DOR Console Development

This repository is all about shaping what the catalog should capture separate from workig out the mechanics of capturing that data from ingest.

(And seriously: who wants to ingest thousands of items just to build up sample data?)

(Also seriously: nothing here tries to implement the repository pattern. Shameless.)

Setup

This project uses uv to manage its setup. As a reminder, dor-py uses poetry. No reasons except for an opportunity to try a rising tool in the ecosystem.

Any verdicts?

Surprisingly, `uv` and `pylance` can get into some kind of inexplicable tiff.

Once you install uv and clone this repository:

$ cd dor-console-py
$ uv sync

...and you should be good to go.

Loading Data

Initalize the SQLite database:

$ uv run dor catalog initialize

To harvest data from DLXS:

# harvest the whole collection
$ uv run dor catalog collection <collid>

# harvest <limit>
$ uv run dor catalog collection <collid> --limit <limit>

# to specify an object_type (default: types:slide)
$ uv run dor catalog collection <collid> --object_type page

Harvesting uses data from the DLXS Image API so the data has the appearance of migrated data.

By default, data will be fetched from quod.lib.umich.edu using curl --- mumble, mumble, cloudflare --- but if you have issues with connecting to quod.lib.umich.edu:

• connect to the Library VPN
• set export DLXS_HOST=roger.quod.lib.umich.edu before running the harvest command

¯\_(ツ)_/¯

The fetched data is cached in tmp/cache, so the harvest can be re-run without re-fetching data from the API.

To re-process a collection, you need to run initialize first to reset the database.

Running the dev server

The application uses FastAPI

# start the server on port 8000
$ uv run dor server start

# start the server on an alternate port
$ uv run dor server start --port <port>

The console is mounted under http://localhost:8000/admin/console/.... These are defined in dor/entrypoints/api/console.py; check that file for what's available.

All the handlers use the catalog service to get data; finding more than one row automatically returns a dor.utils.Page object that supports paginated presentation.

The Page class implements the Digital Collections approach to pagination, focusing on next/previous navigation and assuming you'll set up a form to support jumping to another page.

page.total_items      # total number of items in the query
page.total_pages      # total number of pagination pages, based on limit
page.offset           # the start offset, starts at 0
page.index            # the current "page" 
page.range            # returns a string of "<start>-<end>" of the current page
page.limit            # number of items in each page
page.next_offset      # what's the next offset; -1 if not available
page.previous_offset  # what's the previous offset; -1 if not available
page.items            # query results
page.is_useful        # true if # results > limit

The server uses Jinja2 templates with loop controls. Templates are located in templates (for now).

The server will detect changes to files in the repository.