feat: initial graph-citation-extract v0.1.0 (graph.citation_extract@0.1.0)

Regex-based citation edge extraction from German legal text.
Edges are typed per the Coupette/Katz/Bommarito 2021 schema.

Detected edge types in v0.1.0:

  - cites         "§ N Gesetz"  /  "Art. N Gesetz"
  - verweist_iVm  "i.V.m. §/Art. N"

Output is a { from: eli, citations: [{to, type, raw}] } graph
fragment. raw preserves the matched source text for diagnostics
and Juristen-review.

Pure in-WASM (regex), zero filesystem, zero network.

Reserved for next versions:

  - 0.2: amends, superseded_by (requires title-line analysis)
  - 0.3: transposes (requires cross-document EUR-Lex resolution —
         belongs in a separate graph.resolve step rather than
         here)
  - 0.4: delegates_to_land / delegates_to_kommune

Signed-off-by: flemming-it <sf@flemming.it>

Cargo.toml

@@ -0,0 +1,37 @@
+[package]
+name = "graph_citation_extract"
+version = "0.1.0"
+edition = "2024"
+authors = ["Dr. Stefan Flemming <chain@flemming.ai>"]
+license = "Apache-2.0"
+publish = false
+description = "F∆I module — extract typed citations from legal text into a graph-edge list"
+repository = "https://git.flemming.ws/chain-modules/graph-citation-extract"
+rust-version = "1.85"
+
+[lib]
+crate-type = ["cdylib", "rlib"]
+
+[dependencies]
+chain-module-sdk = { git = "https://git.flemming.ai/fai/chain-module-sdk-rust.git", branch = "main" }
+serde = { version = "1", features = ["derive"] }
+serde_json = "1"
+regex = "1"
+
+[profile.release]
+opt-level = "s"
+lto = true
+codegen-units = 1
+strip = "symbols"
+
+[lints.rust]
+unsafe_op_in_unsafe_fn = "warn"
+
+[lints.clippy]
+unwrap_used = "deny"
+expect_used = "deny"
+panic = "deny"
+print_stdout = "deny"
+print_stderr = "deny"
+
+[workspace]