Developer Area/Import//Export/Interoperability/PebblePad

From Mahara Wiki
< Developer Area‎ | Import//Export‎ | Interoperability
Revision as of 17:56, 9 May 2011 by Brettwilkins (talk | contribs) (Created page with "PebblePad was one of the [http://wiki.cetis.ac.uk/PIOP_PebblePad Portfolio Interoperability Project members]. <div id="section_1"> === Quick Summary === Barring a small issue,…")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

PebblePad was one of the Portfolio Interoperability Project members.

Quick Summary

Barring a small issue, Mahara can import PebblePad exports. Information is lost along the way, and WebFolios are not currently being imported.

PebblePad is able to import Mahara exports, although some information is lost along the way.

Import

PebblePad has provided a service to export Leap2A zip files. Attached to this page is an example export. The example includes some data from a previous Mahara export that was imported into PebblePad for testing purposes.

The PebblePad leap2a.xml file is unfortunately run together on as few lines as possible, which makes it harder to scan with the human eye. Attached to this page is a pretty-ified version of the leap2a.xml file in the export, and the script used to generate it.

As generated, the export has one problem that prevents it being imported into Mahara. Entry portfolio:147 is a file which claims to have out-of-line content in the file "files/a", which doesn't exist in the export. Further investigation shows that the demo PebblePad installation also doesn't know where the file went. PebblePad should probably not include such files in their exports, and Mahara should also handle their non-existence more gracefully than it does.

Once this issue is corrected (see the attached files: leap2a_maharafixes.xml and leap2a_maharafixes.diff), the file is successfully imported into Mahara.

When in doubt, run the PebblePad files through a validator.

Imported Information

PebblePad's export correctly provides an entry representing the portfolio holder. The example provides firstname, lastname and email, and this data is correctly imported into Mahara.

Every other entry in the file is imported by Mahara in some fashion. Most of the information ends up in the blog, either because it wasn't recognised or because PebblePad has explicitly exported the information to end up there. A lot of information that looks like it should be resume information is exported as being a blog, which may harm interoperability with Mahara's resume plugin.

Files in PebblePad's export are imported as files in Mahara. PebblePad doesn't seem to have folders - or maybe isn't exporting folders as folders. A folder from Mahara was imported to PebblePad as a blog, which was then exported again as a blog.

Miscellaneous Observations

Their export does not export all relationships bi-directionally.

Any place where an item has no content in PebblePad, it is exported with the content set to "This item has no content". While this is suggested practice as part of the LEAP standard, it is rather redundant, and Mahara may be taught in future to explicitly ignore content tags where they merely claim to have no content.

Some content is exported as type=text, yet has HTML entities in it (portfolio:110).

Some JPEG files are exported with the content type "images/pjpeg" - this believed to be an MSIE-related file type, and isn't really valid for JPEG images (portfolio:88).

WebFolios

PebblePad are exporting their WebFolios as the Webpage selection type. Mahara is currently not importing Webpage selections as Views.

Export

As of 2009/03/31, PebblePad seems able to import Mahara's export files without fuss.

It converts root level folders from Mahara's Files area into blogs, with all folders/files under it as "blogposts". This is a reasonable approach assuming that PebblePad has no concept of a files area - although it does seem to have the concept of files.

Blogs are imported as expected, with the exception of attachments - perhaps PebblePad's blog does not deal with attachments?

Resume items seem to be imported in a reasonable manner - possibly, some HTML formatting is lost.

Information about the person whose data is being imported seems to be being completely ignored at the moment.

Files