To me, this sounds like a job best done by running a search engine locally
on each laptop, something like Blacklight or VuFind. You would have to
somehow combine the metadata profiles of these varied sources, however, you
couldn't just dump them all together under the same index. So some sort of
basic XSLT/mapping for the different resources into the schema of whatever
tool you're using. I wonder if it would suit your goals to use an app that
can ingest OAI-PMH records from a variety of sources? You might be able to
obtain a variety of different types of records (Hathi Trust comes to mind
as a bit different from the typical academic OAI feed of academic journal
articles), though perhaps nothing quite like Wikipedia, and largely skip
the metadata wrangling.
I'm not sure I would describe this process as "straightforward" but perhaps
it's not so bad, depending on how much time you have for it. Good luck!
On Mon, Nov 11, 2019 at 10:06 AM Kyle Breneman <[log in to unmask]>
> My university has a program that offers classes at a nearby prison, and
> this program is about to get a bunch of new laptops. As many of you know,
> prisons are pretty restrictive and inflexible regarding technology. These
> laptops cannot be connected to the internet, so it is very challenging to
> provide students with opportunities to search across large collections of
> content for research (e.g. Wikipedia, any article database). (A few years
> ago, we briefly considered using a small portable server developed by
> Libraries Without Borders that hosted offline versions of Wikipedia,
> Project Gutenberg, etc., but this solution did not make it past the prison
> Our reference librarian would like to set these new laptops up so that the
> students can have the simulated experience of searching for information,
> even if it is offline.
> 1. His first idea is to simply put a lot of articles, with full-text
> PDFs, into Zotero on each computer and just use the native search
> functionality of Zotero.
> 2. His second idea is to cobble together a handful of disparate
> resources, such as XML files from Medline/PubMed, a version of offline
> Wikpedia, and one or two other collections, then somehow engineer a
> interface that will simultaneously search across these 2-3 separate
> resources (that are stored locally on the laptop) and display results.
> Is there any relatively straightforward way to accomplish #2? (Notice I
> didn't use the adjective *easy.*) What technologies would be involved?
> Kyle Breneman
> University of Baltimore