refactor(tvix): move castore into tvix-castore crate r/6629

This splits the pure content-addressed layers from tvix-store into a
`castore` crate, and only leaves PathInfo related things, as well as the
CLI entrypoint in the tvix-store crate.

Notable changes:
 - `fixtures` and `utils` had to be moved out of the `test` cfg, so they
   can be imported from tvix-store.
 - Some ad-hoc fixtures in the test were moved to proper fixtures in the
   same step.
 - The protos are now created by a (more static) recipe in the protos/
   directory.

The (now two) golang targets are commented out, as it's not possible to
update them properly in the same CL. This will be done by a followup CL
once this is merged (and whitby deployed)

Bug: https://b.tvl.fyi/issues/301

Change-Id: I8d675d4bf1fb697eb7d479747c1b1e3635718107
Reviewed-on: https://cl.tvl.fyi/c/depot/+/9370
Reviewed-by: tazjin <tazjin@tvl.su>
Reviewed-by: flokli <flokli@flokli.de>
Autosubmit: flokli <flokli@flokli.de>
Tested-by: BuildkiteCI
Reviewed-by: Connor Brewster <cbrewster@hey.com>

1 files changed, 72 insertions, 0 deletions

diff --git a/tvix/castore/protos/castore.proto b/tvix/castore/protos/castore.proto
new file mode 100644
index 000000000000..d99df43857ac
--- /dev/null
+++ b/tvix/castore/protos/castore.proto
@@ -0,0 +1,72 @@
+// SPDX-FileCopyrightText: edef <edef@unfathomable.blue>
+// SPDX-License-Identifier: OSL-3.0 OR MIT OR Apache-2.0
+
+syntax = "proto3";
+
+package tvix.castore.v1;
+
+option go_package = "code.tvl.fyi/tvix/castore/protos;castorev1";
+
+// A Directory can contain Directory, File or Symlink nodes.
+// Each of these nodes have a name attribute, which is the basename in that directory
+// and node type specific attributes.
+// The name attribute:
+//  - MUST not contain slashes or null bytes
+//  - MUST not be '.' or '..'
+//  - MUST be unique across all three lists
+// Elements in each list need to be lexicographically ordered by the name
+// attribute.
+message Directory {
+    repeated DirectoryNode directories = 1;
+    repeated FileNode files = 2;
+    repeated SymlinkNode symlinks = 3;
+}
+
+// A DirectoryNode represents a directory in a Directory.
+message DirectoryNode {
+    // The (base)name of the directory
+    bytes name = 1;
+    // The blake3 hash of a Directory message, serialized in protobuf canonical form.
+    bytes digest = 2;
+    // Number of child elements in the Directory referred to by `digest`.
+    // Calculated by summing up the numbers of `directories`, `files` and
+    // `symlinks`, and for each directory, its size field. Used for inode
+    // number calculation.
+    // This field is precisely as verifiable as any other Merkle tree edge.
+    // Resolve `digest`, and you can compute it incrementally. Resolve the
+    // entire tree, and you can fully compute it from scratch.
+    // A credulous implementation won't reject an excessive size, but this is
+    // harmless: you'll have some ordinals without nodes. Undersizing is
+    // obvious and easy to reject: you won't have an ordinal for some nodes.
+    uint32 size = 3;
+}
+
+// A FileNode represents a regular or executable file in a Directory.
+message FileNode {
+    // The (base)name of the file
+    bytes name = 1;
+    // The blake3 digest of the file contents
+    bytes digest = 2;
+    // The file content size
+    uint32 size = 3;
+    // Whether the file is executable
+    bool executable = 4;
+}
+
+// A SymlinkNode represents a symbolic link in a Directory.
+message SymlinkNode {
+    // The (base)name of the symlink
+    bytes name = 1;
+    // The target of the symlink.
+    bytes target = 2;
+}
+
+// A Node is either a DirectoryNode, FileNode or SymlinkNode.
+message Node {
+    oneof node {
+        DirectoryNode directory = 1;
+        FileNode file = 2;
+        SymlinkNode symlink = 3;
+    }
+}
+