For Dr. Reckeweg R 51 Thyreosan
IKSNR 45079 oddb2xml --calc produces a lot less substances.
Adding an example for its atropa belladonna D30, iodum D30, lapis albus D12, lycopus virginicus D12, natrii chloridum D30 ana partes 0.1 ml, excipiens ad solutionem pro 1 ml, corresp. ethanolum 35 % V/V.
Pushed commit Various fixes for --calc. Adds is_active_agent flag
Some IKSNR, eg. 00500 sequence 01 show up after the import under http://oddb-ci2.dyndns.org/de/gcc/drug/reg/00500/seq/04 with NoMethodError undefined method `resolve' for nil:NilClass
. Using pry to debug the problem. Added a work-around in src/util/persistence, which returns nil (and display class/oid of concerned element) when no pointer found. This leads to 4 lines like No pointer in ODDB::ActiveAgent 33039681
when displaying http://oddb-ci2.dyndns.org/de/gcc/drug/reg/00500/seq/04. I found this oid 33039681 in the logs of my imports. Dropping and reloading the database and running the import for 00500 to see, whether I still have the problem.
Must fix the rewritten handling of replacing compositions. We have often differences as we add a lot more substances and solvens as before. Somehow, when adding a new agent, no corresponding substance is set.
Fixing this problem. Now http://oddb-ci2.dyndns.org/de/gcc/drug/reg/00500/seq/04 looks okay after an import, but the changes do not show up in the e-mail message. Also no more nil pointers in active agents.
Pushed commits Reworked parsing compositions in swissmedic and Added parse_compositions
Import does some stuff it should not, but not too much. Takes about 2 hours.