Database Views¶
Classification¶
- Difficulty: Medium
- Skills needed: Python, JavaScript, SQLAlchemy, Svelte.js, PostgreSQL
- Length: Long (~350 hours)
The Problem¶
If a user has views in their database before installing Mathesar, or if they create views through some other client, those views will not be reflected in Mathesar.
Feature Description¶
Mathesar should support read-only views. A user should be able to:
- Look at a view similarly to a table.
- Manipulate a view by sorting, filtering, and grouping on the view page (similarly to the table page)
- Explore the columns of a view in Data Explorer.
UX Design Problems¶
- We need to figure out where/how to list views on the Schema page
- We need to determine which parts of the table page translate to a new “view page”. For example,
- Filtering works,
- Grouping works, but
- Opening the record details page from a view makes no sense.
- We need to determine how to integrate database views into the Data Explorer
- It’s likely this would look something like how a table looks in Data Explorer when it has no links to another table.
Overall, we expect a significant portion of this project to involve dealing with UX concerns before implementing.
Tasks¶
- Add a
DBView
model. - Add any related models needed (potentially
DBViewColumn
, for example) - Set up at least the following read-only endpoints in the API:
/api/db/v0/views/
/api/db/v0/views/<id>/
/api/db/v0/views/<id>/rows/
/api/db/v0/views/<id>/columns/
- Do front end work to create a view page similar to the current table page.
- Do front end work to add views as a selectable object in the Data Explorer.
Expected Outcome¶
Views should be integrated into the UI of Mathesar by the end of the project. In particular, there should be:
- A View page
- A way to query a view in the Data Explorer
- Some list of views in the Schema Page
Application Tips¶
- Demonstrate proficiency with the needed skills.
- Show that you understand the difficulty of this project, and have a grasp of the UX concerns involved.
Out of Scope¶
- Editing data values in views.
- Editing the definition of a view.
- Showing deep information about a DB view, e.g., the data sources for columns.
Resources¶
- https://www.postgresql.org/docs/13/sql-createview.html
Mentors¶
Primary Mentor: Brent Moran Secondary Mentor(s): Rajat Vijay