librarieshacked / catalogues-api

API and front-end for searching all the UK public library catalogues for ISBN results

Libraries Hacked

Public website for Libraries Hacked. Built using a Jekyll theme and GitHub pages

Deployment

Deployed using GitHub pages. GitHub will automatically deploy and build the site when any changes occur on the main branch.

Built With

• Jekyll - The web framework used for creating static sites
• Alembic - Remote Jekyll theme

Authors

• Dave Rowe - Writer - DaveBathnes

Licence

This website uses Alembic theme, which is published under the MIT Licence. If using the theme, please attribute the author and maybe buy them a coffee.

Any content in the website is published under a Creative Commons Attribution 4.0 Licence.

Acknowledgments

• David Darnes for building the theme
• GitHub for hosting
• All the Jekyll contributors

catalogues-api's Issues

Apply custom styles to table

Make the table look good

Make postcode almost mandatory

While a search should be able to be completed without a postcode, it would be good to encourage the use of postcode more as it is less intensive to complete the search,

Add carbon calculator badge

The carbon calculator badge provides a good view of the indicative carbon involved in downloading the webpage.

It also promotes a view of websites having a carbon impact.

Fix bogus results

When searching for an ISBN that has no results it repeatedly returns 3 results from Merton libraries, and 1 from South Ayrshire libraries.

Roadmap for this project?

Hi there

I stumbled across this project while building a scraper to grab books from my local library's Arena web interface.

My intention is to build a better web UI for UK libraries. I'm wondering what your future plans are for this project, and if you're looking for any contributions? It would be great to have a one-stop API for UK library data

Add OpenAPI endpoint

For documentation and testing purposes it would be great to have the default api endpoint (/api) return the OpenAPI interface showing how the API can be used.

Return code as part of services data

So that we can use geography lookups, we should return the ONS local authority code as part of the services list.

Ensure sessions aren't persisted between searches

In some cases, the code uses cookie storage to maintain a session between requests within a single ISBN search. These are then interfering with subsequent searches.

Stopping abruptly when searching Redbridge

After running a search through Redbridge the app stops suddenly and doesn't display results.

Availability being returned as string

In some cases it looks like availability is being returned as a string (e.g. "1") and is breaking the display that sums up how many copies have been found by appending to a string rather than adding up.

Make postcode search more lenient

Postcode search doesn't like lowercase letters

Perhaps using toUpper or similar here would help?
https://github.com/LibrariesHacked/catalogues-api/blob/main/public/js/site.js#L164

Add option for not stopping after finding an available copy

During a postcode-added search the application will stop as soon as it finds an available copy.

This should be made clearer, or an option provided to turn it off.

Provide feedback while searching

During a search we need to provide:

• A searching indicator
• Details of the services that have been searched
• An indication of how long is required to wait

Make the code API only

For better separation of concerns the majority of the connector code to connect with online catalogues is being moved into a separately published npm package https://github.com/LibrariesHacked/catalogues-library

This code should be refactored to use that package. As the final ticket in the 'Get working and maintained' project for this repo.

Deploy to Azure and set up Continuous Integration

I want to deploy to cloud infrastructure and ensure that on any commit into the main branch then it will redeploy and build in Azure.

Refresh packages and front end

General refresh tasks for the front end as well as the packages used.

Upgrade and test for latest library version

Update to the latest catalogues-library (v1.2.0)

Update packages and UI

Refresh on the packages and UI before starting new developments

Add postcode search

When someone enters their postcode the tool should:

1. Validate the postcode and find coordinates
2. Get the services in order of proximity
3. Search 5 services for the items
4. Search further services if no available copies found

Add keyword search

An option instead of entering an ISBN manually should be to search for a title by keyword (including author results).

When selecting a title, the ISBN field will then be populated.

This should be done in a new dialog window to avoid cluttering the current interface.

Is there any way to pre-fill the ISBN on the website?

I post book reviews on my blog, and would like to encourage people to borrow those books from the local library. But I don't know where my readers live.

Would it be possible to use a link like https://www.librarybooks.uk/?isbn=123456789 to pre-fill the ISBN on the website, so users only have to enter their postcode?

Upgrade catalogue to 1.2.4 and other updates

Package updates

Add gzip compression to website and endpoints

Pingdom is reporting a poor score almost purely based on not using any compression.
https://tools.pingdom.com/

We can use code to enable compression. See blog post here:
https://medium.com/@victor.valencia.rico/gzip-compression-with-node-js-cc3ed74196f9

Fix the front end web page

Although this project is mainly an API it does have code to generate a single web page to allow for testing of the ISBN search.

This needs to be fixed. The API could also do with being put in a subdirectory to make a clearer separation - e.g. /api.

A further ticket will provide an OpenAPI documentation page for the API endpoints.

Bug: Running server crashes with error 'Error: Cannot find module './connectors/viewpoint2''

david@Davids-iMac catalogues-api % node server.js
library thing connector loading...
open library connector loading...
spydus connector loading...
enterprise connector loading...
vubis connector loading
prism3 connector loading...
iguana connector loading...
ibistro connector loading...
koha connector loading...
internal/modules/cjs/loader.js:896
  throw err;
  ^

Error: Cannot find module './connectors/viewpoint2'

It seems the issue is the library data JSON file is out of date and references a connector that no longer exists. There is a connector called viewpoint3 however

Only call to required fields from web services

The web service for services has been updated to only return the fields that are used. We can now implement this.

Add clear search functionality

Once a search has been performed there should be an option to clear the search and return the view to the state it was previously.

Any previous postcode should be persisted as this could be useful to the user.

Results not showing

The table render method is failing which is again causing results to not show once a search has been completed.

Refine result buttons

Refine the result views so that the buttons look a bit better.

The results should also be ordered so that the libraries with copies appear at the top.

Update packages and setup

First update is to get packages updated, a small refresh on the website setup (this won't be largely used), and config put in place to integrate with the library data database (though this isn't created yet).

The database will provide access to the catalogues information, as well as a book ISBN lookup,

Add common handle error back in

A lot of functions use a common function called handleErrors. This was removed at the start fo the refactor process but should be added back in.

Iguana returning related title information

In some cases when searching for an item that isn't in stock a related title is shown.

It appears that we are then showing availability for this title, rather than an empty array.

Example: 9781787535213

Refactor code

This code needs to be working again in order to tackle issues.

• Refactor to update to async and await
• Remove all unnecessary connectors
• Test to ensure all connectors are working (not necessarily all lib services)