Differences

This shows you the differences between two versions of the page.

--- devel:hacking [2009/02/09 12:52]
morten update down to, but not including the web interface part
+++ devel:hacking [2013/06/19 07:45]
morten fix jenkins url
@@ Line 1: / Line 1: @@
-:!: **Be warned, this information is becoming very out of date, it is currently being updated** :!:
 ====== Hacker's guide to NAV ======
@@ Line 6: / Line 5: @@
-===== Participating in the community =====
+====== Contributing to NAV ======
 Originally, NAV was a closed source project, initiated by the
 Norwegian University of Science and Technology (NTNU), and eventually
@@ Line 14: / Line 13: @@
 software system.
-While NTNU and UNINETT are still the main contributors to NAV,
+While UNINETT and NTNU are still the main contributors to NAV,
 developing NAV to support the needs of the Norwegian higher education
 community, contributions from third parties is highly appreciated.
 We communicate mainly through mailing lists,
-[[https://launchpad.net/nav/|Launchpad]], and the wiki and Mercurial
+[[https://launchpad.net/nav/|Launchpad]] and the ''#nav'' IRC channel
-repositories hosted at http://metanav.uninett.no/ .  At times, UNINETT
+on //FreeNode//.  At times, UNINETT also arranges workshops and
-also arranges workshops and gatherings for its customers: Norwegian
+gatherings for its customers: Norwegian universities, university
-universities, university colleges and research institutions.
+colleges and research institutions.
 To contribute:
-Go to http://metanav.uninett.no/ and
+Go to http://nav.uninett.no/ and
   * Join the mailing lists.  The //nav-dev// mailing list in
@@ Line 32: / Line 31: @@
     low traffic list. We can only hope this will change ;-)
   * Get a copy of the latest development sources by cloning the
-    Mercurial repository at http://metanav.uninett.no/hg/default/.
+    Mercurial repository at http://nav.uninett.no/hg/default/.
     Most new development takes place on this branch.
   * Take a look at the [[:navprojects|project reports from previous
@@ Line 41: / Line 40: @@
     not hesitate to ask for help on the mailing lists.
-If you wish to contribute code to the project, please tell us about it
+If you wish to contribute code to the project, see the [[#submitting patches]] section.
-on the nav-dev mailing list.  It is always a good idea to check if
-someone is already working on something similar, and to get some
-helpful tips on how to integrate your code with the rest of the
-project.  If you already went ahead and wrote a patch, announce it on
-nav-dev and provide a link to the patch so it can be studied for
-possible inclusion into NAV.
-===== Directory layout =====
+====== Directory layout ======
 A rough guide to the source tree:
-| conf/      | Files related to the autoconf build system. |
+| bin/       | NAV 'binaries'; executable scripts and programs. |
-| doc/       | User and developer documentation, SQL scripts and example NAV configuration files. |
+| contrib/   | User contributed NAV tools. NAV doesn't depend on these, and any maintenance of them is left up to the original developers.  We do not offer support for these tools. |
+| doc/       | User and developer documentation |
+| etc/       | Example/initial configuration files |
+| java/      | Java source code |
+| media/     | Static media such as CSS stylesheets, images and JavaScript to be served by a webserver |
+| packages/  | Stuff to help packaging NAV for various platforms, such as RedHat, CentOS, FreeBSD, Debian and soforth. Much of this is outdated today. |
+| python/    | Python source code |
+| sql/       | SQL schema definitions and installation/sync tools |
+| templates/ | Django HTML templates |
+| tests/     | Automated tests |
 | tools/     | Tool scripts for the build and release processes. |
-| contrib/   | User contributed NAV tools. NAV doesn't depend on these, and any maintenance of them is left up to the original developers.  We do not offer support for these tools. |
-| packages/  | Stuff to help packaging NAV for various platforms, such as RedHat, CentOS, FreeBSD, Debian and soforth. |
-| src/       | Source code to Java subsystems of NAV (here for historic reasons). |
-| subsystem/ | Source code to the rest of NAV - most of it Python.  NAV is loosely divided into subsystems, and each one of these has its own subdirectory in here. |
-| subsystem/lib-python/ | Python libraries & APIs.  Please check what's already there before you roll your own. |
-| subsystem/lib-perl/ | Perl libraries & APIs.  Only a single piece of Perl code remains in NAV; once this has been replaced, this directory will cease to exist. |
-| subsystem/webfront/ | Python libraries for the web interface and front-page handler modules for mod_python. |
-===== Development languages and frameworks =====
+====== Development languages and frameworks ======
-For historic reasons, different parts of NAV are written in different
-programming languages (Perl, Java, PHP and Python).  This has been
-unfortunate in may ways, not at least for the sake of consistency and
-maintenance, but fortunately we have a long-term goal of reducing the
-number of languages and dependencies.  The last few years we've spent
-a significant amount of time rewriting parts that were written in Perl
-and PHP to Python, which is the language we are currently gravitating
-towards.
-Currently (as of February 2009), major parts of NAV are written in
+Historically, NAV was written using multiple programming languages
-Python and Java, while only a single Perl program remains
+(Perl, Java, PHP and Python).  While this has had an unfortunate
-(''makecricketconfig.pl'').
+impact on integration and maintenance over the years, we've managed to
+reduce this to just Python and Java in later years. We have a
+long-term goal to rewrite the remaining Java backend code to Python.
-When contributing patches to existing code, or plugins to existing
+Currently (as of September 2012), NAV consists mostly of Python code,
-subsystems, use the language that subsystem was written in.
+with one remaining backend systems written in Java (eventEngine).
-When writing entirely new subsystems, the following rules apply:
+  * We will only accept new code written in Python (except when it
+    involves patches to the existing Java code).
+  * When you contribute additions to the web interface, use the Django framework.
-  * If your subsystem is a new tool for the web interface, use Python
+If you wish to contribute something really useful that doesn't use
-    and Django.  The legacy parts of the web system interface directly
+Python, we may consider including it in the //contrib// directory.
-    with ''mod_python'', using Cheetah for HTML templating.  Anything
-    written in 2008 and later uses Django.  There is also
-    [[devel:django_introduction|a guide for interfacing Django
-    applications with the legacy web code]].
-  * If your subsystem is a new back-end tool/daemon, please use
+If **YOU** are willing to invest in porting some of the existing Java
-    Python.  The NAV Python API is more complete than for any of the
+code to Python, then you will be celebrated as a NAV hero!
-    other languages, and you will receive a lot for free.
-If these rules are not followed, your patches will not be accepted
+====== Coding style ======
-into NAV (but if they are really good, we will consider including them
-in the //contrib// directory).
-If *YOU* are willing to invest in porting some of the existing
-Java/Perl code to Python, then you will be celebrated as a NAV hero!
-===== Coding style =====
 Much of the legacy NAV code was written without using any coding style
 guidelines.  This has resulted in some chaotic combination of styles,
@@ Line 112: / Line 89: @@
   * For Python code, please refer to PEP-8, "Style Guide for Python
     Code" http://www.python.org/doc/peps/pep-0008/
-FIXME Add info about header guidelines for Python files.
 If you see violations of these guidelines, don't hesitate to fix them.
@@ Line 119: / Line 94: @@
 a separate patch to make your other patches look clean and readable.
+===== Python boilerplate headers =====
+We will generally only accept code into NAV if it is licensed under
+GPL v2, but we may make individual exceptions for code licensed under
+compatible licenses.  Each Python source code file should contain the
+following boilerplate at the top:
-===== Database connections =====
+<code python>
-NOTE: The following is Python-specific, more info should be added for
+#
-the other languages used in NAV.
+# Copyright (C) 2008,2009 Somebody
+#
+# This file is part of Network Administration Visualized (NAV).
+#
+# NAV is free software: you can redistribute it and/or modify it under the
+# terms of the GNU General Public License version 2 as published by the Free
+# Software Foundation.
+#
+# This program is distributed in the hope that it will be useful, but WITHOUT
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+# FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+# more details.  You should have received a copy of the GNU General Public
+# License along with NAV. If not, see <http://www.gnu.org/licenses/>.
+#
+</code>
+If a file uses non-ASCII characters, it **must** be encoded as UTF-8, and an
+encoding statement should be inserted at the top:
+<code python>
+# -*- coding: utf-8 -*-
+</code>
+===== Javascript =====
+When writing javascript code try to focus on modules not pages. If the code is html-related, it should take selectors or objects as input and concern itself solely about those. This makes for much easier testing and reuse. And of course - write the tests first.
+When the module is done you write a controller for the page that plugs the needed plugins to the page elements. This should fail gracefully if the needed elements are not present.
+NAVs javascript uses [[http://requirejs.org/|require.js]] - use this to create modules and specify dependencies.
+Pro tip is to create ''require_config.dev.js'' in ''media/js/'' and add the following configuration to requirejs:
+<code>require.urlArgs = "bust=" +  (new Date()).getTime();</code> This makes sure your not using cached resources in your browser when developing, which browsers loves to do! See [[http://requirejs.org/docs/api.html#config-urlArgs|config-urlArgs]] in requirejs documentation for «details». The ''require_config.dev.js'' is added in global HG ignore.
+==== Accessing resources with ajax requiring authentication ====
+As your authenticated session might have timed out due to idle (no activity), resources will return 500 Internal Error if you do not supply the important ''//X-NAV-AJAX//'' header on your ajax requests.
+So make sure to include the ''**X-NAV-AJAX**'' header so you will get proper HTTP response code in the reponse from your request.
+NAV has a shortcut for fixing this in ''default.js'' (adds a function in the public namespace NAV) which attaches the required handlers for jQuery by doing:
+<code>NAV.addGlobalAjaxHandlers()</code>
+====== Database ======
 NAV uses PostgreSQL as its database backend.  Namespaces (schemas) are
 employed to logically group tables and relations.  NAV versions prior
@@ Line 134: / Line 156: @@
 | ''logger''      | Anything related to NAV's syslog parser/browser system. |
 | ''arnold''      | The port detention system Arnold stores it's data here. |
+| ''radius''      | Radius accounting logs, updated directly by FreeRadius' PostgreSQL module. |
+===== Connecting to the database (Python) ====
+==== Raw SQL ====
 To obtain a connection to the NAV database, use the API accordingly,
@@ Line 154: / Line 180: @@
 database user for a subsystem called ''default'', and also specifies
 the same database user for all known subsystem names.  At present,
-using a subsystem name that is not configured in ''db.conf'' will
+using a subsystem name that is not configured in ''db.conf'' will cause ''nav.db.getConnection()'' to revert to using the ''default'' name.
-raise an exception in ''nav.db.getConnection()''.
-===== Web interface =====
+==== Django models ====
-When programming for NAV's web interface, a few special considerations
+NAV 3.5 and on includes Django models for most database tables.  If no
-need to be made.
+SQL magic is needed to perform your database voodoo, it is recommended
+that you use these models, located in the module ''nav.models''.  You
+do not need to explicitly establish a database connection to use these
+models, as Django takes care of all that.
-==== Mod_python ====
+The models are defined in modules of the ''nav.models'' package.
-NAV uses mod_python to interface with the Apache web server.  See
-http://www.modpython.org/.
-==== Cheetah Templates ====
+===== Changing the schema ====
-The NAV web interface makes extensive use of Cheetah templates for
-generating its HTML output, see http://www.cheetahtemplate.org/ .
-Most of the existing Cheetah templates are to be found in
+The baseline schema is located in ''sql/baseline'' - the ''syncdb.py'' script
-subsystem/webfront/nav/web/templates, although some of NAV's subsystems
+is responsible for running this when creating a new database. To make a schema
-store their templates along with their code in their respective subsystem
+change, you **do not** change the baseline, but go to the ''sql/changes''
-subdirectories. The compiled templates should be placed in the
+directory and create a new schema change script there.
-nav.web.templates package.
-If you are making a new web module for NAV, your module's Cheetah
+Schema change scripts as numbered, using the following pattern:
-template should subclass MainTemplate.tmpl found in
-subsystem/webfront/nav/web/templates/. See other templates for code
-examples of how to inherit from this template.
-==== Database connections in the web interface ====
+  * ''sc.<major>.<minor>.<point>.sql''
-As stated above, use the nav.db.getConnection function to open or
-retrieve an existing database connection.  All NAV web modules share
-the same interpreter and namespace per Apache process, which also
-means that database connections will be shared between the modules
-running in each process.  Therefore, the following conventions apply
-for connections obtained from nav.db.getConnection:
-  * Do not, under any circumstances, retain references to a database
+The ''<major>'' and ''<minor>'' numbers usually correspond to the major and
+minor number of the next NAV release.  The ''<point>'' number is a sequence id
+- pick the next free number when creating a schema change script.
+Remember these points when creating a schema change script:
+  * Create separate change scripts for unrelated schema changes.
+  * Remember to write SQL to //migrate// existing data, if necessary.
+  * Do not use transactional statements - the ''syncdb.py'' script will take
+    care of that.
+To apply your change scripts, just run ''syncdb.py''.  It will look inside the
+''schema_change_log'' table to see which change scripts have already been
+applied, and it will detect your new change script and apply this to the
+database.
+:!: When changing the schema, don't forget to update the Django models in the
+''nav.models'' package.  An integration test exists to verify that the Django
+models can at least be used to run proper SELECTs against the database.
+====== Legacy web code ======
+Legacy web code interfaces directly with
+[[http://www.modpython.org/|mod_python]], and uses
+[[http://www.cheetahtemplate.org/|Cheetah for HTML templating]].
+All Cheetah templates are located in the ''python/nav/web/templates''
+directory.
+===== Legacy database connections in web code =====
+Use the ''nav.db.getConnection()'' call to open or retrieve an
+existing database connection.  All NAV web modules share the same
+interpreter and namespace per Apache process, which also means that
+database connections will be shared between the modules running in
+each process.  Therefore, the following conventions apply for
+connections obtained from ''nav.db.getConnection()'':
+  * **Do not, under any circumstances**, retain references to a database
     connection between client requests.  Make sure to retrieve a new
     connection at the start of each request cycle - the API will cache
@@ Line 195: / Line 245: @@
     modules, retained references may be invalid in the next request
     cycle.
-  * Do not explicitly close database connections.  Although the API
+  * **Do not explicitly close database connections.**  Although the API
     will try to reopen any closed or broken connections, you create
     extra overhead, and you don't play nice with the other web
     modules.
-  * Do not enable/disable autocommit or alter a connection's
+  * **The obtained connections will use an isolation level of //read
-    transaction isolation level, unless you make pretty darn sure to
+    committed//**, i.e. no autocommits.  Be careful to commit the
-    reset them to their original states at the end of a request cycle.
+    current transaction if you modify any data.  A mod_python
-  * NAV 3.0 makes connections autocommit by default, whereas NAV 3.1
+    ''cleanuphandler'' will try to automatically commit all open
-    will not - make sure to commit your transactions when needed.  NAV
+    transactions as the request cycle ends, but this may change
-    will help unfortunate souls by attempting to commit transactions
+    in the future, so you must not rely on it.
-    in a mod_python cleanuphandler, but you should nevertheless
+  * **Do not change the isolation level of a connection** without
-    explicitly call connection.commit to avoid having your
+    restoring it to its original value before the end of the request
-    transactions accidentally rolled back.
+    cycle.
+===== The "death" of mod_python =====
+''mod_python'' is no longer under active development and has been
+placed in the Apache foundation's "Attic".  We do not accept new web
+tools that interface directly with ''mod_python''.
+We do, however, aim to refactor existing mod_python-interfacing code
+into working as Django views.  A few tips for such refactorings:
+  * Each ''mod_python'' handler in NAV mostly performs its own custom
+    URL parsing and view dispatch.  It's best to refactor this into a
+    Django URL configuration and separate view functions first.
+  * Usage of ''mod_python.utils.FieldStorage'' parse URI arguments
+    must be refactored to use the ''POST'', ''GET'' or ''REQUEST''
+    objects of a Django ''HttpRequest''. It's not that hard, as these
+    objects behave like dictionaries, much like the ''FieldStorage''
+    class does.
+  * Conversion from Cheetah to Django templates is not necessary to
+    refactor a mod_python handler into a Django view.  It is desirable
+    to do so in later refactorings, though.
+  * NAV's authentication and authorization scheme hooks into Apache's
+    request cycle using a ''mod_python'' ''headerparserhandler''. It
+    also adds session data as an attribute to the ''mod_python''
+    request object.  Once there are no tools left that interface
+    directly with ''mod_python'', the auth and session parts of NAV
+    must be refactored to work in a pure Django setting before NAV can
+    be free of its dependence on ''mod_python''.
+====== Writing new web code ======
+If you are writing a new web application / tool for NAV, please use
+the Django framework.  [[devel:django_introduction|Here's a quick
+primer on how Django integrates with legacy NAV]].
+====== Version Control ======
+NAV uses [[http://www.selenic.com/mercurial/|Mercurial]] for
+distributed version control.  Official repositories are located at
+http://nav.uninett.no/hg/ .
+===== Guide to the repository jungle =====
+The official repositories represent three types of branches
+==== Unstable (default) ====
+New, bleeding edge development occurs on the
+//[[http://nav.uninett.no/hg/default/|default]]// branch, which is
+considered unstable (although we try to always keep it buildable).
+==== Feature branches ====
+New features that take a while (and a lot of changesets) to implement
+and test will often be published as separate feature branches.  For
+all intents and purposes, the feature branches will look like the
+//default// branch with some added feature.  They will merge changes
+from the //default// branch regularly.  Once a feature is considered
+"ready", the feature branch will be merged onto the default branch.
+==== Stable (series) ====
+Once we are nearing a new series release of NAV (such as 3.5 or 3.6),
+a new [[http://nav.uninett.no/hg/series/|series branch]] is
+created from the //default// branch.  Once this branch is stabilized,
+the first version is tagged and released.  After this point, we accept
+only bug fixes in this branch.  Further point releases in this series
+are tagged on this branch, and all changes are merged back onto the
+//default// branch.
+When someone writes a patch for a bug, this should usually be
+committed to the latest active series branch which is affected by the
+bug.  Once a new series is released, we do not usually maintain the
+older series branches.  We may push bug fixes to these branches, but
+we are unlikely to create a new point release from it.
+===== Push access =====
+Push access to the official repositories is limited to developers
+employed or commisioned by UNINETT.
+====== Testing and Continuous Integration ======
+Much of NAV is **legacy code**, as defined by //Michael C. Feathers//:
+Code that has no tests.  We have been making an effort to introduce
+automated tests into the codebase the past couple of years, and hope
+to improve coverage in time.
+There are no tests for the legacy Java code, but many unit tests and
+integration tests now reside in the ''tests/'' subdirectory.
+===== Running tests =====
+We use ''[[http://pytest.org/|py.test]]'' to run the test suite.  A bundled version is
+included as ''runtests.py'' in the ''python/'' subdirectory, which is
+used to run the unit tests only when a ''make check'' command is
+issued in the ''python/'' subdirectory.
+Some of the test requirements aren't available on the Debian systems
+we use for development, so we often test inside a Python
+//virtualenv//.  A suitable virtualenv for testing (on Debian Lenny)
+can be created thus:
+<code bash>
+virtualenv .env
+. .env/bin/activate
+easy_install pip
+pip install -r tests/requirements.txt
+</code>
+There's also a script to create a test environment, complete with
+database initialization.  This is used by our CI server.  The
+following will configure and build NAV automatically, and install it
+into a directory called ''workspace/build''.  It will also create a
+suitable virtualenv in ''workspace/.env'', which you can activate
+before running tests:
+<code bash>
+export PGDATABASE=testdb
+export PGUSER=testuser
+tests/bootstrap-test-environment.sh workspace
+</code>
+===== Javascript testing =====
+Testing of javascript is in its infancy in NAV. We are currently using [[http://busterjs.org/|buster.js]] as testing toolkit.
+To install buster.js install [[http://nodejs.org/|node]] and then:
+<code>
+npm install -g buster
+</code>
+As we use [[http://requirejs.org/|require.js]] you need the AMD module of buster aswell. Install it in the /media/js directory:
+<code>
+npm install buster-amd
+</code>
+To run the tests you need to
+  - Start a buster server by typing ''buster-server''
+  - Capture browsers by pointing browsers to the buster-server (default localhost:1111)
+  - Go to /media/js
+  - Run the tests by typing ''buster-test''
+All tests are located under ''media/js/tests/''. Create new tests there. For syntax, assertions and related stuff take a look at the tests already there and [[http://busterjs.org/docs/|the buster docs]].
+===== Jenkins =====
+We use //Jenkins// (formerly //Hudson//) for Continuous Integration testing of
+NAV.  All the automated tests are run each time new changesets are pushed to
+the NAV repositories.  Jenkins also runs pylint to create stats on code
+quality.
+Our Jenkins installation is available on https://ci.nav.uninett.no/ .
+===== Tips and tricks =====
+===== Make fixtures for integration testing =====
+<code>
+from django.core import serializers
+from nav.models.manage import Netbox
+fixtures = serializers.serialize("xml", Netbox.objects.all()[:2])
+</code>
+Fixtures can so be used in your integration tests by extending
+the test case DjangoTransactionTestCase in nav.tests.cases
+See nav.tests.integration.l2trace_test for an example on applying
+fixtures for your particular test case.
+Also keep in mind you have to make sure you have the model
+dependency in correct order when importing.
+Example: Netbox contains a location to a Room where it is located,
+you have to make sure Room's are imported first before importing
+Netbox's
+See https://docs.djangoproject.com/en/dev/topics/serialization/
+TODO: Be able to use [[https://docs.djangoproject.com/en/dev/ref/django-admin/#dumpdata-appname-appname-appname-model|django-admin's management command: dumpdata]]
+to create fixtures.
+====== Submitting patches ======
+Unless you are submitting one-off fixes for bugs and small issues,
+please take the time to discuss your change proposals on the
+//nav-dev// mailing list.  This will increase the chances of having
+your patches accepted.
+Base your patches on the relevant Mercurial branches.  If you are
+submitting a patch for an issue that affects the latest stable series,
+base your patch on that series branch.  If you are submitting patches
+containing new features, base them on the default branch.
+There are three common options for submitting patches:
+  * The best way to submit your patches would be using Mercurial. Publish
+    your own Mercurial branch, and mail its URL to the //nav-dev// mailing
+    list.
+  * If unable to host a public Mercurial branch, export your changes
+    as a Mercurial bundle and attach it to an email addressed to the
+    //nav-dev// mailing list.
+  * If you have a single patch to submit, attach it to an email
+    addressed to the //nav-dev// mailing list.  Please **do not
+    patchbomb** the mailing list with multiple emails.

NAV Wiki

User Tools

Site Tools

Differences

Page Tools