structflo-ner 0.1.0__tar.gz

structflo_ner-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,31 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  check:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v4
+
+      - name: Set up Python
+        run: uv python install 3.12
+
+      - name: Install dependencies
+        run: uv sync --dev
+
+      - name: Lint
+        run: uv run ruff check structflo/ tests/
+
+      - name: Format
+        run: uv run ruff format --check structflo/ tests/
+
+      - name: Test
+        run: uv run pytest -q

structflo_ner-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,35 @@
+name: Publish to PyPI
+
+on:
+  push:
+    tags:
+      - "v*"
+
+jobs:
+  publish:
+    runs-on: ubuntu-latest
+    permissions:
+      id-token: write  # trusted publishing
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v4
+
+      - name: Set up Python
+        run: uv python install 3.12
+
+      - name: Install dependencies
+        run: uv sync --dev
+
+      - name: Lint
+        run: uv run ruff check structflo/ tests/
+
+      - name: Test
+        run: uv run pytest -q
+
+      - name: Build
+        run: uv build
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

structflo_ner-0.1.0/.gitignore

@@ -0,0 +1,34 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+
+# Distribution / packaging
+dist/
+build/
+*.egg-info/
+*.egg
+
+# Virtual environments
+.venv/
+venv/
+
+# Testing
+.pytest_cache/
+.coverage
+htmlcov/
+
+# IDEs
+.vscode/
+.idea/
+*.swp
+*.swo
+
+# OS
+.DS_Store
+Thumbs.db
+
+# Environment variables
+.env
+.env.*

structflo_ner-0.1.0/Makefile

@@ -0,0 +1,35 @@
+.PHONY: install lint format check test clean build
+
+## Install all dependencies (including dev)
+install:
+	uv sync --dev
+
+## Run ruff linter
+lint:
+	uv run ruff check structflo/ tests/
+
+## Auto-fix lint issues
+fix:
+	uv run ruff check --fix structflo/ tests/
+
+## Format code with ruff
+format:
+	uv run ruff format structflo/ tests/
+
+## Check formatting + lint (CI-friendly, no changes)
+check:
+	uv run ruff format --check structflo/ tests/
+	uv run ruff check structflo/ tests/
+
+## Run tests
+test:
+	uv run pytest -q
+
+## Remove build artifacts
+clean:
+	rm -rf dist/ build/ *.egg-info .pytest_cache
+	find . -type d -name __pycache__ -exec rm -rf {} +
+
+## Build package
+build:
+	uv build

structflo_ner-0.1.0/PKG-INFO

@@ -0,0 +1,113 @@
+Metadata-Version: 2.4
+Name: structflo-ner
+Version: 0.1.0
+Summary: Drug discovery NER wrapper around LangExtract — zero-config entity extraction for chemistry and biology.
+License: Apache-2.0
+Requires-Python: >=3.10
+Requires-Dist: langextract>=1.1.1
+Provides-Extra: dataframe
+Requires-Dist: pandas>=1.5; extra == 'dataframe'
+Description-Content-Type: text/markdown
+
+# structflo.ner
+
+Drug discovery NER powered by [LangExtract](https://github.com/google/langextract).
+
+Extract compounds, targets, bioactivity data, diseases, and more from scientific text — zero configuration required.
+
+## Install
+
+```bash
+pip install structflo-ner
+# or with uv
+uv add structflo-ner
+
+# optional pandas support
+pip install "structflo-ner[dataframe]"
+```
+
+## Quick start
+
+```python
+from structflo.ner import NERExtractor
+
+extractor = NERExtractor(api_key="YOUR_GEMINI_KEY")
+result = extractor.extract(
+    "Gefitinib (ZD1839) is a first-generation EGFR inhibitor with IC50 = 0.033 µM approved for NSCLC."
+)
+
+print(result.compounds)      # [ChemicalEntity(text='Gefitinib', ...)]
+print(result.targets)        # [TargetEntity(text='EGFR', ...)]
+print(result.bioactivities)  # [BioactivityEntity(text='IC50 = 0.033 µM', ...)]
+print(result.diseases)       # [DiseaseEntity(text='NSCLC', ...)]
+
+df = result.to_dataframe()   # flat pandas DataFrame
+```
+
+## Local models via Ollama
+
+Run extraction entirely on your own hardware — no API key needed:
+
+```python
+extractor = NERExtractor(
+    model_id="gemma3:27b",
+    model_url="http://localhost:11434",
+)
+result = extractor.extract("Sorafenib inhibits VEGFR-2 and RAF kinases.")
+```
+
+Any model served by Ollama works (gemma, llama, mistral, qwen, deepseek, etc.).
+
+## Built-in profiles
+
+| Profile | Entity classes |
+|---|---|
+| `FULL` (default) | compounds, targets, diseases, bioactivities, assays, mechanisms |
+| `CHEMISTRY` | compound names, SMILES, CAS numbers, molecular formulas |
+| `BIOLOGY` | targets, gene names, protein names |
+| `BIOACTIVITY` | bioactivity measurements, assays |
+| `DISEASE` | diseases and clinical indications |
+
+```python
+from structflo.ner import NERExtractor, CHEMISTRY
+
+extractor = NERExtractor(api_key="YOUR_GEMINI_KEY")
+result = extractor.extract(text, profile=CHEMISTRY)
+```
+
+Profiles can be merged:
+
+```python
+from structflo.ner import CHEMISTRY, BIOLOGY
+
+combined = CHEMISTRY.merge(BIOLOGY)
+result = extractor.extract(text, profile=combined)
+```
+
+## Custom profiles
+
+```python
+from structflo.ner import NERExtractor, EntityProfile
+
+my_profile = EntityProfile(
+    name="kinase_inhibitors",
+    entity_classes=["compound_name", "smiles", "target", "bioactivity"],
+    prompt="Extract kinase inhibitor names, SMILES, targets, and potency values.",
+    examples=my_examples,
+)
+result = extractor.extract(text, profile=my_profile)
+```
+
+## Working with results
+
+```python
+result.all_entities()   # flat list of every entity
+result.to_dict()        # plain dictionary
+result.to_dataframe()   # pandas DataFrame (requires structflo-ner[dataframe])
+```
+
+## Notebooks
+
+See the [notebooks/](notebooks/) directory for worked examples:
+
+- **01_quickstart.ipynb** — end-to-end extraction with cloud and local models

structflo_ner-0.1.0/README.md

@@ -0,0 +1,102 @@
+# structflo.ner
+
+Drug discovery NER powered by [LangExtract](https://github.com/google/langextract).
+
+Extract compounds, targets, bioactivity data, diseases, and more from scientific text — zero configuration required.
+
+## Install
+
+```bash
+pip install structflo-ner
+# or with uv
+uv add structflo-ner
+
+# optional pandas support
+pip install "structflo-ner[dataframe]"
+```
+
+## Quick start
+
+```python
+from structflo.ner import NERExtractor
+
+extractor = NERExtractor(api_key="YOUR_GEMINI_KEY")
+result = extractor.extract(
+    "Gefitinib (ZD1839) is a first-generation EGFR inhibitor with IC50 = 0.033 µM approved for NSCLC."
+)
+
+print(result.compounds)      # [ChemicalEntity(text='Gefitinib', ...)]
+print(result.targets)        # [TargetEntity(text='EGFR', ...)]
+print(result.bioactivities)  # [BioactivityEntity(text='IC50 = 0.033 µM', ...)]
+print(result.diseases)       # [DiseaseEntity(text='NSCLC', ...)]
+
+df = result.to_dataframe()   # flat pandas DataFrame
+```
+
+## Local models via Ollama
+
+Run extraction entirely on your own hardware — no API key needed:
+
+```python
+extractor = NERExtractor(
+    model_id="gemma3:27b",
+    model_url="http://localhost:11434",
+)
+result = extractor.extract("Sorafenib inhibits VEGFR-2 and RAF kinases.")
+```
+
+Any model served by Ollama works (gemma, llama, mistral, qwen, deepseek, etc.).
+
+## Built-in profiles
+
+| Profile | Entity classes |
+|---|---|
+| `FULL` (default) | compounds, targets, diseases, bioactivities, assays, mechanisms |
+| `CHEMISTRY` | compound names, SMILES, CAS numbers, molecular formulas |
+| `BIOLOGY` | targets, gene names, protein names |
+| `BIOACTIVITY` | bioactivity measurements, assays |
+| `DISEASE` | diseases and clinical indications |
+
+```python
+from structflo.ner import NERExtractor, CHEMISTRY
+
+extractor = NERExtractor(api_key="YOUR_GEMINI_KEY")
+result = extractor.extract(text, profile=CHEMISTRY)
+```
+
+Profiles can be merged:
+
+```python
+from structflo.ner import CHEMISTRY, BIOLOGY
+
+combined = CHEMISTRY.merge(BIOLOGY)
+result = extractor.extract(text, profile=combined)
+```
+
+## Custom profiles
+
+```python
+from structflo.ner import NERExtractor, EntityProfile
+
+my_profile = EntityProfile(
+    name="kinase_inhibitors",
+    entity_classes=["compound_name", "smiles", "target", "bioactivity"],
+    prompt="Extract kinase inhibitor names, SMILES, targets, and potency values.",
+    examples=my_examples,
+)
+result = extractor.extract(text, profile=my_profile)
+```
+
+## Working with results
+
+```python
+result.all_entities()   # flat list of every entity
+result.to_dict()        # plain dictionary
+result.to_dataframe()   # pandas DataFrame (requires structflo-ner[dataframe])
+```
+
+## Notebooks
+
+See the [notebooks/](notebooks/) directory for worked examples:
+
+- **01_quickstart.ipynb** — end-to-end extraction with cloud and local models

structflo_ner-0.1.0/notebooks/01_quickstart.ipynb

@@ -0,0 +1,262 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# structflo.ner — Quickstart\n",
+    "\n",
+    "This notebook walks through the core features of `structflo.ner`:\n",
+    "\n",
+    "1. Basic extraction with a cloud model (Gemini)\n",
+    "2. Local extraction with Ollama\n",
+    "3. Using built-in profiles\n",
+    "4. Custom profiles\n",
+    "5. Working with results"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Setup\n",
+    "\n",
+    "```bash\n",
+    "uv add structflo-ner\n",
+    "# for DataFrame support\n",
+    "uv add \"structflo-ner[dataframe]\"\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from structflo.ner import NERExtractor, FULL, CHEMISTRY, BIOLOGY, BIOACTIVITY, DISEASE"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 1. Cloud model (Gemini)\n",
+    "\n",
+    "The default model is `gemini-2.5-flash`. Pass your API key or set the `GEMINI_API_KEY` environment variable."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "extractor = NERExtractor(api_key=\"YOUR_GEMINI_KEY\")  # or set GEMINI_API_KEY env var\n",
+    "\n",
+    "text = (\n",
+    "    \"Gefitinib (ZD1839) is a first-generation EGFR tyrosine kinase inhibitor \"\n",
+    "    \"with IC50 = 0.033 µM, approved for non-small cell lung cancer (NSCLC). \"\n",
+    "    \"Its SMILES is COc1cc2ncnc(Nc3ccc(F)c(Cl)c3)c2cc1OCCCN1CCOCC1.\"\n",
+    ")\n",
+    "\n",
+    "result = extractor.extract(text)\n",
+    "result"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 2. Local model via Ollama\n",
+    "\n",
+    "Run extraction on your own hardware — no API key needed.\n",
+    "\n",
+    "Make sure Ollama is running locally:\n",
+    "```bash\n",
+    "ollama serve\n",
+    "ollama pull gemma3:27b\n",
+    "```"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "local_extractor = NERExtractor(\n",
+    "    model_id=\"gemma3:27b\",\n",
+    "    model_url=\"http://localhost:11434\",\n",
+    ")\n",
+    "\n",
+    "result_local = local_extractor.extract(\n",
+    "    \"Sorafenib is a multi-kinase inhibitor targeting VEGFR-2, PDGFR, and RAF with IC50 values of 90 nM, 57 nM, and 6 nM respectively.\"\n",
+    ")\n",
+    "result_local"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 3. Built-in profiles\n",
+    "\n",
+    "Profiles control which entity types are extracted. Use them to focus the model on specific categories."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Extract only chemical entities\n",
+    "chem_result = extractor.extract(text, profile=CHEMISTRY)\n",
+    "print(\"Compounds:\", chem_result.compounds)\n",
+    "print(\"Targets:\", chem_result.targets)  # empty — not in CHEMISTRY profile"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Merge profiles to combine entity types\n",
+    "combined = CHEMISTRY.merge(BIOLOGY)\n",
+    "print(f\"Profile: {combined.name}\")\n",
+    "print(f\"Entity classes: {combined.entity_classes}\")\n",
+    "\n",
+    "combined_result = extractor.extract(text, profile=combined)\n",
+    "print(\"Compounds:\", combined_result.compounds)\n",
+    "print(\"Targets:\", combined_result.targets)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 4. Custom profiles\n",
+    "\n",
+    "Define your own extraction profiles for domain-specific use cases."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from structflo.ner import EntityProfile\n",
+    "\n",
+    "kinase_profile = EntityProfile(\n",
+    "    name=\"kinase_inhibitors\",\n",
+    "    entity_classes=[\"compound_name\", \"smiles\", \"target\", \"bioactivity\"],\n",
+    "    prompt=\"Extract kinase inhibitor names, their SMILES strings, kinase targets, and potency values (IC50, Ki, Kd).\",\n",
+    "    examples=[],  # add your own few-shot examples here for best results\n",
+    ")\n",
+    "\n",
+    "kinase_result = extractor.extract(text, profile=kinase_profile)\n",
+    "kinase_result"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 5. Working with results"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Access typed entity lists\n",
+    "print(\"Compounds:\", result.compounds)\n",
+    "print(\"Targets:\", result.targets)\n",
+    "print(\"Bioactivities:\", result.bioactivities)\n",
+    "print(\"Diseases:\", result.diseases)\n",
+    "print(\"Mechanisms:\", result.mechanisms)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Flat list of all entities\n",
+    "for entity in result.all_entities():\n",
+    "    print(f\"{entity.entity_type:20s} | {entity.text}\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Export to pandas DataFrame\n",
+    "df = result.to_dataframe()\n",
+    "df"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Serialize to dict (useful for JSON export)\n",
+    "import json\n",
+    "\n",
+    "print(json.dumps(result.to_dict(), indent=2))"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## 6. Batch extraction\n",
+    "\n",
+    "Pass a list of texts to extract from multiple documents."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "texts = [\n",
+    "    \"Imatinib inhibits BCR-ABL with IC50 = 0.6 µM in CML.\",\n",
+    "    \"Trastuzumab targets HER2 in breast cancer patients.\",\n",
+    "    \"Remdesivir (GS-5734) is an antiviral with EC50 = 0.77 µM against SARS-CoV-2.\",\n",
+    "]\n",
+    "\n",
+    "results = extractor.extract(texts)\n",
+    "\n",
+    "for i, r in enumerate(results):\n",
+    "    print(f\"\\n--- Text {i+1} ---\")\n",
+    "    for entity in r.all_entities():\n",
+    "        print(f\"  {entity.entity_type:20s} | {entity.text}\")"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "name": "python",
+   "version": "3.12.0"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 4
+}

structflo_ner-0.1.0/pyproject.toml

@@ -0,0 +1,60 @@
+[project]
+name = "structflo-ner"
+dynamic = ["version"]
+description = "Drug discovery NER wrapper around LangExtract — zero-config entity extraction for chemistry and biology."
+readme = "README.md"
+requires-python = ">=3.10"
+license = { text = "Apache-2.0" }
+dependencies = [
+    "langextract>=1.1.1",
+]
+
+[project.optional-dependencies]
+dataframe = ["pandas>=1.5"]
+
+[dependency-groups]
+dev = [
+    "pytest>=7",
+    "pytest-mock>=3",
+    "ruff>=0.8",
+]
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[tool.hatch.version]
+path = "structflo/ner/__init__.py"
+
+[tool.hatch.build.targets.wheel]
+packages = ["structflo"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+
+# --------------------------------------------------------------------------- #
+# Ruff
+# --------------------------------------------------------------------------- #
+[tool.ruff]
+target-version = "py310"
+line-length = 100
+
+[tool.ruff.lint]
+select = [
+    "E",   # pycodestyle errors
+    "W",   # pycodestyle warnings
+    "F",   # pyflakes
+    "I",   # isort
+    "UP",  # pyupgrade
+    "B",   # flake8-bugbear
+    "SIM", # flake8-simplify
+]
+ignore = [
+    "E501", # line length handled by formatter
+]
+
+[tool.ruff.lint.isort]
+known-first-party = ["structflo"]
+
+[tool.ruff.format]
+quote-style = "double"