Revision 54a4ebd2
- Refactored all of ONEDrive to handle paths as lists of path segments. This removed a bunch path splitting and joining logic that was spread out over the code base, enabled handling of paths in their unescaped, plain Unicode form, and removed special logic required for handling DataONE identifier path segments that may contain path separators and other characters that are not valid in paths.
- Added a section on the architecture page about how paths are represented and handled in the ONEDrive code.
- Fixed handling of Unicode identifiers.
- Added escpaping and unescaping of file- and folder names, so that DataONE objects can be displayed unabigously in the filesystem.
- Added requirement for fields to be returnable before they can be considered as candiates for faceting.
- Factored the logic for deriving filename extensions out of ONEDrive and rewrote it to use the new facility in d1_client.
- Started new module that will implement filename escapling rules that take restrictions and conventions that are different between operating systems into account.
- Added function for the Directory object so that it can serialize itself (for logging).
- Removed the "resolver" designator from all the resolver names because the resolvers are now in a separate package.
- Updated tests.
- Changed settings for maximum number of science objects and maximum number of facet values in directories into user configurable settings.
Files
- added
- modified
- copied
- renamed
- deleted
View differences
