Whoosh extension to Flask/SQLAlchemy which used in sina container
Project description
Forked from gyllstromk/Flask-WhooshAlchemy
Flask-WhooshAlchemyPlus is a Flask extension that integrates the text-search functionality of Whoosh with the ORM of SQLAlchemy for use in Flask applications.
Source code and issue tracking at GitHub.
Install
$ pip install flask_whooshalchemyplus
Or:
$ git clone https://github.com/Revolution1/Flask-WhooshAlchemyPlus.git $ cd Flask-WhooshAlchemyPlus && python setup.py install
Quickstart
Let’s set up the environment and create our model:
import flask_whooshalchemyplus # set the location for the whoosh index app.config['WHOOSH_BASE'] = 'path/to/whoosh/base' class BlogPost(db.Model): __tablename__ = 'blogpost' __searchable__ = ['title', 'content'] # these fields will be indexed by whoosh __analyzer__ = SimpleAnalyzer() # configure analyzer; defaults to # StemmingAnalyzer if not specified id = app.db.Column(app.db.Integer, primary_key=True) title = app.db.Column(app.db.Unicode) # Indexed fields are either String, content = app.db.Column(app.db.Text) # Unicode, or Text created = db.Column(db.DateTime, default=datetime.datetime.utcnow) flask_whooshalchemyplus.init_app(app) # initialize
Only two steps to get started:
Set the WHOOSH_BASE to the path for the whoosh index. If not set, it will default to a directory called ‘whoosh_index’ in the directory from which the application is run.
Add a __searchable__ field to the model which specifies the fields (as str s) to be indexed .
set WHOOSH_DISABLED to True to disable whoosh indexing .
Let’s create a post:
db.session.add( BlogPost(title='My cool title', content='This is the first post.') ); db.session.commit()
After the session is committed, our new BlogPost is indexed. Similarly, if the post is deleted, it will be removed from the Whoosh index.
Manually Indexing
By defualt records can be indexed only when the server is running. So if you want to index them manually:
from flask_whooshalchemyplus import index_all index_all(app)
Text Searching
To execute a simple search:
results = BlogPost.query.whoosh_search('cool')
This will return all BlogPost instances in which at least one indexed field (i.e., ‘title’ or ‘content’) is a text match to the query. Results are ranked according to their relevance score, with the best match appearing first when iterating. The result of this call is a (subclass of) sqlalchemy.orm.query.Query object, so you can chain other SQL operations. For example:
two_days_ago = datetime.date.today() - datetime.timedelta(2) recent_matches = BlogPost.query.whoosh_search('first').filter( BlogPost.created >= two_days_ago)
Or, in alternative (likely slower) order:
recent_matches = BlogPost.query.filter( BlogPost.created >= two_days_ago).whoosh_search('first')
We can limit results:
# get 2 best results: results = BlogPost.query.whoosh_search('cool', limit=2)
By default, the search is executed on all of the indexed fields as an OR conjunction. For example, if a model has ‘title’ and ‘content’ indicated as __searchable__, a query will be checked against both fields, returning any instance whose title or content are a content match for the query. To specify particular fields to be checked, populate the fields parameter with the desired fields:
results = BlogPost.query.whoosh_search('cool', fields=('title',))
By default, results will only be returned if they contain all of the query terms (AND). To switch to an OR grouping, set the or_ parameter to True:
results = BlogPost.query.whoosh_search('cool', or_=True)
If you want ordinary text matching result too:
results = BlogPost.query.whoosh_search('cool', like=True)
This acts like whoosh_search('cool') + SQL LIKE '%cool%'
pure_whoosh
If you want the whoosh.index.searcher().search() result:
results = BlogPost.pure_whoosh(self, query, limit=None, fields=None, or_=False)
WhooshDisabled context manager
To disable whoosh indexing temporarily:
with WhooshDisabled(): do sth.
CHANGELOG
v0.7.5 :
feature: add WhooshDisabled context manager
feature: add whoosh_index_all and init_app method
refactory: indexing methods
fix: index error: model has no attribute ‘__searchable__’
v0.7.4 :
Feature: add fuzzy-searching using SQL LIKE
v0.7.3 :
Fix: Chinese analyzer does not take affect
v0.7.2 :
Fix: index_all cannot detect indexable models by itself
v0.7.1 :
Feature: Indexing child module class github issue #43
Feature: Add python3 supprot
Fix: Obey result sorting if caller explicitly uses order_by() on query github pull request #32
Fix: custom query_class usage github pull request #35
Feature: add WHOOSH_DISABLED option to disable whooshalchemyplus at runtime
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
File details
Details for the file Container-WhooshAlchemyPlus-0.7.5.post3.tar.gz
.
File metadata
- Download URL: Container-WhooshAlchemyPlus-0.7.5.post3.tar.gz
- Upload date:
- Size: 10.8 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | 423ecf77954076ae7fa46c638c2b9aa8044fb1df68429fa11bc076231f3624ea |
|
MD5 | a663b567e6b25554817646ba5714befa |
|
BLAKE2b-256 | d3fdfd6b99be36263e8715212c26d9b6a0d1601dbfb3fd5beb6751eb297162b5 |
File details
Details for the file Container_WhooshAlchemyPlus-0.7.5.post3-py2.7.egg
.
File metadata
- Download URL: Container_WhooshAlchemyPlus-0.7.5.post3-py2.7.egg
- Upload date:
- Size: 14.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
File hashes
Algorithm | Hash digest | |
---|---|---|
SHA256 | a60fc2f7c5115e48bcf0a86b8fa8a620130c37dccdf528b48504b880367c3987 |
|
MD5 | 1ec2c389fbf549233ca89e04b79849f5 |
|
BLAKE2b-256 | 8ab2337e8739d4bb6d0cc0a54debb46a994b2b6162e5c748e44f6b57c87969ba |