Skip to content

Latest commit

 

History

History
60 lines (34 loc) · 2.06 KB

README.md

File metadata and controls

60 lines (34 loc) · 2.06 KB

pinboogle

Full text search for pinboard.in pins

Requirements

You should have docker and docker-compose installed. We built using python 3, but things may work on previous versions. You should be using virtualenv anyway.

Executing spider

To install scrapy python project dependencies:

$ pip install -r requirements.txt
$ scrapy crawl --logfile=data/spider.log --loglevel=INFO -o data/data.json -t json -a user=[PINBOARD_USERNAME] -a after=[NUMBER] pinboard

Where PINBOARD_USERNAME is the user name you are registered and NUMBER is from which timestamp you want to fetch pinboard links (use 1 to start from oldest).

Solr

After scraping data and adding it to data folder, start solr container with:

$ docker-compose up -d

Solr will start and precreate a core named 'pinboard' as default. Also ./data folder will be available inside the container at /var/data folder. If you want to explore the container, login with:

$ docker exec -ti pinboogle_solr_1 /bin/bash

You can also access the admin site at http://{DOCKER_HOST}:8983/solr

Importing data

With Solr container running, use the following command to create all the fields needed for Pinboard json we got from scraping task:

$ docker exec -ti pinboogle_solr_1 /var/data/schema_migration.sh

And the next command to import the json file:

$ docker exec -it --user=solr pinboogle_solr_1 bin/post -c pinboard /var/data/[JSON_FILE]

Search Frontend

Search interface is implemented with Flask and it's located on ./web folder.

When you run docker-compose up you will see that another container will be built. To access it, go to your browser at:

http://{DOCKER_HOST}:5000

Everything that you change on web folder will be reflected on container, if you change or add any dependency, rebuild the container with:

$ docker-compose down
$ docker-compose up -d --build

Disclaimer

We built this to play with scrapy, solr, python and docker. Pinboard has a paid subscription if you want to have full text search on your links, if you want high quality results, subscribe to it.