186
105
For further information, see http://bazaar-vcs.org/BzrDevelopment.
191
Preparing a Sandbox for Making Changes to Bazaar
192
================================================
194
Bazaar supports many ways of organising your work. See
195
http://bazaar-vcs.org/SharedRepositoryLayouts for a summary of the
196
popular alternatives.
198
Of course, the best choice for you will depend on numerous factors:
199
the number of changes you may be making, the complexity of the changes, etc.
200
As a starting suggestion though:
202
* create a local copy of the main development branch (bzr.dev) by using
205
bzr branch http://bazaar-vcs.org/bzr/bzr.dev/ bzr.dev
207
* keep your copy of bzr.dev pristine (by not developing in it) and keep
208
it up to date (by using bzr pull)
210
* create a new branch off your local bzr.dev copy for each issue
211
(bug or feature) you are working on.
213
This approach makes it easy to go back and make any required changes
214
after a code review. Resubmitting the change is then simple with no
215
risk of accidentally including edits related to other issues you may
216
be working on. After the changes for an issue are accepted and merged,
217
the associated branch can be deleted or archived as you wish.
220
Navigating the Code Base
221
========================
223
.. Was at <http://bazaar-vcs.org/NewDeveloperIntroduction>
225
Some of the key files in this directory are:
228
The command you run to start Bazaar itself. This script is pretty
229
short and just does some checks then jumps into bzrlib.
232
This file covers a brief introduction to Bazaar and lists some of its
236
Summary of changes in each Bazaar release that can affect users or
240
Installs Bazaar system-wide or to your home directory. To perform
241
development work on Bazaar it is not required to run this file - you
242
can simply run the bzr command from the top level directory of your
243
development copy. Note: That if you run setup.py this will create a
244
'build' directory in your development branch. There's nothing wrong
245
with this but don't be confused by it. The build process puts a copy
246
of the main code base into this build directory, along with some other
247
files. You don't need to go in here for anything discussed in this
251
Possibly the most exciting folder of all, bzrlib holds the main code
252
base. This is where you will go to edit python files and contribute to
256
Holds documentation on a whole range of things on Bazaar from the
257
origination of ideas within the project to information on Bazaar
258
features and use cases. Within this directory there is a subdirectory
259
for each translation into a human language. All the documentation
260
is in the ReStructuredText markup language.
263
Documentation specifically targeted at Bazaar and plugin developers.
264
(Including this document.)
268
Automatically-generated API reference information is available at
269
<http://starship.python.net/crew/mwh/bzrlibapi/>.
271
See also the `Bazaar Architectural Overview
272
<http://doc.bazaar-vcs.org/developers/overview.html>`_.
275
The Code Review Process
276
#######################
278
All code changes coming in to Bazaar are reviewed by someone else.
279
Normally changes by core contributors are reviewed by one other core
280
developer, and changes from other people are reviewed by two core
281
developers. Use intelligent discretion if the patch is trivial.
283
Good reviews do take time. They also regularly require a solid
284
understanding of the overall code base. In practice, this means a small
285
number of people often have a large review burden - with knowledge comes
286
responsibility. No one likes their merge requests sitting in a queue going
287
nowhere, so reviewing sooner rather than later is strongly encouraged.
296
Please put a "cover letter" on your merge request explaining:
298
* the reason **why** you're making this change
300
* **how** this change achieves this purpose
302
* anything else you may have fixed in passing
304
* anything significant that you thought of doing, such as a more
305
extensive fix or a different approach, but didn't or couldn't do now
307
A good cover letter makes reviewers' lives easier because they can decide
308
from the letter whether they agree with the purpose and approach, and then
309
assess whether the patch actually does what the cover letter says.
310
Explaining any "drive-by fixes" or roads not taken may also avoid queries
311
from the reviewer. All in all this should give faster and better reviews.
312
Sometimes writing the cover letter helps the submitter realize something
313
else they need to do. The size of the cover letter should be proportional
314
to the size and complexity of the patch.
317
Reviewing proposed changes
318
==========================
320
Anyone is welcome to review code, and reply to the thread with their
323
The simplest way to review a proposed change is to just read the patch on
324
the list or in Bundle Buggy. For more complex changes it may be useful
325
to make a new working tree or branch from trunk, and merge the proposed
326
change into it, so you can experiment with the code or look at a wider
329
There are three main requirements for code to get in:
331
* Doesn't reduce test coverage: if it adds new methods or commands,
332
there should be tests for them. There is a good test framework
333
and plenty of examples to crib from, but if you are having trouble
334
working out how to test something feel free to post a draft patch
337
* Doesn't reduce design clarity, such as by entangling objects
338
we're trying to separate. This is mostly something the more
339
experienced reviewers need to help check.
341
* Improves bugs, features, speed, or code simplicity.
343
Code that goes in should not degrade any of these aspects. Patches are
344
welcome that only cleanup the code without changing the external
345
behaviour. The core developers take care to keep the code quality high
346
and understandable while recognising that perfect is sometimes the enemy
349
It is easy for reviews to make people notice other things which should be
350
fixed but those things should not hold up the original fix being accepted.
351
New things can easily be recorded in the Bug Tracker instead.
353
It's normally much easier to review several smaller patches than one large
354
one. You might want to use ``bzr-loom`` to maintain threads of related
355
work, or submit a preparatory patch that will make your "real" change
359
Checklist for reviewers
360
=======================
362
* Do you understand what the code's doing and why?
364
* Will it perform reasonably for large inputs, both in memory size and
365
run time? Are there some scenarios where performance should be
368
* Is it tested, and are the tests at the right level? Are there both
369
blackbox (command-line level) and API-oriented tests?
371
* If this change will be visible to end users or API users, is it
372
appropriately documented in NEWS?
374
* Does it meet the coding standards below?
376
* If it changes the user-visible behaviour, does it update the help
377
strings and user documentation?
379
* If it adds a new major concept or standard practice, does it update the
380
developer documentation?
382
* (your ideas here...)
388
From May 2009 on, we prefer people to propose code reviews through
391
* <https://launchpad.net/+tour/code-review>
393
* <https://help.launchpad.net/Code/Review>
395
Anyone can propose or comment on a merge proposal just by creating a
398
There are two ways to create a new merge proposal: through the web
399
interface or by email.
402
Proposing a merge through the web
403
---------------------------------
405
To create the proposal through the web, first push your branch to Launchpad.
406
For example, a branch dealing with documentation belonging to the Launchpad
407
User mbp could be pushed as ::
409
bzr push lp:~mbp/bzr/doc
411
Then go to the branch's web page, which in this case would be
412
<https://code.launchpad.net/~mbp/bzr/doc>. You can simplify this step by just
417
You can then click "Propose for merging into another branch", and enter your
418
cover letter (see above) into the web form. Typically you'll want to merge
419
into ``~bzr/bzr/trunk`` which will be the default; you might also want to
420
nominate merging into a release branch for a bug fix. There is the option to
421
specify a specific reviewer or type of review, and you shouldn't normally
424
Submitting the form takes you to the new page about the merge proposal
425
containing the diff of the changes, comments by interested people, and
426
controls to comment or vote on the change.
428
Proposing a merge by mail
429
-------------------------
431
To propose a merge by mail, send a bundle to ``merge@code.launchpad.net``.
433
You can generate a merge request like this::
435
bzr send -o bug-1234.diff
437
``bzr send`` can also send mail directly if you prefer; see the help.
442
From <https://code.launchpad.net/bzr/+activereviews> you can see all
443
currently active reviews, and choose one to comment on. This page also
444
shows proposals that are now approved and should be merged by someone with
448
Reviews through Bundle Buggy
449
============================
451
The Bundle Buggy tool used up to May 2009 is still available as a review
454
Sending patches for review
455
--------------------------
108
A Closer Look at the Merge & Review Process
109
===========================================
457
111
If you'd like to propose a change, please post to the
458
112
bazaar@lists.canonical.com list with a bundle, patch, or link to a
459
branch. Put ``[PATCH]`` or ``[MERGE]`` in the subject so Bundle Buggy
113
branch. Put '[PATCH]' or '[MERGE]' in the subject so Bundle Buggy
460
114
can pick it out, and explain the change in the email message text.
461
115
Remember to update the NEWS file as part of your change if it makes any
462
116
changes visible to users or plugin developers. Please include a diff
463
117
against mainline if you're giving a link to a branch.
465
You can generate a merge request like this::
467
bzr send -o bug-1234.patch
469
A ``.patch`` extension is recommended instead of .bundle as many mail clients
470
will send the latter as a binary file.
472
``bzr send`` can also send mail directly if you prefer; see the help.
119
You can generate a bundle like this::
121
bzr bundle > mybundle.patch
123
A .patch extension is recommended instead of .bundle as many mail clients
124
will send the latter as a binary file. If a bundle would be too long or your
125
mailer mangles whitespace (e.g. implicitly converts Unix newlines to DOS
126
newlines), use the merge-directive command instead like this::
128
bzr merge-directive http://bazaar-vcs.org http://example.org/my_branch > my_directive.patch
130
See the help for details on the arguments to merge-directive.
474
132
Please do **NOT** put [PATCH] or [MERGE] in the subject line if you don't
475
133
want it to be merged. If you want comments from developers rather than
476
to be merged, you can put ``[RFC]`` in the subject line.
478
If this change addresses a bug, please put the bug number in the subject
479
line too, in the form ``[#1]`` so that Bundle Buggy can recognize it.
481
If the change is intended for a particular release mark that in the
482
subject too, e.g. ``[1.6]``.
483
Anyone can "vote" on the mailing list by expressing an opinion. Core
484
developers can also vote using Bundle Buggy. Here are the voting codes and
134
to be merged, you can put '[RFC]' in the subject line.
136
Anyone is welcome to review code. There are broadly three gates for
139
* Doesn't reduce test coverage: if it adds new methods or commands,
140
there should be tests for them. There is a good test framework
141
and plenty of examples to crib from, but if you are having trouble
142
working out how to test something feel free to post a draft patch
145
* Doesn't reduce design clarity, such as by entangling objects
146
we're trying to separate. This is mostly something the more
147
experienced reviewers need to help check.
149
* Improves bugs, features, speed, or code simplicity.
151
Code that goes in should pass all three. The core developers take care
152
to keep the code quality high and understandable while recognising that
153
perfect is sometimes the enemy of good. (It is easy for reviews to make
154
people notice other things which should be fixed but those things should
155
not hold up the original fix being accepted. New things can easily be
156
recorded in the Bug Tracker instead.)
158
Anyone can "vote" on the mailing list. Core developers can also vote using
159
Bundle Buggy. Here are the voting codes and their explanations.
487
161
:approve: Reviewer wants this submission merged.
488
162
:tweak: Reviewer wants this submission merged with small changes. (No
505
179
outstanding merge requests together with an explanation of the columns.
506
180
Bundle Buggy will also mail you a link to track just your change.
183
Preparing a Sandbox for Making Changes to Bazaar
184
================================================
186
Bazaar supports many ways of organising your work. See
187
http://bazaar-vcs.org/SharedRepositoryLayouts for a summary of the
188
popular alternatives.
190
Of course, the best choice for you will depend on numerous factors:
191
the number of changes you may be making, the complexity of the changes, etc.
192
As a starting suggestion though:
194
* create a local copy of the main development branch (bzr.dev) by using
197
bzr branch http://bazaar-vcs.org/bzr/bzr.dev/ bzr.dev
199
* keep your copy of bzr.dev prestine (by not developing in it) and keep
200
it up to date (by using bzr pull)
202
* create a new branch off your local bzr.dev copy for each issue
203
(bug or feature) you are working on.
205
This approach makes it easy to go back and make any required changes
206
after a code review. Resubmitting the change is then simple with no
207
risk of accidentially including edits related to other issues you may
208
be working on. After the changes for an issue are accepted and merged,
209
the associated branch can be deleted or archived as you wish.
212
Navigating the Code Base
213
========================
215
TODO: List and describe in one line the purpose of each directory
216
inside an installation of bzr.
218
TODO: Refer to a central location holding an up to date copy of the API
219
documentation generated by epydoc, e.g. something like
220
http://starship.python.net/crew/mwh/bzrlibapi/bzrlib.html.
226
The Importance of Testing
227
=========================
229
Reliability is a critical success factor for any Version Control System.
230
We want Bazaar to be highly reliable across multiple platforms while
231
evolving over time to meet the needs of its community.
233
In a nutshell, this is want we expect and encourage:
235
* New functionality should have test cases. Preferably write the
236
test before writing the code.
238
In general, you can test at either the command-line level or the
239
internal API level. See Writing tests below for more detail.
241
* Try to practice Test-Driven Development: before fixing a bug, write a
242
test case so that it does not regress. Similarly for adding a new
243
feature: write a test case for a small version of the new feature before
244
starting on the code itself. Check the test fails on the old code, then
245
add the feature or fix and check it passes.
247
By doing these things, the Bazaar team gets increased confidence that
248
changes do what they claim to do, whether provided by the core team or
249
by community members. Equally importantly, we can be surer that changes
250
down the track do not break new features or bug fixes that you are
253
As of May 2007, Bazaar ships with a test suite containing over 6000 tests
254
and growing. We are proud of it and want to remain so. As community
255
members, we all benefit from it. Would you trust version control on
256
your project to a product *without* a test suite like Bazaar has?
259
Running the Test Suite
260
======================
262
Currently, bzr selftest is used to invoke tests.
263
You can provide a pattern argument to run a subset. For example,
264
to run just the blackbox tests, run::
266
./bzr selftest -v blackbox
268
To skip a particular test (or set of tests), use the --exclude option
269
(shorthand -x) like so::
271
./bzr selftest -v -x blackbox
273
To ensure that all tests are being run and succeeding, you can use the
274
--strict option which will fail if there are any missing features or known
277
./bzr selftest --strict
279
To list tests without running them, use the --list-only option like so::
281
./bzr selftest --list-only
283
This option can be combined with other selftest options (like -x) and
284
filter patterns to understand their effect.
290
In general tests should be placed in a file named test_FOO.py where
291
FOO is the logical thing under test. That file should be placed in the
292
tests subdirectory under the package being tested.
294
For example, tests for merge3 in bzrlib belong in bzrlib/tests/test_merge3.py.
295
See bzrlib/tests/test_sampler.py for a template test script.
297
Tests can be written for the UI or for individual areas of the library.
298
Choose whichever is appropriate: if adding a new command, or a new command
299
option, then you should be writing a UI test. If you are both adding UI
300
functionality and library functionality, you will want to write tests for
301
both the UI and the core behaviours. We call UI tests 'blackbox' tests
302
and they are found in ``bzrlib/tests/blackbox/*.py``.
304
When writing blackbox tests please honour the following conventions:
306
1. Place the tests for the command 'name' in
307
bzrlib/tests/blackbox/test_name.py. This makes it easy for developers
308
to locate the test script for a faulty command.
310
2. Use the 'self.run_bzr("name")' utility function to invoke the command
311
rather than running bzr in a subprocess or invoking the
312
cmd_object.run() method directly. This is a lot faster than
313
subprocesses and generates the same logging output as running it in a
314
subprocess (which invoking the method directly does not).
316
3. Only test the one command in a single test script. Use the bzrlib
317
library when setting up tests and when evaluating the side-effects of
318
the command. We do this so that the library api has continual pressure
319
on it to be as functional as the command line in a simple manner, and
320
to isolate knock-on effects throughout the blackbox test suite when a
321
command changes its name or signature. Ideally only the tests for a
322
given command are affected when a given command is changed.
324
4. If you have a test which does actually require running bzr in a
325
subprocess you can use ``run_bzr_subprocess``. By default the spawned
326
process will not load plugins unless ``--allow-plugins`` is supplied.
332
We make selective use of doctests__. In general they should provide
333
*examples* within the API documentation which can incidentally be tested. We
334
don't try to test every important case using doctests -- regular Python
335
tests are generally a better solution.
337
Most of these are in ``bzrlib/doc/api``. More additions are welcome.
339
__ http://docs.python.org/lib/module-doctest.html
342
Skipping tests and test requirements
343
------------------------------------
345
In our enhancements to unittest we allow for some addition results beyond
346
just success or failure.
348
If a test can't be run, it can say that it's skipped. This is typically
349
used in parameterized tests - for example if a transport doesn't support
350
setting permissions, we'll skip the tests that relating to that. ::
353
return self.branch_format.initialize(repo.bzrdir)
354
except errors.UninitializableFormat:
355
raise tests.TestSkipped('Uninitializable branch format')
357
Raising TestSkipped is a good idea when you want to make it clear that the
358
test was not run, rather than just returning which makes it look as if it
361
Several different cases are distinguished:
364
Generic skip; the only type that was present up to bzr 0.18.
367
The test doesn't apply to the parameters with which it was run.
368
This is typically used when the test is being applied to all
369
implementations of an interface, but some aspects of the interface
370
are optional and not present in particular concrete
371
implementations. (Some tests that should raise this currently
372
either silently return or raise TestSkipped.) Another option is
373
to use more precise parameterization to avoid generating the test
377
**(Not implemented yet)**
378
The test can't be run because of an inherent limitation of the
379
environment, such as not having symlinks or not supporting
383
The test can't be run because a dependency (typically a Python
384
library) is not available in the test environment. These
385
are in general things that the person running the test could fix
386
by installing the library. It's OK if some of these occur when
387
an end user runs the tests or if we're specifically testing in a
388
limited environment, but a full test should never see them.
391
The test exists but is known to fail, for example because the
392
code to fix it hasn't been run yet. Raising this allows
393
you to distinguish these failures from the ones that are not
394
expected to fail. This could be conditionally raised if something
395
is broken on some platforms but not on others.
397
We plan to support three modes for running the test suite to control the
398
interpretation of these results. Strict mode is for use in situations
399
like merges to the mainline and releases where we want to make sure that
400
everything that can be tested has been tested. Lax mode is for use by
401
developers who want to temporarily tolerate some known failures. The
402
default behaviour is obtained by ``bzr selftest`` with no options, and
403
also (if possible) by running under another unittest harness.
405
======================= ======= ======= ========
406
result strict default lax
407
======================= ======= ======= ========
408
TestSkipped pass pass pass
409
TestNotApplicable pass pass pass
410
TestPlatformLimit pass pass pass
411
TestDependencyMissing fail pass pass
412
KnownFailure fail pass pass
413
======================= ======= ======= ========
416
Test feature dependencies
417
-------------------------
419
Rather than manually checking the environment in each test, a test class
420
can declare its dependence on some test features. The feature objects are
421
checked only once for each run of the whole test suite.
423
For historical reasons, as of May 2007 many cases that should depend on
424
features currently raise TestSkipped.)
428
class TestStrace(TestCaseWithTransport):
430
_test_needs_features = [StraceFeature]
432
This means all tests in this class need the feature. The feature itself
433
should provide a ``_probe`` method which is called once to determine if
436
These should generally be equivalent to either TestDependencyMissing or
437
sometimes TestPlatformLimit.
443
Known failures are when a test exists but we know it currently doesn't
444
work, allowing the test suite to still pass. These should be used with
445
care, we don't want a proliferation of quietly broken tests. It might be
446
appropriate to use them if you've committed a test for a bug but not the
447
fix for it, or if something works on Unix but not on Windows.
450
Testing exceptions and errors
451
-----------------------------
453
It's important to test handling of errors and exceptions. Because this
454
code is often not hit in ad-hoc testing it can often have hidden bugs --
455
it's particularly common to get NameError because the exception code
456
references a variable that has since been renamed.
458
.. TODO: Something about how to provoke errors in the right way?
460
In general we want to test errors at two levels:
462
1. A test in ``test_errors.py`` checking that when the exception object is
463
constructed with known parameters it produces an expected string form.
464
This guards against mistakes in writing the format string, or in the
465
``str`` representations of its parameters. There should be one for
466
each exception class.
468
2. Tests that when an api is called in a particular situation, it raises
469
an error of the expected class. You should typically use
470
``assertRaises``, which in the Bazaar test suite returns the exception
471
object to allow you to examine its parameters.
473
In some cases blackbox tests will also want to check error reporting. But
474
it can be difficult to provoke every error through the commandline
475
interface, so those tests are only done as needed -- eg in response to a
476
particular bug or if the error is reported in an unusual way(?) Blackbox
477
tests should mostly be testing how the command-line interface works, so
478
should only test errors if there is something particular to the cli in how
479
they're displayed or handled.
485
The Python ``warnings`` module is used to indicate a non-fatal code
486
problem. Code that's expected to raise a warning can be tested through
489
The test suite can be run with ``-Werror`` to check no unexpected errors
492
However, warnings should be used with discretion. It's not an appropriate
493
way to give messages to the user, because the warning is normally shown
494
only once per source line that causes the problem. You should also think
495
about whether the warning is serious enought that it should be visible to
496
users who may not be able to fix it.
499
Interface implementation testing and test scenarios
500
---------------------------------------------------
502
There are several cases in Bazaar of multiple implementations of a common
503
conceptual interface. ("Conceptual" because
504
it's not necessary for all the implementations to share a base class,
505
though they often do.) Examples include transports and the working tree,
506
branch and repository classes.
508
In these cases we want to make sure that every implementation correctly
509
fulfils the interface requirements. For example, every Transport should
510
support the ``has()`` and ``get()`` and ``clone()`` methods. We have a
511
sub-suite of tests in ``test_transport_implementations``. (Most
512
per-implementation tests are in submodules of ``bzrlib.tests``, but not
513
the transport tests at the moment.)
515
These tests are repeated for each registered Transport, by generating a
516
new TestCase instance for the cross product of test methods and transport
517
implementations. As each test runs, it has ``transport_class`` and
518
``transport_server`` set to the class it should test. Most tests don't
519
access these directly, but rather use ``self.get_transport`` which returns
520
a transport of the appropriate type.
522
The goal is to run per-implementation only tests that relate to that
523
particular interface. Sometimes we discover a bug elsewhere that happens
524
with only one particular transport. Once it's isolated, we can consider
525
whether a test should be added for that particular implementation,
526
or for all implementations of the interface.
528
The multiplication of tests for different implementations is normally
529
accomplished by overriding the ``test_suite`` function used to load
530
tests from a module. This function typically loads all the tests,
531
then applies a TestProviderAdapter to them, which generates a longer
532
suite containing all the test variations.
538
Some utilities are provided for generating variations of tests. This can
539
be used for per-implementation tests, or other cases where the same test
540
code needs to run several times on different scenarios.
542
The general approach is to define a class that provides test methods,
543
which depend on attributes of the test object being pre-set with the
544
values to which the test should be applied. The test suite should then
545
also provide a list of scenarios in which to run the tests.
547
Typically ``multiply_tests_from_modules`` should be called from the test
548
module's ``test_suite`` function.
551
Essential Domain Classes
552
########################
554
Introducing the Object Model
555
============================
557
The core domain objects within the bazaar model are:
567
Transports are explained below. See http://bazaar-vcs.org/Classes/
568
for an introduction to the other key classes.
573
The ``Transport`` layer handles access to local or remote directories.
574
Each Transport object acts like a logical connection to a particular
575
directory, and it allows various operations on files within it. You can
576
*clone* a transport to get a new Transport connected to a subdirectory or
579
Transports are not used for access to the working tree. At present
580
working trees are always local and they are accessed through the regular
581
Python file io mechanisms.
586
Transports work in URLs. Take note that URLs are by definition only
587
ASCII - the decision of how to encode a Unicode string into a URL must be
588
taken at a higher level, typically in the Store. (Note that Stores also
589
escape filenames which cannot be safely stored on all filesystems, but
590
this is a different level.)
592
The main reason for this is that it's not possible to safely roundtrip a
593
URL into Unicode and then back into the same URL. The URL standard
594
gives a way to represent non-ASCII bytes in ASCII (as %-escapes), but
595
doesn't say how those bytes represent non-ASCII characters. (They're not
596
guaranteed to be UTF-8 -- that is common but doesn't happen everywhere.)
598
For example if the user enters the url ``http://example/%e0`` there's no
599
way to tell whether that character represents "latin small letter a with
600
grave" in iso-8859-1, or "latin small letter r with acute" in iso-8859-2
601
or malformed UTF-8. So we can't convert their URL to Unicode reliably.
603
Equally problematic if we're given a url-like string containing non-ascii
604
characters (such as the accented a) we can't be sure how to convert that
605
to the correct URL, because we don't know what encoding the server expects
606
for those characters. (Although this is not totally reliable we might still
607
accept these and assume they should be put into UTF-8.)
609
A similar edge case is that the url ``http://foo/sweet%2Fsour`` contains
610
one directory component whose name is "sweet/sour". The escaped slash is
611
not a directory separator. If we try to convert URLs to regular Unicode
612
paths this information will be lost.
614
This implies that Transports must natively deal with URLs; for simplicity
615
they *only* deal with URLs and conversion of other strings to URLs is done
616
elsewhere. Information they return, such as from ``list_dir``, is also in
617
the form of URL components.
626
We have a commitment to 6 months API stability - any supported symbol in a
627
release of bzr MUST NOT be altered in any way that would result in
628
breaking existing code that uses it. That means that method names,
629
parameter ordering, parameter names, variable and attribute names etc must
630
not be changed without leaving a 'deprecated forwarder' behind. This even
631
applies to modules and classes.
633
If you wish to change the behaviour of a supported API in an incompatible
634
way, you need to change its name as well. For instance, if I add an optional keyword
635
parameter to branch.commit - that's fine. On the other hand, if I add a
636
keyword parameter to branch.commit which is a *required* transaction
637
object, I should rename the API - i.e. to 'branch.commit_transaction'.
639
When renaming such supported API's, be sure to leave a deprecated_method (or
640
_function or ...) behind which forwards to the new API. See the
641
bzrlib.symbol_versioning module for decorators that take care of the
642
details for you - such as updating the docstring, and issuing a warning
643
when the old api is used.
645
For unsupported API's, it does not hurt to follow this discipline, but it's
646
not required. Minimally though, please try to rename things so that
647
callers will at least get an AttributeError rather than weird results.
650
Deprecation decorators
651
----------------------
653
``bzrlib.symbol_versioning`` provides decorators that can be attached to
654
methods, functions, and other interfaces to indicate that they should no
657
To deprecate a static method you must call ``deprecated_function``
658
(**not** method), after the staticmethod call::
661
@deprecated_function(zero_ninetyone)
662
def create_repository(base, shared=False, format=None):
664
When you deprecate an API, you should not just delete its tests, because
665
then we might introduce bugs in them. If the API is still present at all,
666
it should still work. The basic approach is to use
667
``TestCase.applyDeprecated`` which in one step checks that the API gives
668
the expected deprecation message, and also returns the real result from
669
the method, so that tests can keep running.
508
671
Coding Style Guidelines
509
#######################
672
=======================
514
677
``hasattr`` should not be used because it swallows exceptions including
515
678
``KeyboardInterrupt``. Instead, say something like ::
806
932
being phased out.
809
Object string representations
810
=============================
812
Python prints objects using their ``__repr__`` method when they are
813
written to logs, exception tracebacks, or the debugger. We want
814
objects to have useful representations to help in determining what went
817
If you add a new class you should generally add a ``__repr__`` method
818
unless there is an adequate method in a parent class. There should be a
821
Representations should typically look like Python constructor syntax, but
822
they don't need to include every value in the object and they don't need
823
to be able to actually execute. They're to be read by humans, not
824
machines. Don't hardcode the classname in the format, so that we get the
825
correct value if the method is inherited by a subclass. If you're
826
printing attributes of the object, including strings, you should normally
827
use ``%r`` syntax (to call their repr in turn).
829
Try to avoid the representation becoming more than one or two lines long.
830
(But balance this against including useful information, and simplicity of
833
Because repr methods are often called when something has already gone
834
wrong, they should be written somewhat more defensively than most code.
835
The object may be half-initialized or in some other way in an illegal
836
state. The repr method shouldn't raise an exception, or it may hide the
837
(probably more useful) underlying exception.
842
return '%s(%r)' % (self.__class__.__name__,
849
A bare ``except`` statement will catch all exceptions, including ones that
850
really should terminate the program such as ``MemoryError`` and
851
``KeyboardInterrupt``. They should rarely be used unless the exception is
852
later re-raised. Even then, think about whether catching just
853
``Exception`` (which excludes system errors in Python2.5 and later) would
860
All code should be exercised by the test suite. See the `Bazaar Testing
861
Guide <http://doc.bazaar-vcs.org/developers/testing.html>`_ for detailed
862
information about writing tests.
871
We don't change APIs in stable branches: any supported symbol in a stable
872
release of bzr must not be altered in any way that would result in
873
breaking existing code that uses it. That means that method names,
874
parameter ordering, parameter names, variable and attribute names etc must
875
not be changed without leaving a 'deprecated forwarder' behind. This even
876
applies to modules and classes.
878
If you wish to change the behaviour of a supported API in an incompatible
879
way, you need to change its name as well. For instance, if I add an optional keyword
880
parameter to branch.commit - that's fine. On the other hand, if I add a
881
keyword parameter to branch.commit which is a *required* transaction
882
object, I should rename the API - i.e. to 'branch.commit_transaction'.
884
(Actually, that may break code that provides a new implementation of
885
``commit`` and doesn't expect to receive the parameter.)
887
When renaming such supported API's, be sure to leave a deprecated_method (or
888
_function or ...) behind which forwards to the new API. See the
889
bzrlib.symbol_versioning module for decorators that take care of the
890
details for you - such as updating the docstring, and issuing a warning
891
when the old API is used.
893
For unsupported API's, it does not hurt to follow this discipline, but it's
894
not required. Minimally though, please try to rename things so that
895
callers will at least get an AttributeError rather than weird results.
898
Deprecation decorators
899
----------------------
901
``bzrlib.symbol_versioning`` provides decorators that can be attached to
902
methods, functions, and other interfaces to indicate that they should no
903
longer be used. For example::
905
@deprecated_method(deprecated_in((0, 1, 4)))
907
return self._new_foo()
909
To deprecate a static method you must call ``deprecated_function``
910
(**not** method), after the staticmethod call::
913
@deprecated_function(deprecated_in((0, 1, 4)))
914
def create_repository(base, shared=False, format=None):
916
When you deprecate an API, you should not just delete its tests, because
917
then we might introduce bugs in them. If the API is still present at all,
918
it should still work. The basic approach is to use
919
``TestCase.applyDeprecated`` which in one step checks that the API gives
920
the expected deprecation message, and also returns the real result from
921
the method, so that tests can keep running.
923
Deprecation warnings will be suppressed for final releases, but not for
924
development versions or release candidates, or when running ``bzr
925
selftest``. This gives developers information about whether their code is
926
using deprecated functions, but avoids confusing users about things they
1016
In general tests should be placed in a file named test_FOO.py where
1017
FOO is the logical thing under test. That file should be placed in the
1018
tests subdirectory under the package being tested.
1020
For example, tests for merge3 in bzrlib belong in bzrlib/tests/test_merge3.py.
1021
See bzrlib/tests/test_sampler.py for a template test script.
1023
Tests can be written for the UI or for individual areas of the library.
1024
Choose whichever is appropriate: if adding a new command, or a new command
1025
option, then you should be writing a UI test. If you are both adding UI
1026
functionality and library functionality, you will want to write tests for
1027
both the UI and the core behaviours. We call UI tests 'blackbox' tests
1028
and they are found in ``bzrlib/tests/blackbox/*.py``.
1030
When writing blackbox tests please honour the following conventions:
1032
1. Place the tests for the command 'name' in
1033
bzrlib/tests/blackbox/test_name.py. This makes it easy for developers
1034
to locate the test script for a faulty command.
1036
2. Use the 'self.run_bzr("name")' utility function to invoke the command
1037
rather than running bzr in a subprocess or invoking the
1038
cmd_object.run() method directly. This is a lot faster than
1039
subprocesses and generates the same logging output as running it in a
1040
subprocess (which invoking the method directly does not).
1042
3. Only test the one command in a single test script. Use the bzrlib
1043
library when setting up tests and when evaluating the side-effects of
1044
the command. We do this so that the library api has continual pressure
1045
on it to be as functional as the command line in a simple manner, and
1046
to isolate knock-on effects throughout the blackbox test suite when a
1047
command changes its name or signature. Ideally only the tests for a
1048
given command are affected when a given command is changed.
1050
4. If you have a test which does actually require running bzr in a
1051
subprocess you can use ``run_bzr_subprocess``. By default the spawned
1052
process will not load plugins unless ``--allow-plugins`` is supplied.
1058
We have a rich collection of tools to support writing tests. Please use
1059
them in preference to ad-hoc solutions as they provide portability and
1060
performance benefits.
1065
The ``TreeBuilder`` interface allows the construction of arbitrary trees
1066
with a declarative interface. A sample session might look like::
1068
tree = self.make_branch_and_tree('path')
1069
builder = TreeBuilder()
1070
builder.start_tree(tree)
1071
builder.build(['foo', "bar/", "bar/file"])
1072
tree.commit('commit the tree')
1073
builder.finish_tree()
1075
Please see bzrlib.treebuilder for more details.
1080
The ``BranchBuilder`` interface allows the creation of test branches in a
1081
quick and easy manner. A sample session::
1083
builder = BranchBuilder(self.get_transport().clone('relpath'))
1084
builder.build_commit()
1085
builder.build_commit()
1086
builder.build_commit()
1087
branch = builder.get_branch()
1089
Please see bzrlib.branchbuilder for more details.
1094
We make selective use of doctests__. In general they should provide
1095
*examples* within the API documentation which can incidentally be tested. We
1096
don't try to test every important case using doctests -- regular Python
1097
tests are generally a better solution.
1099
Most of these are in ``bzrlib/doc/api``. More additions are welcome.
1101
__ http://docs.python.org/lib/module-doctest.html
1106
Currently, bzr selftest is used to invoke tests.
1107
You can provide a pattern argument to run a subset. For example,
1108
to run just the blackbox tests, run::
1110
./bzr selftest -v blackbox
1112
To skip a particular test (or set of tests), use the --exclude option
1113
(shorthand -x) like so::
1115
./bzr selftest -v -x blackbox
1117
To list tests without running them, use the --list-only option like so::
1119
./bzr selftest --list-only
1121
This option can be combined with other selftest options (like -x) and
1122
filter patterns to understand their effect.
1066
1125
Handling Errors and Exceptions
1067
1126
==============================
1705
1756
As well as prioritizing bugs and nominating them against a
1706
1757
target milestone, Launchpad lets core developers offer to mentor others in
1767
To start a new release cycle:
1769
#. Send mail to the list with the key dates, who will be the release
1770
manager, and the main themes or targetted bugs. Ask people to nominate
1771
objectives, or point out an high-risk things that are best done early,
1772
or that interact with other changes.
1774
#. Add a new "series" in Launchpad at <https://launchpad.net/bzr/+addseries>. There is one
1775
series for every *x.y* release.
1777
Weekly Status Updates
1778
---------------------
1780
TODO: Things to cover:
1782
* Early communication to downstream teams (e.g. Launchpad) about changes in dependencies.
1783
* Reminder re lifecycle and where we're up to right now
1784
* Summary of recent successes and pending work
1785
* Reminder re release objectives
1786
* Reminder re things needing attention, e.g. bug triage, reviews, testing of certain things, etc.
1792
TODO: Get material from http://bazaar-vcs.org/FeatureFreeze.
1796
Making a Release or Release Candidate
1797
-------------------------------------
1799
.. Was previously at http://bazaar-vcs.org/ReleaseChecklist
1801
.. TODO: Still needs more clarity on what's in a RC versus a final
1804
.. TODO: Too much of this is manual but could be automated...
1806
This is the procedure for making a new bzr release:
1808
#. If the release is the first candidate, make a new branch in PQM. (Contact RobertCollins for this step).
1810
Register the branch at https://launchpad.net/products/bzr/+addbranch
1812
#. Run the automatic test suite and any non-automated tests. (For example, try a download over http; these should eventually be scripted though not automatically run.). Try to have all optional dependencies installed so that there are no tests skipped. Also make sure that you have the c extensions compiled (``make`` or ``python setup.py build_ext -i``).
1814
#. In the release branch, update ``version_info`` in ``./bzrlib/__init__.py``
1816
#. Add the date and release number to ``./NEWS``.
1818
#. Update the release number in the README. (It's not there as of 0.15, but please check).
1820
#. Commit these changes to the release branch, using a command like::
1822
bzr commit -m "(jam) Release 0.12rc1."
1824
The diff before you commit will be something like::
1826
=== modified file 'NEWS'
1827
--- NEWS 2006-10-23 13:11:17 +0000
1828
+++ NEWS 2006-10-23 22:50:50 +0000
1831
+bzr 0.12rc1 2006-10-23
1836
=== modified file 'bzrlib/__init__.py'
1837
--- bzrlib/__init__.py 2006-10-16 01:47:43 +0000
1838
+++ bzrlib/__init__.py 2006-10-23 22:49:46 +0000
1840
# Python version 2.0 is (2, 0, 0, 'final', 0)." Additionally we use a
1841
# releaselevel of 'dev' for unreleased under-development code.
1843
-version_info = (0, 12, 0, 'dev', 0)
1844
+version_info = (0, 12, 0, 'candidate', 1)
1846
if version_info[3] == 'final':
1847
version_string = '%d.%d.%d' % version_info[:3]
1849
#. Send the changes to PQM, to update the official master branch.
1851
#. When PQM succeeds, pull down the master release branch.
1853
#. Merge the release branch back into the trunk. Check that changes in NEWS were merged into the right sections. If it's not already done, advance the version number in bzr and bzrlib/__init__.py Submit this back into pqm for bzr.dev.
1855
#. Make a distribution directory by running e.g. ``bzr export /tmp/bzr-<version>/`` in the working directory.
1857
#. Run make in /tmp/bzr-<version>. This creates the extensions from the pyrex source.
1859
#. Run the test suite in the distribution directory
1861
#. Run ``setup.py install`` --root=prefix to do a test install into your system directory, home directory, or some other prefix. Check the install worked and that the installed version is usable. (run the bzr script from the installed path with PYTHONPATH set to the site-packages directory it created). i.e. ::
1863
python setup.py install --root=installed
1864
PYTHONPATH=installed/usr/lib/python2.4/site-packages installed/usr/bin/bzr
1866
#. Clean the tree to get rid of .pyc files etc: make clean && rm -rf build && rm bzrlib/_*.c bzrlib/_*.so
1868
#. Generate the reference documentation in text format: make doc/en/user-reference/bzr_man.txt.
1870
#. Change back to your original branch and then run: make clean && make to create the compiled pyrex extensions. You then need to copy the .c files over to the exported directory.
1872
``find . -name "*.c"`` will tell you which files you need.
1874
#. Create the release tarball::
1876
cd /tmp && tar czf bzr-<version>.tar.gz bzr-<version>
1878
#. Sign the tarball with e.g. ``gpg --detach-sign -a bzr-0.10rc1.tar.gz``
1881
Publishing the release
1882
----------------------
1884
Now you have the releasable product. The next step is making it
1885
available to the world.
1887
#. In <https://launchpad.net/bzr/> click the "Release series" for this
1888
series, to take you to e.g. <https://launchpad.net/bzr/1.1>. Then
1889
click "Register a release", and add information about this release.
1891
#. Within that release, upload the source tarball and the GPG signature.
1893
(These used to also be uploaded to
1894
<sftp://escudero.ubuntu.com/srv/bazaar.canonical.com/www/releases/src>
1895
but that's not accessible to all developers, and gets some mime types
1898
#. Link from http://bazaar-vcs.org/Download to the tarball and signature.
1900
#. Update http://doc.bazaar-vcs.org/ to have a directory of documentation
1901
for this release. (Controlled by the ``update-bzr-docs`` script on
1902
escudero, and also update the ``latest`` symlink in
1903
``/srv/bazaar.canonical.com/doc/``.)
1905
#. Announce on the `Bazaar home page`__
1907
__ http://bazaar-vcs.org/
1910
Announcing the release
1911
----------------------
1913
Now that the release is publicly available, tell people about it.
1915
#. Announce to ``bazaar-announce`` and ``bazaar`` mailing lists.
1916
The announce mail will look something like this:
1918
| Subject: bzr 0.11 release candidate 1
1920
| INTRO HERE. Mention the release number and date, and why the release. (i.e. release candidate for testing, final release of a version, backport/bugfix etc).
1923
| http://bazaar-vcs.org/releases/src/bzr-VERSION.tar.gz
1924
| and GPG signature:
1925
| http://bazaar-vcs.org/releases/src/bzr-VERSION.tar.gz.sig
1927
| DESCRIBE-CHANGES-IN-OVERVIEW-HERE
1929
| DESCRIBE-when the next release will be (if there is another - i.e. this is a release candidate)
1931
| Many thanks to all the contributors to this release! I've included the
1932
| contents of NEWS for VERSION below:
1934
To generate the data from NEWS, just copy and paste the relevant news section and clean it up as appropriate. The main clean-up task is to confirm that all major changes are indeed covered. This can be done by running ``bzr log`` back to the point when the branch was opened and cross checking the changes against the NEWS entries.
1936
(RC announcements should remind plugin maintainers to update their plugins.)
1938
* For point releases (i.e. a release candidate, or an incremental fix to a released version) take everything in the relevant NEWS secion : for 0.11rc2 take everything in NEWS from the bzr 0.11rc2 line to the bzr 0.11rc1 line further down.
1940
* For major releases (i.e. 0.11, 0.12 etc), take all the combined NEWS sections from within that version: for 0.11 take all of the 0.11 specific section, plus 0.11rc2, plus 0.11rc1 etc.
1942
#. Update the `news side menu`__ -- this currently requires downloading the file, editing it, deleting it, and uploading a replacement.
1944
__ http://bazaar-vcs.org/site/menu?action=AttachFile&do=view&target=news.html
1946
#. Update the IRC channel topic. Use the ``/topic`` command to do this, ensuring the new topic text keeps the project name, web site link, etc.
1948
#. Announce on http://freshmeat.net/projects/bzr/
1950
This should be done for both release candidates and final releases. If you do not have a Freshmeat account yet, ask one of the existing admins.
1952
#. Update http://en.wikipedia.org/wiki/Bzr -- this should be done for final releases but not Release Candidates.
1954
#. Package maintainers should update packages when they see the
1959
#. Post to http://mail.python.org/mailman/listinfo/python-announce-list for major releases
1961
#. Update the python package index: <http://pypi.python.org/pypi/bzr> - best
1964
python setup.py register
1966
Remember to check the results afterwards.
1969
Making Win32 installers
1970
-----------------------
1972
**XXX:** This information is now probably obsolete, as Alexander uploads
1973
direct to Launchpad. --mbp 20080116
1975
Alexander Belchenko has been very good about getting packaged installers compiled (see Win32ReleaseChecklist for details). He generally e-mails John Arbash Meinel when they are ready. This is just a brief checklist of what needs to be done.
1977
#. Download and verify the sha1 sums and gpg signatures. Frequently the sha1 files are in dos mode, and need to be converted to unix mode (strip off the trailing ``\r``) before they veryify correctly.
1979
#. Upload to the Launchpad page for this release.
1981
#. Upload to escudero (to the b.c.c/www/releases/win32 directory) using sftp, lftp or rsync
1983
#. Cat the contents of the .sha1 files into the SHA1SUM.
1985
#. Update the SHA1SUM and MD5SUM files using something like ``md5sum bzr-0.14.0.win32.exe >> MD5SUM``. Make sure you use append (>>) rather than overwrite (>).
1987
#. Verify once again that everything is correct with ``sha1sum -c SHA1SUM`` and ``md5sum -c MD5SUM``.
1989
#. Update ``.htaccess`` so that the 'bzr-latest.win32.exe' links point to the latest release. This is not done for candidate releases, only for final releases. (example: bzr-0.14.0, but not bzr-0.14.0rc1).
1991
#. Make sure these urls work as expected:
1993
http://bazaar-vcs.org/releases/win32/bzr-latest.win32-py2.5.exe
1995
http://bazaar-vcs.org/releases/win32/bzr-latest.win32-py2.5.exe.asc
1997
http://bazaar-vcs.org/releases/win32/bzr-latest.win32-py2.4.exe
1999
http://bazaar-vcs.org/releases/win32/bzr-latest.win32-py2.4.exe.asc
2001
http://bazaar-vcs.org/releases/win32/bzr-setup-latest.exe
2003
http://bazaar-vcs.org/releases/win32/bzr-setup-latest.exe.asc
2005
They should all try to download a file with the correct version number.
2007
#. Update http://bazaar-vcs.org/Download to indicate the newly available versions.
2009
#. Update http://bazaar-vcs.org/WindowsDownloads to have the correct version number as well as the correct sha1sum displayed.
2012
The Bazaar PPA archive
2013
----------------------
2015
We build Ubuntu ``.deb`` packages for Bazaar as an important part of the release
2016
process. These packages are hosted in a `Personal Package Archive (PPA)`__ on
2017
Launchpad, at <https://launchpad.net/~bzr/+archive>.
2019
__ https://help.launchpad.net/PPAQuickStart
2021
We build packages for every supported Ubuntu release
2022
<https://wiki.ubuntu.com/Releases>. Packages need no longer be updated
2023
when the release passes end-of-life because all users should have then
2026
The ``debian/`` directory containing the packaging information is kept in
2027
branches on Launchpad, named like
2028
<https://code.launchpad.net/~bzr/bzrtools/packaging-dapper>
2030
Updating the PPA for a new release
2031
----------------------------------
2033
Preconditions for building these packages:
2035
* You must have a Launchpad account and be a member of the `~bzr`__ team
2037
__ https://edge.launchpad.net/~bzr/+members>
2039
* You must have a GPG key registered to your Launchpad account.
2041
* Configure ``dput`` to upload to our PPA with this section in your
2045
fqdn = ppa.launchpad.net
2047
incoming = ~bzr/ubuntu
2049
allow_unsigned_uploads = 0
2051
* You need a Ubuntu (or probably Debian) machine, and ::
2053
sudo apt-get install build-essential devscripts dput
2055
Here is the process; there are some steps which should be automated in
2058
#. You will need a working directory for each supported release, such as
2059
``~/bzr/Packaging/dapper``
2061
#. Download the official tarball of the release to e.g. ``~/bzr/Releases``
2063
#. Copy the original tarball into your per-disto directory, then untar it
2064
and if necessary rename it::
2066
cp -l ~/bzr/Releases/bzrtools-1.3.0.tar.gz bzrtools_1.3.0.orig.tar.gz
2067
tar xfvz bzrtools_1.3.0.orig.tar.gz
2068
mv bzrtools bzrtools-1.3.0
2070
#. Change into that directory and check out the packaging branch::
2074
bzr+ssh://bazaar.launchpad.net/~bzr/bzrtools/packaging-dapper \
2077
#. For Bazaar plugins, change the ``debian/control`` file to express a
2078
dependency on the correct version of ``bzr``.
2080
For bzrtools this is typically::
2082
Build-Depends-Indep: bzr (>= 1.3~), rsync
2083
Depends: ${python:Depends}, bzr (>= 1.3~), bzr (<< 1.4~), patch
2085
#. Make a new ``debian/changelog`` entry for the new release,
2086
either by using ``dch`` or just editing the file::
2088
dch -v '1.3.0-1~bazaar1~dapper1' -D dapper
2090
dch will default to the distro you're working in and this isn't checked
2091
against the version number (which is just our conversion). So make
2094
Make sure you have the correct email address for yourself, version
2095
number, and distribution. It should look something like this::
2097
> bzrtools (1.3.0-1~bazaar1~dapper1) dapper; urgency=low
2099
> * New upstream release.
2101
> -- John Sample <sample@example.com> Mon, 31 Mar 2008 12:36:27 +1100
2103
If you need to upload the package again to fix a problem, normally you
2104
should increment the last number in the version number, following the
2105
distro name. Make sure not to omit the initial ``-1``, and make sure
2106
that the distro name in the version is consistent with the target name
2107
outside the parenthesis.
2109
#. Commit these changes into the packaging branch::
2111
bzr ci -m '1.3.0-1~bazaar1~dapper1: New upstream release.' debian
2113
#. Build a source package::
2117
This will create a ``.changes`` file in the per-distro directory,
2118
and should invoke gpg to sign it with your key.
2119
Check that file is reasonable: it should be uploading to the intended
2120
distribution, have a .orig file included, and the right version number.
2122
#. Upload into the PPA::
2124
dput bzr-ppa ../bzrtools__1.3.0-1\~bazaar1\~dapper1_source.changes
2126
Don't forget the ``bzr-ppa`` component or dput will try to upload into
2127
the main archive by default. You can disable this by adding this
2128
section to your ``.dput.cf``::
2131
fqdn = SPECIFY.A.PPA.NAME
2133
#. You should soon get an "upload accepted" mail from Launchpad, which
2134
means that your package is waiting to be built. You can then track its
2135
progress in <https://launchpad.net/~bzr/+archive> and
2136
<https://launchpad.net/~bzr/+archive/+builds>.
added
removed
doc/developers/HACKING.txt
