This is the development-oriented readme. If you want to write notes for end users, please put them in dist/README.
The "core" stuff is in desktop/core/, whereas installable apps live in apps/. Please place third-party dependencies in the app's ext-py/ directory.
The typical directory structure for inside an application includes:
- for Python code
- for configuration (.ini) files to be installed
- for static HTML and js resources
- for data to be put through a template engine
- for helpful notes
The python code is structured simply as module/package.py, where module may be "filebrowser" or "jobsub". Because it is unlikely that there are going to be huge conflicts, we're going without a deep nested hierarchy.
core/src/desktop/urls.py contains the current layout for top-level URLs. For the URLs within your application, you should make your own urls.py which will be automatically rooted at /yourappname/ in the global namespace. See apps/hello/src/hello/urls.py for an example.
On your host system, you need to have the python "virtualenv" package installed.
Also, you'll need these library development packages installed on your system:
- python-simplejson (for the crepo tool)
- MacOS (mac port):
- simplejson (easy_install)
You need to have crepo installed, and preferably on your path. If it is not on your path, set the environment variable CREPO to point to crepo.py from that distribution. You can clone crepo from http://github.com/cloudera/crepo.git somewhere else on your system.
To build and get the core server running:
$ export HADOOP_HOME=<path-to-hadoop-home> $ git clone http://github.com/cloudera/hue.git $ cd hue $ make apps $ build/env/bin/hue runserver_plus
To start the helper daemons:
$ build/env/bin/hue beeswax_server $ build/env/bin/hue jobsubd
Now Hue should be running on http://localhost:8000.
In order to start up a pseudodistributed cluster with the plugins enabled, run:
$ ./tools/scripts/configure-hadoop.sh all
- 1: What does "Exception: no app!" mean?
Your template has an error in it. Check for messages from the server that look like:
INFO:root:Processing exception: Unclosed tag 'if'. Looking for one of: else, endif
- 2: What do I do if I get "There was an error launching ..."?
- Turn on debugging by issuing dbug.cookie() in a Firebug console.
If you need to name your urls (http://docs.djangoproject.com/en/dev/topics/http/urls/#naming-url-patterns) because there's ambiguity in the view, be sure to prefix the name with the application name. The url name namespace is global. So jobsub.list is fine, but list is not.
Hue is using Django 1.1, which supports the notion of URL namespaces: http://docs.djangoproject.com/en/dev/topics/http/urls/#url-namespaces. We have yet to move over our URLs to this construct. Brownie points for the developer who takes this on.
Right now, we check in the generated thrift code. To generate the code, you'll need the thrift binary. Compile it like so:
$ git clone http://github.com/dreiss/thrift.git $ cd thrift $ ./bootstrap.sh $ ./configure --with-py=no --with-java=no --with-perl=no --prefix=$HOME/pub
We exclude python, java, and perl because they don't like to install in prefix. If you look around at configure's --help, there are environment variables that determine where those runtime bindings are installed.
$ make && make install
When preparing .thrift files, you can use she-bangs to generate the python bindings like so:
#!/usr/bin/env thrift -r --gen py:new_style -o ../../../
This file is in reStructuredText. You may run rst2html README.rst > README.html to produce a HTML.