Create the conda environment and use it:
conda create -n componentsDB --file requirements.txt
conda activate componentsDB
Testing uses unittest module, should be already available.
python testing.py
Tests are run and if everything is ok no errors is issued.
If you want to see how the code is transformed to SQL calls,
change the VERBOSITY flag in dbclasses.py file.
The file dbclasses.py defines a set of components objects
that are stored in the database. There are two components
type each with one parameter (a number): Quadrupole and
Crystal.
Components can be grouped together to form a compound object.
To see an example of how to fill and query the database,
check the testing.py file. In particular check the methods:
test1_add* to see how to add components to the database
starting from python instances of objects and test2_read*
functions to see how to query the database.
The database is of type sqlite3 and is in memory, modify
the funciton dbclasses.initialize to store the database
in a single file.
When objects are persistified in the database from python two possible approaches are possible:
- Store every data in a single giant table that will contain all data from all possible components type in separate columns (as it is now the AD database)
- Store each component type in a separate table.
According to blog posts the former should be more efficient
in terms of CPU speed, but the latter is much better for DB
design and makes everything readable. However switching
between the two is a very simple matter, change the value
of the variable USE_SEPARATE_TABLES in dblcasses.py
file.