Content Mirror and Pylons
I was experimenting with Pylons and Content Mirror. The result, a proof of concept product called Plyne.
Following posts from Carlos De La Guardia and Kapil Thangavelu (aka hazmat) i started to play with ContentMirror. ContentMirror serializes content from a Plone site to a relational database. Once the Plone content is in a relational database, it is possible to write a small application to consume those data. The result, a lightweight alternative for Plone deployment. In my case i wanted to use Pylons and i was surprised how easy was everything.
According documentation, Content Mirror supports MySql and Postgresql at least, but I'm too lazy to setup a relational database, so i started using sqlite. Anyway my laziness goes beyond that, I even wanted not to think how to do that and i found a document at eletztrick.de with the lines to setup ContentMirror and Sqlite. Cool!.
Following those very simple steps, i got Plone's content in a relational database. The generated tables are very intuitive. For example, in a Plone default's site you'll get something like:
table|content|content|3|CREATE TABLE content ( content_id INTEGER NOT NULL, id VARCHAR(256) NOT NULL, uid VARCHAR(36) NOT NULL, portal_type VARCHAR(64), status VARCHAR(64), type VARCHAR(64), container_id INTEGER, title TEXT, description TEXT, subject TEXT, location TEXT, contributors TEXT, creators TEXT, creation_date TIMESTAMP, modification_date TIMESTAMP, effectivedate TIMESTAMP, expirationdate TIMESTAMP, language TEXT, rights TEXT, PRIMARY KEY (content_id), FOREIGN KEY(container_id) REFERENCES content (content_id) ) ..... table|atdocument|atdocument|22|CREATE TABLE atdocument ( content_id INTEGER NOT NULL, text TEXT, presentation BOOLEAN, tablecontents BOOLEAN, PRIMARY KEY (content_id), FOREIGN KEY(content_id) REFERENCES content (content_id) ON DELETE CASCADE )
Just to show two of them. The complete SQL is available here. That's a default Plone site, Content Mirror also supports custom archetypes but I haven't test that yet. Using those tables we dump the database. Here's available a dump for sqlite3.
Regarding the Pylons app... there's not much to say, but it is a proof of concept project, some lines aren't very nice but my idea was to test what is possible to do in a couple of hours and it works(tm). The result is available here.