Re: dbfetch medlinedb errors when using pubmedxml or medlinexml

On 12/2/16, 3:25 PM, "[email protected] on behalf of Carnë Draug" <[email protected] on behalf of [email protected]> wrote:

>On 2 December 2016 at 20:02, Fields, Christopher J
><[email protected]> wrote:
>> That error does make me think it has to do with the recent NCBI
>> switch-over to HTTPS; the issue may be primarily on the EBI dbfetch
>> though (maybe a redirect to grab the appropriate format directly from
>> NCBI, which is now failing?).
>
>When did the switch over happened?  I found a planed date for 30/09/2016 [1]
>but this still worked on October 30 [2] and I got the first failure report
>from cpantests on November 24 [3].

Sometime in November

>I originally though the cause to be same but unless that happens on the
>EBI servers, I don't know how.  I did notice that the code was making
>requests using an undocumented [4] format "medlinexml".  Actually, all
>formats are failing other than default (seems to be medlinefull),
>medlinefull, and medlineref.
>
>  ## works
>  https://www.ebi.ac.uk/Tools/dbfetch/dbfetch?db=medline;id=2859121;format=medlinefull
>  ## fails
>  https://www.ebi.ac.uk/Tools/dbfetch/dbfetch?db=medline;id=2859121;format=pubmedxml
>
>Carnë
>
>[1] https://www.ncbi.nlm.nih.gov/news/06-10-2016-ncbi-https/
>[2] http://www.cpantesters.org/cpan/report/957a89f2-9ec4-11e6-a23c-a658b607aa42
>[3] http://www.cpantesters.org/cpan/report/92af4674-b218-11e6-8dde-c5e727772ce6
>[4] https://www.ebi.ac.uk/Tools/dbfetch/dbfetch/dbfetch.databases#medline

Not sure, I’m hoping EBI can shed light on that.  But the timing is about right with the NCBI switch.

chris

_______________________________________________
Bioperl-l mailing list
[email protected]
http://mailman.open-bio.org/mailman/listinfo/bioperl-l