- Change rclone/fs interfaces to accept context.Context
- Update interface implementations to use context.Context
- Change top level usage to propagate context to lover level functions
Context propagation is needed for stopping transfers and passing other
request-scoped values.
The purpose of this is to make it easier to maintain and eventually to
allow the rclone backends to be re-used in other projects without
having to use the rclone configuration system.
The new code layout is documented in CONTRIBUTING.
Now --dump-flag is written as --dump flag. This is a comma separated list which can contain
* headers - HTTP headers as before
* bodies - HTTP bodies as before
* requests - HTTP request bodies
* responses - HTTP response bodies
* auth - HTTP auth
* filters - Filter rexeps
Leave --dump-headers and --dump-bodies for the time being but remove
the other --dump-* flags as they aren't used very often.
Optional interfaces are becoming more important in rclone,
--track-renames and --backup-dir both rely on them.
Up to this point rclone has used interface upgrades to define optional
behaviour on Fs objects. However when one Fs object wraps another it
is very difficult for this scheme to work accurately. rclone has
relied on specific error messages being returned when the interface
isn't supported - this is unsatisfactory because it means you have to
call the interface to see whether it is supported.
This change enables accurate detection of optional interfaces by use
of a Features struct as returned by an obligatory Fs.Features()
method. The Features struct contains flags and function pointers
which can be tested against nil to see whether they can be used.
As a result crypt and hubic can accurately reflect the capabilities of
the underlying Fs they are wrapping.
* Make move command check for overlapping remotes and refuse to run
* Do copy/delete rather than all the copies then all the deletes
* Doesn't purge the source - this was unexpected behaviour see #512 and #416
* Add -list-retries flag to test suite to control retries
This changes the semantics of `move` slightly. However it now errs on
the side of not deleting stuff.
Gives more accurate error propagation, control of depth of recursion
and short circuit recursion where possible.
Most of the the heavy lifting is done in the "fs" package, making file
system implementations a bit simpler.
This commit contains some code originally by Klaus Post.
Fixes#316
* Make all integration tests start with an empty remote
* Add an -individual flag so this can be a different bucket/container/directory
* Fix up tests after changing the hashers
* Add sha1sum test
* Make directory checking in tests sleep more to fix acd inconsistencies
* Factor integration tests to make more maintainable
* Ensure remote writes have a fstest.CheckItems() before use
* this fixes eventual consistency on the directory listings later
* Call fs.Stats.ResetCounters() before every fs.Sync()
Note that the tests shouldn't be run concurrently as fs.Config is global state.
* Convert all paths to UNC paths on Windows.
* Update local filesystem to always use UNC paths.
* Change tests, so they can work with Windows character replacements.
* Remove "/" suffix on paths.
* Always use path/filepath
* Define Mover interface to move a single object
* Define DirMover interface to move a directory
* Implement DirMove operation
* Add `rclone move` command
* Tests for Dir Move
To Do
* Implement Move, DirMover in local, drive, dropbox
* unit test for Mover
* unit test for DirMover
This means that dropbox no longer stores MD5SUMs and modified times.
Fixup the tests so that blank MD5SUMs are ignored, and that if
Precision is set to a fs.ModTimeNotSupported, ModTimes can be ignored too.
This opens the door for other FSs which don't support metadata easily.
