update docs

Author: HideBa

src/rust/static-btree/docs/implementation_plan.md

@@ -117,6 +117,30 @@ The structure **allows duplicate keys**.  `lower_bound` returns **all** offsets
 
 All search routines keep **only a small, fixed number of nodes in memory**.  When descending, a node is fetched via `reader.seek()` + `reader.read_exact()` into a scratch buffer.  The maximum resident memory during search is therefore `(H + 2) × B × size_of<Entry<K>>`, typically a few kilobytes even for large trees.
 
+### 3.9 Query Operators
+To support a richer query API, the static B+Tree will expose all common comparison operators on keys:
+  * **Eq**: exact match ⇒ all offsets where `key == target`.
+  * **Ne**: not equal ⇒ all offsets where `key != target`.
+  * **Gt**: greater than ⇒ all offsets where `key > target`.
+  * **Ge**: greater or equal ⇒ all offsets where `key >= target`.
+  * **Lt**: less than ⇒ all offsets where `key < target`.
+  * **Le**: less or equal ⇒ all offsets where `key <= target`.
+
+#### Design
+1. Create `query.rs` defining:
+   ```rust
+   #[derive(Copy, Clone, Debug)]
+   pub enum Comparison { Eq, Ne, Gt, Ge, Lt, Le }
+   ```
+2. In `StaticBTree<K, R>` add:
+   - `fn query(&mut self, cmp: Comparison, key: &K) -> Result<Vec<Offset>>`
+   - Convenience methods: `find_eq`, `find_ne`, `find_gt`, `find_ge`, `find_lt`, `find_le`.
+3. Implementation Strategy:
+   * **Eq** → call `lower_bound`, return matching offsets or empty vec.
+   * **Ne** → combine results of `<` and `>` scans.
+   * **Gt/Ge/Lt/Le** → use `lower_bound_index`/`upper_bound_index` to compute start/end, then scan leaf layer entries via `read_entry`.
+4. Each operator incurs at most `O(log_B N)` node reads plus a sequential leaf scan.
+
 ## 4. Public Rust API
 
 ```rust

src/rust/static-btree/docs/progress.md

@@ -14,7 +14,7 @@ This file tracks the incremental progress of the `static-btree` crate inside **F
 |---|-----------|-------|--------|
 | 1 | Core infrastructure | • Define `Key` trait<br>• Implement primitive + custom key types<br>• Implement `Entry` struct | `[x]` Done |
 | 2 | Implementation plan | • Draft initial policy<br>• Review feedback & iterate | `[x]` Updated  ✅ (see implementation_plan.md) |
-| 3 | Tree search API     | • Design `StaticBTree` struct & public API<br>• Lower‑bound & range search handling duplicates<br>• Streaming node reads | `[~]` In progress |
+| 3 | Tree search API     | • Design `StaticBTree` struct & public API<br>• Lower‑bound & range search handling duplicates<br>• Streaming node reads<br>• Extended comparison operators via `query.rs` (Eq, Ne, Gt, Ge, Lt, Le) | `[~]` In progress |
 | 4 | Builder             | • `StaticBTreeBuilder` to serialize trees<br>• Construction algorithm following policy | `[x]` Done |
 | 5 | Async / HTTP query  | • `http_stream_query` mirroring packed_rtree<br>• Feature‑gated under `http` | `[ ]` |
 | 6 | Testing & Benchmarks| • Unit tests for all key types & duplicate cases<br>• Criterion benchmark suite | `[~]` In progress |
@@ -28,11 +28,12 @@ This file tracks the incremental progress of the `static-btree` crate inside **F
 
 ## Next Steps
 
-1. Implement loop‑based `lower_bound` search loading nodes on‑demand.
-2. Add contiguous‑duplicate gathering logic across node boundaries.
-3. Integrate `StaticBTreeBuilder` construction following the layer‑by‑layer algorithm.
-4. Write unit tests (start with u32 and duplicate scenarios).
-5. Prototype `http_stream_query` using packed_rtree's client abstraction.
+1. Implement extended query API: add `query.rs` with `Comparison` enum and methods (`query`, `find_eq`, `find_ne`, `find_gt`, `find_ge`, `find_lt`, `find_le`).
+2. Implement loop‑based `lower_bound` search loading nodes on‑demand.
+3. Add contiguous‑duplicate gathering logic across node boundaries if necessary.
+4. Integrate `StaticBTreeBuilder` construction following the layer‑by‑layer algorithm.
+5. Write unit tests for all new operator methods.
+6. Prototype `http_stream_query` using packed_rtree's client abstraction.
 
 ## Task Guidelines for Contributors & LLMs