7
(The current version of this document is available in the file ``HACKING``
8
in the source tree, or at
9
http://doc.bazaar-vcs.org/bzr.dev/developers/HACKING.htm)
15
Exploring the Bazaar Platform
16
=============================
18
Before making changes, it's a good idea to explore the work already
19
done by others. Perhaps the new feature or improvement you're looking
20
for is available in another plug-in already? If you find a bug,
21
perhaps someone else has already fixed it?
23
To answer these questions and more, take a moment to explore the
24
overall Bazaar Platform. Here are some links to browse:
26
* The Plugins page on the Wiki - http://bazaar-vcs.org/BzrPlugins
28
* The Bazaar product family on Launchpad - https://launchpad.net/bazaar
30
* Bug Tracker for the core product - https://bugs.launchpad.net/bzr/
32
* Blueprint Tracker for the core product - https://blueprints.launchpad.net/bzr/
34
If nothing else, perhaps you'll find inspiration in how other developers
35
have solved their challenges.
38
Planning and Discussing Changes
39
===============================
41
There is a very active community around Bazaar. Mostly we meet on IRC
42
(#bzr on irc.freenode.net) and on the mailing list. To join the Bazaar
43
community, see http://bazaar-vcs.org/BzrSupport.
45
If you are planning to make a change, it's a very good idea to mention it
46
on the IRC channel and/or on the mailing list. There are many advantages
47
to involving the community before you spend much time on a change.
50
* you get to build on the wisdom on others, saving time
52
* if others can direct you to similar code, it minimises the work to be done
54
* it assists everyone in coordinating direction, priorities and effort.
56
In summary, maximising the input from others typically minimises the
57
total effort required to get your changes merged. The community is
58
friendly, helpful and always keen to welcome newcomers.
61
Bazaar Development in a Nutshell
62
================================
64
Looking for a 10 minute introduction to submitting a change?
65
See http://bazaar-vcs.org/BzrGivingBack.
67
TODO: Merge that Wiki page into this document.
70
Understanding the Development Process
71
=====================================
73
The development team follows many best-practices including:
75
* a public roadmap and planning process in which anyone can participate
77
* time based milestones everyone can work towards and plan around
79
* extensive code review and feedback to contributors
81
* complete and rigorous test coverage on any code contributed
83
* automated validation that all tests still pass before code is merged
84
into the main code branch.
86
The key tools we use to enable these practices are:
88
* Launchpad - https://launchpad.net/
90
* Bazaar - http://bazaar-vcs.org/
92
* Bundle Buggy - http://bundlebuggy.aaronbentley.com/
94
* Patch Queue Manager - https://launchpad.net/pqm/
96
For further information, see http://bazaar-vcs.org/BzrDevelopment.
99
A Closer Look at the Merge & Review Process
100
===========================================
102
If you'd like to propose a change, please post to the
103
bazaar@lists.canonical.com list with a bundle, patch, or link to a
104
branch. Put '[PATCH]' or '[MERGE]' in the subject so Bundle Buggy
105
can pick it out, and explain the change in the email message text.
106
Remember to update the NEWS file as part of your change if it makes any
107
changes visible to users or plugin developers. Please include a diff
108
against mainline if you're giving a link to a branch.
110
You can generate a bundle like this::
112
bzr bundle > mybundle.patch
114
A .patch extension is recommended instead of .bundle as many mail clients
115
will send the latter as a binary file. If a bundle would be too long or your
116
mailer mangles whitespace (e.g. implicitly converts Unix newlines to DOS
117
newlines), use the merge-directive command instead like this::
119
bzr merge-directive http://bazaar-vcs.org http://example.org/my_branch > my_directive.patch
121
See the help for details on the arguments to merge-directive.
123
Please do **NOT** put [PATCH] or [MERGE] in the subject line if you don't
124
want it to be merged. If you want comments from developers rather than
125
to be merged, you can put '[RFC]' in the subject line.
127
Anyone is welcome to review code. There are broadly three gates for
130
* Doesn't reduce test coverage: if it adds new methods or commands,
131
there should be tests for them. There is a good test framework
132
and plenty of examples to crib from, but if you are having trouble
133
working out how to test something feel free to post a draft patch
136
* Doesn't reduce design clarity, such as by entangling objects
137
we're trying to separate. This is mostly something the more
138
experienced reviewers need to help check.
140
* Improves bugs, features, speed, or code simplicity.
142
Code that goes in should pass all three. The core developers take care
143
to keep the code quality high and understandable while recognising that
144
perfect is sometimes the enemy of good. (It is easy for reviews to make
145
people notice other things which should be fixed but those things should
146
not hold up the original fix being accepted. New things can easily be
147
recorded in the Bug Tracker instead.)
149
Anyone can "vote" on the mailing list. Core developers can also vote using
150
Bundle Buggy. Here are the voting codes and their explanations.
152
-1 really don't want it in current form
153
-0 somewhat uncomfortable
154
+0 comfortable but resubmission after changes requested
155
+1 conditional good to go after some minor changes
158
+1 conditional is used as a way to avoid another submit/review cycle for
159
patches that need small changes.
161
If a change gets two +1 votes from core reviewers, and no
162
vetos, then it's OK to come in. Any of the core developers can bring it
163
into the bzr.dev trunk and backport it to maintenance branches if required.
164
The Release Manager will merge the change into the branch for a pending
165
release, if any. As a guideline, core developers usually merge their own
166
changes and volunteer to merge other contributions if they were the second
167
reviewer to agree to a change.
169
To track the progress of proposed changes, use Bundle Buggy. See
170
http://bundlebuggy.aaronbentley.com/help for a link to all the
171
outstanding merge requests together with an explanation of the columns.
172
Bundle Buggy will also mail you a link to track just your change.
175
Preparing a Sandbox for Making Changes to Bazaar
176
================================================
178
Bazaar supports many ways of organising your work. See
179
http://bazaar-vcs.org/SharedRepositoryLayouts for a summary of the
180
popular alternatives.
182
Of course, the best choice for you will depend on numerous factors:
183
the number of changes you may be making, the complexity of the changes, etc.
184
As a starting suggestion though:
186
* create a local copy of the main development branch (bzr.dev) by using
189
bzr branch http://bazaar-vcs.org/bzr/bzr.dev/ bzr.dev
191
* keep your copy of bzr.dev prestine (by not developing in it) and keep
192
it up to date (by using bzr pull)
194
* create a new branch off your local bzr.dev copy for each issue
195
(bug or feature) you are working on.
197
This approach makes it easy to go back and make any required changes
198
after a code review. Resubmitting the change is then simple with no
199
risk of accidentially including edits related to other issues you may
200
be working on. After the changes for an issue are accepted and merged,
201
the associated branch can be deleted or archived as you wish.
204
Navigating the Code Base
205
========================
207
TODO: List and describe in one line the purpose of each directory
208
inside an installation of bzr.
210
TODO: Refer to a central location holding an up to date copy of the API
211
documentation generated by epydoc, e.g. something like
212
http://starship.python.net/crew/mwh/bzrlibapi/bzrlib.html.
218
The Importance of Testing
219
=========================
221
Reliability is a critical success factor for any Version Control System.
222
We want Bazaar to be highly reliable across multiple platforms while
223
evolving over time to meet the needs of its community.
225
In a nutshell, this is want we expect and encourage:
227
* New functionality should have test cases. Preferably write the
228
test before writing the code.
230
In general, you can test at either the command-line level or the
231
internal API level. See Writing tests below for more detail.
233
* Try to practice Test-Driven Development: before fixing a bug, write a
234
test case so that it does not regress. Similarly for adding a new
235
feature: write a test case for a small version of the new feature before
236
starting on the code itself. Check the test fails on the old code, then
237
add the feature or fix and check it passes.
239
By doing these things, the Bazaar team gets increased confidence that
240
changes do what they claim to do, whether provided by the core team or
241
by community members. Equally importantly, we can be surer that changes
242
down the track do not break new features or bug fixes that you are
245
As of May 2007, Bazaar ships with a test suite containing over 6000 tests
246
and growing. We are proud of it and want to remain so. As community
247
members, we all benefit from it. Would you trust version control on
248
your project to a product *without* a test suite like Bazaar has?
251
Running the Test Suite
252
======================
254
Currently, bzr selftest is used to invoke tests.
255
You can provide a pattern argument to run a subset. For example,
256
to run just the blackbox tests, run::
258
./bzr selftest -v blackbox
260
To skip a particular test (or set of tests), use the --exclude option
261
(shorthand -x) like so::
263
./bzr selftest -v -x blackbox
265
To list tests without running them, use the --list-only option like so::
267
./bzr selftest --list-only
269
This option can be combined with other selftest options (like -x) and
270
filter patterns to understand their effect.
276
In general tests should be placed in a file named test_FOO.py where
277
FOO is the logical thing under test. That file should be placed in the
278
tests subdirectory under the package being tested.
280
For example, tests for merge3 in bzrlib belong in bzrlib/tests/test_merge3.py.
281
See bzrlib/tests/test_sampler.py for a template test script.
283
Tests can be written for the UI or for individual areas of the library.
284
Choose whichever is appropriate: if adding a new command, or a new command
285
option, then you should be writing a UI test. If you are both adding UI
286
functionality and library functionality, you will want to write tests for
287
both the UI and the core behaviours. We call UI tests 'blackbox' tests
288
and they are found in ``bzrlib/tests/blackbox/*.py``.
290
When writing blackbox tests please honour the following conventions:
292
1. Place the tests for the command 'name' in
293
bzrlib/tests/blackbox/test_name.py. This makes it easy for developers
294
to locate the test script for a faulty command.
296
2. Use the 'self.run_bzr("name")' utility function to invoke the command
297
rather than running bzr in a subprocess or invoking the
298
cmd_object.run() method directly. This is a lot faster than
299
subprocesses and generates the same logging output as running it in a
300
subprocess (which invoking the method directly does not).
302
3. Only test the one command in a single test script. Use the bzrlib
303
library when setting up tests and when evaluating the side-effects of
304
the command. We do this so that the library api has continual pressure
305
on it to be as functional as the command line in a simple manner, and
306
to isolate knock-on effects throughout the blackbox test suite when a
307
command changes its name or signature. Ideally only the tests for a
308
given command are affected when a given command is changed.
310
4. If you have a test which does actually require running bzr in a
311
subprocess you can use ``run_bzr_subprocess``. By default the spawned
312
process will not load plugins unless ``--allow-plugins`` is supplied.
318
We make selective use of doctests__. In general they should provide
319
*examples* within the API documentation which can incidentally be tested. We
320
don't try to test every important case using doctests -- regular Python
321
tests are generally a better solution.
323
Most of these are in ``bzrlib/doc/api``. More additions are welcome.
325
__ http://docs.python.org/lib/module-doctest.html
328
Skipping tests and test requirements
329
------------------------------------
331
In our enhancements to unittest we allow for some addition results beyond
332
just success or failure.
334
If a test can't be run, it can say that it's skipped. This is typically
335
used in parameterized tests - for example if a transport doesn't support
336
setting permissions, we'll skip the tests that relating to that. ::
339
return self.branch_format.initialize(repo.bzrdir)
340
except errors.UninitializableFormat:
341
raise tests.TestSkipped('Uninitializable branch format')
343
Raising TestSkipped is a good idea when you want to make it clear that the
344
test was not run, rather than just returning which makes it look as if it
347
A subtly different case is a test that should run, but can't run in the
348
current environment. This covers tests that can only run in particular
349
operating systems or locales, or that depend on external libraries. Here
350
we want to inform the user that they didn't get full test coverage, but
351
they possibly could if they installed more libraries. These are expressed
352
as a dependency on a feature so we can summarise them, and so that the
353
test for the feature is done only once. (For historical reasons, as of
354
May 2007 many cases that should depend on features currently raise
355
TestSkipped.) The typical use is::
357
class TestStrace(TestCaseWithTransport):
359
_test_needs_features = [StraceFeature]
361
which means all tests in this class need the feature. The feature itself
362
should provide a ``_probe`` method which is called once to determine if
369
Known failures are when a test exists but we know it currently doesn't
370
work, allowing the test suite to still pass. These should be used with
371
care, we don't want a proliferation of quietly broken tests. It might be
372
appropriate to use them if you've committed a test for a bug but not the
373
fix for it, or if something works on Unix but not on Windows.
376
Testing exceptions and errors
377
-----------------------------
379
It's important to test handling of errors and exceptions. Because this
380
code is often not hit in ad-hoc testing it can often have hidden bugs --
381
it's particularly common to get NameError because the exception code
382
references a variable that has since been renamed.
384
.. TODO: Something about how to provoke errors in the right way?
386
In general we want to test errors at two levels:
388
1. A test in ``test_errors.py`` checking that when the exception object is
389
constructed with known parameters it produces an expected string form.
390
This guards against mistakes in writing the format string, or in the
391
``str`` representations of its parameters. There should be one for
392
each exception class.
394
2. Tests that when an api is called in a particular situation, it raises
395
an error of the expected class. You should typically use
396
``assertRaises``, which in the Bazaar test suite returns the exception
397
object to allow you to examine its parameters.
399
In some cases blackbox tests will also want to check error reporting. But
400
it can be difficult to provoke every error through the commandline
401
interface, so those tests are only done as needed -- eg in response to a
402
particular bug or if the error is reported in an unusual way(?) Blackbox
403
tests should mostly be testing how the command-line interface works, so
404
should only test errors if there is something particular to the cli in how
405
they're displayed or handled.
408
Essential Domain Classes
409
########################
411
Introducing the Object Model
412
============================
414
The core domain objects within the bazaar model are:
424
Transports are explained below. See http://bazaar-vcs.org/Classes/
425
for an introduction to the other key classes.
430
The ``Transport`` layer handles access to local or remote directories.
431
Each Transport object acts like a logical connection to a particular
432
directory, and it allows various operations on files within it. You can
433
*clone* a transport to get a new Transport connected to a subdirectory or
436
Transports are not used for access to the working tree. At present
437
working trees are always local and they are accessed through the regular
438
Python file io mechanisms.
443
Transports work in URLs. Take note that URLs are by definition only
444
ASCII - the decision of how to encode a Unicode string into a URL must be
445
taken at a higher level, typically in the Store. (Note that Stores also
446
escape filenames which cannot be safely stored on all filesystems, but
447
this is a different level.)
449
The main reason for this is that it's not possible to safely roundtrip a
450
URL into Unicode and then back into the same URL. The URL standard
451
gives a way to represent non-ASCII bytes in ASCII (as %-escapes), but
452
doesn't say how those bytes represent non-ASCII characters. (They're not
453
guaranteed to be UTF-8 -- that is common but doesn't happen everywhere.)
455
For example if the user enters the url ``http://example/%e0`` there's no
456
way to tell whether that character represents "latin small letter a with
457
grave" in iso-8859-1, or "latin small letter r with acute" in iso-8859-2
458
or malformed UTF-8. So we can't convert their URL to Unicode reliably.
460
Equally problematic if we're given a url-like string containing non-ascii
461
characters (such as the accented a) we can't be sure how to convert that
462
to the correct URL, because we don't know what encoding the server expects
463
for those characters. (Although this is not totally reliable we might still
464
accept these and assume they should be put into UTF-8.)
466
A similar edge case is that the url ``http://foo/sweet%2Fsour`` contains
467
one directory component whose name is "sweet/sour". The escaped slash is
468
not a directory separator. If we try to convert URLs to regular Unicode
469
paths this information will be lost.
471
This implies that Transports must natively deal with URLs; for simplicity
472
they *only* deal with URLs and conversion of other strings to URLs is done
473
elsewhere. Information they return, such as from ``list_dir``, is also in
474
the form of URL components.
483
We have a commitment to 6 months API stability - any supported symbol in a
484
release of bzr MUST NOT be altered in any way that would result in
485
breaking existing code that uses it. That means that method names,
486
parameter ordering, parameter names, variable and attribute names etc must
487
not be changed without leaving a 'deprecated forwarder' behind. This even
488
applies to modules and classes.
490
If you wish to change the behaviour of a supported API in an incompatible
491
way, you need to change its name as well. For instance, if I add an optional keyword
492
parameter to branch.commit - that's fine. On the other hand, if I add a
493
keyword parameter to branch.commit which is a *required* transaction
494
object, I should rename the API - i.e. to 'branch.commit_transaction'.
496
When renaming such supported API's, be sure to leave a deprecated_method (or
497
_function or ...) behind which forwards to the new API. See the
498
bzrlib.symbol_versioning module for decorators that take care of the
499
details for you - such as updating the docstring, and issuing a warning
500
when the old api is used.
502
For unsupported API's, it does not hurt to follow this discipline, but it's
503
not required. Minimally though, please try to rename things so that
504
callers will at least get an AttributeError rather than weird results.
507
Coding Style Guidelines
508
=======================
510
Please write PEP-8__ compliant code.
512
One often-missed requirement is that the first line of docstrings
513
should be a self-contained one-sentence summary.
515
__ http://www.python.org/peps/pep-0008.html
521
* Imports should be done at the top-level of the file, unless there is
522
a strong reason to have them lazily loaded when a particular
523
function runs. Import statements have a cost, so try to make sure
524
they don't run inside hot functions.
526
* Module names should always be given fully-qualified,
527
i.e. ``bzrlib.hashcache`` not just ``hashcache``.
533
Functions, methods or members that are "private" to bzrlib are given
534
a leading underscore prefix. Names without a leading underscore are
535
public not just across modules but to programmers using bzrlib as an
536
API. As a consequence, a leading underscore is appropriate for names
537
exposed across modules but that are not to be exposed to bzrlib API
540
We prefer class names to be concatenated capital words (``TestCase``)
541
and variables, methods and functions to be lowercase words joined by
542
underscores (``revision_id``, ``get_revision``).
544
For the purposes of naming some names are treated as single compound
545
words: "filename", "revno".
547
Consider naming classes as nouns and functions/methods as verbs.
549
Try to avoid using abbreviations in names, because there can be
550
inconsistency if other people use the full name.
556
``revision_id`` not ``rev_id`` or ``revid``
558
Functions that transform one thing to another should be named ``x_to_y``
559
(not ``x2y`` as occurs in some old code.)
565
Python destructors (``__del__``) work differently to those of other
566
languages. In particular, bear in mind that destructors may be called
567
immediately when the object apparently becomes unreferenced, or at some
568
later time, or possibly never at all. Therefore we have restrictions on
569
what can be done inside them.
571
0. Never use a __del__ method without asking Martin/Robert first.
573
1. Never rely on a ``__del__`` method running. If there is code that
574
must run, do it from a ``finally`` block instead.
576
2. Never ``import`` from inside a ``__del__`` method, or you may crash the
579
3. In some places we raise a warning from the destructor if the object
580
has not been cleaned up or closed. This is considered OK: the warning
581
may not catch every case but it's still useful sometimes.
587
In some places we have variables which point to callables that construct
588
new instances. That is to say, they can be used a lot like class objects,
589
but they shouldn't be *named* like classes:
591
> I think that things named FooBar should create instances of FooBar when
592
> called. Its plain confusing for them to do otherwise. When we have
593
> something that is going to be used as a class - that is, checked for via
594
> isinstance or other such idioms, them I would call it foo_class, so that
595
> it is clear that a callable is not sufficient. If it is only used as a
596
> factory, then yes, foo_factory is what I would use.
602
Several places in Bazaar use (or will use) a registry, which is a
603
mapping from names to objects or classes. The registry allows for
604
loading in registered code only when it's needed, and keeping
605
associated information such as a help string or description.
611
To make startup time faster, we use the ``bzrlib.lazy_import`` module to
612
delay importing modules until they are actually used. ``lazy_import`` uses
613
the same syntax as regular python imports. So to import a few modules in a
616
from bzrlib.lazy_import import lazy_import
617
lazy_import(globals(), """
626
revision as _mod_revision,
628
import bzrlib.transport
632
At this point, all of these exist as a ``ImportReplacer`` object, ready to
633
be imported once a member is accessed. Also, when importing a module into
634
the local namespace, which is likely to clash with variable names, it is
635
recommended to prefix it as ``_mod_<module>``. This makes it clearer that
636
the variable is a module, and these object should be hidden anyway, since
637
they shouldn't be imported into other namespaces.
640
Modules versus Members
641
~~~~~~~~~~~~~~~~~~~~~~
643
While it is possible for ``lazy_import()`` to import members of a module
644
when using the ``from module import member`` syntax, it is recommended to
645
only use that syntax to load sub modules ``from module import submodule``.
646
This is because variables and classes can frequently be used without
647
needing a sub-member for example::
649
lazy_import(globals(), """
650
from module import MyClass
654
return isinstance(x, MyClass)
656
This will incorrectly fail, because ``MyClass`` is a ``ImportReplacer``
657
object, rather than the real class.
660
Passing to Other Variables
661
~~~~~~~~~~~~~~~~~~~~~~~~~~
663
It also is incorrect to assign ``ImportReplacer`` objects to other variables.
664
Because the replacer only knows about the original name, it is unable to
665
replace other variables. The ``ImportReplacer`` class will raise an
666
``IllegalUseOfScopeReplacer`` exception if it can figure out that this
667
happened. But it requires accessing a member more than once from the new
668
variable, so some bugs are not detected right away.
674
The null revision is the ancestor of all revisions. Its revno is 0, its
675
revision-id is ``null:``, and its tree is the empty tree. When referring
676
to the null revision, please use ``bzrlib.revision.NULL_REVISION``. Old
677
code sometimes uses ``None`` for the null revision, but this practice is
684
Processing Command Lines
685
------------------------
687
bzrlib has a standard framework for parsing command lines and calling
688
processing routines associated with various commands. See builtins.py
689
for numerous examples.
692
Standard Parameter Types
693
------------------------
695
There are some common requirements in the library: some parameters need to be
696
unicode safe, some need byte strings, and so on. At the moment we have
697
only codified one specific pattern: Parameters that need to be unicode
698
should be checked via ``bzrlib.osutils.safe_unicode``. This will coerce the
699
input into unicode in a consistent fashion, allowing trivial strings to be
700
used for programmer convenience, but not performing unpredictably in the
701
presence of different locales.
707
(The strategy described here is what we want to get to, but it's not
708
consistently followed in the code at the moment.)
710
bzrlib is intended to be a generically reusable library. It shouldn't
711
write messages to stdout or stderr, because some programs that use it
712
might want to display that information through a GUI or some other
715
We can distinguish two types of output from the library:
717
1. Structured data representing the progress or result of an
718
operation. For example, for a commit command this will be a list
719
of the modified files and the finally committed revision number
722
These should be exposed either through the return code or by calls
723
to a callback parameter.
725
A special case of this is progress indicators for long-lived
726
operations, where the caller should pass a ProgressBar object.
728
2. Unstructured log/debug messages, mostly for the benefit of the
729
developers or users trying to debug problems. This should always
730
be sent through ``bzrlib.trace`` and Python ``logging``, so that
731
it can be redirected by the client.
733
The distinction between the two is a bit subjective, but in general if
734
there is any chance that a library would want to see something as
735
structured data, we should make it so.
737
The policy about how output is presented in the text-mode client
738
should be only in the command-line tool.
745
Bazaar has online help for various topics through ``bzr help COMMAND`` or
746
equivalently ``bzr command -h``. We also have help on command options,
747
and on other help topics. (See ``help_topics.py``.)
749
As for python docstrings, the first paragraph should be a single-sentence
750
synopsis of the command.
752
The help for options should be one or more proper sentences, starting with
753
a capital letter and finishing with a full stop (period).
755
All help messages and documentation should have two spaces between
762
In general tests should be placed in a file named test_FOO.py where
763
FOO is the logical thing under test. That file should be placed in the
764
tests subdirectory under the package being tested.
766
For example, tests for merge3 in bzrlib belong in bzrlib/tests/test_merge3.py.
767
See bzrlib/tests/test_sampler.py for a template test script.
769
Tests can be written for the UI or for individual areas of the library.
770
Choose whichever is appropriate: if adding a new command, or a new command
771
option, then you should be writing a UI test. If you are both adding UI
772
functionality and library functionality, you will want to write tests for
773
both the UI and the core behaviours. We call UI tests 'blackbox' tests
774
and they are found in ``bzrlib/tests/blackbox/*.py``.
776
When writing blackbox tests please honour the following conventions:
778
1. Place the tests for the command 'name' in
779
bzrlib/tests/blackbox/test_name.py. This makes it easy for developers
780
to locate the test script for a faulty command.
782
2. Use the 'self.run_bzr("name")' utility function to invoke the command
783
rather than running bzr in a subprocess or invoking the
784
cmd_object.run() method directly. This is a lot faster than
785
subprocesses and generates the same logging output as running it in a
786
subprocess (which invoking the method directly does not).
788
3. Only test the one command in a single test script. Use the bzrlib
789
library when setting up tests and when evaluating the side-effects of
790
the command. We do this so that the library api has continual pressure
791
on it to be as functional as the command line in a simple manner, and
792
to isolate knock-on effects throughout the blackbox test suite when a
793
command changes its name or signature. Ideally only the tests for a
794
given command are affected when a given command is changed.
796
4. If you have a test which does actually require running bzr in a
797
subprocess you can use ``run_bzr_subprocess``. By default the spawned
798
process will not load plugins unless ``--allow-plugins`` is supplied.
804
We have a rich collection of tools to support writing tests. Please use
805
them in preference to ad-hoc solutions as they provide portability and
806
performance benefits.
811
The ``TreeBuilder`` interface allows the construction of arbitrary trees
812
with a declarative interface. A sample session might look like::
814
tree = self.make_branch_and_tree('path')
815
builder = TreeBuilder()
816
builder.start_tree(tree)
817
builder.build(['foo', "bar/", "bar/file"])
818
tree.commit('commit the tree')
819
builder.finish_tree()
821
Please see bzrlib.treebuilder for more details.
826
The ``BranchBuilder`` interface allows the creation of test branches in a
827
quick and easy manner. A sample session::
829
builder = BranchBuilder(self.get_transport().clone('relpath'))
830
builder.build_commit()
831
builder.build_commit()
832
builder.build_commit()
833
branch = builder.get_branch()
835
Please see bzrlib.branchbuilder for more details.
840
We make selective use of doctests__. In general they should provide
841
*examples* within the API documentation which can incidentally be tested. We
842
don't try to test every important case using doctests -- regular Python
843
tests are generally a better solution.
845
Most of these are in ``bzrlib/doc/api``. More additions are welcome.
847
__ http://docs.python.org/lib/module-doctest.html
852
Currently, bzr selftest is used to invoke tests.
853
You can provide a pattern argument to run a subset. For example,
854
to run just the blackbox tests, run::
856
./bzr selftest -v blackbox
858
To skip a particular test (or set of tests), use the --exclude option
859
(shorthand -x) like so::
861
./bzr selftest -v -x blackbox
863
To list tests without running them, use the --list-only option like so::
865
./bzr selftest --list-only
867
This option can be combined with other selftest options (like -x) and
868
filter patterns to understand their effect.
871
Handling Errors and Exceptions
872
==============================
874
Commands should return non-zero when they encounter circumstances that
875
the user should really pay attention to - which includes trivial shell
878
Recommended values are:
881
1. Conflicts in merge-like operations, or changes are present in
882
diff-like operations.
883
2. Unrepresentable diff changes (i.e. binary files that we cannot show
885
3. An error or exception has occurred.
887
Errors are handled through Python exceptions. Exceptions should be defined
888
inside bzrlib.errors, so that we can see the whole tree at a glance.
890
We broadly classify errors as either being either internal or not,
891
depending on whether ``internal_error`` is set or not. If we think it's our
892
fault, we show a backtrace, an invitation to report the bug, and possibly
893
other details. This is the default for errors that aren't specifically
894
recognized as being caused by a user error. Otherwise we show a briefer
895
message, unless -Derror was given.
897
Many errors originate as "environmental errors" which are raised by Python
898
or builtin libraries -- for example IOError. These are treated as being
899
our fault, unless they're caught in a particular tight scope where we know
900
that they indicate a user errors. For example if the repository format
901
is not found, the user probably gave the wrong path or URL. But if one of
902
the files inside the repository is not found, then it's our fault --
903
either there's a bug in bzr, or something complicated has gone wrong in
904
the environment that means one internal file was deleted.
906
Many errors are defined in ``bzrlib/errors.py`` but it's OK for new errors
907
to be added near the place where they are used.
909
Exceptions are formatted for the user by conversion to a string
910
(eventually calling their ``__str__`` method.) As a convenience the
911
``._fmt`` member can be used as a template which will be mapped to the
912
error's instance dict.
914
New exception classes should be defined when callers might want to catch
915
that exception specifically, or when it needs a substantially different
918
Exception strings should start with a capital letter and should not have a
919
final fullstop. If long, they may contain newlines to break the text.
925
When you change bzrlib, please update the relevant documentation for the
926
change you made: Changes to commands should update their help, and
927
possibly end user tutorials; changes to the core library should be
928
reflected in API documentation.
933
If you make a user-visible change, please add a note to the NEWS file.
934
The description should be written to make sense to someone who's just
935
a user of bzr, not a developer: new functions or classes shouldn't be
936
mentioned, but new commands, changes in behaviour or fixed nontrivial
937
bugs should be listed. See the existing entries for an idea of what
940
Within each release, entries in the news file should have the most
941
user-visible changes first. So the order should be approximately:
943
* changes to existing behaviour - the highest priority because the
944
user's existing knowledge is incorrect
945
* new features - should be brought to their attention
946
* bug fixes - may be of interest if the bug was affecting them, and
947
should include the bug number if any
948
* major documentation changes
949
* changes to internal interfaces
951
People who made significant contributions to each change are listed in
952
parenthesis. This can include reporting bugs (particularly with good
953
details or reproduction recipes), submitting patches, etc.
958
The docstring of a command is used by ``bzr help`` to generate help output
959
for the command. The list 'takes_options' attribute on a command is used by
960
``bzr help`` to document the options for the command - the command
961
docstring does not need to document them. Finally, the '_see_also'
962
attribute on a command can be used to reference other related help topics.
967
Functions, methods, classes and modules should have docstrings
968
describing how they are used.
970
The first line of the docstring should be a self-contained sentence.
972
For the special case of Command classes, this acts as the user-visible
973
documentation shown by the help command.
975
The docstrings should be formatted as reStructuredText_ (like this
976
document), suitable for processing using the epydoc_ tool into HTML
979
.. _reStructuredText: http://docutils.sourceforge.net/rst.html
980
.. _epydoc: http://epydoc.sourceforge.net/
989
The copyright policy for bzr was recently made clear in this email (edited
990
for grammatical correctness)::
992
The attached patch cleans up the copyright and license statements in
993
the bzr source. It also adds tests to help us remember to add them
994
with the correct text.
996
We had the problem that lots of our files were "Copyright Canonical
997
Development Ltd" which is not a real company, and some other variations
998
on this theme. Also, some files were missing the GPL statements.
1000
I want to be clear about the intent of this patch, since copyright can
1001
be a little controversial.
1003
1) The big motivation for this is not to shut out the community, but
1004
just to clean up all of the invalid copyright statements.
1006
2) It has been the general policy for bzr that we want a single
1007
copyright holder for all of the core code. This is following the model
1008
set by the FSF, which makes it easier to update the code to a new
1009
license in case problems are encountered. (For example, if we want to
1010
upgrade the project universally to GPL v3 it is much simpler if there is
1011
a single copyright holder). It also makes it clearer if copyright is
1012
ever debated, there is a single holder, which makes it easier to defend
1013
in court, etc. (I think the FSF position is that if you assign them
1014
copyright, they can defend it in court rather than you needing to, and
1015
I'm sure Canonical would do the same).
1016
As such, Canonical has requested copyright assignments from all of the
1019
3) If someone wants to add code and not attribute it to Canonical, there
1020
is a specific list of files that are excluded from this check. And the
1021
test failure indicates where that is, and how to update it.
1023
4) If anyone feels that I changed a copyright statement incorrectly, just
1024
let me know, and I'll be happy to correct it. Whenever you have large
1025
mechanical changes like this, it is possible to make some mistakes.
1027
Just to reiterate, this is a community project, and it is meant to stay
1028
that way. Core bzr code is copyright Canonical for legal reasons, and
1029
the tests are just there to help us maintain that.
1032
Miscellaneous Topics
1033
####################
1038
Bazaar has a few facilities to help debug problems by going into pdb_, the
1041
.. _pdb: http://docs.python.org/lib/debugger-commands.html
1043
If the ``BZR_PDB`` environment variable is set
1044
then bzr will go into pdb post-mortem mode when an unhandled exception
1047
If you send a SIGQUIT signal to bzr, which can be done by pressing
1048
Ctrl-\\ on Unix, bzr will go into the debugger immediately. You can
1049
continue execution by typing ``c``. This can be disabled if necessary
1050
by setting the environment variable ``BZR_SIGQUIT_PDB=0``.
1057
Integer identifier for a revision on the main line of a branch.
1058
Revision 0 is always the null revision; others are 1-based
1059
indexes into the branch's revision history.
1062
Unicode and Encoding Support
1063
============================
1065
This section discusses various techniques that Bazaar uses to handle
1066
characters that are outside the ASCII set.
1071
When a ``Command`` object is created, it is given a member variable
1072
accessible by ``self.outf``. This is a file-like object, which is bound to
1073
``sys.stdout``, and should be used to write information to the screen,
1074
rather than directly writing to ``sys.stdout`` or calling ``print``.
1075
This file has the ability to translate Unicode objects into the correct
1076
representation, based on the console encoding. Also, the class attribute
1077
``encoding_type`` will effect how unprintable characters will be
1078
handled. This parameter can take one of 3 values:
1081
Unprintable characters will be represented with a suitable replacement
1082
marker (typically '?'), and no exception will be raised. This is for
1083
any command which generates text for the user to review, rather than
1084
for automated processing.
1085
For example: ``bzr log`` should not fail if one of the entries has text
1086
that cannot be displayed.
1089
Attempting to print an unprintable character will cause a UnicodeError.
1090
This is for commands that are intended more as scripting support, rather
1091
than plain user review.
1092
For exampl: ``bzr ls`` is designed to be used with shell scripting. One
1093
use would be ``bzr ls --null --unknows | xargs -0 rm``. If ``bzr``
1094
printed a filename with a '?', the wrong file could be deleted. (At the
1095
very least, the correct file would not be deleted). An error is used to
1096
indicate that the requested action could not be performed.
1099
Do not attempt to automatically convert Unicode strings. This is used
1100
for commands that must handle conversion themselves.
1101
For example: ``bzr diff`` needs to translate Unicode paths, but should
1102
not change the exact text of the contents of the files.
1105
``bzrlib.urlutils.unescape_for_display``
1106
----------------------------------------
1108
Because Transports work in URLs (as defined earlier), printing the raw URL
1109
to the user is usually less than optimal. Characters outside the standard
1110
set are printed as escapes, rather than the real character, and local
1111
paths would be printed as ``file://`` urls. The function
1112
``unescape_for_display`` attempts to unescape a URL, such that anything
1113
that cannot be printed in the current encoding stays an escaped URL, but
1114
valid characters are generated where possible.
1120
The ``bzrlib.osutils`` module has many useful helper functions, including
1121
some more portable variants of functions in the standard library.
1123
In particular, don't use ``shutil.rmtree`` unless it's acceptable for it
1124
to fail on Windows if some files are readonly or still open elsewhere.
1125
Use ``bzrlib.osutils.rmtree`` instead.
1131
We write some extensions in C using pyrex. We design these to work in
1134
* User with no C compiler
1135
* User with C compiler
1138
The recommended way to install bzr is to have a C compiler so that the
1139
extensions can be built, but if no C compiler is present, the pure python
1140
versions we supply will work, though more slowly.
1142
For developers we recommend that pyrex be installed, so that the C
1143
extensions can be changed if needed.
1145
For the C extensions, the extension module should always match the
1146
original python one in all respects (modulo speed). This should be
1147
maintained over time.
1149
To create an extension, add rules to setup.py for building it with pyrex,
1150
and with distutils. Now start with an empty .pyx file. At the top add
1151
"include 'yourmodule.py'". This will import the contents of foo.py into this
1152
file at build time - remember that only one module will be loaded at
1153
runtime. Now you can subclass classes, or replace functions, and only your
1154
changes need to be present in the .pyx file.
1156
Note that pyrex does not support all 2.4 programming idioms, so some
1157
syntax changes may be required. I.e.
1159
- 'from foo import (bar, gam)' needs to change to not use the brackets.
1160
- 'import foo.bar as bar' needs to be 'import foo.bar; bar = foo.bar'
1162
If the changes are too dramatic, consider
1163
maintaining the python code twice - once in the .pyx, and once in the .py,
1164
and no longer including the .py file.
1167
Making Installers for OS Windows
1168
================================
1169
To build a win32 installer, see the instructions on the wiki page:
1170
http://bazaar-vcs.org/BzrWin32Installer
1174
vim: ft=rst tw=74 ai
added
removed
doc/developers/HACKING
