode.on.net>
1197
smtp_server = mail.internode.on.net:25
1199
Then add these lines for the relevant branches in ``locations.conf``::
1201
post_commit_to = bazaar-commits@lists.canonical.com
1202
post_commit_mailer = smtplib
1204
While attending a sprint, RobertCollins' Dbus plugin is useful for the
1205
same reason. See the documentation within the plugin for information on
1206
how to set it up and configure it.
1212
Setting Up Your Workspace for Reviews
1213
-------------------------------------
1215
TODO: Incorporate John Arbash Meinel's detailed email to Ian C on the
1216
numerous ways of setting up integration branches.
1219
The Review Checklist
1220
--------------------
1222
See `A Closer Look at the Merge & Review Process`_
1223
for information on the gates used to decide whether code can be merged
1224
or not and details on how review results are recorded and communicated.
1227
The Importance of Timely Reviews
1228
--------------------------------
1230
Good reviews do take time. They also regularly require a solid
1231
understanding of the overall code base. In practice, this means a small
1232
number of people often have a large review burden - with knowledge comes
1233
responsibility. No one like their merge requests sitting in a queue going
1234
nowhere, so reviewing sooner rather than later is strongly encouraged.
1243
Of the many workflows supported by Bazaar, the one adopted for Bazaar
1244
development itself is known as "Decentralized with automatic gatekeeper".
1245
To repeat the explanation of this given on
1246
http://bazaar-vcs.org/Workflows:
1249
In this workflow, each developer has their own branch or
1250
branches, plus read-only access to the mainline. A software gatekeeper
1251
(e.g. PQM) has commit rights to the main branch. When a developer wants
1252
their work merged, they request the gatekeeper to merge it. The gatekeeper
1253
does a merge, a compile, and runs the test suite. If the code passes, it
1254
is merged into the mainline.
1256
In a nutshell, here's the overall submission process:
1258
#. get your work ready (including review except for trivial changes)
1259
#. push to a public location
1260
#. ask PQM to merge from that location
1263
At present, PQM always takes the changes to merge from a branch
1264
at a URL that can be read by it. For Bazaar, that means a public,
1265
typically http, URL.
1267
As a result, the following things are needed to use PQM for submissions:
1269
#. A publicly available web server
1270
#. Your OpenPGP key registered with PQM (contact RobertCollins for this)
1271
#. The PQM plugin installed and configured (not strictly required but
1272
highly recommended).
1275
Selecting a Public Branch Location
1276
----------------------------------
1278
If you don't have your own web server running, branches can always be
1279
pushed to Launchpad. Here's the process for doing that:
1281
Depending on your location throughout the world and the size of your
1282
repository though, it is often quicker to use an alternative public
1283
location to Launchpad, particularly if you can set up your own repo and
1284
push into that. By using an existing repo, push only needs to send the
1285
changes, instead of the complete repository every time. Note that it is
1286
easy to register branches in other locations with Launchpad so no benefits
1287
are lost by going this way.
1290
For Canonical staff, http://people.ubuntu.com/~<user>/ is one
1291
suggestion for public http branches. Contact your manager for information
1292
on accessing this system if required.
1294
It should also be noted that best practice in this area is subject to
1295
change as things evolve. For example, once the Bazaar smart server on
1296
Launchpad supports server-side branching, the performance situation will
1297
be very different to what it is now (Jun 2007).
1300
Configuring the PQM Plug-In
1301
---------------------------
1303
While not strictly required, the PQM plugin automates a few things and
1304
reduces the chance of error. Before looking at the plugin, it helps to
1305
understand a little more how PQM operates. Basically, PQM requires an
1306
email indicating what you want it to do. The email typically looks like
1309
star-merge source-branch target-branch
1313
star-merge http://bzr.arbash-meinel.com/branches/bzr/jam-integration http://bazaar-vcs.org/bzr/bzr.dev
1315
Note that the command needs to be on one line. The subject of the email
1316
will be used for the commit message. The email also needs to be ``gpg``
1317
signed with a key that PQM accepts.
1319
The advantages of using the PQM plugin are:
1321
#. You can use the config policies to make it easy to set up public
1322
branches, so you don't have to ever type the full paths you want to merge
1325
#. It checks to make sure the public branch last revision matches the
1326
local last revision so you are submitting what you think you are.
1328
#. It uses the same public_branch and smtp sending settings as bzr-email,
1329
so if you have one set up, you have the other mostly set up.
1331
#. Thunderbird refuses to not wrap lines, and request lines are usually
1332
pretty long (you have 2 long URLs in there).
1334
Here are sample configuration settings for the PQM plugin. Here are the
1335
lines in bazaar.conf::
1338
email = Joe Smith <joe.smith@internode.on.net>
1339
smtp_server=mail.internode.on.net:25
1341
And here are the lines in ``locations.conf`` (or ``branch.conf`` for
1342
dirstate-tags branches)::
1344
[/home/joe/bzr/my-integration]
1345
push_location = sftp://joe-smith@bazaar.launchpad.net/%7Ejoe-smith/bzr/my-integration/
1346
push_location:policy = norecurse
1347
public_branch = http://bazaar.launchpad.net/~joe-smith/bzr/my-integration/
1348
public_branch:policy = appendpath
1349
pqm_email = Bazaar PQM <pqm@bazaar-vcs.org>
1350
pqm_branch = http://bazaar-vcs.org/bzr/bzr.dev
1352
Note that the push settings will be added by the first ``push`` on
1353
a branch. Indeed the preferred way to generate the lines above is to use
1354
``push`` with an argument, then copy-and-paste the other lines into
1361
Here is one possible recipe once the above environment is set up:
1363
#. pull bzr.dev => my-integration
1364
#. merge patch => my-integration
1365
#. fix up any final merge conflicts (NEWS being the big killer here).
1371
The ``push`` step is not required if ``my-integration`` is a checkout of
1374
Because of defaults, you can type a single message into commit and
1375
pqm-commit will reuse that.
1378
Tracking Change Acceptance
1379
--------------------------
1381
The web interface to PQM is https://pqm.bazaar-vcs.org/. After submitting
1382
a change, you can visit this URL to confirm it was received and placed in
1385
When PQM completes processing a change, an email is sent to you with the
1389
Reviewing Blueprints
1390
====================
1392
Blueprint Tracking Using Launchpad
1393
----------------------------------
1395
New features typically require a fair amount of discussion, design and
1396
debate. For Bazaar, that information is often captured in a so-called
1397
"blueprint" on our Wiki. Overall tracking of blueprints and their status
1398
is done using Launchpad's relevant tracker,
1399
https://blueprints.launchpad.net/bzr/. Once a blueprint for ready for
1400
review, please announce it on the mailing list.
1402
Alternatively, send an email begining with [RFC] with the proposal to the
1403
list. In some cases, you may wish to attach proposed code or a proposed
1404
developer document if that best communicates the idea. Debate can then
1405
proceed using the normal merge review processes.
1408
Recording Blueprint Review Feedback
1409
-----------------------------------
1411
Unlike its Bug Tracker, Launchpad's Blueprint Tracker doesn't currently
1412
(Jun 2007) support a chronological list of comment responses. Review
1413
feedback can either be recorded on the Wiki hosting the blueprints or by
1414
using Launchpad's whiteboard feature.
1423
As the two senior developers, Martin Pool and Robert Collins coordinate
1424
the overall Bazaar product development roadmap. Core developers provide
1425
input and review into this, particularly during sprints. It's totally
1426
expected that community members ought to be working on things that
1427
interest them the most. The roadmap is valuable though because it provides
1428
context for understanding where the product is going as a whole and why.
1431
Using Releases and Milestones in Launchpad
1432
------------------------------------------
1434
TODO ... (Exact policies still under discussion)
1440
Keeping on top of bugs reported is an important part of ongoing release
1441
planning. Everyone in the community is welcome and encouraged to raise
1442
bugs, confirm bugs raised by others, and nominate a priority. Practically
1443
though, a good percentage of bug triage is often done by the core
1444
developers, partially because of their depth of product knowledge.
1446
With respect to bug triage, core developers are encouraged to play an
1447
active role with particular attention to the following tasks:
1449
* keeping the number of unconfirmed bugs low
1450
* ensuring the priorities are generally right (everything as critical - or
1451
medium - is meaningless)
1452
* looking out for regressions and turning those around sooner rather than later.
1455
As well as prioritizing bugs and nominating them against a
1456
target milestone, Launchpad lets core developers offer to mentor others in
1461
vim: ft=rst tw=74 ai
added
removed
doc/developers/HACKING.txt
