A package for importing and exporting content from Plone
An XML import/export add-on for Plone.
collective.plone.gsxml is a egg, so installation is easy if you use buildout. You just need to add collective.plone.gsxml to your plone part’s eggs AND zcml option.
A example buildout is available here:
This package needs lxml, see the buildout example.
After installation, you’ll get two new items in Plone’s action menu, one for import, and one for export. For additional stuff, please see the detailled documentation below.
- Exports virtually all Archetypes-based documents
- exports and keeps references
- exports binary data as separate file
- works with ZOPE blobs
- sends custom events which you can subscribe to prior and after import/export
- uses the ZCA to fetch a serializer for a content type – the default adapter uses the Plone Marshaller
This package relies pretty much on the Marshall product for plone (which is shipped with Plone). This product, is, while offering great functionality, a bit convoluted and does not allow to hook in using the ZCA.
Also, this package tries to export references, and this is currently done using pickles, which is not safe. References should be exported by using adapters defined by those who actually use the references and know how to export them (it’s impossible to do this in a generic way IMHO).
- exports references using pickles
- messes with the internals of the Marshal product due to lack of hooks
- uses pickles, this is not secure
- manifest XML is not yet parsed
- convoluted code. The code of this package needs cleanup. This will be done in due course.
- This package out-of-the-box export AT based content only, but you can provide your own serializer as adapter
- This package does NOT export dynamically marked interfaces
- This package does NOT export annotations on content
You need the branch version of Products.Marshall to handle the DataGridFields
change line 44 of Products.eXtremeManagement.content.PoiTask to:
def getAssignees(self): managers = set() try: for issue in self.getRefs('task_issues'): managers.add(issue.getResponsibleManager()) except: pass return sorted(list(managers))
because the references to the POI issues are fixed on the end of import and are not available when the object is recreated
- vim: set ft=rst ts=4 sw=4 expandtab tw=78 :
- Fixed reference handling after import [ramonski]
- Fixed tests and moved them into the source package. [seletz]
- Added fix for PloneArticle. Thanks to Freshmilk Entertainment GmbH for sponsoring this. [seletz]
- Added a events notified for ObjectWillBeExportedEvent, ObjectExportedEvent, ObjectWillBeImportedEvent and ObjectImportedEvent. [seletz]
- Added lxml and libxml2-python to the buildout. Added ZOPE-enabled ipython to the buildout. [seletz]
- Changes package layout to have the sources under src – this allows to have the buildout include development packages during development [seletz]
- Fix bug wrt. handling of blob fields. These are now treated like the file fields. [seletz]
- Added plone.app.blob to buildout to test for #10 [seletz]
- Switched buildout to plone 3.2a1 [seletz]
- More importview work. [seletz]
- Added missing file. [seletz]
- Fix typo [seletz]
To test exporting some content, we’ll create content:
>>> _ = self.folder.invokeFactory("Folder", "content") >>> folder = self.folder["content"] >>> _ = folder.invokeFactory("Document", "doc1", title="A Document") >>> _ = folder.invokeFactory("Document", "doc2", title="A second Document") >>> folder.keys() ['doc1', 'doc2']
Now we create a export context – this will act like a container holding the exported data. We’ll use the TarballExportContext here:
>>> from collective.plone.gsxml.context import TarballExportContext >>> export_context = TarballExportContext()
Now we create a exporter object, which will use the export context created above to store the export data, and use the Marshaller to actually marshall content. We initialize the exporter with the root folder which we want to export:
>>> from collective.plone.gsxml.content import XMLContentFSExporter >>> exporter = XMLContentFSExporter(folder)
Now we’re able to start the export. We supply the export context and a folder inside the export context (thus, one export context can be used multiple times by supplying different root folders). We also specify that this is a root-level export:
>>> exporter.export(export_context, "structure", True)
Ok, that’s it. The archive stream may now be fetched from the export context:
>>> archive = export_context.getArchiveStream() >>> archive.seek(0)
To test importing, we’ll delete the objects exported above:
>>> folder.manage_delObjects(['doc1', 'doc2']) >>> folder.keys() 
Now we create a import context, which will hold our exported stream, and a importer, which will use that import context to read marshalled data from:
>>> from collective.plone.gsxml.context import TarballImportContext >>> import_context = TarballImportContext(archive)
The importer – we initialize the importer with the target folder:
>>> from collective.plone.gsxml.content import XMLContentFSImporter >>> importer = XMLContentFSImporter(folder)
Now we start the import. We supply the import context and the root folder within the import context. Also, we indicate that this is indeed a root-level import:
>>> importer.import_(import_context, "structure", True)
We’re now able to get the documents again:
>>> folder.keys() ['doc1', 'doc2'] >>> folder.doc1.Title() 'A Document'
(in no particular order. These are the nicks on #plone)