c4lj2doaj.cgi
Eric Lease Morgan
I have written a set of libraries/scripts providing a read-only API
against the Code4Lib Journal (WordPress) database, and in turn all
but figuring out how to send our Journal's metadata to the Directory
of Open Access Journals through a script called c4lj2doaj.cgi. Please
try:
http://infomotions.com/C4LJ/c4lj2doaj.cgi
For you Perl hackers in the crowd, please consider reading some of
the details, downloading the tar ball, and lending your eyeballs to
the code, here:
If nothing serious is discovered, then us editor types can go the
next step making our metadata more accessible.
--
Eric Lease Morgan
Ed Summers
//Ed
Eric Lease Morgan
On May 24, 2008, at 7:41 PM, Ed Summers wrote:
>> http://infomotions.com/C4LJ/c4lj2doaj.cgi
>
> I'm a bit confused isn't this what Atom and RSS are for?
Atom and RSS are not going to feed the DOAJ, because the DOAJ wants a
specific XML flavor, but I suppose it might be possible to munge RSS
into "DOAJ XML", maybe.
--
Eric
Ed Summers
for the intrusion.
//Ed
Jonathan Rochkind
Is there any reason we can't use this XML generated for issue 2 to feed
DOAJ right now, and get that metadata in? I see two things: one, I
think DOAJ has a metadata field for "article number", no? I could be
wrong. If they do, we use the WordPress ID found in the URL as the
'article number' (as opposed to page number) for citations for our
articles.
Secondly, I think DOAJ wants individual authors in separate XML
elements, not lumped together, yes? If our current metadata makes it
hard to do this automated, can someone just hand-edit the XML to do that
before submitting it to DOAJ?
In general, i would really really like us to have a 'rule' that article
metadata gets uploaded to DOAJ within a week of publication--whether
it's with this nice automated solution, or by hand, either way. If the
automated solution isn't ready, let's upload it by hand. Again, I
volunteer to do that manual entry if the coordinating editor wants me
to. But I think it's very important to get that stuff updated soon--one
of our journal goals is timeliness after all! It's uploading to the DOAJ
that gets us in EBSCO, in Google Scholar, etc. So I'd like to get this
issue 2 metadata in DOAJ asap, and subsequent issue metadata in within a
week of publication. Can we go ahead with issue 2?
Jonathan
--
Jonathan Rochkind
Digital Services Software Engineer
The Sheridan Libraries
Johns Hopkins University
410.516.8886
rochkind (at) jhu.edu
Eric Lease Morgan
On May 27, 2008, at 12:08 PM, Jonathan Rochkind wrote:
> Is there any reason we can't use this XML generated for issue 2 to
> feed
> DOAJ right now, and get that metadata in?
I will make a sincere effort to upload the metadata today, but first
I need to be reminded of our secret DOAJ username/password. I will
then do my best to create a wiki page outlining the process, unless
one already exists. In which case I will update the page.
--
Eric "If-Then-Else-Unless-Otherwise" Morgan
Jodi Schneider
http://wiki.code4lib.org/index.php/Code4Lib_Journal_Entries_in_Directory_of_Open_Access_Journals
-Jodi
Eric Lease Morgan
On May 27, 2008, at 12:08 PM, Jonathan Rochkind wrote:
> Is there any reason we can't use this XML generated for issue 2 to
> feed
> DOAJ right now, and get that metadata in?
I believe I have both updated the Directory with Issue #2's metadata
as well as updated the wiki (with the briefest of) instructions. [1]
If we want to use an XML file to update the Directory without too
much intervention, then a few things need to happen:
1. The Perl library needs to access the live
WordPress MySQL database. This can happen
remotely or on the machine where WordPress
is hosted.
2. Ideally, a CGI script would be installed
someplace such that if it were given an
issue number it would generate the XML.
3. Finally, the XML my current CGI script
generates is well-formed and valid, but
the author information needs to be munged.
To get around this obstacle, we may
consider changing the way we do data-entry
into WordPress regarding our authors. First
name last? Last name first? Email address?
Affiliation? Etc.
I am going out of town staring bright and early tomorrow morning, and
I won't be back until Tuesday. Next week I will see about getting up
with Jonathan B. regarding Item #1, above. We can then tackle Items
#2 and #3.
--
Eric Lease Morgan
Jonathan Rochkind
Also, I'm still a bit concerned about filling out the "record number"
field in the submitted metadata with the article's wordpress ID (in
URL). We decided that this would be the 'record number' for citation
purposes of our articles. Not a huge disaster if it's not there I guess,
but would be nice. Any chance of adding this to the script, or of
editors hand-adding it in the future?
Jonathan
--
Eric Lease Morgan
On May 28, 2008, at 2:21 PM, Jonathan Rochkind wrote:
> Also, I'm still a bit concerned about filling out the "record number"
> field in the submitted metadata with the article's wordpress ID (in
> URL). We decided that this would be the 'record number' for citation
> purposes of our articles. Not a huge disaster if it's not there I
> guess,
> but would be nice. Any chance of adding this to the script, or of
> editors hand-adding it in the future?
??? I don't understand. Please rephrase the question.
--
Earache
Jonathan Rochkind
don't know what's obvious and what's not!
Traditionally, articles are cited with a page number. However, many
online only publications don't really have page numbers, and
increasingly instead assign articles an "article number" or "record
number" for citation purposes (including open url linking citations).
The DOAJ metadata indeed provides a "record number" field for this purpose.
When we were creating the WordPress platform structure for the Journal,
we decided that our articles indeed should have an article number for
citation purposes. And the simplest thing to do would be to just call
the WordPress ID the "article number". Ie,
http://journal.code4lib.org/articles/71 => Article/record number: 71.
[ That means there will be gaps in our article numbers, and the article
numbers won't restart with each issue--but while that might be
unaesthetic, we decided that was fine, it still met the basic needs of
an article number (which is unique within an issue for citation
purposes. Nothing says it can't be unique within the publication too,
and nothing says it has to be sequential). I could find the thread in
the google group where we discussed this, but I'm not feeling that
scholarly. ]
I think it's important that this article/record number gets entered into
the DOAJ metadata in the field provided for it. Both for citation
purposes, and because in the future this might be used for automated
linking purposes. Once we add something to do DOAJ and it gets
subsequently indexed in other places, it's very difficult for us ever to
change it, so we should include all the metadata we have that might be
necessary for future needs too. I think article number is included there.
Make sense?
Jonathan
--
Eric Lease Morgan
On May 28, 2008, at 2:35 PM, Jonathan Rochkind wrote:
> ...
>>
> I think it's important that this article/record number gets entered
> into
> the DOAJ metadata in the field provided for it. Both for citation
> purposes, and because in the future this might be used for automated
> linking purposes. Once we add something to do DOAJ and it gets
> subsequently indexed in other places, it's very difficult for us
> ever to
> change it, so we should include all the metadata we have that might be
> necessary for future needs too. I think article number is included
> there.
>
> Make sense?
Yes, this makes sense, boy, you sure do type fast, and I think the
record numbers are already there. See all the records associated with
Issue #2. Record number is listed as "Start page" at the article level:
http://www.doaj.org/restricted?func=listJournalArticles&issn=19405758
While my public CGI interface to our XML version of the Journal's
database does not list record numbers, I have tweaked my local
version to include that information, and I believe it has been
successfully uploaded to the DOAJ.
--
Eric
Jonathan Rochkind
Jonathan
--
Ron Peterson
Ron
--
Eric Lease Morgan
Eric Lease Morgan
On May 28, 2008, at 4:19 PM, Ron Peterson wrote:
> I noticed that the author affiliations were left out of the article
> entries for issue 2, did we decide to leave them out? Also, some
> garbage has crept into the data somewhere in the process, so the
> abstracts for a couple of the articles have extraneous characters
> in them (e.g., http://www.doaj.org/doaj?
> func=abstract&id=269453&recNo=7&toc=1 ).
Good call.
Those extraneous characters are "smart" quotes, and I deleted records
and resubmitted the XML file. Let's hope those quotes go away.
Regarding the affiliations, you are correct. The affiliations aren't
in the database except as author notes, and as I alluded to
previously, I think we ought to reconsider the way we enter author
names into the WordPress postings so things like affiliations and
email addresses can more easily extracted from the database and
ultimately sent to the Directory.
--
Eric "Off To California" Morgan