Go to file
TheAMM 60d1570fb5 Preprocess ES terms for better literal matching
This commit adds a new .exact subfield to display_name, which holds a
barely-filtered version of the original title we can do "literal"
matching against. This is not real substring matching, but quoting
terms now actually does something!

Implements a simple preprocessor for the search terms to extract quoted
parts from the search terms, optionally prefixed with - to negate them.
The preprocessor will create a query that'll join all three query-types:
the simple_query_string, must-phrases and must-not-phrases.
2018-04-13 15:04:02 +03:00
.github Words (#355) 2017-09-01 18:14:11 -04:00
configs moved some files 2017-05-17 00:53:54 -07:00
info_dicts Move bencoded info dicts from mysql torrent_info table to info_dict directory. DB change! 2018-02-02 20:39:02 +01:00
migrations site-specific changes for new tracker (#453) 2018-02-12 15:52:35 -08:00
nyaa Preprocess ES terms for better literal matching 2018-04-13 15:04:02 +03:00
tests Refactor into an app factory [2 of 2] (#322) 2017-08-01 21:02:08 +03:00
torrents Move bencoded info dicts from mysql torrent_info table to info_dict directory. DB change! 2018-02-02 20:39:02 +01:00
utils remove broken offset option from infodict_mysql2file.py 2018-02-03 21:05:24 +01:00
.gitattributes Add .gitattributes (#310) 2017-07-26 18:57:46 +03:00
.gitignore Move bencoded info dicts from mysql torrent_info table to info_dict directory. DB change! 2018-02-02 20:39:02 +01:00
.travis.yml Nyaa development helper (tool) (#324) 2017-08-06 00:04:38 +03:00
LICENSE Add license (GPLv3) 2017-05-13 01:03:42 +03:00
README.md Nyaa development helper (tool) (#324) 2017-08-06 00:04:38 +03:00
WSGI.py Refactor into an app factory [2 of 2] (#322) 2017-08-01 21:02:08 +03:00
config.example.py Improve and tidy up email blacklist regexes (Hotmail) (#438) 2018-01-27 01:55:35 +02:00
create_es.sh hooked up ES... 90% done, need to figure out how to generate magnet URIs 2017-05-15 23:51:58 -07:00
db_create.py Refactor into an app factory [2 of 2] (#322) 2017-08-01 21:02:08 +03:00
db_migrate.py Refactor into an app factory [2 of 2] (#322) 2017-08-01 21:02:08 +03:00
dev.py Nyaa development helper (tool) (#324) 2017-08-06 00:04:38 +03:00
es_mapping.yml Preprocess ES terms for better literal matching 2018-04-13 15:04:02 +03:00
es_sync_config.example.json Update import_to_es.py to index both torrent flavors, rename sync_es config 2017-05-28 02:12:48 +03:00
import_to_es.py Pad info_hash in ElasticSearch sync scripts 2018-02-25 15:12:35 +02:00
lint.sh Nyaa development helper (tool) (#324) 2017-08-06 00:04:38 +03:00
requirements.txt Update email verification, add Mailgun backend (#380) 2017-10-07 17:31:32 -07:00
run.py Refactor into an app factory [2 of 2] (#322) 2017-08-01 21:02:08 +03:00
setup.cfg Nyaa development helper (tool) (#324) 2017-08-06 00:04:38 +03:00
sync_es.py Pad info_hash in ElasticSearch sync scripts 2018-02-25 15:12:35 +02:00
trackers.txt Remove tracker limit and always add our trackers (#417) 2017-11-22 00:02:22 -08:00
uwsgi.ini Initial commit. 2017-05-12 20:51:49 +02:00

README.md

NyaaV2 Build Status

Setting up for development

This project uses Python 3.6. There are features used that do not exist in 3.5, so make sure to use Python 3.6.
This guide also assumes you 1) are using Linux and 2) are somewhat capable with the commandline.
It's not impossible to run Nyaa on Windows, but this guide doesn't focus on that.

Code Quality:

  • Before we get any deeper, remember to follow PEP8 style guidelines and run ./dev.py lint before committing to see a list of warnings/problems.
    • You may also use ./dev.py fix && ./dev.py isort to automatically fix some of the issues reported by the previous command.
  • Other than PEP8, try to keep your code clean and easy to understand, as well. It's only polite!

Running Tests

The tests folder contains tests for the the nyaa module and the webserver. To run the tests:

  • Make sure that you are in the python virtual environment.
  • Run ./dev.py test while in the repository directory.

Setting up Pyenv

pyenv eases the use of different Python versions, and as not all Linux distros offer 3.6 packages, it's right up our alley.

Setting up MySQL/MariaDB database

You may use SQLite but the current support for it in this project is outdated and rather unsupported.

  • Enable USE_MYSQL flag in config.py
  • Install latest mariadb by following instructions here https://downloads.mariadb.org/mariadb/repositories/
    • Tested versions: mysql Ver 15.1 Distrib 10.0.30-MariaDB, for debian-linux-gnu (x86_64) using readline 5.2
  • Run the following commands logged in as your root db user (substitute for your own config.py values if desired):
    • CREATE USER 'test'@'localhost' IDENTIFIED BY 'test123';
    • GRANT ALL PRIVILEGES ON *.* TO 'test'@'localhost';
    • FLUSH PRIVILEGES;
    • CREATE DATABASE nyaav2 DEFAULT CHARACTER SET utf8 COLLATE utf8_bin;

Finishing up

  • Run python db_create.py to create the database and import categories
    • Follow the advice of db_create.py and run ./db_migrate.py stamp head to mark the database version for Alembic
  • Start the dev server with python run.py
  • When you are finished developing, deactivate your virtualenv with pyenv deactivate or source deactivate (or just close your shell session)

You're now ready for simple testing and development!
Continue below to learn about database migrations and enabling the advanced search engine, Elasticsearch.

Database migrations

  • Database migrations are done with flask-Migrate, a wrapper around Alembic.
  • If someone has made changes in the database schema and included a new migration script:
    • If your database has never been marked by Alembic (you're on a database from before the migrations), run ./db_migrate.py stamp head before pulling the new migration script(s).
      • If you already have the new scripts, check the output of ./db_migrate.py history instead and choose a hash that matches your current database state, then run ./db_migrate.py stamp <hash>.
    • Update your branch (eg. git fetch && git rebase origin/master)
    • Run ./db_migrate.py upgrade head to run the migration. Done!
  • If you have made a change in the database schema:
    • Save your changes in models.py and ensure the database schema matches the previous version (ie. your new tables/columns are not added to the live database)
    • Run ./db_migrate.py migrate -m "Short description of changes" to automatically generate a migration script for the changes
      • Check the script (migrations/versions/...) and make sure it works! Alembic may not able to notice all changes.
    • Run ./db_migrate.py upgrade to run the migration and verify the upgrade works.
      • (Run ./db_migrate.py downgrade to verify the downgrade works as well, then upgrade again)

Setting up and enabling Elasticsearch

Installing Elasticsearch

Setting up ES

  • Run ./create_es.sh to create the indices for the torrents: nyaa and sukebei
    • The output should show acknowledged: true twice
  • Stop the Nyaa app if you haven't already
  • Run python import_to_es.py to import all the torrents (on nyaa and sukebei) into the ES indices.
    • This may take some time to run if you have plenty of torrents in your database.

Enable the USE_ELASTIC_SEARCH flag in config.py and (re)start the application.
Elasticsearch should now be functional! The ES indices won't be updated "live" with the current setup, continue below for instructions on how to hook Elasticsearch up to MySQL binlog.

However, take note that binglog is not necessary for simple ES testing and development; you can simply run import_to_es.py from time to time to reindex all the torrents.

Enabling MySQL Binlogging

  • Edit your MariaDB/MySQL server configuration and add the following under [mariadb]:
    log-bin
    server_id=1
    log-basename=master1
    binlog-format=row
    
  • Restart MariaDB/MySQL (sudo service mysql restart)
  • Copy the example configuration (es_sync_config.example.json) as es_sync_config.json and adjust options in it to your liking (verify the connection options!)
  • Connect to mysql as root
    • Verify that the result of SHOW VARIABLES LIKE 'binlog_format'; is ROW
    • Execute GRANT REPLICATION SLAVE ON *.* TO 'username'@'localhost'; to allow your configured user access to the binlog

Setting up sync_es.py

sync_es.py keeps the Elasticsearch indices updated by reading the binlog and pushing the changes to the ES indices.

  • Make sure es_sync_config.json is configured with the user you grated the REPLICATION permissions
  • Run import_to_es.py and copy the outputted JSON into the file specified by save_loc in your es_sync_config.json
  • Run sync_es.py as-is or, for actual deployment, set it up as a service and run it, preferably as the system/root
    • Make sure sync_es.py runs within the venv with the right dependencies!

You're done! The script should now be feeding updates from the database to Elasticsearch.
Take note, however, that the specified ES index refresh interval is 30 seconds, which may feel like a long time on local development. Feel free to adjust it or poke Elasticsearch yourself!