This change introduces versioning to the identity document format.
The `Version` type is introduced to capture this. Importantly, it is very
restricted in its construction and API. `Version::new` provides a smart
constructor to ensure that it is only ever constructed within the expected
bounds of `0 < v <= IDENTITY_VERSION`, where `IDENTITY_VERSION` is the constant
for the current version.
Note that serializing of the `Doc::version` is skipped to preserve the `Oid`
outputs of all the tests – which means that the first `version` is always
implicit and is not really expected to exist in the serialized forms of the
identity document.
Ensure that remote branches can be delete via the `-d` flag. This
includes the fact that one can delete the `refs/heads/patches`
branches, as long as they are owned by the current operator.
The `missing_delegate_default_branch` test was flakey and would fail checking if
Bob had the `rid` in storage.
This can be improved by waiting to ensure that Bob has a `RefsFetched` event
before performing the check.
The `Editor` is very useful for correctly opening a text editor and making changes
to some initial input. The current use of `Editor` is only for getting input for
commets.
However, it would also be useful for opening up a text editor on some other
existing files or text, for example, in the command `rad config edit`.
The `Editor` struct was changed to have two new options, `truncate` and
`cleanup`, to allow the user of the struct to dictate whether existing text is
truncated, and if the underlying file should be remove.
The original `new` method is now named `comment`, mimicing the existing
construction of a temporary `RAD_COMMENT` file.
The new version of `new` accepts any file for the `Editor` and will open it
without truncating or removing the file.
`Editor` is now used for the `rad config edit` command.
To make it easier for caching COBs across all the repositories in storage, add
an option `--storage` that iterates over all the repositories and performs the
cache operation for each.
This is meant to record in Radicle that automated "jobs" have been run
on a repository. The job might be, for example, a CI run that builds
the software and runs its tests. The recording is done in a "job COB",
which can be in three states:
* fresh -- when the job has been triggered, but has not yet started running
* running -- when it's running
* finished -- when it's finished running
The finished state also records whether the job was successful or if
it failed.
A job COB records the git commit that it runs on, and the "run ID" of
the job. The run ID represents an external identifier, such as one
assigned by a remote CI system. The COB can optionally store a URL
related to the job, such as to the build log from CI.
Add the "rad job" subcommand to create and manage job COBs. For
production use, the COBs should be created and updated using
`heartwood` as a library instead of running "rad job", if at all
possible. However, "rad job list" lists the jobs that have been run on
the repository.
The COB type name is xyz.radicle.beta.job, until we've gained
experience with the COB and its implementation and are ready to
declare it stable.
Signed-off-by: Lars Wirzenius <liw@liw.fi>
When fetching, the validation logic would prune a delegate if it was missing the
default branch. Originally, this was done to prevent any broken state, but later
it was realised that this can be too restrictive and would not allow nodes to
recover into an improved state.
Remove this check so thtat nodes can still fetch from each other when the
default branch is missing.
An e2e test is added to show that nodes can still fetch from each other and
still receive other references, even when a delegate is missing their default
branch.
We were rejecting patches that were once valid, due to a rule that was
added some time ago. Move that rule one level up, in the write path,
instead of in the read path.
Adds `set`, `add`, `remove`, `delete` to `rad config`.
Note that the changes will affect the configuration file, but not the
running instances. For changes to be reflected in the running instances
the node would need to be restarted.
The underlying `Dag` for a COB will use the `Ord` implementation of
whatever is provided as the key. This means that the lexicographical
ordering of the `Oid` SHA will be used, which in turn means that if
the SHAs change in our tests, then the ordering will be broken for
concurrent updates.
To prevent this from happening, a variant of the `prune` method is
introduced that takes an ordering of the keys and values – this method
being called `prune_by`.
This allows `prune` to be replaced with `prune_by` in the
`ChangeGraph::evaluate` method. The ordering used is first comparing
the `Entry::timestamp` and then the `EntryId` (`Oid`).
The aim of this change is to make the `Doc` type more safe to use by approaching
the design via [Parse don't validate][[0]] approach.
The problem with the previous approach was that all field were `pub` and thus a
`Doc<Verified>` could easily be mutated and serialized. Granted, the code that
used the serialization would tend to verify the `Doc` first, however, this
approach *ensures* that only a verified `Doc` can be serialized. It also meant
that trying to add new data that would follow the parse approach would require
more generic parameters on top of the existing `PhantomData` parameter, i.e. we
need to do something like: `Doc<RawField, V> -> Doc<ValidField, V>`.
The new approach splits the type into two separate types: `RawDoc` and `Doc`.
The former is allowed to be mutated at will, and uses types that are less
strict. The latter is the valid type that can only be constructed by validating
a `RawDoc` (or the `initial` constructor). The `Doc` type's fields can then only
be accessed by read-only methods.
Solves the problems above by only allowing mutations to `RawDoc`, as well as,
new fields being added to `RawDoc` which are then validated via
`RawDoc::verified`.
[0]: https://lexi-lambda.github.io/blog/2019/11/05/parse-don-t-validate/
We ensure that a `rad/root` ref is included in the signed refs file
under `rad/sigrefs` for all remotes. This prevents a certain kind of
"grafting" attack where signed refs can be copied between repositories,
by having the peer sign over the identity root together with the data refs.
When verifying signed refs, we ensure that the ref is present and points
to an identity branch root that matches the repository identity containing
the signed refs.
Alternatives: lots of alternatives were considered, but this one doesn't
introduce any changes to the signing. The `rad/id/root` name was
considered but is invalid due to `rad/id`.
Add a fetch refspec to fetch tags from a Radicle remote. The
`--no-tag` option is also included. This is to ensure that it is easy
to collaborate by having tags placed in the remote namespace of
another peer. However, if the namespace is the default `rad` remote,
then tags are expected to be allowed -- this is to pave way for the
feature of canonical tags.
In the tests, the `--tags` option is removed -- and is generally
recommended that it is not used. This is to ensure that it does not
override the `--no-tag` option.
To facilitate edit actions, take URIs instead of the actual blobs. This
means API callers don't have to load all the blobs just for them to be
re-hashed when an edit action is submitted.
There are some peculiarities when dealing with the `Identity` COB since
the embed is the identity document itself. We handle that special case.
Though refs of private repositories are not announced publicly, they can
sometimes be relayed to nodes that are not in the allow list.
We fix this by always checking the visibility of a repository before
sending a refs announcement of it to a peer.
Discovered-by: Adrian Duke <adrian.duke@gmail.com>
Due to the way review editing was implemented, we were losing comments
on previous review edits.
Here, we add a new `review.edit` action that is specifically for
editing, and make the existing `review` action fail silently (for
backwards compatibility) in case it is used when there is already a
review.
We also simplify the review data structure by only keeping track of one
review per author per revision, instead of all edits. Later, if needed,
it will be possible to keep track of all review edits.
When creating a new upstream for the patch workflow, the remote name is always
assumed to be `rad`.
It's possible for users to use another remote name but still use a `rad://` URL
for pushing. To allow for this, the function that creates the upstream entry now
takes a remote name. In the remote helper, `rad patch checkout`, and `rad patch
set` the remote can be optionally specified and fall back to the `rad` remote
name as a default.
Signed-off-by: Fintan Halpenny <fintan.halpenny@gmail.com>
X-Clacks-Overhead: GNU Terry Pratchett
In the `is_authorized` helper, the logic checks the policy and the visibility of
the repository.
If the policy is set to block, the function can return before getting the
repository and the identity document. This improves the check, since the
repository and identity document may be missing if the repository is blocked, so
it would return a different error other than the expected unauthorized error.
When pushing to the default branch and updating the canonical head,
check to see if any merged patches are reverted due to this change.
If so, make sure the COB cache reflects this.
When two `review` actions were posted by the same author, for the same
revision, an error was returned, cancelling evaluation for that author.
This caused certain patches to for eg. lose their merge status.
To remedy this, we *allow* multiple `review` actions per revision, but
simply take the last one as the "current" review. Since this makes the
`review.edit` action redundant, we remove that action completely. This
is safe, as none of the tools created review edits.
Fintan Halpenny
Lars Wirzenius
cloudhead
Johannes Kuehlewindt
Alexis Sellier
Lorenz Leutgeb