Tried to fix it. Running watir und unit tests before pushing commit.
Watir tests had problems because of the obsole fachinfo_search, some GTINs for the epha-interaction tests had to be changed, too. Also we still have the problem that a logged in user gets errors when going to home_companies.
The watir tests had many errors, because the helper method search_by_trademark did no work when the Instant search was active. Fixed this problem.
Also had to fix changes introduced with the redesign (e.g. a new-line between Art der Suche: and Plus). Also remarked that a wrong permission in /var/www/oodb.org/data/download lead to non working status pages.
Pushed the following commits:
The mobile LNF does not center correctly the log under https://i.ch.oddb.org. Could not reproduce this problem here.
Must I align the google_adsense on the right?
Reorganizing the welcomehead by suppressing the sponsorhead and placing always 3 elements on the top. Replacing left, center or right by a non break space if nothing has to be displayed.
Adding CSS-classes welcome(left|center|right).
Things begin to look fine, but https://oddb-ci2.dyndns.org/de/gcc/home_interactions/ now has way too big logo, as it is the only page using LogoHead directly and not via WelcomeHead. Fixed this problem. Too. Running watir tests this night. Will review commit tomorrow before pushing the changes.
Running update_medwin_companies fails, because the Download URL and XML have changed. We must replace it by using src/plugin/refdata_partner.rb.
Running the jobs/import_refdata_partners and analysing the import log. Added gln/name of created and update companies for easier analysis.
Besides importing companies the medwin plugin also checked all packages. Is this not already done? Yes. update_trade_status is never called via jobs or src/util/updater.rb. Removing update_medwin_companies from the swissmedic_followers in src/util/updater.rb. Removing jobs/import_medwin_companies, too.
Must replace plugin/medreg_doctors.rb, too.
Via refdata we get only the following information about a doctor (see http://refdatabase.refdata.ch/Schemas/Output/Partner_out.xsd):
Before we could use medreg and extract also the following fields:
What shall we do with entries with the role PrivPra, e.g.
<ITEM xmlns="http://refdatabase.refdata.ch/Partner_out" DT="2017-11-12T00:00:00">
<PTYPE>JUR</PTYPE>
<GLN>7601002528122</GLN>
<STATUS>A</STATUS>
<STDATE>2017-11-10T14:21:31.69</STDATE>
<LANG>DE</LANG>
<DESCR1>Arztpraxis Frohsinn AG</DESCR1>
<DESCR2/>
<ROLE>
<TYPE>PrivPra</TYPE>
<STREET>Zürcherstrasse</STREET>
<STRNO>25</STRNO>
<ZIP>8730</ZIP>
<CITY>Uznach</CITY>
<CTN>SG</CTN>
<CNTRY>CH</CNTRY>
</ROLE>
</ITEM>
The doctors have the following possible roles
I will read and parse all NAT and JUR persons at once to simplify the parsing.