- api: extract check_api_version_header() helper, eliminating duplicated
header-checking logic in parse_response_status()
- api: simplify parse_rfc5988() using split_once and let-else
- api: propagate JSON decode errors in fetch_paginated() instead of
silently swallowing them
- api: add connect/request timeouts via a shared build_client() helper;
all handlers now use a configured client instead of reqwest::get()
- api: fix stale log trace name get_manifest -> get_digest
- commands: promote inline response structs to module-level for clarity
- commands: fix etag stripping logic (was using wrong quote/apostrophe
pattern; now correctly strips RFC 7232 double-quotes)
- commands: simplify iterator chains in catalog/tags handlers
- error: simplify ResponseHeaderParseError from Box<dyn Error> to String
- main: fix stale log trace name make_registry_url -> parse_registry_arg
- main: use as_deref().unwrap_or() instead of allocating via to_owned()
- cli: remove unused imports and #![allow(unused_imports)] attribute
The types may have similar names but are distinct types and should be used consistently. This changes
the code to use the header and status code information directly from the `reqwests` library instead of
mixing the two.
Signed-off-by: Anthony Oteri <anthony.oteri@gmail.com>
This changes the effective license from a simple Apache-2.0 license to a
dual license consisting of both Apache-2.0 and the MIT license to be more
compatible for use in GPL Code.
The MIT license requires reproducing countless copies of the same copyright
header with diffferent names in the copyright field, for every MIT library
in use. The Apache license does not have this drawback. However, this is not
the primary motivation for creating these issues. The Apache license also has
protections from patent trolls and explicit contribuition licensing clause.
However the Apache license is incompatible with GPLv2. This is why Rust is
dual-licensed as MIT/Apache (the "primary" license being Apache, MIT only for
GPLv2 compat), and doing so would be wise for this project. This also makes
this crate suitable for inclusion and unrestricted sharing in the Rust
standard distribution and other projects using dual MIT/Apache.
If the Docker Registry API allows for images to be deleted, issuing the
command `delete <image> <tag>` will result in the tag being removed from
the server.
A limitation of this change is that it will cause any unreferenced tags
to become orphaned, requiring the garbage collector within the docker
registry server to clean these up. This is because there is no way to
efficiently determine if each of the layers are reused by other tags or
images on the server.
Implement delete logic
Fix broken unit tests
Refactor the existing code architecture to be a little flatter to support adding the show and delete handlers. Currently these two handlers are just stubbed functions, but the CLI looks about right.
The official Docker Registry V2 API calls the commmand to list the repositories
the "catalog", so to keep things as consistent as possible and not invent new
names for things that have meaning in the domain, stick with the same naming
convention.
Refactor out several smaller single-responsibility methods from the
current `handle_list` method, including: `fetch_all` and `parse_rfc5988`.
Also provide some better documentation around that module.
Anthony Oteri
Anthony Oteri
Anthony Oteri