GitLab

fixed the reference to the sample_type

some queries were written multiple times to manage the different
sample_set tables. These are now generic.

listing sample_sets in edit was broken due to the changes implemented
with the generic table. This commit resovles that issue

Fixing some remaining issues with the file controller linked to recent
changes with sample_sets

With the new model it is easier to test some parts of the queries.
Therefore it is good to do so. Also the new controller required testing.

Previously patient and run lists has filtering. Generic sample_sets now
have filtering too.

With the recent changes, going back to a multi-table system, it was
necessary to add a table to contain some extra information to generic
sample sets

Due to the amount of data being applied there was a considerable amount
of overhead involved in creating an object for each row in the query.

In order to attempt to preserve some of the generic code, the models
have been converted to helper classes.

For these checks type must be specified in order to avoid checking all
possible affected tables as it is done in other methods.

Python objects contain dicts by default. A slight performance
improvement is to declare the slots attribute which contains the list of
all attributes in the object.
This disabled dynamic attribute allocation, but in this case we would
like to save a little on performance/execution time.

Added a model for sequence files.

Added a model to be able to view runs in the generic sample_set
controllers

This patient model is designed to work with the generic sample_set views
and controllers

Some field sorting names did not match the model in the view and had
some code converting them in the controller.
All sorting now matches the name of the model field.

Making things as generic as possible. The view now relies on a set of
methods provided by the model in order to generate each row in the table

In order to make it more clear what kind of sample set is being viewed

Sample type was hard-coded into class although it is set in creation
form. This duality was not clean and could have been a source of issues

in order to use super(A, self) in subclasses, the object must explicitly
inherit from object.

groups are only displayed to admin users, therefore, there is no need to
load the data if the user is not admin.

sample_set.py -> SampleSet.py
models are immported in lexicographical order, so we need to rename this
file in order to ensure it is loaded before any models that might
inherit from it.

zmodel_facotry.py -> z_model_factory.py
just a typo

In order to have a generic sorting mechanism this allos us to easily
access attributes without calling getattr in our controllers

In the all route, the mechanism for defaulting to a type when none are
given was not working. This way is no type is given, it will default to
sample_set

removed a debug print