aman/kg-scr
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

added scaffold struct for the prj

Browse Source
This commit is contained in:
Aman Nalakath
2026-03-24 14:58:58 +01:00
parent 95655686f9
commit cbfcf1e315
26 changed files with 6655 additions and 312 deletions
Download Patch File Download Diff File Expand all files Collapse all files

81
.gitignore vendored Normal file
View File

@@ -0,0 +1,81 @@
# Byte-compiled / optimized / DLL files
__pycache__/
*.py[cod]
*$py.class
# C extensions
*.so
# Distribution / packaging
.Python
build/
develop-eggs/
dist/
downloads/
eggs/
.eggs/
lib/
lib64/
parts/
sdist/
var/
wheels/
*.egg-info/
.installed.cfg
*.egg
# PyInstaller
*.manifest
*.spec
# Installer logs
pip-log.txt
pip-delete-this-directory.txt
# Unit test / coverage reports
htmlcov/
.tox/
.nox/
.coverage
.coverage.*
.cache
nosetests.xml
coverage.xml
*.cover
*.py,cover
.hypothesis/
.pytest_cache/
# Translations
*.mo
*.pot
# Virtual environments
venv/
env/
ENV/
.venv
# IDEs
.vscode/
.idea/
*.swp
*.swo
*~
# Jupyter Notebook
.ipynb_checkpoints
# IPython
profile_default/
ipython_config.py
# pyenv
.python-version
# OS
.DS_Store
Thumbs.db
# Project specific
data/

312
functions.ipynb
View File

@@ -1,312 +0,0 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": null,
"id": "e58ed372",
"metadata": {},
"outputs": [],
"source": [
"import platform\n",
"from pathlib import Path\n",
"import pytesseract\n",
"from PIL import Image\n",
"from txtai.embeddings import Embeddings"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "10a7eff9",
"metadata": {},
"outputs": [],
"source": [
"def_paths = {\n",
" \"Darwin\": Path.home() / \"Desktop\",\n",
" \"Windows\": Path.home() / \"Pictures\" / \"Screenshots\",\n",
" \"Linux\": Path.home() / \"Pictures\",\n",
"}\n",
"\n",
"sc_pathpatterns = {\n",
" \"Darwin\": [\"SCR*.png\", \"Screenshot*.png\"],\n",
" \"Windows\": [\"Screenshot*.png\"],\n",
" \"Linux\": [\"Screenshot*.png\", \"scrot*.png\", \"screenshot*.png\"],\n",
"}\n",
"\n",
"def get_screenshots(path: str | Path | None = None):\n",
" if path is None:\n",
" path = def_paths.get(platform.system(), Path.home())\n",
" os_name = platform.system()\n",
" patterns = sc_pathpatterns.get(os_name, [\"SCR*.png\"]) # assume mac\n",
" path = Path(path)\n",
" results = []\n",
" for pattern in patterns:\n",
" # results.extend(path.glob(pattern))\n",
" results.extend(str(f.absolute()) for f in path.glob(pattern))\n",
" \n",
" return sorted(set(results))\n",
"def extract_text(path, limit: int = 50):\n",
" text_scr = []\n",
" for img in get_screenshots(path):\n",
" image = Image.open(img) \n",
" text = pytesseract.image_to_string(image)\n",
" text_scr.append(text)\n",
" return text_scr"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e73d6386",
"metadata": {},
"outputs": [],
"source": [
"txt = extract_text(\"/Users/Aman/Pictures\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "a613a361",
"metadata": {},
"outputs": [],
"source": [
"# create embeddings\n",
"embeddings = Embeddings({\n",
" \"path\": \"sentence-transformers/all-MiniLM-L6-v2\",\n",
" \"content\": True\n",
"})\n",
"\n",
"# do indexing\n",
"embeddings.index(txt)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "7982d00e",
"metadata": {},
"outputs": [],
"source": [
"#print(txt)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "6f94de70",
"metadata": {},
"outputs": [],
"source": [
"# embeddings search\n",
"print(\"%-20s %s\" % (\"Query\", \"Best Match\"))\n",
"print(\"-\" * 50)\n",
"\n",
"for query in [\"genome\"]:\n",
" results = embeddings.search(query, 100)\n",
" for r in results:\n",
" print(r[\"text\"])"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "fa9b189d",
"metadata": {},
"outputs": [],
"source": [
"#results"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "cb5ee81e",
"metadata": {},
"outputs": [],
"source": [
"#embeddings.save(\"index\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "c263aee0",
"metadata": {},
"outputs": [],
"source": [
"embeddings = Embeddings()\n",
"embeddings.load(\"index\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "10c81e27",
"metadata": {},
"outputs": [],
"source": [
"import os\n",
"os.environ[\"OPENROUTER_API_KEY\"] = \"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\""
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "e9519cf2",
"metadata": {},
"outputs": [],
"source": [
"from txtai import LLM"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "98164787",
"metadata": {},
"outputs": [],
"source": [
"#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"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "58bce2ae",
"metadata": {},
"outputs": [],
"source": [
"OPENROUTER_API_KEY = os.getenv(\"OPENROUTER_API_KEY\")\n",
"OPENROUTER_BASE_URL = os.getenv(\"OPENROUTER_API_BASE\", \"https://openrouter.ai/api/v1\")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "8e20bf7e",
"metadata": {},
"outputs": [],
"source": [
"messages = \"What is Hi-C and how does it work?\""
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "41f0f066",
"metadata": {},
"outputs": [],
"source": [
"import litellm\n",
"\n",
"response = litellm.completion(\n",
" model=\"openrouter/minimax/minimax-m2.5:free\",\n",
" messages=[\n",
" {\"role\": \"user\", \"content\": \"How do population size fluctuations affect effective population size??\"}\n",
" ]\n",
")"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "8caf0ff4",
"metadata": {},
"outputs": [],
"source": [
"#print(response)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "879c7011",
"metadata": {},
"outputs": [],
"source": [
"# Just the answer\n",
"print(\"Answer:\", response.choices[0].message.content)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "b2f7af13",
"metadata": {},
"outputs": [],
"source": [
"# The reasoning/thinking\n",
"print(\"Reasoning:\", response.choices[0].message.reasoning_content)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "0e29bc4c",
"metadata": {},
"outputs": [],
"source": [
"# Token usage\n",
"print(\"Tokens used:\", response.usage.total_tokens)"
]
},
{
"cell_type": "code",
"execution_count": null,
"id": "4c5ca3c7",
"metadata": {},
"outputs": [],
"source": [
"# do embedding search\n",
"question = \"How do population size fluctuations affect effective population size?\"\n",
"results = embeddings.search(question, 3)\n",
"context = \"\\n\\n\".join([r[\"text\"] for r in results]) # pass to llm\n",
"\n",
"# verify\n",
"print(\"Retrieved from docs\")\n",
"for r in results:\n",
" print(f\"[Score: {r['score']:.3f}] {r['text'][:150]}...\")\n",
" print()\n",
"\n",
"# send with context\n",
"response = litellm.completion(\n",
" model=\"openrouter/minimax/minimax-m2.5:free\",\n",
" messages=[\n",
" {\n",
" \"role\": \"system\",\n",
" \"content\": \"Answer ONLY using the provided context. Cite which parts you're drawing from. If the context doesn't cover something, say 'not in my documents'.\"\n",
" },\n",
" {\n",
" \"role\": \"user\",\n",
" \"content\": f\"Context from my documents:\\n{context}\\n\\nQuestion: {question}\"\n",
" }\n",
" ]\n",
")\n",
"print(\"\\nllm ans\")\n",
"print(response.choices[0].message.content)"
]
}
],
"metadata": {
"kernelspec": {
"display_name": "base",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.12.7"
}
},
"nbformat": 4,
"nbformat_minor": 5
}

4
kg_ocr/__init__.py Normal file
View File

@@ -0,0 +1,4 @@
from .ocr import get_screenshots, extract_text
from .embeddings import create_and_index, query_embedding
__all__ = ["get_screenshots", "extract_text", "create_and_index", "query_embedding"]

0
kg_ocr/cli/__init__.py Normal file
View File

3
kg_ocr/cli/build_graph.py Normal file
View File

@@ -0,0 +1,3 @@
# to do
# add entrypoint to setuppy

1
kg_ocr/cli/export_graph.py Normal file
View File

@@ -0,0 +1 @@
# to do

1
kg_ocr/cli/process_screenshots.py Normal file
View File

@@ -0,0 +1 @@
# to do

3
kg_ocr/embeddings/__init__.py Normal file
View File

@@ -0,0 +1,3 @@
from .indexer import create_and_index, query_embedding
__all__ = ["create_and_index", "query_embedding"]

1
kg_ocr/embeddings/config_loader.py Normal file
View File

@@ -0,0 +1 @@
# in future prefer a config, especially for graph traversal

23
kg_ocr/embeddings/indexer.py Normal file
View File

@@ -0,0 +1,23 @@
from txtai.embeddings import Embeddings
def create_and_index(
data: list[str], model: str = "sentence-transformers/all-MiniLM-L6-v2"
) -> Embeddings:
"""Create and index embeddings from text."""
embeddings = Embeddings({
"path": model,
"content": True,
"hybrid": True,
"scoring": "bm25",
})
embeddings.index(data)
return embeddings
def query_embedding(
embeddings: Embeddings, query: str, limit: int = 100
) -> list[str]:
"""Search embeddings and return matching texts."""
results = embeddings.search(query, limit)
return [r["text"] for r in results]

0
kg_ocr/export/__init__.py Normal file
View File

0
kg_ocr/export/neo4j_exporter.py Normal file
View File

0
kg_ocr/graph/__init__.py Normal file
View File

1
kg_ocr/graph/analyzer.py Normal file
View File

@@ -0,0 +1 @@
# also anomaly detection here

0
kg_ocr/graph/builder.py Normal file
View File

4
kg_ocr/ocr/__init__.py Normal file
View File

@@ -0,0 +1,4 @@
from .extractor import get_screenshots
from .batch_processor import extract_text
__all__ = ["get_screenshots", "extract_text"]

7
kg_ocr/ocr/batch_processor.py Normal file
View File

@@ -0,0 +1,7 @@
from PIL import Image
import pytesseract
def extract_text(images: list[str]) -> list[str]:
"""OCR a list of image paths into text."""
return [pytesseract.image_to_string(Image.open(img)) for img in images]

15
kg_ocr/ocr/constants.py Normal file
View File

@@ -0,0 +1,15 @@
from pathlib import Path
import platform
def_paths = {
"Darwin": Path.home() / "Desktop",
"Windows": Path.home() / "Pictures" / "Screenshots",
"Linux": Path.home() / "Pictures",
}
sc_pathpatterns = {
"Darwin": ["SCR*.png", "Screenshot*.png"],
"Windows": ["Screenshot*.png"],
"Linux": ["Screenshot*.png", "scrot*.png", "screenshot*.png"],
}

17
kg_ocr/ocr/extractor.py Normal file
View File

@@ -0,0 +1,17 @@
import platform
from pathlib import Path
from typing import Optional
from .constants import def_paths, sc_pathpatterns
def get_screenshots(path: Optional[str | Path] = None) -> list[str]:
"""Find screenshot files for the current OS."""
if path is None:
path = def_paths.get(platform.system(), Path.home())
path = Path(path)
patterns = sc_pathpatterns.get(platform.system(), ["SCR*.png"])
results = []
for pattern in patterns:
results.extend(str(f.absolute()) for f in path.glob(pattern))
return sorted(set(results))

0
ocr_sc.ipynb → notebooks/01_ocr_sc.ipynb
View File

3654
notebooks/02_functions_legacy.ipynb Normal file
View File

File diff suppressed because it is too large Load Diff

2827
notebooks/03_flow.ipynb Normal file
View File

File diff suppressed because it is too large Load Diff

13
setup.py Normal file
View File

@@ -0,0 +1,13 @@
from setuptools import setup, find_packages
setup(
name="kg_ocr",
packages=find_packages(),
python_requires=">=3.10",
install_requires=[
"pytesseract",
"Pillow",
"txtai",
"sentence-transformers",
],
)

0
tests/test_graph.py Normal file
View File

0
tests/test_indexer.py Normal file
View File

0
tests/test_ocr.py Normal file
View File