kotti/kotti_es

Kotti ElasticSearch implementation

Kotti

Kotti is a high-level, Pythonic web application framework based on Pyramid and SQLAlchemy. It includes an extensible Content Management System called the Kotti CMS (see below).

Kotti is most useful when you are developing applications that

• have complex security requirements,
• use workflows, and/or
• work with hierarchical data.

Built on top of a number of best-of-breed software components, most notably Pyramid and SQLAlchemy, Kotti introduces only a few concepts of its own, thus hopefully keeping the learning curve flat for the developer.

Kotti CMS

Kotti CMS is a content management system that's heavily inspired by Plone. Its main features are:

• User-friendliness: editors can edit content where it appears; thus the edit interface is contextual and intuitive
• WYSIWYG editor: includes a rich text editor
• Responsive design: Kotti builds on Bootstrap, which looks good both on desktop and mobile
• Templating: you can extend the CMS with your own look & feel with almost no programming required
• Add-ons: install a variety of add-ons and customize them as well as many aspects of the built-in CMS by use of an INI configuration file
• Security: the advanced user and permissions management is intuitive and scales to fit the requirements of large organizations
• Internationalized: the user interface is fully translatable, Unicode is used everywhere to store data

License

Kotti is offered under the BSD-derived Repoze Public License.

Install

See installation instructions.

Support and Documentation

Read Kotti's extensive documentation on Read the Docs.

If you have questions or need help, you can post on our mailing list / forum or join us on IRC: #kotti on irc.freenode.net.

If you think you found a bug, open an issue on our Github bugtracker.

Development

Kotti is actively developed and maintained. We adhere to high quality coding standards, have an extensive test suite with high coverage and use continuous integration.

Contributions are always welcome, read our contribution guidelines and visit our Github repository.

kotti_es's Issues

blacklist types

Add a resource types blacklist as ini setting. Useful if you want to omit from elastic search indexing certain resource types

Partial update indexing

Avoid reindexing the whole object if only a property has changed (eg: path after a rename).
See #17

Add support for asynchronous indexing

The very first implementation comes with synchronous indexing after a successfull commit related to a insert, update or delete for IContent types.

This is acceptable for simple sites but it may introduce performance problems in a site with a high write ratio.
The first connection to ES requires quite long time.
Probably we should add optional support for asynchronous indexing, executed by a worker that reads from a queue service fed by our event handlers

Add test coverage badge

Files indexing

Support PDF, doc, odt, etc file contents indexing

Refactor two-phase indexing

No more annotate objects on request (only id and operation)

Add tags to indexer

Add travisCI support

Improve README file

Explain better what kotti_es does, how to configure hooks, distinguish among optional and mandatory configuration (thank you @Tibi)

Fix insert/delete tests

pyramid_es status

This package depends on a pyramid_es fork: https://github.com/truelab/pyramid_es
It should be considered almost stable since it is already used on a heavy production site without any problem but it cannot be published until the changes will be merged with the original pyramid_es and releasedon PyPI.

Why? The original pyramid_es does not support models not pyramid_es aware: your models have to inherit from a pyramid_es mixin class so you cannot use pyramid_es as it is when your models already exist because they are part of an existing framework or third party released.

There are two possible ways:
1- [preferred] merge our fork into the original pyramid_es. We need the @storborg opinion! Link: storborg/pyramid_es#8
2- fork pyramid_es and publish a custom renamed version

Option 1 is the preferred way obviously.

Fix our custom kotti.search_content

The kotti.search_contentprovided by kotti_es is broken. It returns a list of dicts but the search results template expects another kind of data structure

add example elastic search installation

Via buildout

Add language to indexer

console scripts (rebuild_es and clear_rebuild_es)

Console scripts should be added in order to invoke a global reindex or a clear and rebuild

Fix folder rename

The folder rename (with name changed) should reindex all nested objects paths.

Enable ACL indexing

Enable Zope-style hierarchical ACLs in filtered queries

Add workflow state

Exception during resource rename

This exception is raised:

InvalidRequestError: This session is in 'committed' state; no further SQL can be emitted within this transaction.

only if you try to rename the object id (@@rename view). If you change the title only, no exception occurs

Fix tests with real databases

Tests goes well only with sqlite.