summaryrefslogtreecommitdiff
path: root/crates/tor-keymgr
Commit message (Collapse)AuthorAgeFilesLines
...
| * KeySpecifier d-a macro: Say what we produce, in the docsIan Jackson2024-01-041-1/+17
| |
| * KeySpecifier: Make KeySpecifierPattern a traitIan Jackson2024-01-043-11/+21
| | | | | | | | | | | | So now .arti_pattern() is a trait method. arti#1151 item 3.
| * KeySpecifier d-a macro: Abolish prefix fnsIan Jackson2024-01-042-55/+0
| | | | | | | | | | | | | | | | arti_prefix is no longer used. Abolish prefix too. arti#1151 item 7. Closes #1147.
| * KeySpecifier d-a macro: Introduce SomeSpecifierPattern and use dynIan Jackson2024-01-042-40/+72
| | | | | | | | | | | | | | | | This is a better API. The implementation has less monomorphisation. arti#1151 item 2 and the patterns part of 4. Closes #1126.
| * KeySpecifier d-a macro: Use dyn for arti_pathIan Jackson2024-01-041-11/+153
| | | | | | | | | | | | | | | | | | This reduces monomorphisation. arti#1151 item 4. The commentary reflects a decision not to do arti#1151 item 9 (using fmt signature for to_component) at this time.
| * KeySpecifier tests: add a TimePeriod testIan Jackson2024-01-042-2/+37
| |
| * KeySpecifierComponent: Add .fmt_pretty() methodIan Jackson2024-01-041-1/+21
| | | | | | | | | | | | | | Part of arti#1151 item 10. Discussion in https://gitlab.torproject.org/tpo/core/arti/-/issues/1151#note_2977278
| * KeySpecifier d-a macro: Move to its own moduleIan Jackson2024-01-043-318/+336
| | | | | | | | This is going to expand and gain private helper methods.
| * KeySpecifier d-a macro: insist on structsIan Jackson2024-01-031-1/+1
| | | | | | | | | | We don't handle enums and we don't want to try. Put this here as documentation of this intent.
| * KeySpecifier d-a macro: rename from KeySpecifierDefaultIan Jackson2024-01-032-11/+11
| | | | | | | | arti#1151 item 1.
* | Remove an unneeded allowIan Jackson2024-01-021-2/+0
| | | | | | | | Now d-a #27 is fixed, we don't need this.
* | Update to derive-adhoc 0.8Ian Jackson2024-01-021-1/+1
|/ | | | | As recommended in the d-a changelog, I'm not calling this a breaking change to our APIs.
* rustdoc: linkify a url properlyIan Jackson2024-01-021-2/+2
|
* clippy: An an allow for fallible ed25519 conversionIan Jackson2024-01-021-0/+5
| | | | | I'm not sure about this. Leaving it this way seems the most conservative choice for now.
* tor-keymgr: Wrap a long doc comment line.Gabriela Moldovan2023-12-141-3/+3
|
* tor-keymgr: Update docs referencing outdated SSH algo names.Gabriela Moldovan2023-12-142-3/+3
|
* tor-keymgr: Remove the TODOs about adding a protocol name registry.Gabriela Moldovan2023-12-141-68/+3
| | | | Diziet addressed these in #1049
* tor-keymgr: Use spec.torproject.org for the SSH algo name domain.Gabriela Moldovan2023-12-145-15/+15
| | | | | | | | | Note: any existing x25519 or expanded ed25519 keys you might have in the keystore will become invalid (your keystore will appear corrupt, so you will need to manually delete them if you want to continue using the onion service they were originally generated for). Part of #1108
* arti-client: use sub_builder for ArtiNativeKeystoreConfigNick Mathewson2023-12-131-0/+2
| | | | | | | | | The sub_builder pattern changes `StorageConfigBuilder` so that instead of holding an `Option<ArtiNativeKeystoreConfig>`, it holds an `ArtiNativeKeystoreConfigBuilder`. This makes it a little more ergonomic to use from Rust, and lets us use defaults for the builder fields so that we can make them optional in our configuration.
* tor-keymgr: ArtiPath[Component]: Manually reformat derive listsIan Jackson2023-12-071-32/+8
|
* tor-keymgr: ArtiPath: impl serdeIan Jackson2023-12-071-0/+3
|
* tor-keymgr: ArtiPathComponent: Use unqualified namesIan Jackson2023-12-071-3/+3
| | | | This makes this like the list for ArtiPath.
* tor-keymgr: ArtiPath[Component]: Don't DerefMutIan Jackson2023-12-071-3/+1
| | | | That would allow construction of invalid paths.
* tor-keymgr: ArtiPath[Component]: Add some more test aspectsIan Jackson2023-12-071-1/+10
| | | | | | | | | Test that the FromStr and TryFrom impls give the same answers as new(). These tests dodn't even compile before Properly validate in FromStr because the derive-more generated version has a wrong error type.
* tor-keymgr: ArtiPath[Component]: Properly validate in FromStrIan Jackson2023-12-071-2/+11
| | | | | | | As pointed out in https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/1796#note_2974513 derive-more just parses like the inner type, so an unvalidated String. That is wrong.
* tor-keymgr: ArtiPath[Component]: Use d-a to generate various methodsIan Jackson2023-12-071-30/+36
| | | | Now ArtiPath too is TryFrom<String> and AsRef<str>.
* tor-keymgr: ArtiPath: Introduce validate_strIan Jackson2023-12-071-3/+8
| | | | This makes it more like ArtiPathComponent and will allow more unification
* tor-keymgr: ArtiPath tests: Add a test for a multi-component pathIan Jackson2023-12-071-2/+6
|
* tor-keymgr: ArtiPath tests: separate path from component testsIan Jackson2023-12-071-2/+4
| | | | | | Rename VALID_ARTI_PATH_COMPONENTS from _PATHS. These are tested for validity as components. I want to add a test for a multi-component path, so make a place to put it.
* tor-keymgr: ArtiPath::new: Rename a confusing variableIan Jackson2023-12-071-2/+2
| | | | Rebinding inner locally like this confused me.
* tor-keymgr: ArtiPathSyntaxError: impl EqIan Jackson2023-12-071-1/+1
|
* tor-keymgr: arti_path tests: Add two TODOsIan Jackson2023-12-071-0/+3
| | | | These don't seem critical to me.
* tor-keymgr: path errors: Add some xrefsIan Jackson2023-12-072-1/+9
|
* tor-keymgr: ArtiPathSyntaxError: rename from ArtiPathError (fmt)Ian Jackson2023-12-072-11/+43
|
* tor-keymgr: ArtiPathSyntaxError: rename from ArtiPathErrorIan Jackson2023-12-075-38/+37
| | | | | The previous name was ambiguous, and confusable with KeyPathError (of which it is, basically, a variant).
* tor-keymgr: Abolish unneeded StdResult alias in arti_pathIan Jackson2023-12-071-6/+4
|
* tor-keymgr: Move ArtiPath etc. into new module (docs links)Ian Jackson2023-12-071-2/+6
| | | | Fix docs link breakage.
* tor-keymgr: Move ArtiPath etc. into new moduleIan Jackson2023-12-073-191/+200
| | | | Minimal changes to still compile and pass tests.
* tor-keymgr: KeySpecifierComponent::to_component: rename from as_componentIan Jackson2023-12-071-7/+7
| | | | | | Normally as_... is a cheap and non-consuming (reference to reference) conversion. to_ is better here since we don't consume, but we do allocate.
* tor-keymgr: Make KeySpecifierComponent::as_component fallible (fmt)Ian Jackson2023-12-071-3/+8
|
* tor-keymgr: Make KeySpecifierComponent::as_component fallibleIan Jackson2023-12-071-28/+35
| | | | | | | | | | This allows a principled implentation via Display. The alternative would be to have runtime panics whenever someone uses KeySpecifierComponentViaDisplayFromStr but the Display impl can produce wrong syntax (that doesn't fit in an ArtiPathComponent). Bug seems like the right error since implemnting KeySpecifierComponent implies that every value is representable in an ArtiPathComponent.
* tor-keymgr: Introduce InvalidKeyPathComponentValue errors (fmt)Ian Jackson2023-12-071-2/+1
|
* tor-keymgr: Introduce InvalidKeyPathComponentValue errorsIan Jackson2023-12-072-14/+67
| | | | | | | | This is a variant in KeyPathError, which contains the actual wrong value, and the key name. But it is also a unit struct to return from the trait impl. This replaces ArtiPathError::InvalidDenotator.
* tor-keymgr: Have KeySpecifierComponent::from_component take a refIan Jackson2023-12-071-6/+6
| | | | | Call sites are shortly going to need the component on the error path, so this function ought probably not to consume it.
* tor-keymgr: Make two errors not be CopyIan Jackson2023-12-071-1/+1
| | | | | | This is a hostage to fortune. (In a previous version of !1796 I wanted to make it untrue.)
* tor-keymgr: KeySpecifierDefault: role: allow runtime valueIan Jackson2023-12-071-15/+72
| | | | Make it possible for the role to be a field.
* tor-keymgr: KeySpecifierDefault: Use meta as strIan Jackson2023-12-071-3/+6
| | | | This saves on some stringify! and improves error checking.
* tor-keymgr: KeySpecifierDefault: fixed_path_component featureIan Jackson2023-12-071-0/+29
|
* tor-keymgr: KeySpecifierDefault: recast field loopIan Jackson2023-12-071-1/+2
| | | | | We're going to do another per-field thing that doesn't want this condition.
* tor-keymgr: KeySpecifierDefault: ctor_path featureIan Jackson2023-12-071-8/+47
|
generated by cgit v1.2.3 (git 2.46.0) at 2026-08-28 10:49:18 +0000