tensor-grep 0.1.0__tar.gz

tensor_grep-0.1.0/.github/workflows/benchmark.yml

@@ -0,0 +1,13 @@
+name: Benchmarks
+on:
+  pull_request:
+    paths: ['src/**', 'tests/performance/**']
+jobs:
+  benchmark:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with: { python-version: '3.11' }
+      - run: pip install -e ".[dev]"
+      - run: pytest tests/performance/ -v --tb=short

tensor_grep-0.1.0/.github/workflows/release.yml

@@ -0,0 +1,102 @@
+name: Release
+on:
+  push:
+    tags:
+      - 'v*'
+
+jobs:
+  build-binaries:
+    name: Build on ${{ matrix.os }}
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
+    
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+          
+      - name: Install dependencies
+        run: |
+          pip install -e ".[dev]"
+          pip install nuitka
+          
+      - name: Build Binary
+        run: python build_binaries.py
+        
+      - name: Rename Artifact (Windows)
+        if: runner.os == 'Windows'
+        run: mv tg.exe tg-windows-amd64.exe
+        
+      - name: Rename Artifact (Linux)
+        if: runner.os == 'Linux'
+        run: mv tg.bin tg-linux-amd64
+        
+      - name: Rename Artifact (macOS)
+        if: runner.os == 'macOS'
+        run: mv tg.bin tg-macos-amd64
+        
+      - name: Upload Artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: binary-${{ runner.os }}
+          path: tg-*
+
+  create-release:
+    needs: build-binaries
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Download Artifacts
+        uses: actions/download-artifact@v4
+        with:
+          path: artifacts
+          
+      - name: Create GitHub Release
+        uses: softprops/action-gh-release@v2
+        with:
+          files: artifacts/**/tg-*
+          generate_release_notes: true
+
+  publish-npm:
+    needs: create-release
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Setup Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: '20'
+          registry-url: 'https://registry.npmjs.org'
+          
+      - name: Publish NPM Package
+        working-directory: npm
+        run: npm publish --access public
+        env:
+          NODE_AUTH_TOKEN: ${{ secrets.NPM_TOKEN }}
+
+  publish-docs:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+      
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: '3.11'
+          
+      - name: Install mkdocs
+        run: pip install mkdocs-material
+        
+      - name: Deploy Docs
+        run: mkdocs gh-deploy --force

tensor_grep-0.1.0/.gitignore

@@ -0,0 +1,31 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+
+# Distribution / packaging
+dist/
+build/
+*.egg-info/
+
+# Pytest / Coverage
+.pytest_cache/
+.coverage
+htmlcov/
+
+# Mypy
+.mypy_cache/
+
+# Ruff
+.ruff_cache/
+
+# IDEs
+.vscode/
+.idea/

tensor_grep-0.1.0/.hypothesis/constants/0321b4d9a56ecd5d

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\formatters\json_fmt.py
+# hypothesis_version: 6.151.9
+
+['file', 'line_number', 'matches', 'text', 'total_files', 'total_matches']

tensor_grep-0.1.0/.hypothesis/constants/08b4847abc9ccb4c

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\core\pipeline.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/09619f2e7b2664f4

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\backends\cpu_backend.py
+# hypothesis_version: 6.151.9
+
+['latin-1', 'r', 'utf-8']

tensor_grep-0.1.0/.hypothesis/constants/097541b56a59acb5

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\io\reader_fallback.py
+# hypothesis_version: 6.151.9
+
+['.gz', 'latin-1', 'rt', 'utf-8']

tensor_grep-0.1.0/.hypothesis/constants/0aa14fe3f4d66eb8

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\gpu\device_detect.py
+# hypothesis_version: 6.151.9
+
+[1024, '/run/WSL', 'linux', 'win32']

tensor_grep-0.1.0/.hypothesis/constants/1987ea0e0b593c3b

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\core\pipeline.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/244a932d4f46a9e8

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_dstorage.py
+# hypothesis_version: 6.151.9
+
+['dstorage_gpu.Tensor', 'win32']

tensor_grep-0.1.0/.hypothesis/constants/259b95774007f88d

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\formatters\table_fmt.py
+# hypothesis_version: 6.151.9
+
+['File\tLine\tMatch']

tensor_grep-0.1.0/.hypothesis/constants/26f238250b8499f1

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\gpu\device_detect.py
+# hypothesis_version: 6.151.9
+
+[1024, '/run/WSL', 'linux', 'win32']

tensor_grep-0.1.0/.hypothesis/constants/2d172376ce08cd3e

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\backends\cybert_backend.py
+# hypothesis_version: 6.151.9
+
+[0.1, 0.8, 0.9, 'INT64', 'bert-base-uncased', 'confidence', 'cybert', 'error', 'info', 'input_ids', 'label', 'localhost:8000', 'logits', 'np', 'warn']

tensor_grep-0.1.0/.hypothesis/constants/2fec718b7d5b449f

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\backends\cybert_backend.py
+# hypothesis_version: 6.151.9
+
+[0.1, 0.8, 0.9, 'INT64', 'bert-base-uncased', 'confidence', 'cybert', 'error', 'info', 'input_ids', 'label', 'localhost:8000', 'logits', 'np', 'warn']

tensor_grep-0.1.0/.hypothesis/constants/3881db1d5f997c09

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\gpu\memory_manager.py
+# hypothesis_version: 6.151.9
+
+[0.8]

tensor_grep-0.1.0/.hypothesis/constants/468c441a699c5e31

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_cudf.py
+# hypothesis_version: 6.151.9
+
+['cudf.Series']

tensor_grep-0.1.0/.hypothesis/constants/54e4b49940b89cd5

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\gpu\memory_manager.py
+# hypothesis_version: 6.151.9
+
+[0.8]

tensor_grep-0.1.0/.hypothesis/constants/5a48abcf717d7583

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\backends\cpu_backend.py
+# hypothesis_version: 6.151.9
+
+['latin-1', 'r', 'utf-8']

tensor_grep-0.1.0/.hypothesis/constants/6524b0fcb7a55341

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_kvikio.py
+# hypothesis_version: 6.151.9
+
+['r']

tensor_grep-0.1.0/.hypothesis/constants/67abd2ccbf381759

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\core\query_analyzer.py
+# hypothesis_version: 6.151.9
+
+['anomaly', 'classify', 'detect', 'extract entities']

tensor_grep-0.1.0/.hypothesis/constants/6b442457f5745159

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\backends\base.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/6dfa700ee8bb8fa5

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_fallback.py
+# hypothesis_version: 6.151.9
+
+['.gz', 'latin-1', 'rt', 'utf-8']

tensor_grep-0.1.0/.hypothesis/constants/72f590d9256fc7a6

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\core\result.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/740a6fb9febec20c

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\formatters\json_fmt.py
+# hypothesis_version: 6.151.9
+
+['file', 'line_number', 'matches', 'text', 'total_files', 'total_matches']

tensor_grep-0.1.0/.hypothesis/constants/74ea32f85ecfd56c

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\backends\base.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/a3fd68449bfef454

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\formatters\base.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/ab596ae2844fd274

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\base.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/ac1d131da99aea9b

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\io\reader_cudf.py
+# hypothesis_version: 6.151.9
+
+['cudf.Series']

tensor_grep-0.1.0/.hypothesis/constants/b0714817c49af4f1

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_fallback.py
+# hypothesis_version: 6.151.9
+
+['.gz', 'latin-1', 'rt', 'utf-8']

tensor_grep-0.1.0/.hypothesis/constants/b0f490b7b2d6bf29

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_kvikio.py
+# hypothesis_version: 6.151.9
+
+['r']

tensor_grep-0.1.0/.hypothesis/constants/c30bf139df59b92f

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\formatters\base.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/c63b8b09cb09eb60

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_cudf.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/cabbe58a6de53f9b

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\backends\cudf_backend.py
+# hypothesis_version: 6.151.9
+
+[512, 1024]

tensor_grep-0.1.0/.hypothesis/constants/cba1b20bea830cc3

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\formatters\csv_fmt.py
+# hypothesis_version: 6.151.9
+
+['file', 'line_number', 'text']

tensor_grep-0.1.0/.hypothesis/constants/d7e5661845384844

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\formatters\csv_fmt.py
+# hypothesis_version: 6.151.9
+
+['file', 'line_number', 'text']

tensor_grep-0.1.0/.hypothesis/constants/da39a3ee5e6b4b0d

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\__init__.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/dff29feff30d7749

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\gpu\memory_manager.py
+# hypothesis_version: 6.151.9
+
+[0.8]

tensor_grep-0.1.0/.hypothesis/constants/e2c8c4005a3c9dad

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_dstorage.py
+# hypothesis_version: 6.151.9
+
+['win32']

tensor_grep-0.1.0/.hypothesis/constants/f06fa362a827014e

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\formatters\ripgrep_fmt.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/f167f3c3dea5b11a

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\formatters\ripgrep_fmt.py
+# hypothesis_version: 6.151.9
+
+[]

tensor_grep-0.1.0/.hypothesis/constants/f46aa3ceacc645fd

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\formatters\table_fmt.py
+# hypothesis_version: 6.151.9
+
+['File\tLine\tMatch']

tensor_grep-0.1.0/.hypothesis/constants/f5078de4866a65c6

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\tensor_grep\backends\cudf_backend.py
+# hypothesis_version: 6.151.9
+
+[512, 1024]

tensor_grep-0.1.0/.hypothesis/constants/f9ba50acb0fd2077

@@ -0,0 +1,4 @@
+# file: C:\dev\projects\cudf-grep\src\cudf_grep\io\reader_cudf.py
+# hypothesis_version: 6.151.9
+
+['cudf.Series']

tensor_grep-0.1.0/.hypothesis/patches/2026-02-24--4ebab487.patch

@@ -0,0 +1,16 @@
+From HEAD Mon Sep 17 00:00:00 2001
+From: Hypothesis 6.151.9 <no-reply@hypothesis.works>
+Date: Wed, 25 Feb 2026 04:56:08
+Subject: [PATCH] Hypothesis: add explicit examples
+
+---
+--- ./tests\property\test_tokenizer_props.py
++++ ./tests\property\test_tokenizer_props.py
+@@ -4,6 +4,7 @@
+ pytestmark = pytest.mark.property
+ 
+ @given(st.text(min_size=1, max_size=10000, alphabet=st.characters(blacklist_categories=("Cs",))))
++@example(text="0").via("discovered failure")
+ def test_tokenizer_never_crashes_on_valid_text(text):
+     from cudf_grep.backends.cybert_backend import tokenize
+     tokens = tokenize([text])

tensor_grep-0.1.0/PKG-INFO

@@ -0,0 +1,32 @@
+Metadata-Version: 2.4
+Name: tensor-grep
+Version: 0.1.0
+Requires-Python: >=3.11
+Requires-Dist: rich>=13.0
+Requires-Dist: typer[all]>=0.12
+Provides-Extra: ast
+Requires-Dist: tree-sitter-javascript; extra == 'ast'
+Requires-Dist: tree-sitter-python; extra == 'ast'
+Requires-Dist: tree-sitter>=0.22; extra == 'ast'
+Provides-Extra: dev
+Requires-Dist: hypothesis>=6.100; extra == 'dev'
+Requires-Dist: mutmut>=3.0; extra == 'dev'
+Requires-Dist: mypy>=1.11; extra == 'dev'
+Requires-Dist: pytest-asyncio>=0.24; extra == 'dev'
+Requires-Dist: pytest-cov>=5.0; extra == 'dev'
+Requires-Dist: pytest-mock>=3.14; extra == 'dev'
+Requires-Dist: pytest-snapshot>=0.9; extra == 'dev'
+Requires-Dist: pytest>=8.0; extra == 'dev'
+Requires-Dist: ruff>=0.6; extra == 'dev'
+Provides-Extra: gpu
+Requires-Dist: cudf-cu12; extra == 'gpu'
+Requires-Dist: kvikio-cu12; extra == 'gpu'
+Requires-Dist: torch-geometric>=2.5.0; extra == 'gpu'
+Requires-Dist: torch>=2.0; extra == 'gpu'
+Provides-Extra: gpu-win
+Requires-Dist: dstorage-gpu>=1.0; extra == 'gpu-win'
+Requires-Dist: torch-geometric>=2.5.0; extra == 'gpu-win'
+Requires-Dist: torch>=2.0; extra == 'gpu-win'
+Provides-Extra: nlp
+Requires-Dist: transformers>=4.40; extra == 'nlp'
+Requires-Dist: tritonclient[all]; extra == 'nlp'

tensor_grep-0.1.0/README.md

@@ -0,0 +1,87 @@
+# tensor-grep (tg)
+
+**The GPU-Accelerated Semantic Log Parsing CLI**
+
+`tensor-grep` combines the raw regex speed of traditional tools like `ripgrep` with the semantic understanding of Transformer AI networks (`cyBERT`), parallelized across multiple GPUs using NVIDIA RAPIDS `cuDF`.
+
+## Features
+* **Drop-in Replacement:** Supports 70+ `ripgrep` CLI flags (e.g., `-i`, `-v`, `-C`, `-g`, `-t`).
+* **AST-Grep Parity (NEW):** Structural code searching via PyTorch Geometric Graph Neural Networks (GNNs). Run `tg run`, `tg scan`, `tg lsp` natively on your GPU!
+* **Multi-GPU Scaling:** Automatically detects and shards massive log files across dual, quad, or enterprise GPU arrays.
+* **Semantic NLP Classification:** Utilize cyBERT to classify logs contextually (e.g. identify "ERROR" severity without explicit regexes) in a single pass.
+* **CPU Fallback Resiliency:** Works gracefully on Windows, macOS, and CPU-only systems using a resilient Python Regex backend.
+
+---
+
+## 💻 Hardware & Software Requirements
+
+`tensor-grep` runs on any machine with Python 3.11+ using its highly-optimized CPU fallback. However, to unlock its 3x-10x GPU-accelerated speeds, your system must meet these requirements:
+
+* **Hardware:**
+  * NVIDIA GPU (GTX 10-Series or newer, RTX 30/40/50 series recommended)
+  * Minimum 4GB VRAM (8GB+ recommended for massive logs)
+* **Software / Drivers:**
+  * **NVIDIA Display Drivers:** v535.xx or newer
+  * **CUDA Toolkit:** 12.0 or newer (CUDA 12.4 highly recommended)
+* **Python Environments:**
+  * **Linux / WSL2:** Requires NVIDIA RAPIDS `cuDF` (`cudf-cu12`) for maximum throughput via instant `fork()` process spanning.
+  * **Windows Native:** Requires PyTorch with CUDA 12 support (`torch==2.5.1+cu124`). Note that PyTorch `spawn()` on Windows adds a ~10-second initial overhead, so for files <50MB, `tg` intelligently routes to the CPU backend instead.
+
+---
+
+## 🚀 GPU Acceleration Setup (CRITICAL)
+
+To achieve the 3x-10x performance gains over traditional CPU tools, `tensor-grep` utilizes NVIDIA's RAPIDS suite (`cuDF`) on Linux/WSL2, and falls back to an optimized native **PyTorch Tensor** pipeline when running natively on Windows.
+
+### Windows Native GPU Support (No WSL2 Required)
+If you do not want to use WSL2 and want to run `tensor-grep` natively from PowerShell/CMD while still utilizing your GPU, you can use `uv` (the fast Python package manager) to dynamically provision an isolated Python 3.12 environment with CUDA bindings:
+
+```powershell
+# Run using uv to automatically pull PyTorch CUDA 12.4 hooks securely on Windows
+uv run --python 3.12 --extra-index-url https://download.pytorch.org/whl/cu124 --index-strategy unsafe-best-match --with "torch==2.5.1+cu124" tg search "ERROR" /var/logs
+```
+`tensor-grep` will automatically detect Windows + PyTorch and dispatch workloads to the `TorchBackend`. 
+
+#### ⚠️ Windows PyTorch Spawn Overhead
+Because Windows Python `multiprocessing` requires `spawn()` rather than Linux's `fork()`, the PyTorch CUDA context takes ~11 seconds to initialize across multiple worker processes on Windows. 
+- For small files (< 50MB), `tensor-grep` automatically bypasses the GPU on Windows to avoid this delay, routing to an optimized `CPUBackend` instead.
+- For massive logs (> 200MB), the 11s Windows spawn overhead is absorbed by the sheer throughput of the GPU matrix math.
+
+### Linux / Windows WSL2 (Maximum Enterprise Performance) 🚀
+For absolute maximum performance using raw CUDA C++ string bindings (`cuDF`), **run tensor-grep inside WSL2 or Linux.**
+Because Linux uses `fork()`, process initialization is practically instantaneous, meaning you will actually see sub-`0.02s` speeds across your dual GPUs!
+
+```bash
+# If using a RAPIDS conda environment:
+conda activate rapids-24.04
+
+# Or using uv to pull the linux cuDF wheels directly:
+uv run --python 3.12 --extra-index-url https://pypi.nvidia.com --with "cudf-cu12" python run_benchmarks.py
+```
+
+Once installed, `tensor-grep` will automatically detect `cuDF`, discover your GPUs, and route all regex and string operations directly to your video cards' VRAM using the `CuDFBackend`.
+
+### 3. Install tensor-grep
+```bash
+pip install tensor-grep
+```
+
+Once installed, `tensor-grep` will automatically detect `cuDF`, discover your GPUs, and route all regex and string operations directly to your video cards' VRAM.
+
+## Usage
+
+```bash
+# Standard regex search (GPU Accelerated)
+tg search "Exception.*timeout" /var/logs
+
+# Context lines, case-insensitive, ripgrep parity
+tg search -i -C 2 "database" /var/logs
+
+# AI Semantic Classification
+tg classify /var/logs/syslog.log --format json
+
+# AST Structural Code Search (ast-grep parity via PyTorch GNNs)
+tg run --ast --lang python "if ($A) { return $B; }" ./src
+tg scan -c sgconfig.yml
+tg lsp
+```