mirrors/arti.git/crates/tor-netdoc/src/doc/netstatus, branch arti-v2.4.0mirror of https://gitlab.torproject.org/tpo/core/arti
http://git.dilluti0n.com/mirrors/arti.git/atom?h=arti-v2.4.02026-05-28T10:38:40Ztor-netdoc: implement encoding for votes' `m` items2026-05-28T10:38:40ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-18T15:39:58Zurn:sha1:5948c4b6cd230b04a7eda15a73585ffb87b6c125tor-netdoc: derive Ord for RouterStatusMdDigestsVote2026-05-28T10:38:40ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-18T15:21:38Zurn:sha1:b25729975d9b619842c82139b47a9d837794eb92
Sorting by the applicable consensus methods set seems reasonable.
The spec doesn't state the order for this. I think that's fine.
We can't expect to repro the same consensus with different software,
and we will produce stable output.
tor-netdoc: Rename RelayWeights to RelayWeightsItem2026-05-27T14:38:26ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-18T16:18:26Zurn:sha1:5e9498a6423efd0e0f2f9ac400cbac9fe5845dc3
Prompted by
https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3991#note_3413168
tor-netdoc: Introduce RelayWeights and use it in RouterStatus2026-05-27T14:38:26ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-14T11:14:18Zurn:sha1:307822741164adfc4fd4439d2ad3157e41d91eb2
For encoding, we need to represent the raw parameters.
This change is carefully arranged so that when the retain unknown feature is
disabled (ie, in clients), the per-router data structure remains the
same.
tor-netdoc: Rename FixedString to ConstantString2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-18T16:44:11Zurn:sha1:c5ae62289528b33be9cb52c78147935c7d2a33c7
As per
https://gitlab.torproject.org/tpo/core/arti/-/merge_requests/3985#note_3413182
I left the word "fixed" in some of the docs and error messages, where
it seemed to make sense.
tor-netdoc: Apply deferred rustfmt import churn2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-13T14:07:27Zurn:sha1:1fc2b947a92958eeb4d04ecb7ddcc4f2e6465e06tor-netdoc: NetworkStatusVersionItem, replacing poc's tuple2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-12T20:20:16Zurn:sha1:bfa47db273bede961d4ac14426fb3e4c0c5f48a2
Defining a struct gives us a place to hang the documentation, and
avoids an accidentally-exhaustive tuple.
tor-netdoc: VarietyKeyword, replacing poc's flavour type2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-12T10:09:20Zurn:sha1:1f8919e69927bd5227755892c56ce623bec7f671
This is more code than the poc version but much of the extra is
documentation, and it seems much less magical this way.
We still need a handwritten parser for plain consensuses, sadly.
tor-netdoc: Promote/adjust poc's NetworkStatusSignatures2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-12T11:26:32Zurn:sha1:c4ed766e476cb029e48ac878a42cae04cb23a2e4
The order of arguments of poc's ns_type macro is different to that of
the real code. This is very confusing!
However, the type is correct, once we've added the singleton signature
for votes.
tor-netdoc: use DirectorySignaturesHashesAccu in SignatureGroup2026-05-27T12:43:10ZIan Jacksonijackson@chiark.greenend.org.uk2026-05-11T16:04:47Zurn:sha1:ff905a71c1dff3e71f5c4cbaf8f681de6de311b4
This will let us (re)use the existing signature checking code with the
parse2 type, since we'll be able to make a SignatureGroup out of a
parse2'd UnverifiedNetworkStatus.