34 releases (5 breaking)
✓ Uses Rust 2018 edition
|new 0.5.9||Mar 25, 2020|
|0.5.6||Feb 29, 2020|
|0.0.8||Dec 12, 2019|
|0.0.7||Nov 30, 2019|
|0.0.5||May 29, 2019|
#17 in HTTP client
448 downloads per month
Used in 2 crates
The code makes some assumptions around the behaviour of bodhi servers. If those are wrong (either because the implementation is wrong, or because the behaviour changed), then that is a bug. Also, there have been instances where the schemas of responses and accepted requests has changed. If that's the case, then that's also a bug. If you encounter (de)serialization issues, please open a bug noting the bodhi server version, and paste the failure message (which should contain the reason for the (de)serialization failure, e.g. missing / renamed fields).
This library tries to do error handling where reasonable, but passes server- or network-related errors through to the caller. For example, if bodhi is under heavy load, it sometimes returns garbage responses with empty bodies. On the other hand, there might just be a persistent network issue that makes a server request fail or time out. These are not handled by the library, but are transparently wrapped and returned. If necessary, the request can be retried by the caller. This library only implements a simple retry logic for network-related failures, not for client or server issues.
GETrequests are implemented, and all actual API responses should successfully deserialize.
POSTrequests are implemented for creating and editing items, except for creating and editing releases (since I have no way of testing that).
Tests should pass for every commit that gets pushed to master. However, currently tests either require internet access to check some assumptions for server behaviour, or require test data which is too big to be committed into git.
Tests are controlled with the following feature flags:
offline-tests: tests able to run offline, without prerequisites (enabled by default)
online-tests: tests that require internet access (for checking bodhi server behavior)
data-tests: tests that require data files (data needs to be downloaded separately, but tests can run offline)
examples directory contains a few example applications to test and showcase some of the crate's functionality.
It might be helpful to enable the
debug feature for local testing, since that enables some debugging statements at
various places in the code (mostly HTTP responses from
It might also be helpful to enable building documentation for private items with
cargo doc --document-private-items
for locally browsing all documented items.