| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
|
|
| |
This commit adds a hash accumulator for router descriptor signature
hashes. We will need this because router descriptors have overlapping
signatures.
|
| |
|
|
| |
Part of #2492 phase 2.
|
| |
|
|
|
| |
This avoids much repetition (and consequent conflicts as code is
added).
|
| | |
|
| |
|
|
|
|
|
|
|
| |
plain-consensus
"incomplete" is correct since encoded authcert depends on votes.
"plain-consensus" is going backwards, but stripping "plain-consensus"
gates from everything will be very intrusive.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
| |
Unconditionally enable everything previously gated with parse2 or
encode.
|
| | |
|
| |
|
|
|
|
| |
Part of #2492. We're going to make encode unconditional, and remove
all the cfg-gates for it. Code currently gated by encode depends on
rand.
|
| |\
| |
| |
| |
| | |
Introduce router descriptor types
See merge request tpo/core/arti!3908
|
| | |
| |
| |
| | |
Replaces "B16" by "B16U".
|
| | |
| |
| | |
Replaces "FixedB16" by "FixedB16U"
|
| | | |
|
| | |
| |
| |
| | |
This commit adds spec links to all relevant routerdesc types.
|
| | |
| |
| |
| | |
This commit adds a specification link to the RouterDesc struct.
|
| | |
| |
| |
| |
| |
| |
| | |
This commit replaces `pub use routerdesc::*` with `pub mod routerdesc`
alongside making the required changes in types.rs to not directly
prelude these types but to introduce them in their own module, as that
is the primary advantage of having these types there.
|
| | |
| |
| |
| |
| | |
This commit implements ItemArgumentParseable for RelayPlatform which we
will need in the parse2 version of RouterDesc eventually.
|
| | |
| |
| |
| |
| |
| |
| |
| | |
We will need this type in public fields of the router descriptor data
type at one point, most notably for the `fingerprint` item.
No need to update semver.md because this type is, for whatever reason,
already contained there (in main)?
|
| | |
| |
| |
| |
| | |
This commit adds a test for the recent parse2 accumulator to see whether
it accumulates the policies as expected.
|
| | |
| |
| |
| |
| |
| | |
This commit implements NetdocParseableFields for AddrPolicy, which will
be required, because this field accumulates exit policies in a firewall
like order, making the use of it necessary.
|
| | |
| |
| |
| |
| | |
This commit implements NormalItemArgument for AddrPortPattern because we
will need it later for an accumulator.
|
| | |
| |
| |
| |
| | |
This commit derives PartialEq and Eq for a few types in addrpolicy.rs,
as we will need these for a few unit tests later on.
|
| | |
| |
| |
| |
| | |
The inner Vec ought to be better of as a BTreeSet as that would save us
dedup and sort calls.
|
| | |
| |
| |
| |
| | |
We will need this later for outputting RelayFamilyIds in a fashion
compatible with the current API.
|
| | |
| |
| |
| |
| | |
We will need this later for outputting RelayFamilyIds in a fashion
compatible with the current API.
|
| | |
| |
| |
| |
| |
| | |
This commit adds the ExtraInfoDigests structure storing the, if present,
mandatory SHA-1 hash and optional SHA-256 hash of the extra-info
document associated with the router descriptor it is referenced from.
|
| | |
| |
| |
| |
| |
| |
| | |
This commit adds RouterDescIntroItem which will serve as the
introduction item for router descriptors, that is, a line with the
keyword "router" followed by the nickname, the IPv4 address, and a few
ports.
|
| | |
| |
| |
| |
| | |
This derives Eq for Nickname because we will need it for unit tests that
will be added later on.
|
| | |
| |
| |
| |
| |
| | |
This type comes from the specification and indicates overload
information about a relay. We will use it in router descriptors at one
point, hence why we are adding it already.
|
| | |
| |
| |
| |
| | |
This commit adds a test to verify the 10x4 character requirement for
SpFingerprint.
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| | |
This commit introduces FixedB16U which is a FixedB64 pedant for B16U.
It will be required for ExtraInfoDigests.
|
| | |
| |
| |
| |
| | |
This commit adds a round-trip test for NumericBoolean to ensure that 0
is false and 1 is true.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
This commit adds a new type, NumericBoolean, which implements
NormalItemArgument in order to be parseable with parse2.
It works very simple, namely it can be used to decode/encode a boolean
as a `0` (false) or `1` (true) which is found in a few places around
tor-netdoc.
The reason why this is its own type instead of implementing
NormalItemValue for the bool primitive is, that booleans are special
numerical types and to be future-proof against other oddities, a
separate wrapper type makes most sense.
|
| | |
| |
| |
| |
| |
| | |
Part of #2492 phase 1.
No semver note because this is an experimental feature.
|
| |/
|
|
|
|
| |
Part of #2492 phase 1.
No semver note because this is an undocumented transitional phase.
|
| |\
| |
| |
| |
| | |
Fix new warnings from Rust 1.95
See merge request tpo/core/arti!3902
|
| | | |
|
| | |
| |
| |
| |
| | |
See torspec#401 and
https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3864#note_3397715
|
| | |
| |
| |
| |
| | |
As per
https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3864#note_3397714
|
| | |
| |
| |
| |
| |
| |
| | |
Now it follows, rather than precedes, the parse2_impls and
encode_impls modules that it imports things from.
Pure code motion apart from deleting the todo.
|
| | | |
|
| | | |
|
| | | |
|