Wordpress Conversion

0 views
Skip to first unread message

Dennis Whiteman

unread,
Apr 14, 2008, 2:27:22 PM4/14/08
to Scriblio
I've been following this project for more than a year now and see that
it has progressed nicely.

Knowing that this would eventually get done, we chose to implement our
collection in Wordpress beginning with version 2.1 and now up to
version 2.5. We implemented our books as posts under major categories
with items such as authors, illustrators, subjects, etc. as postmeta
fields. When Wordpress 2.3 came out, we moved the subjects into tags.

I've been through the scriblio docs and am working with a dev version
of our site, but can't seem to import any marc records so I can figure
out how the titles are imported, which would allow me to convert our
existing structure to something that would better use the features
offered by Scriblio. I get the marc import and can see the titles, but
nothing ever seems to imported (even though I reach a point that seems
to have finished the import). It's also not totally clear that books
are/can be entered as posts in addtion to working out other import
methods.

Anyway, some pointers on how the individual titles are implemented in
Wordpress might direct me toward converting our site to Scriblio. You
can view our standard Wordpress install here...

<http://wowlit.org/>

Thanks in advance,

Dennis Whiteman
--
FastPipe Media, Inc.

Casey Bisson

unread,
Apr 16, 2008, 4:58:43 PM4/16/08
to scri...@googlegroups.com

Dennis,

The WOW project looks awesome.

There's another thread going that might help explain why the import
isn't working quite as you expect. See my response here:

http://groups.google.com/group/scriblio/t/bb396da016eeba75

But your real question is about how the data is structured in Scriblio.

All the metadata is stored in WordPress' tag/taxonomy system that was
released in WP 2.3 (I'll only be using the word "taxonomy" in the
sense that WP uses it). By default, WP uses a few taxonomies,
including post_tag and post_category. Scriblio extends that with other
taxonomies that represent author, subject, and whatever else is useful.

As a simple shorthand, that metadata can be expressed in terms of
fieldname=value, and gathering up those field/value pairs is one of
the two things the importer does (the other is to create an HTML
representation of the record). Once a record is imported, though, you
can see and manipulate the metadata using the normal post editor.

Using this record as an example:

http://library.plymouth.edu/read/222334

If you go to edit it, you'll see a box for "bSuite Machine Tags" with
the following contents (well, I see it, and you would to if you had
this record in your site):

auth=monninger, joseph
format=book
isbn=1586421158
isbn=9781586421151
lccn=2006012828
pubyear=2006
sourceid=bb1347215
subj=1910
subj=1914-1981
subj=biography
subj=boxers (sports)
subj=boxing matches
subj=galento, domenick anthony,
subj=louis, joe,
subj=new york
subj=new york (state)
subj=psu faculty publications
subj=united states
title=two ton : one fight, one night : tony galento v. joe lo

The last item ("title=two...") demonstrates one limitation of this:
tag names are limited around 50 characters, but I should point out
that these tags are meant mostly for accessing the record, the user
representation of the record comes from the HTML rendering generated
by the importer. If you make changes in the "bSuite Machine Tags"
field, then save the record, those changes will affect how you can
search for and find the record, but won't be reflected in the
post_content.

This site, a collection of photographs from northern NH paper
companies from about 1885 through 1965, contains all original
cataloging done using field=value pairs:

http://beyondbrownpaper.plymouth.edu/browse

Even the parent/child/reverse relationships you see in the record
below are expressed that way:

http://beyondbrownpaper.plymouth.edu/item/14524 ("child=http://URI/of/child/image/
")

I browsed the WOW site a bit and I think you'd be able to convert the
existing content there. The bSuite Tag Importer might offer an example
of how you could parse the post content and generate tags/taxonomy
data from what you find.

It's an interesting project and I'd love to hear more about it and how
you think Scriblio might work (or not work) in that context,

--Casey

Reply all
Reply to author
Forward
0 new messages