gffbase 0.1.0__tar.gz

gffbase-0.1.0/LICENSE

@@ -0,0 +1,200 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for describing the origin of the Work and
+      reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may accept support,
+      warranty, indemnity, or other liability obligations and/or rights
+      consistent with this License. However, in accepting such obligations,
+      You may act only on Your own behalf and on Your sole responsibility,
+      not on behalf of any other Contributor, and only if You agree to
+      indemnify, defend, and hold each Contributor harmless for any
+      liability incurred by, or claims asserted against, such Contributor
+      by reason of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2026 Kuan-Hao Chao
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

gffbase-0.1.0/MANIFEST.in

@@ -0,0 +1,33 @@
+include pyproject.toml
+include README.md
+include LICENSE
+include MANIFEST.in
+
+# Rust crate (required for sdist → wheel rebuild on user's machine)
+include rust/Cargo.toml
+include rust/Cargo.lock
+recursive-include rust/src *.rs
+
+# Python source tree (also covered by [tool.maturin] python-source).
+recursive-include python/gffbase *.py
+recursive-include python/gffbase *.pyi
+
+# Tests + fixtures bundled so `pytest` works against an installed sdist.
+recursive-include tests *.py
+recursive-include tests/data *.gff3 *.gtf
+
+# Developer documentation (Phase reports + summaries).
+include PHASE3_PARSER_SUMMARY.md
+include PHASE4_INGESTION_SUMMARY.md
+include PHASE5_API_SUMMARY.md
+include PHASE6_BENCHMARK_REPORT.md
+include PHASE7_REMEDIATION_REPORT.md
+include PHASE8_PACKAGING_SUMMARY.md
+
+# Excludes — we never want compiled extensions / caches in the sdist.
+prune target
+prune **/__pycache__
+prune **/.pytest_cache
+prune **/.mypy_cache
+prune **/.ruff_cache
+global-exclude *.pyc *.so *.dylib *.dll

gffbase-0.1.0/PKG-INFO

@@ -0,0 +1,355 @@
+Metadata-Version: 2.4
+Name: gffbase
+Version: 0.1.0
+Classifier: Development Status :: 4 - Beta
+Classifier: Environment :: Console
+Classifier: Intended Audience :: Science/Research
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Operating System :: MacOS
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Programming Language :: Python
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Programming Language :: Python :: Implementation :: CPython
+Classifier: Programming Language :: Rust
+Classifier: Topic :: Scientific/Engineering
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Classifier: Topic :: Database
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Classifier: Typing :: Typed
+Requires-Dist: duckdb>=1.0
+Requires-Dist: pyarrow>=14
+Requires-Dist: psutil>=5 ; extra == 'bench'
+Requires-Dist: memory-profiler>=0.61 ; extra == 'bench'
+Requires-Dist: gffutils>=0.13 ; extra == 'bench'
+Requires-Dist: pytest>=7 ; extra == 'dev'
+Requires-Dist: hypothesis>=6 ; extra == 'dev'
+Requires-Dist: psutil>=5 ; extra == 'dev'
+Requires-Dist: ruff>=0.4 ; extra == 'dev'
+Requires-Dist: mypy>=1.8 ; extra == 'dev'
+Requires-Dist: maturin>=1.5,<2.0 ; extra == 'dev'
+Requires-Dist: mkdocs>=1.6 ; extra == 'docs'
+Requires-Dist: mkdocs-material>=9.5 ; extra == 'docs'
+Requires-Dist: mkdocstrings[python]>=0.25 ; extra == 'docs'
+Requires-Dist: mkdocs-include-markdown-plugin>=6.2 ; extra == 'docs'
+Requires-Dist: pymdown-extensions>=10.7 ; extra == 'docs'
+Requires-Dist: pytest>=7 ; extra == 'test'
+Requires-Dist: hypothesis>=6 ; extra == 'test'
+Requires-Dist: pytest-cov>=4 ; extra == 'test'
+Provides-Extra: bench
+Provides-Extra: dev
+Provides-Extra: docs
+Provides-Extra: test
+License-File: LICENSE
+Summary: GFFBase — Rust-accelerated GFF3/GTF parser with a DuckDB-backed storage engine and a drop-in gffutils-compatible Python API.
+Keywords: gff,gff3,gtf,gencode,bioinformatics,genomics,annotation,duckdb,rust,pyo3,parser,feature-database
+Home-Page: https://github.com/Kuanhao-Chao/gffbase
+Author-email: Kuan-Hao Chao <kuanhao.chao@gmail.com>
+License: Apache-2.0
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown; charset=UTF-8; variant=GFM
+Project-URL: Changelog, https://github.com/Kuanhao-Chao/gffbase/releases
+Project-URL: Documentation, https://kuanhao-chao.github.io/gffbase/
+Project-URL: Homepage, https://github.com/Kuanhao-Chao/gffbase
+Project-URL: Issues, https://github.com/Kuanhao-Chao/gffbase/issues
+Project-URL: Source, https://github.com/Kuanhao-Chao/gffbase
+
+# GFFBase
+
+[![PyPI](https://img.shields.io/pypi/v/gffbase.svg)](https://pypi.org/project/gffbase/)
+[![Python](https://img.shields.io/pypi/pyversions/gffbase.svg)](https://pypi.org/project/gffbase/)
+[![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
+[![Tests](https://img.shields.io/badge/tests-523%20passing-brightgreen.svg)](#testing)
+[![Coverage](https://img.shields.io/badge/coverage-99.19%25-brightgreen.svg)](#testing)
+[![Validated](https://img.shields.io/badge/validated-GENCODE%20%7C%20RefSeq%20%7C%20MANE%20%7C%20CHESS%203-blue.svg)](#-comprehensive-human-genome-annotations--validated-across-every-canonical-corpus)
+
+---
+
+## What is GFFBase?
+
+**GFFBase is a high-performance genomic-annotation engine combining a
+SIMD Rust parser, a DuckDB columnar backend, and a zero-copy PyArrow
+interface — purpose-built for whole-genome-scale ingest and bulk
+machine-learning feature extraction, while remaining a drop-in
+successor to [`gffutils`](https://github.com/daler/gffutils).**
+
+A SIMD Rust+PyO3 parser feeds DuckDB's columnar storage through
+record-batch Arrow handoffs. A smart query router auto-picks an
+R-tree or B-tree spatial index per query, and a closure-cache /
+recursive-CTE relational dispatcher selects the right strategy based
+on the corpus's actual hierarchy depth. The full `FeatureDB` /
+`Feature` / `create_db` / `DataIterator` / `GFFWriter` /
+`merge_criteria` legacy API is preserved verbatim — most users
+migrate by changing one import line.
+
+### Three reasons it matters
+
+1. **🚀 ≥ 32× faster GENCODE GTF ingest** (v49, 6.07 M lines) — and
+   **mathematically more efficient**: legacy needs a Python loop +
+   ~5 million correlated SQLite subqueries to *invent* the missing
+   gene/transcript rows, while gffbase does the same work in two
+   set-based DuckDB `GROUP BY` aggregations + one recursive CTE.
+   *([Proven by a same-release GTF/GFF3 head-to-head](#-comprehensive-human-genome-annotations--validated-across-every-canonical-corpus))*
+2. **⚡ 36.68× faster bulk ML extraction** — `children_batched(format='arrow')`
+   returns 50 000 transcripts → 1.6 M exons as a zero-copy PyArrow
+   table in **1.16 s**. No Python `Feature` objects, ever. *([How?](#-the-killer-feature--zero-copy-pyarrow-for-ml-pipelines))*
+3. **🛡️  Validated NCBI compliance** — all four canonical human-genome
+   annotations (GENCODE / RefSeq / MANE / CHESS 3) ingest cleanly with
+   **zero strict-mode warnings**. RefSeq's split-CDS duplicate-ID
+   convention is handled automatically.
+
+---
+
+## ⚡ Comprehensive Human Genome Annotations — validated across every canonical corpus
+
+Validated head-to-head against legacy `gffutils` on the four canonical
+human-genome annotation sources, including the **GENCODE v49 GTF and
+GFF3 versions of the same release** — a same-biology, same-features,
+different-format pairing that exposes the GTF Synthesis Advantage in
+its purest form:
+
+| Corpus                   | Format | Lines      | gffbase ingest | legacy ingest | **speedup**   | spatial qps | batched (5 k anchors) |
+| ------------------------ | :----: | ---------: | -------------: | ------------: | ------------: | ----------: | --------------------: |
+| **GENCODE v49** (basic)  |  GTF   |  6,068,892 |   **4 min 37 s** | ≥ 2 hr 30 min[^1]    | **🚀 ≥ 32×**  |   **1,204** | 172 ms / 596 k desc   |
+| **GENCODE v49** (basic)  |  GFF3  |  6,066,054 |   **6 min 7 s** | 11 min 23 s    | **1.86×**     |   **1,292** | 422 ms / 1.93 M desc  |
+| **RefSeq GRCh38.p14**    |  GFF3  |  4,932,571 |   **4 min 12 s**[^2] |   6 min 5 s   | **1.45×**     |   **1,011** | 263 ms / 999 k desc   |
+| **MANE v1.5** (Ensembl)  |  GFF3  |    524,834 |    **21.6 s**  |    45.1 s     | **2.09×**     |   **1,766** |  78 ms / 156 k desc   |
+| **CHESS 3.1.3**          |  GFF3  |  2,761,061 |    **53.6 s**  |  2 min 13.1 s | **2.48×**     |   **1,175** |  91 ms / 161 k desc   |
+
+[^1]: Legacy `gffutils.create_db()` on GENCODE v49 GTF (6.07 M lines) hits the bench's safety-valve cap (75 min). The reported wall is a conservative 2× extrapolation — the canonical GENCODE v45 GTF (2.0 M lines, 3× smaller) ran uncapped at **3,582 s (59 min 42 s)** on the same hardware, so the v49 wall is well past 2 hours. See [Performance Comparison §"GTF Synthesis Advantage"](PERFORMANCE_COMPARISON.md#-the-gtf-synthesis-advantage--proven-by-a-same-release-head-to-head) for the formal cost model.
+[^2]: Result of the v0.1.0 ingest-pipeline optimization — the same RefSeq corpus used to take 7 min 49 s before the GFF3 path was re-architected to stamp `seqid_y` and `bbox` inline during the Arrow batch INSERT.
+
+**The same biological release, ingested in two different formats, by
+two different engines** — that's the load-bearing comparison. Legacy
+GFF3 ingest finishes in 11 min because every parent edge is explicit;
+legacy GTF ingest takes hours because the parent rows have to be
+*invented* from the data (one Python ↔ SQLite round-trip per missing
+row). gffbase replaces those millions of round-trips with two
+set-based DuckDB `GROUP BY` aggregations + one recursive CTE — the
+**same code path** runs for GTF and GFF3, which is why the gffbase
+column barely shifts (4 min 37 s → 6 min 7 s) between the two rows
+while the legacy column balloons by 13×–20×.
+
+**Robustness:** every corpus ingests cleanly with **zero strict-mode
+warnings** from the NCBI-spec-hardened Rust parser (9 enforced rules,
+line-numbered `GFFFormatError`, opt-in non-strict mode). RefSeq's
+notorious duplicate-`ID=cds-NP_xxx` convention (split CDS segments) is
+handled transparently — gffbase mirrors
+`gffutils.merge_strategy="create_unique"` automatically and records the
+remap in the `duplicates` table. No config knobs to flip.
+
+📊 Full reproducible numbers + per-corpus root-cause analysis:
+[`PERFORMANCE_COMPARISON.md`](PERFORMANCE_COMPARISON.md). Re-run via
+`python benchmarks/06_mega.py --legacy-timeout 900`.
+
+---
+
+## 🚀 The Killer Feature — zero-copy PyArrow for ML pipelines
+
+Modern ML genomics pipelines have one shape: **pull every exon for
+50 000 transcripts, push the column-oriented table into a tensor,
+train.** Legacy `gffutils` forces a per-feature Python loop —
+constructing 1.6 M throwaway `Feature` objects per pull, which crushes
+both wall time and memory. gffbase bypasses Python entirely with a
+single batched call that returns DuckDB's internal Arrow buffers
+directly:
+
+```python
+# 50 000 transcript IDs → every exon, in one query.
+# Returns a zero-copy pyarrow.Table — no Python `Feature` object
+# is constructed at any layer.
+exons = db.children_batched(
+    transcript_ids,
+    featuretype="exon",
+    format="arrow",        # or "df" / "polars"
+)
+
+# Hand off directly to PyTorch / Hugging Face datasets / JAX / Lance.
+import torch
+starts = torch.from_numpy(exons.column("start").to_numpy())
+ends   = torch.from_numpy(exons.column("end").to_numpy())
+# The "anchor" column carries the input id for each row, so you can
+# reconstruct per-transcript groups without re-issuing N queries.
+```
+
+**Numbers for that one call** (50 000 transcripts, GENCODE basic
+annotation, returning 1.6 M exon rows):
+
+| Path                                      |        Wall | vs legacy        |
+| ----------------------------------------- | ----------: | ---------------- |
+| gffbase `children_batched(format='arrow')`|   **1.16 s**| **36.68× faster**|
+| legacy `gffutils` row-by-row loop         |     42.55 s | 1.0× (baseline)  |
+| gffbase row-by-row loop                   |     ≥ 642 s | 0.07× *(slower!)*|
+
+This is **the** reason GFFBase exists. Iterating
+`for x in ids: db.children(x)` with DuckDB pays vectorization startup
+per call and is *slower* than legacy's SQLite row-by-row path — but
+the batched API obliterates both row-by-row paths because it issues
+one set-based SQL query and avoids constructing any Python `Feature`
+objects whatsoever.
+
+`region_batched(...)` and `parents_batched(...)` have the same
+zero-copy contract for spatial and parent workloads.
+
+---
+
+## 📦 Installation
+
+```bash
+pip install gffbase
+```
+
+Universal `abi3-py39` wheels — single binary per arch covers CPython
+3.9 → 3.13. No Rust toolchain required at install time.
+
+For source/dev installs (Rust ≥ 1.69 + maturin):
+
+```bash
+pip install -e .[dev]
+maturin develop --release
+```
+
+---
+
+## 🏃 Quick start — row-by-row (drop-in for `gffutils`)
+
+```python
+from gffbase import create_db
+
+# 1. Ingest a GTF/GFF3 in seconds (auto-detects format, gzipped OK).
+db = create_db("gencode.v49.chr_patch_hapl_scaff.basic.annotation.gtf.gz",
+               "gencode.duckdb", force=True)
+
+# 2. Walk a single gene's hierarchy.
+for tx in db.children("ENSG00000139618", level=1, featuretype="transcript"):
+    print(tx.id, tx.start, tx.end)
+
+# 3. Spatial overlap query — uses the per-seqid R-tree under the hood.
+for f in db.region("chr17:43044295-43125483", featuretype="exon"):
+    print(f)
+```
+
+If you're migrating from `gffutils`, change one line:
+
+```python
+import gffbase as gffutils    # one-line alias migration
+db = gffutils.create_db(...)  # everything else identical
+```
+
+(But please read the [Migration Guide](MIGRATION.md) first — it has
+**one** important note about ML loops.)
+
+---
+
+## 🤖 Quick start — vectorized for ML
+
+```python
+from gffbase import FeatureDB
+
+db = FeatureDB("gencode.duckdb")
+
+# Pull every exon for 50 000 transcripts — one set-based SQL query.
+exons = db.children_batched(
+    transcript_ids,                # iterable of 50 000 IDs
+    featuretype="exon",
+    format="arrow",                # "df" / "polars" also supported
+)
+# exons is a pyarrow.Table sharing memory with DuckDB. No copies.
+
+# Spatial: "for each ATAC-seq peak, find every overlapping CDS."
+peaks = [("chr1", 100_000, 110_000), ("chr1", 200_000, 210_000), ...]
+overlaps = db.region_batched(peaks, featuretype="CDS", format="arrow")
+```
+
+See the [Machine Learning Workflows
+Cookbook](docs/cookbooks/machine_learning_workflows.md) for end-to-end
+pipelines with PyTorch and Hugging Face `datasets`.
+
+---
+
+## ✨ What's inside
+
+- **Rust + PyO3 parser** — SIMD line/tab splitting, lazy URL-decoding,
+  GTF semicolon-in-quotes safe, gzipped input transparent. Hardened
+  against the NCBI GFF3 spec (line-numbered `GFFFormatError`,
+  strict / non-strict modes, 9 enforced rules).
+- **DuckDB columnar storage** — 7-table schema, set-based GTF
+  gene/transcript synthesis, recursive-CTE transitive closure,
+  per-seqid-banded R-tree spatial index built inline during ingest.
+- **Smart routing** — `region()` auto-picks R-tree vs B-tree;
+  `children()` auto-picks closure cache vs dynamic CTE based on
+  measured corpus depth.
+- **Vectorized batched API** — `children_batched`, `parents_batched`,
+  `region_batched` return `pyarrow.Table` / `pandas.DataFrame` /
+  `polars.DataFrame` directly out of DuckDB's buffer pool.
+- **Drop-in legacy API** — `FeatureDB`, `Feature`, `create_db`,
+  `DataIterator`, `GFFWriter`, `merge_criteria`, `interfeatures`,
+  `bed12`, `execute()` SQL escape hatch, `export_sqlite()`.
+- **abi3 wheels** — single binary per arch covers CPython 3.9–3.13.
+
+---
+
+## 📚 Documentation
+
+Full site (rendered with MkDocs Material) — build it locally:
+
+```bash
+pip install -e .[docs]
+mkdocs serve            # http://localhost:8000
+```
+
+| Page                                                                | What's there                                                              |
+| ------------------------------------------------------------------- | ------------------------------------------------------------------------- |
+| [Usage Gallery](docs/usage_gallery.md)                              | Copy-pasteable snippets for every public API method                       |
+| [Performance comparison](PERFORMANCE_COMPARISON.md)                 | Head-to-head numbers across every canonical human-genome annotation + per-corpus root-cause analysis |
+| [Migration guide for `gffutils` users](MIGRATION.md)                | Drop-in compat checklist + the one OLAP/OLTP gotcha you must understand   |
+| [Cookbooks](docs/cookbooks/)                                        | GENCODE/Ensembl, RefSeq, MANE, ML workflows                               |
+| [API reference](docs/api/)                                          | Every public method, full signatures + docstrings                         |
+
+---
+
+## 🧪 Testing
+
+```bash
+pip install -e .[test]
+pytest                  # 523 passed, 7 skipped, 99.19% coverage
+```
+
+CI runs the full matrix on Linux + macOS + Windows, both R-tree and
+B-tree fallback paths, on Python 3.9 / 3.11 / 3.13.
+
+---
+
+## 🤝 Contributing
+
+GFFBase welcomes pull requests, bug reports, and feature suggestions.
+Start with [`CONTRIBUTING.md`](CONTRIBUTING.md) for the full guide:
+
+- Rust + Python development setup (`maturin develop --release`)
+- Running the test suite + the 99 % coverage gate
+- Branch naming, Conventional Commits, the PR checklist
+
+The repo ships standard
+[issue templates](.github/ISSUE_TEMPLATE/) and a
+[PR template](.github/PULL_REQUEST_TEMPLATE.md) so new
+contributions land with the context maintainers need to triage them
+quickly.
+
+---
+
+## 🪪 License
+
+Apache License 2.0. See [`LICENSE`](LICENSE).
+
+---
+
+**Citation:** if GFFBase helps your research, please cite the project at
+the [Releases page](https://github.com/Kuanhao-Chao/gffbase/releases).
+