From: Enno B. <enn...@gm...> - 2013-06-02 18:22:53
|
Benny, > You should look at latest version in trunk of the csv: > http://sourceforge.net/p/gramps/code/HEAD/tree/branches/geps/gep-018-evidence-style/data/evidencestyle/evidence_style.csv > > It already is much changed and reduced. There are still 324 fields > though, which is much (but is half of before though I think!), but now > they at least don't overlap anymore. Big thanks for that. It's still a lot, and I will probably switch off the whole set in my Gramps, once I move to 4.1, but I really like the cleaning up, and I'm sure that it will benefit EE users. I want to add that I do share some of Tim's concerns. I recognize fields from 3 different Gramps entities in the CSV, i.e. repository, source, and citation, and I truly wonder how we can map all this. I can imagine that it's nice for users to have all fields in a single window, like the one you put in the second screen shot here, http://www.gramps-project.org/wiki/index.php?title=GEPS_018:_Evidence_style_sources#In_branch_-_testing_of_ideas but model wise, I prefer to have repository, source, and citations attributes to be stored in the proper place, i.e. close to where they are stored now. When implemented right, that storage can be used to fill relevant fields from source or repository just like Tim suggested. > I already edited that csv for 4 evenings, I'm a bit fed up now. Understood. > 2. Gedcom mapping. For all fields in a type L reference, the Gedcom > column should contain A, T or P or nothing. > For export to GEDCOM this will then go to author, Title and Pub, info > field. Nice! > 3. I added a new type at the end of the csv: GEDCOM. This is the > current style used in Gramps, and will be the default for now on new > sources, and on upgrade/import. It's easy, see srcattrtype.py in GEPS > branch: > > > GEDCOM: { > 'F': [ > ('', AUTHOR, ',', EMPTY, False, False, EMPTY, EMPTY), > ('', TITLE, ',', STYLE_QUOTE, False, False, EMPTY, EMPTY), > ('', PUB_INFO, '.', EMPTY, False, False, EMPTY, EMPTY), > ('', DATE, ' -', EMPTY, False, False, EMPTY, EMPTY), > ('', PAGE6S9, '.', EMPTY, False, False, EMPTY, EMPTY), > ], > 'L': [ > ('', AUTHOR, '.', EMPTY, False, False, EMPTY, GED_AUTHOR), > ('', TITLE, '.', STYLE_QUOTE, False, False, EMPTY, > GED_TITLE), > ('', PUB_INFO, '', EMPTY, False, False, EMPTY, > GED_PUBINF), > ], > 'S': [ > ('', AUTHOR, ',', EMPTY, False, False, EMPTY, EMPTY), > ('', DATE, ' -', EMPTY, False, False, EMPTY, EMPTY), > ('', PAGE6S9, '.', EMPTY, False, False, EMPTY, EMPTY), > ], > }, Nice too, maybe add repository field(s) to the F version. > 2/ Just like I added a GEDCOM style, we can add some GRAMPS001, ... > types that we consider best practice. It suffices to send me a patch > of the csv file with typical Dutch, French, ... templates. Do use the > available templates as base for this though. When personal circumstances allow, I will give that a try for Dutch and FS sources. > I'll now concentrate on the GUI part. I think the template things are > sufficiently cleaned up to make a working version. Well, as a non US user, I do like to get rid of all ESM stuff alltogether, and use Dutch and FS templates only, but I do like the built-in tree view that you created to make source types work. thank you, Enno |