Books search engine powered by Elasticserch.
Built on Falcon minimalistic Python framework and Celery task queue.
Packed in Docker container.
For Python requirements see
./requirements.txt
Works only with books in PDF format! May not work with non-latin alphabet.
(optional) Specify path to folder with books by replacing /path/to/folder/ in docker-compose.yml
...
web:
build: .
command: ./run_web.sh
volumes:
- ./src:/falcon/bse
- /path/to/folder/:/opt/books/
ports:
- "8000:8000"
depends_on:
- elasticsearch
...
Run build command
docker-compose build
Run composition
docker-compose up
(optional) Add books to Elasticsearch
docker exec bse_web_1 /bin/bash -c 'python es.py /opt/books/'
Go to admin page and create index, add books, perform advanced search etc.
http://localhost:8000/admin
Example of advanced search query (Don't forget to replace query value "python" on your own):
{
"fields": ["title", "file_name"],
"query": {
"nested": {
"path": "content",
"query": {
"match": {"content.text": "python"}
},
"inner_hits": {
"fields": ["content.page_number"],
"size": 1000000,
"sort": [
{"content.page_number": {"order": "asc"}}
],
"highlight": {
"number_of_fragments": 10000,
"pre_tags": ["<b>"],
"post_tags": ["</b>"],
"fields": {
"content.text": {}
}
}
}
}
}
}
Go to page and do search
http://localhost:8000/
Observe logs
docker exec bse_web_1 cat requests.log
Cleanup
docker-compose down --rmi all
MIT