From ehs at pobox.com Sun Jul 2 07:21:48 2006 From: ehs at pobox.com (Edward Summers) Date: Sun Jul 2 07:22:57 2006 Subject: [gcs-pcs-list] unapi validator Message-ID: <9F022D19-9CCA-4B72-A17C-E48B9256443A@pobox.com> I've take a stab at updating the validator [1] and the attendant ruby client library to v1.0. This amounted to: - grokking the new format element with attributes rather than children - look for identifiers in with unapi-id class names rather than elements with unapi-uri class names. - no longer look for 415 but 406 when an invalid format is requested of a valid identifier - no longer look for a 200 status code in responses - updated all the ruby library's methods to use 'id' instead of 'uri' I've thrown a mjgiarlo blog entry at it and everything seems OK except for a content-type that's coming back slightly different from what is advertised: Valid or invalid haiku feedback welcome. //Ed From ehs at pobox.com Sun Jul 2 07:42:40 2006 From: ehs at pobox.com (Edward Summers) Date: Sun Jul 2 07:43:45 2006 Subject: [gcs-pcs-list] unapi validator In-Reply-To: <9F022D19-9CCA-4B72-A17C-E48B9256443A@pobox.com> References: <9F022D19-9CCA-4B72-A17C-E48B9256443A@pobox.com> Message-ID: Whoops I forgot: [1] http://validator.unapi.info From leftwing at alumni.rutgers.edu Sun Jul 2 17:44:10 2006 From: leftwing at alumni.rutgers.edu (Michael J. Giarlo) Date: Sun Jul 2 17:46:20 2006 Subject: [gcs-pcs-list] unapi validator In-Reply-To: <9F022D19-9CCA-4B72-A17C-E48B9256443A@pobox.com> References: <9F022D19-9CCA-4B72-A17C-E48B9256443A@pobox.com> Message-ID: <22dbc4ae0607021444g2410d37eo2ca6c794e4ec39bc@mail.gmail.com> On 7/2/06, Edward Summers wrote: > > I've thrown a mjgiarlo blog entry at it and everything seems OK > except for a content-type that's coming back slightly different from > what is advertised: Content-types in the un-appy WordPress plug-in have all been fixed up Thanks, edsu. -Mike -------------- next part -------------- An HTML attachment was scrubbed... URL: http://cipolo.med.yale.edu/pipermail/gcs-pcs-list/attachments/20060702/0102878f/attachment.htm From mat at extracts.de Mon Jul 24 14:22:59 2006 From: mat at extracts.de (Matthias Steffens) Date: Mon Jul 24 14:24:09 2006 Subject: [gcs-pcs-list] refbase unAPI implementation Message-ID: Hi everybody, I just joined this list and wanted to let you know that I was impressed with the simplicity of the unAPI spec (which I discovered recently). Thanks for developing this little gem! I'm the lead developer of refbase, an open-source bibliographic web application (). refbase can export records to common bibliographic formats (such as Endnote, RIS, BibTeX, MODS, SRW+MODS) or output records in citation format as HTML, RTF, PDF, LaTeX, Markdown (structured text) or plain text. Given all these discrete output formats, I was interested to expose them in a more machine-readable way. So I tried to implement the unAPI service and it would be great if someone here could help me with a few related problems, outlined below. Thanks in advance! I tried to make use of the validator at but it seems as if the validator might not be able to handle identifiers which happen to be permanent record URLs that redirect to another URL, is this correct? But maybe I did something wrong... E.g., my unAPI implementation doesn't return appropriate HTML status codes yet. Would my case (where a refbase response to UNAPI?id=IDENTIFIER&format=FORMAT redirects twice) require a status response of "302 Found", "303 See Other" or "307 Temporary Redirect"? If so, which one? Permanent record URLs in a refbase database look like this: I did also try to paste the record's raw URL (which is generated after the redirect) into the validators form, but this gives me errors like: "NoMethodError in Validate#results" Is there some other way how I could test if my implementation confirms to the unAPI spec? If not, it would be great if somebody could take a quick look at my unAPI server. You can test my unAPI implementation at: When called, it should return all output formats supported by refbase (well, almost all of them, RSS and ODF/OpenDocument is currently missing from the list): Btw, does somebody know of a better documentation link for the Endnote format? Also, should I better omit the empty docs attribute for 'text'? All web pages have an autodiscovery link in the header: Single records are identified on web pages like this: I.e., calls to the unAPI server that include an identifier and format look like this: although I guess that the identifier strings in the above examples need to be URL encoded, right? So the correct URL would be: Is this correct? I realize that output formats such as RTF, PDF or LaTeX may make less sence for single records but, firstly, I wanted to expose as many of the supported formats as possible. Secondly, the refbase unAPI service happily accepts identifiers which point to multiple records, e.g., try these unAPI requests: Would this be still a valid use of the unAPI service? I appreciate your help regarding my above questions or any other suggestions you can provide. Thanks, Matthias From daniel.chudnov at yale.edu Tue Jul 25 10:54:53 2006 From: daniel.chudnov at yale.edu (Daniel Chudnov) Date: Tue Jul 25 10:55:26 2006 Subject: [gcs-pcs-list] list move Message-ID: <20060725145451.GA20910@sildin.med.yale.edu> For a variety of reasons I need to move this list to another server. Does anyone object to using google groups? If not, I'll plan to move all the current subscriptions over as of this weekend. -Dan -- Daniel Chudnov Yale Center for Medical Informatics (203) 737-5789 From daniel.chudnov at yale.edu Thu Jul 27 10:16:42 2006 From: daniel.chudnov at yale.edu (Daniel Chudnov) Date: Thu Jul 27 10:17:18 2006 Subject: [gcs-pcs-list] list move this sunday a.m. EST Message-ID: <20060727141642.GA17601@sildin.med.yale.edu> Having heard no concerns about moving this list to google groups, I will go ahead as planned and move the list permanently this coming Sunday a.m., EST (or is it EDT, I can never remember). Expect little or no downtime, and, expect a new welcome message to come your way sometime Sunday. Do *not* unsubscribe from this list first if you want to be on the new one... I will use the current subscriber list here to seed the new list. If you're not the old list, you won't be on the new one. I'll do my best to preserve digest preferences, but you might need to do some twiddling to get everything the way you want. -Dan -- Daniel Chudnov Yale Center for Medical Informatics (203) 737-5789