Gutenfinder: search widget for Project Gutenberg

Gutenfinder is a little project to make a search widget for Project Gutenberg.

It is under development.

CLI examples:

Serve a webapp on localhost, port 8080:

$ gutenfinder web serve \
	--postgres="user=username dbname=gutenfinder host=dbhost sslmode=disable" \
	--elastic="http://gutenfinder_dev_elastic:9200"

Search for eBooks on the command line:

$ gutenfinder search --term beancurd
{
	"id": 9,
	"term": "beancurd",
	"result_count": 1,
	"items": [
		{
			"id": 467,
			"ebook_id": 724,
			"ebook_title": "Have We No Rights? A frank discussion of the \"rights\" of missionaries",
			"html_snippet": "be made of a couple of\r\nplanks laid on sawhorses, and you may have to eat boiled rice, greens,\r\nand \u003cem\u003ebeancurd\u003c/em\u003e"
		}
	]
}

Results are returned based on titles, the full text of the book, and any other properties currently indexed.

Getting started

Dependencies

Install the following dependencies:

• Install Docker
• Install Docker Compose

Due to an issue with the dockerized Elasticsearch, the following tweak must be made on Linux:

sudo sysctl -w vm.max_map_count=262144

This setting will be lost on reboot, but can be permanently set by adding it to /etc/etc/sysctl.conf

Initialize the project

$ ./script/init
$ ./script/download

This will download a lot of data from a Project Gutenberg archive.

Run with docker

The following sequence of commands will build and launch the app, run migrations, read the eBook catalogue, and index the eBooks on Elasticsearch.

$ ./script/docker-compose-up
$ ./script/db-migrate
$ ./script/catalogue-load-xml
$ ./script/catalogue-load-text

Credits

Johnny Morrice

I am not affiliated with Project Gutenberg in any way. This is just for fun.

License

MIT-style, see the LICENSE file.