Tracking user views on content items marked as must-read
Project description
Track reads on content objects in Plone.
Features
Mark objects as must-read
Keep a record of first reads of content objects per user
Query if a specific user has read a specific content object
List top-x of content objects by user reads in a specific time window
Limitations
This is not a install-and-forget plugin for Plone.
This product does not track reads out of the box. It merely provides a backend you can use for doing so.
Development of this backend was sponsored by Quaive. Quaive has it’s own frontend integration on top of this backend to cater for the specific use cases Quaive needs. We hope that this generic backend is useful for other Plone projects as well.
Rationale
If you’d want a naive implementation to track reads, you could simply create a behavior that stores a list of user ids on every content object.
Obviously that would soon destroy your site with database writes.
Instead, this backend is designed to:
Be compatible with async scheduling, even if it does not provide async itself.
Be flexible to support multiple policy scenarios, without having to rewrite or fork the whole backend.
Use a pluggable SQL backend instead of the ZODB, both to offload writes and to make it easier to run reporting queries.
Architecture
You’re forgiven for thinking the architecture below is overly complex. Please see the rationale above.
Not included in collective.mustread is the frontend and async part:
[ user browser ] -> [ view ] -> [ async queue ]
The backend implementation in this package provides the following:
[ @@mustread-hit ] -> [behavior] -> [database store]
Let’s narrate that starting at the database end.
Database
The database storage provides a rich API as specified in collective.mustread.interfaces.ITracker.
Note that the API design already contains specifications for some future extensions that have not been implemented yet.
The SQL store is derived from the collective.auditlog implementation. We’ve designed collective.mustread to be compatible with collective.auditlog to the point where we’ll even re-use the database connector from auditlog if possible.
The database connection is configured via a registry record collective.mustread.interfaces.IMustReadSettings. You typically want this to contain the same value as your auditlog configuration.
Behaviors
We provide two behaviors:
IMaybeMustRead basically only provides a checkbox where you can specify whether a content object is ‘must-read’.
ITrackReadEnabled activates view tracking on a content object. We track views even if IMaybeMustRead does not mark the object as ‘must-read’. The reason for this is we’d like to track popular content even if the items are not compulsory.
You’d typically activate both these behaviors on the content types you’d like to track.
These behaviors are not activated by default - the :default install profile only provides a browser layer and configures the database connector. It’s up to you to choose and implement your tracking policy in your own projects.
The behaviors provide a flex point where you can implement different tracking policies. You could create a behavior that only tracks reads for certain groups of users for example. You can easily do that by creating a new behavior in a few lines of code, with some extra business logic, which then re-uses our extensive read tracking API for the heavy lifting.
View
A helper view @mustread-hit is available for all ITrackReadEnabledMarker i.e. all objects with the ITrackReadEnabled behavior activated. Hitting that view will store a read record in the database for the content object.
In Quaive we will hit this view from our async stack.
You could conceivably, instead of this view, provide a viewlet that accesses the tracking behavior and API. Just be aware that doing all of that full sync is a risk. YMMV.
There’s also a special debugging view @@mustread-hasread which will tell you whether the user you’re logged in as, has read the object you’re calling this view on.
Installation
Install collective.mustread by adding it to your buildout:
[buildout] ... eggs = collective.mustread
and then running bin/buildout
Or use the built-in buildout:
virtualenv . bin/pip install -r requirements.txt bin/buildout bootstrap bin/buildout
Using collective.mustread
The minimal steps required to actually use collective.mustread in your own project:
Install collective.mustread and configure a database connector. The default connector is a in-memory database which is not suitable for production.
Activate the IMaybeMustRead and ITrackReadEnabled behaviors on the content types you’d like to track, via GenericSetup. Or roll your own custom behaviors.
For these content types, hit ${context/absolute_url}/@@mustread-hit when viewing the content. Ideally you’ll use some kind of async queue at this stage.
Use the tracker API to query the database and adjust your own browser views based on your own business logic. The recommended way to obtain the tracker is:
from collective.mustread.interfaces import ITracker from zope.component import getUtility tracker = getUtility(ITracker)
Contribute
Issue Tracker: https://github.com/collective/collective.mustread/issues
Source Code: https://github.com/collective/collective.mustread
Support
If you are having issues, please let us know via the issue tracker.
License
The project is licensed under the GPLv2.
Contributors
Guido A.J. Stevens, guido.stevens@cosent.net
Changelog
1.0.1 (2016-12-28)
Provide verbose error logging when database is not accessible [gyst]
Trivial testing change [gyst]
1.0 (2016-11-24)
Initial release. [gyst]
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
Hashes for collective.mustread-1.0.1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 35fb7ac29013984d16e248dc8bcd445867be01f7c33bd4a0e8f2e458204f541f |
|
MD5 | 43b23185f47cfcf8e4d102ab650589c6 |
|
BLAKE2b-256 | e750a4a416ae593fb3b271b83ae423d25950854aac3bd78a71160499febea69a |