1
=============================
2
Bazaar Architectural Overview
3
=============================
5
This document describes the key classes and concepts within Bazaar. It is
6
intended to be useful to people working on the Bazaar codebase, or to
7
people writing plugins. People writing plugins may also like to read the
8
guide to `Integrating with Bazaar <integration.html>`_ for some specific
11
If you have any questions, or if something seems to be incorrect, unclear
12
or missing, please talk to us in ``irc://irc.freenode.net/#bzr``, or write
13
to the Bazaar mailing list.
22
When using bzrlib within the ``bzr`` program (for instance as a bzr
23
plugin), bzrlib's global state is already available for use.
28
To use bzrlib outside of ``bzr`` some global state needs to be setup.
29
bzrlib needs ways to handle user input, passwords, a place to emit
30
progress bars, logging setup appropriately for your program. The easiest
31
way to set all this up in the same fashion ``bzr`` does is to call
32
``bzrlib.initialize``. This returns a context manager within which bzrlib
33
functions will work correctly. See the pydoc for ``bzrlib.initialize`` for
34
more information. In Python 2.4 the ``with`` keyword is not supported and
35
so you need to use the context manager manually::
37
# This sets up your ~/.bzr.log, ui factory and so on and so forth. It is
38
# not safe to use as a doctest.
39
library_state = bzrlib.initialize()
40
library_state.__enter__()
45
library_state.__exit__(None, None, None)
54
The ``Transport`` layer handles access to local or remote directories.
55
Each Transport object acts as a logical connection to a particular
56
directory, and it allows various operations on files within it. You can
57
*clone* a transport to get a new Transport connected to a subdirectory or
60
Transports are not used for access to the working tree. At present
61
working trees are always local and they are accessed through the regular
62
Python file I/O mechanisms.
67
Transports work in terms of URLs. Take note that URLs are by definition
68
only ASCII - the decision of how to encode a Unicode string into a URL
69
must be taken at a higher level, typically in the Store. (Note that
70
Stores also escape filenames which cannot be safely stored on all
71
filesystems, but this is a different level.)
73
The main reason for this is that it's not possible to safely roundtrip a
74
URL into Unicode and then back into the same URL. The URL standard
75
gives a way to represent non-ASCII bytes in ASCII (as %-escapes), but
76
doesn't say how those bytes represent non-ASCII characters. (They're not
77
guaranteed to be UTF-8 -- that is common but doesn't happen everywhere.)
79
For example, if the user enters the URL ``http://example/%e0``, there's no
80
way to tell whether that character represents "latin small letter a with
81
grave" in iso-8859-1, or "latin small letter r with acute" in iso-8859-2,
82
or malformed UTF-8. So we can't convert the URL to Unicode reliably.
84
Equally problematic is if we're given a URL-like string containing
85
(unescaped) non-ASCII characters (such as the accented a). We can't be
86
sure how to convert that to a valid (i.e. ASCII-only) URL, because we
87
don't know what encoding the server expects for those characters.
88
(Although it is not totally reliable, we might still accept these and
89
assume that they should be put into UTF-8.)
91
A similar edge case is that the URL ``http://foo/sweet%2Fsour`` contains
92
one directory component whose name is "sweet/sour". The escaped slash is
93
not a directory separator, but if we try to convert the URL to a regular
94
Unicode path, this information will be lost.
96
This implies that Transports must natively deal with URLs. For simplicity
97
they *only* deal with URLs; conversion of other strings to URLs is done
98
elsewhere. Information that Transports return, such as from ``list_dir``,
99
is also in the form of URL components.
105
A workingtree is a special type of Tree that's associated with a working
106
directory on disk, where the user can directly modify the files.
110
* Maintaining a WorkingTree on disk at a file path.
111
* Maintaining the basis inventory (the inventory of the last commit done)
112
* Maintaining the working inventory.
113
* Maintaining the pending merges list.
114
* Maintaining the stat cache.
115
* Maintaining the last revision the working tree was updated to.
116
* Knows where its Branch is located.
121
* an MutableInventory
122
* local access to the working tree
127
A Branch is a key user concept - its a single line of history that one or
128
more people have been committing to.
130
A Branch is responsible for:
132
* Holding user preferences that are set in a Branch.
133
* Holding the 'tip': the last revision to be committed to this Branch. (And the revno of that revision.)
134
* Knowing how to open the Repository that holds its history.
135
* Allowing write locks to be taken out to prevent concurrent alterations to the branch.
138
* URL access to its base directory.
139
* A Transport to access its files.
140
* A Repository to hold its history.
145
Repositories store committed history: file texts, revisions, inventories,
146
and graph relationships between them. A repository holds a bag of
147
revision data that can be pointed to by various branches:
149
* Maintains storage of various history data at a URL:
151
* Revisions (Must have a matching inventory)
153
* Inventories for each Revision. (Must have all the file texts available).
156
* Synchronizes concurrent access to the repository by different
157
processes. (Most repository implementations use a physical
158
mutex only for a short period, and effectively support multiple readers
164
A repository can be configured to refer to a list of "fallback"
165
repositories. If a particular revision is not present in the original
166
repository, it refers the query to the fallbacks.
168
Compression deltas don't span physical repository boundaries. So the
169
first commit to a new, empty repository with fallback repositories will
170
store a full text of the inventory, and of every new file text.
172
At runtime, repository stacking is actually configured by the branch, not
173
the repository. So doing ``a_bzrdir.open_repository()``
174
gets you just the single physical repository, while
175
``a_bzrdir.open_branch().repository`` gets one configured with a stacking.
176
Therefore, to permanently change the fallback repository stored on disk,
177
you must use ``Branch.set_stacked_on_url``.
179
Changing away from an existing stacked-on URL will copy across any
180
necessary history so that the repository remains usable.
182
A repository opened from an HPSS server is never stacked on the server
183
side, because this could cause complexity or security problems with the
184
server acting as a proxy for the client. Instead, the branch on the
185
server exposes the stacked-on URL and the client can open that.