flake8 reported 49 findings and pyright 10, and two of them were real
bugs rather than matters of style. The rest were unused imports and
four functions over the complexity limit.
- Unlock a DOI folder with the id the DOI response carries, rather
than an attribute the model does not have, which raised
`AttributeError` on every locked DOI folder.
- Size the `ls` Size column from the sub-folder on the row rather than
from its parent, which pushed the later columns out of line.
- Share the path resolution and the destination checks between `cp`
and `mv`, and split the recursive download and the `ls` row
printing, bringing all four functions under the complexity limit.
- Accept a client built without a connection, which `config` and
`version` rely on, and report the reason if one is then asked for.
- Declare the config protocol's constructor for the type checker
alone, so the protocol keeps its guard against being instantiated.
- Remove 41 unused imports, and let flake8 accept black's spacing.
- Point pyright at the project's own environment, without which it
resolved no dependency and reported 124 findings that were not real.
- Cover the two fixes and the shared `cp`/`mv` paths with new tests.
The login cache is shared by every mdrs process, but the refresh was
guarded by a lock that only reaches inside one. Concurrent runs each
sent the same refresh token, and a provider that rotates them accepts
the first and refuses the rest.
- hold a lock that spans processes across the whole read-refresh-
write, checking cheaply first so ordinary requests never take it
- write the cache through a temporary file: opening it for writing
truncates it, and a reader landing in that window found it empty
and threw the session away
- take the lock for every write, not just the refresh, so a login
running beside one cannot be silently reverted
- wait for a busy lock on Windows rather than giving up after the ten
attempts msvcrt allows, and retry the rename it refuses while a
reader still holds the file open
- bound the refresh request on its own, so a provider that goes quiet
cannot hold the lock indefinitely
- accept bare hostnames such as localhost, store URLs without the
trailing slash, and join download paths through one helper
Decouple CLI commands from internal helper logic and consolidate the
core file transfer operations in the service layer to improve library
portability.
- Make MdrsClient subclass MdrsService to inherit resource resolution.
- Remove all deprecated helper methods from BaseCommand.
- Move core upload and download logic to a new transfer module.
- Refactor all CLI commands to route actions through MdrsClient.
- Eliminate circular imports between client and CLI command modules.
To improve the tool's portability as a Python library, the core logic
has been decoupled from the CLI interface. This allows developers to
programmatically interact with MDRS without relying on CLI-specific
argument parsing or local file-based caches.
- Introduce `MdrsClient` service layer to handle core operations.
- Abstract authentication state using `CacheInterface` and `InMemoryCache`.
- Migrate all CLI commands to utilize `MdrsClient` for execution.
- Separate `Doi` data model from API responses and move to `models/doi.py`.
- Update `README.md` to include Python API usage examples.
- Bump package version to 1.3.17.
Support accessing repositories using DOI strings with optional subpaths
across ls, download, metadata, and file-metadata commands.
- Implement GET v3/doi/{id}/ API model and client calls
- Parse and resolve DOI paths into respective folder and files
- Extract common folder and file resolution logic to shared helpers
- Update README with example DOI-based shell commands
orrisroot
pw-serizawa