Feature/49 dynamic schema generation #69
Conversation
nlp4all/models/DataSource.py
Outdated
| id = Column(Integer, primary_key=True) | ||
| user = Column(Integer, ForeignKey("user.id")) | ||
| data_source_id = Column(String(80), unique=True, nullable=False) | ||
| user_id = Column(Integer, nullable=False) |
There was a problem hiding this comment.
ok I def shouldn't need a user id here, that's already in the user relationship, fixing now
There was a problem hiding this comment.
one further comment here, that column has been removed, but we might need a column or columns at some point to store information about the data source, like which columns are used in analysis or filtering etc, but that most likely belongs in the controller / model of the specific analysis, not in the data source / data source manager.
|
all suggestions, comments, improvements, criticisms welcome! :D |
nlp4all/models/datasource_manager.py
Outdated
| os.makedirs(ds_dir) | ||
| # generate a unique filename from a hash of the data source id | ||
| # this is to prevent a user from accessing another user's data | ||
| ds_id_hash = hashlib.sha256(self._data_source_name.encode()).hexdigest() |
There was a problem hiding this comment.
I think I should make the hash not based on the data source name, because the user might change it (if we let them)...
The ID might be better, but it's also predictable, this might not matter too much because the data directory should never be publicly accessible, so I think we should change it to ID rather than name, as ID should be unique.
There was a problem hiding this comment.
Is there any circumstance why the ID of the user may change? some migration? Docker reasons?
There was a problem hiding this comment.
not in any way that will affect things, migrations won't change user IDs, and independent installations can have different IDs but they will also have their own data sources and their own users so it won't be an issue
Fixes #49
So, we now have a fluent, SQLAlchemy compatible interface to dynamic schemas per user per datasource. These are stored in unique SQLAlchemy SQLite database files.