From: Doug B. <dou...@gm...> - 2011-10-09 16:34:02
|
On Sun, Oct 9, 2011 at 11:15 AM, Doug Blank <dou...@gm...> wrote: > 4) What would be more useful now is an "idempotent" XML export. That > is, if one could export -> import -> export, do a diff between before > and after, and there would be no differences, then we could have some > assurance that all data made the round trip. Currently, that doesn't > quite work. I'm working on that. I've made a change in trunk in the XML export: all main objects are sorted on their handles, and then written out [1]. This should have minimal impact on export, but allows diffs made on XML files to be much more useful. For example, a diff should now show only real changes to the data. Also useful for checking regressions. Things to consider: 1) change all exports to do the same; 2) does this fix all of the irrelevant orderings? 3) Change example Gramps XMLs in trunk to be in handle-order for regression testing. -Doug [1] - http://www.gramps-project.org/bugs/view.php?id=4365 > -Doug > |