chore: update README and remove unused code

Author: DarkingLee

README.md

@@ -1,14 +1,14 @@
 # HarborX — Blob Direct Write & High-Performance SQL
 
-HarborX is a high-performance data engine for Web3. It ingests incremental **blob** data from L2/rollup ecosystems, writes directly into **columnar formats** (Arrow/Parquet), and exposes a **standard SQL** interface that runs either fully in the browser (DuckDB-WASM) or on your backend. The long-term goal is a verifiable pipeline with **ZK proofs** (ZKSQL) for trusted computation and cross-chain verification.
+HarborX is a high-performance data engine for Web3. It ingests incremental **blob** data from L2/rollup ecosystems, writes directly into **columnar formats** (Arrow/Parquet), and exposes a **standard SQL** interface. The long-term goal is a verifiable pipeline with **ZK proofs** (ZKSQL) for trusted computation and cross-chain verification.
 
 ## Key Features
 
 - **Blob → Columnar (Direct Write)**
 
     Pull real blob payloads and write straight to Arrow/Parquet—no heavyweight node sync or custom ETL needed. Supports small, incremental updates for low latency.
 
-- **SQL Anywhere (Frontend-Only or Backend)**
+- **SQL Anywhere**
 
     Query latest and historical data with SQL. The PoC ships a **pure-frontend** demo (DuckDB-WASM) that reads Arrow/Parquet over HTTP, no server code required.
 
@@ -34,7 +34,7 @@ pip install -e .[cli]
 
 ```
 
-## Option A — Use Existing Static Data (most stable for PoC)
+## Option A — Use Existing Static Data
 
 Commit your prepared dataset under `apps/web/data/` (including `manifest.json`) and run:
 
@@ -44,7 +44,7 @@ harborx serve --dir apps/web --port 8080
 
 ```
 
-## Option B — Fetch a Tiny Real Dataset (Blobscan)
+## Option B — Fetch a Tiny Real Dataset
 
 If you want to refresh the PoC data from the public API:
 
@@ -84,9 +84,9 @@ LIMIT 50;
 
 ---
 
-# Unified CLI
+# CLI
 
-All PoC commands are available via the single `harborx` entrypoint:
+All commands are available via the single `harborx` entrypoint:
 
 - **Fetch blobs + write Arrow/Parquet + manifest**
 
@@ -96,37 +96,26 @@ All PoC commands are available via the single `harborx` entrypoint:
 
     ```
 
-- **Serve static web demo (correct MIME types)**
+- **Serve static web demo**
 
     ```bash
     harborx serve --dir apps/web --port 8080
     
     ```
-    
-- **Tidy repo (plan vs apply)**
-    
-    ```bash
-    harborx tidy
-    harborx tidy --apply
-    # (optional) aggressive consolidation into bench/
-    harborx tidy --apply --aggressive
-    
-    ```
-    
 
 ---
 
-# Frontend-Only Demo
+# Frontend Demo
 
 - Location: `apps/web/`
 - Data folder: `apps/web/data/`
 - Manifest: `apps/web/data/manifest.json` (lists Arrow/Parquet files)
 
-The app resolves file paths **relative to the manifest**. If you ever see 404s like `/data/data/...`, it means both manifest & code added the `data/` prefix. Fix either the manifest (no `data/` prefix) **or** the app’s normalization (strip `data/` if present)—don’t do both.
+The app resolves file paths **relative to the manifest**. 
 
 ---
 
-# Repository Layout (Recommended)
+# Repository Layout
 
 ```base
 harborx/       # unified CLI + data channel (blobscan, tools)
@@ -137,25 +126,8 @@ legacy/        # archived/older code and experiments
 
 ```
 
-Generated artifacts (e.g., large datasets) are typically ignored by Git—except in PoC static mode, where you intentionally commit a tiny `apps/web/data/` for Pages.
-
----
-
-# Roadmap
-
-- **ZKSQL integration**: verifiable write & query proofs (SNARK-friendly schemas).
-- **Connectors**: more sources (e.g., rollup-specific indexers) and push-based ingest.
-- **Scaling knobs**: distributed query backends, tiered storage, caching layers.
-- **Dev-friendly packaging**: Docker + Compose for “click-to-run” deployments.
-
 ---
 
 # License
 
-MIT (PoC). See `LICENSE`.
-
----
-
-**Questions / Feedback?**
-
-File an issue or ping us with repro steps and logs (CLI args, browser console error, small manifest). Happy to help you get HarborX running smoothly.
+Apache-2.0. See `LICENSE`.

harborx/starknet.py harborx/rollup.pyharborx/starknet.py renamed to harborx/rollup.py

@@ -19,80 +19,8 @@ def build_manifest(root:str="apps/web", data:str="data", include_parquet:bool=Fa
         json.dump({"arrow":sorted(arrow), **({"parquet":sorted(parquet)} if include_parquet else {})}, fp, indent=2)
     print(f"[manifest] wrote {len(arrow)} arrow file(s){' and '+str(len(parquet))+' parquet file(s)' if include_parquet else ''} at {data_dir}")
 
-def tidy_repo(apply:bool=False, aggressive:bool=False):
-    plan = []
-
-    def mv(src, dst):
-        if os.path.exists(src):
-            plan.append(("move", src, dst))
-
-    def rm(path):
-        if os.path.exists(path):
-            plan.append(("remove", path))
-
-    # apps/web & docs
-    mv("poc-e2e-blob-sql/packages/web-demo", "apps/web")
-    mv("poc-e2e-blob-sql/packages/docs", "docs")
-
-    # bench/legacy
-    if aggressive:
-        mv("bench", "bench/sqlite-synthetic")
-        mv("poc-sqlite-benchmark", "bench/sqlite-bench")
-    else:
-        mv("bench", "legacy/bench")
-        mv("poc-sqlite-benchmark", "legacy/sqlite-benchmark")
-
-    # misc legacy buckets
-    mv("web", "legacy/web")
-    mv("data", "legacy/data")
-    mv("lake", "legacy/lake")
-    mv("poc-e2e-blob-sql/packages/ingestor", "legacy/ingestor")
-
-    # nested mistake cleanup
-    nested = "poc-e2e-blob-sql/packages/web-demo/poc-e2e-blob-sql"
-    if os.path.exists(nested):
-        plan.append(("remove", nested))
-
-    print("=== Proposed repo layout ===")
-    print("harborx/                 # unified CLI + data channel")
-    print("apps/web/                # pure-frontend demo (DuckDB-WASM)")
-    print("docs/                    # docs")
-    print("bench/                   # benchmarks (optional)")
-    print("legacy/                  # archived code")
-    print(".gitignore               # ignore data/")
-    print("pyproject.toml           # entrypoint: harborx")
-    print()
-
-    for op, a, b in plan:
-        if op == "move":
-            print(f" - MOVE  {a}  ->  {b}")
-        else:
-            print(f" - RM    {a}")
-
-    if not apply:
-        print("\n(dry-run) Nothing changed. To apply:\n  harborx tidy --apply [--aggressive]")
-        return
-
-    for op, a, b in plan:
-        if op == "move":
-            os.makedirs(os.path.dirname(b), exist_ok=True)
-            if os.path.exists(b):
-                base = os.path.basename(a.rstrip('/\\'))
-                dst = os.path.join(b, f"_migrated_{base}")
-                print(f"   ! {b} exists, moving into {dst}")
-                shutil.move(a, dst)
-            else:
-                shutil.move(a, b)
-        elif op == "remove":
-            if os.path.isdir(a):
-                shutil.rmtree(a)
-            else:
-                os.remove(a)
-    print("\n[tidy] repo re-organization complete.")
-
 if __name__ == "__main__":
     ap = argparse.ArgumentParser()
     ap.add_argument("--apply", action="store_true")
     ap.add_argument("--aggressive", action="store_true")
-    args = ap.parse_args()
-    tidy_repo(apply=args.apply, aggressive=args.aggressive)
+    args = ap.parse_args()