What this package do?

IMS VDEX is a standard for exchanging vocabularies. collective.vdexvocabulary bridges between vdex vocabularies and zope vocabularies (zope.schema.vocabularies), so you can easily use it in systems like Plone/ Zope, Pyramid or any other Zope-Toolkit supporting system.

Whats so special about it?

collective.vdexvocabulary contains two different types of vocabularies:

1. If you have big vocabularies with a lot of relations, like +10.000 terms with +30.000 relations, so this would be perfect use case to use the VdexVocabulary type of collective.vdexvocabulary.
2. If you have tree-like vocabularies this is perfect too. TreeVocabulary supports nested/ hierachical vocabularies.

Also there is other stuff which not supported by other vocabulary packages (i.e. for Plone/Zope):

• i18n support. IMS VDEX supports translations within the VDEX-XML-File, both vocabulary types are supporting this way of translations.
• proper order also with unicode characters (if zope.ucol is installed, vdexvocabulary only). If VDEX is order-sigificant the order given by vdex file is taken (supported by both vocabulary types).
• easy registration using zcml
• relations as it specified in IMS VDEX standard (for now only VdexVocabulary)

How do I use it?

Configuration

In your configure.zcml add:

And to register a vdex vocabulary simply add line bellow pointing to file containing vdex vocabulary:

or for tree vocabularies:

To make registration of vocabularies even easier you can also register several vocabularies and just point to directory:

or for tree vocabularies:

vdex files in path-to/my-vdex-vocabularies directory should have ending .vdex to be recognized by vdex:vocabulary ZCML directive.

Sometimes you dont want VDEX-files inside your code tree. Therefore an environment variable can be given defining the base directory:

or for tree vocabularies:

Before running the code with environment variable relative filenames/directories one has to set the environmant variable, i.e. do an export VDEX_BASE_DIR=/home/joe/vdex/ in order to make it look for the vdex at /home/joe/vdex/my-vocabulary.vdex.

Usage in Code

Vocabularies are named utiltiies in a zope-toolkit manner. The name is taken from the vdex file, its the vocabIdentifier.

Given a vocabulary with the name beeeurope (see tree example below) one has to get the utility using the zope component architecture way:

>>> from zope.component import getUtility
>>> from zope.schema.interfaces import IVocabularyFactory
>>> factory = zope.component.getUtility(IVocabularyFactory, 'beeeurope')

The factory returns on call a vocabulary. It expects a context, which can be None in our case. If you are in an application server pass here your current context. In case of flat vocabularies this is used to detect the language, for tree vocabularies it is ignored, here an more advanced method is used to support i18n:

>>> context = None
>>> vocabulary = factory(context)

Now you can use the vocabulary:

>>> for term in vocabulary:
...     print term.value
...     print term.token
...     print term.title
...     print term.description

How to use tree-vocabularies

Once looked up as shown above traversing the tree is easy. It works as defined in zope.schema.interfaces.ITreeVocabulary. The term is also the key for the sublevel:

>>> def printlevel(leveldict, ident=0):
...     for term in leveldict:
...         print indent * '  ' + term.title
...         printlevel(leveldict[term], indent+1)

Hint: collective.dynatree uses this kind of vocabularies and can be used as an example for own implementations too.

How to access relations (from code)

Relations are defined by ISO2788.

To get listing of BMW car models from above VDEX example you have to:

>>> from zope.schema.vocabulary import getVocabularyRegistry

>>> vr = getVocabularyRegistry()
>>> car_manufacturers = vr.get(self.context, 'your.package.car_manufacturers')
>>> car_models = vr.get(self.context, 'your.package.car_models')

>>> bmw = car_manufacturers.getTerm('bmw')
>>> bmw_car_models = bmw.related.get('NT', [])

Example VDEX file

Flat with with relations

Example of car manufacturers list (car_manufacturers.vdex).:

    your.package.car_manufacturers
    
        ford
        
            Ford
            Una miedra de coche
        
    
    
        bmw
        
            BMW
            Be-eMe-uWe, mierda
        
    

    
        bmw
        very-special-bmw-model
        NT
    

    ...

List of car models (car_models.vdex).:

    your.package.car_models

    
        very-special-bmw-model
        
            Very special BMW model
            Un modelo de Be-eMe-uWe
        
    

    
        very-special-bmw-model
        bmw
        BT
    

...

Hierachical Tree

example of a tree vocabulary:

  beeeurope
  
    European Honey Bees
  
  
    nwe
    
      North-west of Europe
    
    
      nwe.1
      
        A. m. iberica
      
    
    
      nwe.2
      
        A. m. intermissa
      
    
    
      nwe.3
      
        A. m. lihzeni
      
    
    
      nwe.4
      
        A. m. mellifera
      
    
    
      nwe.5
      
        A. m. sahariensis
      
    
  
  
    swe
    
      South-west of Europe
    
    
      swe.1
      
        A. m. carnica
      
    
    
   
      swe.2
      
        A. m. cecropia
      
    
    
      swe.3
      
        A. m. ligustica
      
    
    
      swe.4
      
        A. m. macedonica
      
    
    
      swe.5
      
        A. m. ruttneri
      
    
    
      swe.6
      
        A. m. sicula

Where can I complain / help / send rum?

Home + Source

Report Issues

Send rum
contact for more info

Credit

• Rok Garbas, , <>, Author
• Seantis gmbh, Thank you for initial idea with seantis.vdex where got the idea and then reimplement and extend it.
• Jens W Klein, , <>, Cleanup, Treevocabulary/ i18n-support

TODO

• fetch vocab(s) via url (new directive)
• load vocabs view entry_points
• store vocabs (or changed vocabs in zodb), will probably also need diff and merge option
• write test and get decent test coverage
• write documentation
• make ZCML optional
• make through the web vdex editor (this would probably need sponsoring)
• add relation support to TreeVocabulary

Changelog

0.3 (2020-06-19)

• Support for Python 3. [thet]
• Reformat the codebase using black and isort to meet current coding guidelines. [thet]
• Remove bootstrap.py. Buildout is installed via pip nowadays. [thet]
• Add default_language option when registering vocabularies. Default language is used when the site language is not available in a vdex file. Not specifying this still means a default of 'en'. [tmog]

0.2.2 (2015-09-01)

• Newer lxml doesn't allow accessing _root anymore. This switches to getroot(). [pilz]

0.2.1 (2015-06-22)

• Plone 4.3 compatibility [ale-rt]
• lxml compatibility (imsvdex dropped elementtree in favour of lxml) [ale-rt]
• entry point for z3c.autoinclude [ale-rt]

0.2 (2014-11-03)

• A bunch of refactoring in order to add a new vocab type: TreeVocabulary. As the name suggests, treevocabulary supports zope.schema.interfaces.ITreeVocabulary. It has better i18n-support using own i18n-domains for the caption and description of a term. [jensens]

0.1.2 (2014-01-07)

• don't use context to determine current language, but use getSite. context may be adapter or other object without acquisition (eg. in forms with ignoreContext=True). [naro]
• depend on "setuptools", not "distribute" [nutjob]

0.1.1 (2010-10-11)

• added History, How to access relations (from code) and Example VDEX file section to README. [garbas]
• moved code to . [garbas]
• BUG(Fixed): when vdex file was loaded it failed if there were not terms. [garbas]

0.1 (2010-06-23)

• add documentation and clean up code a little bit. [garbas]

0.1a1 (2010-04-29)

• initial release. [garbas]