ai-parrot 0.3.0__cp311-cp311-manylinux_2_28_x86_64.whl → 0.3.3__cp311-cp311-manylinux_2_28_x86_64.whl

ai_parrot-0.3.3.dist-info/METADATA

@@ -0,0 +1,318 @@
+Metadata-Version: 2.1
+Name: ai-parrot
+Version: 0.3.3
+Summary: Live Chatbots based on Langchain chatbots and Agents     Integrated into Navigator Framework or used into aiohttp applications.
+Home-page: https://github.com/phenobarbital/ai-parrot
+Author: Jesus Lara
+Author-email: jesuslara@phenobarbital.info
+License: MIT
+Project-URL: Source, https://github.com/phenobarbital/ai-parrot
+Project-URL: Tracker, https://github.com/phenobarbital/ai-parrot/issues
+Project-URL: Documentation, https://github.com/phenobarbital/ai-parrot/
+Project-URL: Funding, https://paypal.me/phenobarbital
+Project-URL: Say Thanks!, https://saythanks.io/to/phenobarbital
+Keywords: asyncio,asyncpg,aioredis,aiomcache,langchain,chatbot,agents
+Platform: POSIX
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Environment :: Web Environment
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Topic :: Software Development :: Build Tools
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+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 :: Only
+Classifier: Framework :: AsyncIO
+Requires-Python: >=3.10.12
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: Cython==3.0.11
+Requires-Dist: accelerate==0.34.2
+Requires-Dist: langchain>=0.2.6
+Requires-Dist: langchain-community>=0.2.6
+Requires-Dist: langchain-core>=0.2.32
+Requires-Dist: langchain-experimental==0.0.62
+Requires-Dist: langchainhub==0.1.15
+Requires-Dist: langchain-text-splitters==0.2.2
+Requires-Dist: langchain-huggingface==0.0.3
+Requires-Dist: huggingface-hub==0.23.5
+Requires-Dist: llama-index==0.10.20
+Requires-Dist: llama-cpp-python==0.2.56
+Requires-Dist: bitsandbytes==0.43.3
+Requires-Dist: Cartopy==0.22.0
+Requires-Dist: chromadb==0.4.24
+Requires-Dist: datasets==2.18.0
+Requires-Dist: faiss-cpu==1.8.0
+Requires-Dist: fastavro==1.9.4
+Requires-Dist: gunicorn==21.2.0
+Requires-Dist: jq==1.7.0
+Requires-Dist: rank-bm25==0.2.2
+Requires-Dist: matplotlib==3.8.3
+Requires-Dist: numba==0.59.0
+Requires-Dist: querysource>=3.12.10
+Requires-Dist: safetensors>=0.4.3
+Requires-Dist: sentence-transformers==3.0.1
+Requires-Dist: tabulate==0.9.0
+Requires-Dist: tiktoken==0.7.0
+Requires-Dist: tokenizers==0.19.1
+Requires-Dist: unstructured==0.14.3
+Requires-Dist: unstructured-client==0.18.0
+Requires-Dist: youtube-transcript-api==0.6.2
+Requires-Dist: selenium==4.18.1
+Requires-Dist: webdriver-manager==4.0.1
+Requires-Dist: transitions==0.9.0
+Requires-Dist: sentencepiece==0.2.0
+Requires-Dist: duckduckgo-search==5.3.0
+Requires-Dist: google-search-results==2.4.2
+Requires-Dist: google-api-python-client>=2.86.0
+Requires-Dist: gdown==5.1.0
+Requires-Dist: weasyprint==61.2
+Requires-Dist: markdown2==2.4.13
+Requires-Dist: fastembed==0.3.4
+Requires-Dist: yfinance==0.2.40
+Requires-Dist: youtube-search==2.1.2
+Requires-Dist: wikipedia==1.4.0
+Requires-Dist: mediawikiapi==1.2
+Requires-Dist: pyowm==3.3.0
+Requires-Dist: O365==2.0.35
+Requires-Dist: stackapi==0.3.1
+Requires-Dist: timm==1.0.9
+Requires-Dist: torchvision==0.19.1
+Provides-Extra: all
+Requires-Dist: langchain-milvus==0.1.1; extra == "all"
+Requires-Dist: milvus==2.3.5; extra == "all"
+Requires-Dist: pymilvus==2.4.4; extra == "all"
+Requires-Dist: groq==0.11.0; extra == "all"
+Requires-Dist: langchain-groq==0.1.4; extra == "all"
+Requires-Dist: llama-index-llms-huggingface==0.2.7; extra == "all"
+Requires-Dist: langchain-google-vertexai==1.0.8; extra == "all"
+Requires-Dist: langchain-google-genai==1.0.8; extra == "all"
+Requires-Dist: google-generativeai==0.7.2; extra == "all"
+Requires-Dist: vertexai==1.60.0; extra == "all"
+Requires-Dist: google-cloud-aiplatform>=1.60.0; extra == "all"
+Requires-Dist: grpc-google-iam-v1==0.13.0; extra == "all"
+Requires-Dist: langchain-openai==0.1.21; extra == "all"
+Requires-Dist: openai==1.40.8; extra == "all"
+Requires-Dist: llama-index-llms-openai==0.1.11; extra == "all"
+Requires-Dist: langchain-anthropic==0.1.23; extra == "all"
+Requires-Dist: anthropic==0.34.0; extra == "all"
+Provides-Extra: analytics
+Requires-Dist: annoy==1.17.3; extra == "analytics"
+Requires-Dist: gradio-tools==0.0.9; extra == "analytics"
+Requires-Dist: gradio-client==0.2.9; extra == "analytics"
+Requires-Dist: streamlit==1.37.1; extra == "analytics"
+Requires-Dist: simsimd==4.3.1; extra == "analytics"
+Requires-Dist: opencv-python==4.10.0.84; extra == "analytics"
+Provides-Extra: anthropic
+Requires-Dist: langchain-anthropic==0.1.11; extra == "anthropic"
+Requires-Dist: anthropic==0.25.2; extra == "anthropic"
+Provides-Extra: crew
+Requires-Dist: colbert-ai==0.2.19; extra == "crew"
+Requires-Dist: vanna==0.3.4; extra == "crew"
+Requires-Dist: crewai[tools]==0.28.8; extra == "crew"
+Provides-Extra: google
+Requires-Dist: langchain-google-vertexai==1.0.10; extra == "google"
+Requires-Dist: langchain-google-genai==1.0.10; extra == "google"
+Requires-Dist: vertexai==1.65.0; extra == "google"
+Provides-Extra: groq
+Requires-Dist: groq==0.11.0; extra == "groq"
+Requires-Dist: langchain-groq==0.1.9; extra == "groq"
+Provides-Extra: hunggingfaces
+Requires-Dist: llama-index-llms-huggingface==0.2.7; extra == "hunggingfaces"
+Provides-Extra: loaders
+Requires-Dist: pymupdf==1.24.4; extra == "loaders"
+Requires-Dist: pymupdf4llm==0.0.1; extra == "loaders"
+Requires-Dist: pdf4llm==0.0.6; extra == "loaders"
+Requires-Dist: PyPDF2==3.0.1; extra == "loaders"
+Requires-Dist: pdfminer.six==20231228; extra == "loaders"
+Requires-Dist: pdfplumber==0.11.0; extra == "loaders"
+Requires-Dist: GitPython==3.1.42; extra == "loaders"
+Requires-Dist: opentelemetry-sdk==1.24.0; extra == "loaders"
+Requires-Dist: rapidocr-onnxruntime==1.3.15; extra == "loaders"
+Requires-Dist: pytesseract==0.3.10; extra == "loaders"
+Requires-Dist: python-docx==1.1.0; extra == "loaders"
+Requires-Dist: python-pptx==0.6.23; extra == "loaders"
+Requires-Dist: docx2txt==0.8; extra == "loaders"
+Requires-Dist: pytube==15.0.0; extra == "loaders"
+Requires-Dist: pydub==0.25.1; extra == "loaders"
+Requires-Dist: markdownify==0.12.1; extra == "loaders"
+Requires-Dist: yt-dlp==2024.4.9; extra == "loaders"
+Requires-Dist: moviepy==1.0.3; extra == "loaders"
+Requires-Dist: mammoth==1.7.1; extra == "loaders"
+Requires-Dist: paddlepaddle==2.6.1; extra == "loaders"
+Requires-Dist: paddlepaddle-gpu==2.6.1; extra == "loaders"
+Requires-Dist: paddleocr==2.8.1; extra == "loaders"
+Requires-Dist: ftfy==6.2.3; extra == "loaders"
+Requires-Dist: librosa==0.10.1; extra == "loaders"
+Requires-Dist: XlsxWriter==3.2.0; extra == "loaders"
+Requires-Dist: xformers==0.0.27.post2; extra == "loaders"
+Provides-Extra: milvus
+Requires-Dist: langchain-milvus>=0.1.4; extra == "milvus"
+Requires-Dist: milvus==2.3.5; extra == "milvus"
+Requires-Dist: pymilvus==2.4.6; extra == "milvus"
+Provides-Extra: openai
+Requires-Dist: langchain-openai==0.1.21; extra == "openai"
+Requires-Dist: openai==1.40.3; extra == "openai"
+Requires-Dist: llama-index-llms-openai==0.1.11; extra == "openai"
+Requires-Dist: tiktoken==0.7.0; extra == "openai"
+Provides-Extra: qdrant
+Requires-Dist: qdrant-client==1.8.0; extra == "qdrant"
+
+# AI Parrot: Python package for creating Chatbots
+This is an open-source Python package for creating Chatbots based on Langchain and Navigator.
+This README provides instructions for installation, development, testing, and releasing Parrot.
+
+## Installation
+
+**Creating a virtual environment:**
+
+This is recommended for development and isolation from system-wide libraries.
+Run the following command in your terminal:
+
+Debian-based systems installation:
+   ```
+   sudo apt install gcc python3.11-venv python3.11-full python3.11-dev libmemcached-dev zlib1g-dev build-essential libffi-dev unixodbc unixodbc-dev libsqliteodbc libev4 libev-dev
+   ```
+
+   For Qdrant installation:
+   ```
+   docker pull qdrant/qdrant
+   docker run -d -p 6333:6333 -p 6334:6334 --name qdrant -v $(pwd)/qdrant_storage:/qdrant/storage:z qdrant/qdrant
+   ```
+
+For VertexAI, creates a folder on "env" called "google" and copy the JSON credentials file into it.
+
+   ```bash
+   make venv
+   ```
+
+   This will create a virtual environment named `.venv`. To activate it, run:
+
+   ```bash
+   source .venv/bin/activate  # Linux/macOS
+   ```
+
+   Once activated, install Parrot within the virtual environment:
+
+   ```bash
+   make install
+   ```
+   The output will remind you to activate the virtual environment before development.
+
+   **Optional** (for developers):
+   ```bash
+   pip install -e .
+   ```
+
+## Start http server
+```bash
+python run.py
+```
+
+## Development Setup
+
+This section explains how to set up your development environment:
+
+1. **Install development requirements:**
+
+   ```bash
+   make setup
+   ```
+
+   This installs development dependencies like linters and test runners mentioned in the `docs/requirements-dev.txt` file.
+
+2. **Install Parrot in editable mode:**
+
+   This allows you to make changes to the code and test them without reinstalling:
+
+   ```bash
+   make dev
+   ```
+
+   This uses `flit` to install Parrot in editable mode.
+
+
+### Usage (Replace with actual usage instructions)
+
+*Once you have set up your development environment, you can start using Parrot.*
+
+#### Test with Code ChatBOT
+* Set environment variables for:
+ [google]
+   GOOGLE_API_KEY=apikey
+   GOOGLE_CREDENTIALS_FILE=.../credentials.json
+   VERTEX_PROJECT_ID=vertex-project
+   VERTEX_REGION=region
+
+* Run the chatbot:
+    ```bash
+    python examples/test_agent.py
+    ```
+
+### Testing
+
+To run the test suite:
+
+```bash
+make test
+```
+
+This will run tests using `coverage` to report on code coverage.
+
+
+### Code Formatting
+
+To format the code with black:
+
+```bash
+make format
+```
+
+
+### Linting
+
+To lint the code for style and potential errors:
+
+```bash
+make lint
+```
+
+This uses `pylint` and `black` to check for issues.
+
+
+### Releasing a New Version
+
+This section outlines the steps for releasing a new version of Parrot:
+
+1. **Ensure everything is clean and tested:**
+
+   ```bash
+   make release
+   ```
+
+   This runs `lint`, `test`, and `clean` tasks before proceeding.
+
+2. **Publish the package:**
+
+   ```bash
+   make release
+   ```
+
+   This uses `flit` to publish the package to a repository like PyPI. You'll need to have publishing credentials configured for `flit`.
+
+
+### Cleaning Up
+
+To remove the virtual environment:
+
+```bash
+make distclean
+```
+
+
+### Contributing
+
+We welcome contributions to Parrot! Please refer to the CONTRIBUTING.md file for guidelines on how to contribute.

{ai_parrot-0.3.0.dist-info → ai_parrot-0.3.3.dist-info}/RECORD

@@ -1,10 +1,10 @@
 parrot/__init__.py,sha256=eTkAkHeJ5BBDG2fxrXA4M37ODBJoS1DQYpeBAWL2xeI,387
 parrot/conf.py,sha256=-9bVGC7Rf-6wpIg6-ojvU4S_G1wBLUCVDt46KEGHEhM,4257
-parrot/exceptions.cpython-311-x86_64-linux-gnu.so,sha256=gDwsnUlOlwphVU97XaqG5e7BJs_PWPKdwgwDsjyVZIg,361200
+parrot/exceptions.cpython-311-x86_64-linux-gnu.so,sha256=VNyBh3uLxGQgB0l1bkWjQDqYUN2ZAvRmV12AqQijV9Q,361184
 parrot/manager.py,sha256=NhzXoWxSgtoWHpmYP8cV2Ujq_SlvCbQYQBaohAeL2TM,5935
 parrot/models.py,sha256=RsVQCqhSXBKRPcu-BCga9Y1wyvENFXDCuq3_ObIKvAo,13452
 parrot/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-parrot/version.py,sha256=FGfo76wlKkd0zQ9C4qQseeRf_HKXNQSmd2Lfu8iOctk,373
+parrot/version.py,sha256=pbGrvnHWVk2vkgFh0ab5xc4-svi5oC7IvapZz06YLpM,373
 parrot/chatbots/__init__.py,sha256=ypskCnME0xUv6psBEGCEyXCrD0J0ULHSllpVmSxqb4A,200
 parrot/chatbots/abstract.py,sha256=CmDn3k4r9uKImOZRN4L9zxLbCdC-1MPUAorDlfZT-kA,26421
 parrot/chatbots/asktroc.py,sha256=gyWzyvpAnmXwXd-3DEKoIJtAxt6NnP5mUZdZbkFky8s,604
@@ -56,7 +56,7 @@ parrot/loaders/excel.py,sha256=Y1agxm-jG4AgsA2wlPP3p8uBH40wYW1KM2ycTTLKUm4,12441
 parrot/loaders/github.py,sha256=CscyUIqoHTytqCbRUUTcV3QSxI8XoDntq5aTU0vdhzQ,2593
 parrot/loaders/image.py,sha256=A9KCXXoGuhDoyeJaascY7Q1ZK12Kf1ggE1drzJjS3AU,3946
 parrot/loaders/json.py,sha256=6B43k591OpvoJLbsJa8CxJue_lAt713SCdldn8bFW3c,1481
-parrot/loaders/pdf.py,sha256=flGlUf9dLAD2Uh8MkvLP27OU1nvroeHU2HM5a3rBH3M,7996
+parrot/loaders/pdf.py,sha256=wGwFnsUmeQqtqk3L2vYt2DkW09LUODUJN-xLjuAa-do,17826
 parrot/loaders/pdfchapters.py,sha256=YhA8Cdx3qXBR0vuTVnQ12XgH1DXT_rp1Tawzh4V2U3o,5637
 parrot/loaders/pdffn.py,sha256=gA-vJEWUiIUwbMxP8Nmvlzlcb39DVV69vGKtSzavUoI,4004
 parrot/loaders/pdfimages.py,sha256=4Q_HKiAee_hALBsG2qF7PpMgKP1AivHXhmcsCkUa9eE,7899
@@ -68,7 +68,7 @@ parrot/loaders/repo.py,sha256=vBqBAnwU6p3_DCvI9DVhi1Bs8iCDYHwFGp0P9zvGRyw,3737
 parrot/loaders/rtd.py,sha256=oKOC9Qn3iwulYx5BEvXy4_kusKRsy5RLYNHS-e5p-1k,1988
 parrot/loaders/txt.py,sha256=AeGroWffFT--7TYlTSTr5Av5zAr8YIp1fxt8r5qdi-A,2802
 parrot/loaders/video.py,sha256=pl5Ho69bp5vrWMqg5tLbsnHUus1LByTDoL6NPk57Ays,2929
-parrot/loaders/videolocal.py,sha256=QjCoiDREkpSyyVQ8yDZefzV2g24Gz4VUF4Eiei6v-dY,3791
+parrot/loaders/videolocal.py,sha256=3EASzbettSO2tboTe3GndR4p6Nihwj6HGZoiPXekYo0,4302
 parrot/loaders/vimeo.py,sha256=zOvOOIiaZr_bRswJFI7uIMKISgALOxcSim9ZRUFY1Fc,4114
 parrot/loaders/web.py,sha256=3x06JNpfTGFtvSBPAEBVoVdZkpVXePcJeMtj61B2xJk,8867
 parrot/loaders/web_base.py,sha256=5SjQddT0Vhne8C9s30iU3Ex_9O1PJ8kyDmy8EdhGBo0,4380
@@ -94,17 +94,17 @@ parrot/tools/wikipedia.py,sha256=oadBTRAupu2dKThEORSHqqVs4u0G9lWOltbP6vSZgPE,199
 parrot/tools/zipcode.py,sha256=knScSvKgK7bHxyLcBkZFiMs65e-PlYU2_YhG6mohcjU,6440
 parrot/utils/__init__.py,sha256=vkBIvfl9-0NRLd76MIZk4s49PjtF_dW5imLTv_UOKxM,101
 parrot/utils/toml.py,sha256=CVyqDdAEyOj6AHfNpyQe4IUvLP_SSXlbHROYPeadLvU,302
-parrot/utils/types.cpython-311-x86_64-linux-gnu.so,sha256=kdox48-JUzj92QP6amGOCTIEQhrBUMn6qzrhX1u17CY,791912
+parrot/utils/types.cpython-311-x86_64-linux-gnu.so,sha256=jghuq8bBlgGDjkb88Efi5l9cgR5KZL_qO7yxglGNsTA,791256
 parrot/utils/uv.py,sha256=Mb09bsi13hhi3xQDBjEhCf-U1wherXl-K4-BLcSvqtc,308
 parrot/utils/parsers/__init__.py,sha256=l82uIu07QvSJ8Xt0d_seei9n7UUL8PE-YFGBTyNbxSI,62
-parrot/utils/parsers/toml.cpython-311-x86_64-linux-gnu.so,sha256=vdQTxL4AyxinDpoDVk0Syx-ycDL02OmXESJOtiVFl0A,451056
+parrot/utils/parsers/toml.cpython-311-x86_64-linux-gnu.so,sha256=gEnv6QGF6DtxExEdVTdNx48j90wPYKBLyCH1UCRj4MQ,451088
 resources/users/__init__.py,sha256=sdXUV7h0Oogcdru1RrQxbm9_RcMjftf0zTWqvxBVpO8,151
 resources/users/handlers.py,sha256=BGzqBvPY_OaIF_nONWX4b_B5OyyBrdGuSihIsdlFwjk,291
 resources/users/models.py,sha256=glk7Emv7QCi6i32xRFDrGc8UwK23_LPg0XUOJoHnwRU,6799
 settings/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 settings/settings.py,sha256=9ueEvyLNurUX-AaIeRPV8GKX1c4YjDLbksUAeqEq6Ck,1854
-ai_parrot-0.3.0.dist-info/LICENSE,sha256=vRKOoa7onTsLNvSzJtGtMaNhWWh8B3YAT733Tlu6M4o,1070
-ai_parrot-0.3.0.dist-info/METADATA,sha256=84sQiAphAjSB9xEqCw1DWLUfwbDciz-1hp8cekb_-MA,10410
-ai_parrot-0.3.0.dist-info/WHEEL,sha256=tFO7F0mawMNWa_NzTDA1ygqZBeMykVNIr04O5Zxk1TE,113
-ai_parrot-0.3.0.dist-info/top_level.txt,sha256=qHoO4BhYDfeTkyKnciZSQtn5FSLN3Q-P5xCTkyvbuxg,26
-ai_parrot-0.3.0.dist-info/RECORD,,
+ai_parrot-0.3.3.dist-info/LICENSE,sha256=vRKOoa7onTsLNvSzJtGtMaNhWWh8B3YAT733Tlu6M4o,1070
+ai_parrot-0.3.3.dist-info/METADATA,sha256=LHLvoMsy1VvMlC33Kl2RKhwnFgj40lMpPDVmVWYj1m8,10624
+ai_parrot-0.3.3.dist-info/WHEEL,sha256=UQ-0qXN3LQUffjrV43_e_ZXj2pgORBqTmXipnkj0E8I,113
+ai_parrot-0.3.3.dist-info/top_level.txt,sha256=qHoO4BhYDfeTkyKnciZSQtn5FSLN3Q-P5xCTkyvbuxg,26
+ai_parrot-0.3.3.dist-info/RECORD,,

{ai_parrot-0.3.0.dist-info → ai_parrot-0.3.3.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (72.2.0)
+Generator: setuptools (74.1.2)
 Root-Is-Purelib: false
 Tag: cp311-cp311-manylinux_2_28_x86_64
 

parrot/loaders/pdf.py

@@ -2,10 +2,26 @@ from collections.abc import Callable
 from pathlib import Path, PurePath
 from typing import Any
 from io import BytesIO
+import re
+import ftfy
 import fitz
 import pytesseract
+from paddleocr import PaddleOCR
+import torch
+import cv2
+from transformers import (
+    # DonutProcessor,
+    # VisionEncoderDecoderModel,
+    # VisionEncoderDecoderConfig,
+    # ViTImageProcessor,
+    # AutoTokenizer,
+    LayoutLMv3ForTokenClassification,
+    LayoutLMv3Processor
+)
+from pdf4llm import to_markdown
 from PIL import Image
 from langchain.docstore.document import Document
+from navconfig import config
 from .basepdf import BasePDF
 
 
@@ -31,6 +47,29 @@ class PDFLoader(BasePDF):
             **kwargs
         )
         self.parse_images = kwargs.get('parse_images', False)
+        self.page_as_images = kwargs.get('page_as_images', False)
+        if self.page_as_images is True:
+            # # Load the processor and model from Hugging Face
+            # self.image_processor = DonutProcessor.from_pretrained(
+            #     "naver-clova-ix/donut-base-finetuned-docvqa"
+            # )
+            # self.image_model = VisionEncoderDecoderModel.from_pretrained(
+            #     "naver-clova-ix/donut-base-finetuned-docvqa",
+
+            # )
+            # Load the processor and model from Hugging Face
+            self.image_processor = LayoutLMv3Processor.from_pretrained(
+                "microsoft/layoutlmv3-base",
+                apply_ocr=True
+            )
+            self.image_model = LayoutLMv3ForTokenClassification.from_pretrained(
+                # "microsoft/layoutlmv3-base-finetuned-funsd"
+                "HYPJUDY/layoutlmv3-base-finetuned-funsd"
+            )
+            # Set device to GPU if available
+            self.image_device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+            self.image_model.to(self.image_device)
+
         # Table Settings:
         self.table_settings = {
             #"vertical_strategy": "text",
@@ -42,6 +81,134 @@ class PDFLoader(BasePDF):
         if table_settings:
             self.table_settings.update(table_settings)
 
+    def explain_image(self, image_path):
+        """Function to explain the image."""
+        # with open(image_path, "rb") as image_file:
+        #     image_content = image_file.read()
+
+        # Open the image
+        image = cv2.imread(image_path)
+        task_prompt = "<s_docvqa><s_question>{user_input}</s_question><s_answer>"
+        question = "Extract Questions about Happily Greet"
+        prompt = task_prompt.replace("{user_input}", question)
+
+        decoder_input_ids = self.image_processor.tokenizer(
+            prompt,
+            add_special_tokens=False,
+            return_tensors="pt",
+        ).input_ids
+
+        pixel_values = self.image_processor(
+            image,
+            return_tensors="pt"
+        ).pixel_values
+
+        # Send inputs to the appropriate device
+        pixel_values = pixel_values.to(self.image_device)
+        decoder_input_ids = decoder_input_ids.to(self.image_device)
+
+        outputs = self.image_model.generate(
+            pixel_values,
+            decoder_input_ids=decoder_input_ids,
+            max_length=self.image_model.decoder.config.max_position_embeddings,
+            pad_token_id=self.image_processor.tokenizer.pad_token_id,
+            eos_token_id=self.image_processor.tokenizer.eos_token_id,
+            bad_words_ids=[[self.image_processor.tokenizer.unk_token_id]],
+            # use_cache=True
+            return_dict_in_generate=True,
+        )
+
+        sequence = self.image_processor.batch_decode(outputs.sequences)[0]
+
+
+        sequence = sequence.replace(
+            self.image_processor.tokenizer.eos_token, ""
+        ).replace(
+            self.image_processor.tokenizer.pad_token, ""
+        )
+        # remove first task start token
+        sequence = re.sub(r"<.*?>", "", sequence, count=1).strip()
+        # Print the extracted sequence
+        print("Extracted Text:", sequence)
+
+        print(self.image_processor.token2json(sequence))
+
+        # Format the output as Markdown (optional step)
+        markdown_text = self.format_as_markdown(sequence)
+        print("Markdown Format:\n", markdown_text)
+
+        return None
+
+    def convert_to_markdown(self, text):
+        """
+        Convert the cleaned text into a markdown format.
+        You can enhance this function to detect tables, headings, etc.
+        """
+        # For example, we can identify sections or headers and format them in Markdown
+        markdown_text = text
+        # Detect headings and bold them
+        markdown_text = re.sub(r"(^.*Scorecard.*$)", r"## \1", markdown_text)
+        # Convert lines with ":" to a list item (rough approach)
+        markdown_text = re.sub(r"(\w+):", r"- **\1**:", markdown_text)
+        # Return the markdown formatted text
+        return markdown_text
+
+    def clean_tokenized_text(self, tokenized_text):
+        """
+        Clean the tokenized text by fixing encoding issues and formatting, preserving line breaks.
+        """
+        # Fix encoding issues using ftfy
+        cleaned_text = ftfy.fix_text(tokenized_text)
+
+        # Remove <s> and </s> tags (special tokens)
+        cleaned_text = cleaned_text.replace("<s>", "").replace("</s>", "")
+
+        # Replace special characters like 'Ġ' and fix multiple spaces, preserving new lines
+        cleaned_text = cleaned_text.replace("Ġ", " ")
+
+        # Avoid collapsing line breaks, but still normalize multiple spaces
+        # Replace multiple spaces with a single space, but preserve line breaks
+        cleaned_text = re.sub(r" +", " ", cleaned_text)
+
+        return cleaned_text.strip()
+
+    def extract_page_text(self, image_path) -> str:
+        # Open the image
+        image = Image.open(image_path).convert("RGB")
+
+        # Processor handles the OCR internally, no need for words or boxes
+        encoding = self.image_processor(image, return_tensors="pt", truncation=True)
+        encoding = {k: v.to(self.image_device) for k, v in encoding.items()}
+
+        # Forward pass
+        outputs = self.image_model(**encoding)
+        logits = outputs.logits
+
+        # Get predictions
+        predictions = logits.argmax(-1).squeeze().tolist()
+        labels = [self.image_model.config.id2label[pred] for pred in predictions]
+
+        # Get the words and boxes from the processor's OCR step
+        words = self.image_processor.tokenizer.convert_ids_to_tokens(
+            encoding['input_ids'].squeeze().tolist()
+        )
+        boxes = encoding['bbox'].squeeze().tolist()
+
+        # Combine words and labels, preserving line breaks based on vertical box position
+        extracted_text = ""
+        last_box = None
+        for word, label, box in zip(words, labels, boxes):
+            if label != 'O':
+                # Check if the current word is on a new line based on the vertical position of the box
+                if last_box and abs(box[1] - last_box[1]) > 10:  # A threshold for line breaks
+                    extracted_text += "\n"  # Add a line break
+
+                extracted_text += f"{word} "
+                last_box = box
+        cleaned_text = self.clean_tokenized_text(extracted_text)
+        markdown_text = self.convert_to_markdown(cleaned_text)
+        return markdown_text
+
     def _load_pdf(self, path: Path) -> list:
         """
         Load a PDF file using the Fitz library.
@@ -56,6 +223,32 @@ class PDFLoader(BasePDF):
             self.logger.info(f"Loading PDF file: {path}")
             pdf = fitz.open(str(path))  # Open the PDF file
             docs = []
+            try:
+                md_text = to_markdown(pdf) # get markdown for all pages
+                _meta = {
+                    "url": f'{path}',
+                    "source": f"{path.name}",
+                    "filename": path.name,
+                    "type": 'pdf',
+                    "question": '',
+                    "answer": '',
+                    "source_type": self._source_type,
+                    "data": {},
+                    "summary": '',
+                    "document_meta": {
+                        "title": pdf.metadata.get("title", ""),
+                        "creationDate": pdf.metadata.get("creationDate", ""),
+                        "author": pdf.metadata.get("author", ""),
+                    }
+                }
+                docs.append(
+                    Document(
+                        page_content=md_text,
+                        metadata=_meta
+                    )
+                )
+            except Exception:
+                pass
             for page_number in range(pdf.page_count):
                 page = pdf[page_number]
                 text = page.get_text()
@@ -79,12 +272,7 @@ class PDFLoader(BasePDF):
                         "summary": summary,
                         "document_meta": {
                             "title": pdf.metadata.get("title", ""),
-                            # "subject": pdf.metadata.get("subject", ""),
-                            # "keywords": pdf.metadata.get("keywords", ""),
                             "creationDate": pdf.metadata.get("creationDate", ""),
-                            # "modDate": pdf.metadata.get("modDate", ""),
-                            # "producer": pdf.metadata.get("producer", ""),
-                            # "creator": pdf.metadata.get("creator", ""),
                             "author": pdf.metadata.get("author", ""),
                         }
                     }
@@ -96,9 +284,10 @@ class PDFLoader(BasePDF):
                     )
                 # Extract images and use OCR to get text from each image
                 # second: images
+                file_name = path.stem.replace(' ', '_').replace('.', '').lower()
                 if self.parse_images is True:
+                    # extract any images in page:
                     image_list = page.get_images(full=True)
-                    file_name = path.stem.replace(' ', '_').replace('.', '').lower()
                     for img_index, img in enumerate(image_list):
                         xref = img[0]
                         base_image = pdf.extract_image(xref)
@@ -181,7 +370,68 @@ class PDFLoader(BasePDF):
                             )
                 except Exception as exc:
                     print(exc)
+                # fourth: page as image
+                if self.page_as_images is True:
+                    # Convert the page to a Pixmap (which is an image)
+                    mat = fitz.Matrix(2, 2)
+                    pix = page.get_pixmap(dpi=300, matrix=mat) # Increase DPI for better resolution
+                    img_name = f'{file_name}_page_{page_num}.png'
+                    img_path = self._imgdir.joinpath(img_name)
+                    if img_path.exists():
+                        img_path.unlink(missing_ok=True)
+                    self.logger.notice(
+                        f"Saving Page {page_number} as Image on {img_path}"
+                    )
+                    pix.save(
+                        img_path
+                    )
+                    # TODO passing the image to a AI visual to get explanation
+                    # Get the extracted text from the image
+                    text = self.extract_page_text(img_path)
+                    url = f'/static/images/{img_name}'
+                    image_meta = {
+                        "url": url,
+                        "source": f"{path.name} Page.#{page_num}",
+                        "filename": path.name,
+                        "index": f"{path.name}:{page_num}",
+                        "question": '',
+                        "answer": '',
+                        "type": 'page',
+                        "data": {},
+                        "summary": '',
+                        "document_meta": {
+                            "image_name": img_name,
+                            "page_number": f"{page_number}"
+                        },
+                        "source_type": self._source_type
+                    }
+                    docs.append(
+                        Document(page_content=text, metadata=image_meta)
+                    )
             pdf.close()
             return docs
         else:
             return []
+
+    def get_ocr(self, img_path) -> list:
+        # Initialize PaddleOCR with table recognition
+        self.ocr_model = PaddleOCR(
+            lang='en',
+            det_model_dir=None,
+            rec_model_dir=None,
+            rec_char_dict_path=None,
+            table=True,
+            # use_angle_cls=True,
+            # use_gpu=True
+        )
+        result = self.ocr_model.ocr(img_path, cls=True)
+
+        # extract tables:
+        # The result contains the table structure and content
+        tables = []
+        for line in result:
+            if 'html' in line[1]:
+                html_table = line[1]['html']
+                tables.append(html_table)
+
+        print('TABLES > ', tables)

parrot/loaders/videolocal.py

@@ -105,3 +105,16 @@ class VideoLocalLoader(BaseVideoLoader):
                     if set(item.parts).isdisjoint(self.skip_directories):
                         documents.extend(self.load_video(item))
         return self.split_documents(documents)
+
+    def extract(self) -> list:
+        documents = []
+        if self.path.is_file():
+            docs = self.load_video(self.path)
+            documents.extend(docs)
+        if self.path.is_dir():
+            # iterate over the files in the directory
+            for ext in self._extension:
+                for item in self.path.glob(f'*{ext}'):
+                    if set(item.parts).isdisjoint(self.skip_directories):
+                        documents.extend(self.load_video(item))
+        return documents

parrot/version.py

@@ -3,7 +3,7 @@
 __title__ = "ai-parrot"
 __description__ = "Live Chatbots based on Langchain chatbots and Agents \
     Integrated into Navigator Framework or used into aiohttp applications."
-__version__ = "0.3.0"
+__version__ = "0.3.3"
 __author__ = "Jesus Lara"
 __author_email__ = "jesuslarag@gmail.com"
 __license__ = "MIT"

ai_parrot-0.3.0.dist-info/METADATA

@@ -1,319 +0,0 @@
-Metadata-Version: 2.1
-Name: ai-parrot
-Version: 0.3.0
-Summary: Live Chatbots based on Langchain chatbots and Agents     Integrated into Navigator Framework or used into aiohttp applications.
-Home-page: https://github.com/phenobarbital/ai-parrot
-Author: Jesus Lara
-Author-email: jesuslara@phenobarbital.info
-License: MIT
-Project-URL: Source, https://github.com/phenobarbital/ai-parrot
-Project-URL: Tracker, https://github.com/phenobarbital/ai-parrot/issues
-Project-URL: Documentation, https://github.com/phenobarbital/ai-parrot/
-Project-URL: Funding, https://paypal.me/phenobarbital
-Project-URL: Say Thanks!, https://saythanks.io/to/phenobarbital
-Keywords: asyncio,asyncpg,aioredis,aiomcache,langchain,chatbot,agents
-Platform: POSIX
-Classifier: Development Status :: 4 - Beta
-Classifier: Intended Audience :: Developers
-Classifier: Operating System :: POSIX :: Linux
-Classifier: Environment :: Web Environment
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Topic :: Software Development :: Build Tools
-Classifier: Topic :: Software Development :: Libraries :: Python Modules
-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 :: Only
-Classifier: Framework :: AsyncIO
-Requires-Python: >=3.10.12
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: Cython ==3.0.9
-Requires-Dist: pymupdf ==1.24.4
-Requires-Dist: pymupdf4llm ==0.0.1
-Requires-Dist: pdf4llm ==0.0.6
-Requires-Dist: PyPDF2 ==3.0.1
-Requires-Dist: pdfminer.six ==20231228
-Requires-Dist: pdfplumber ==0.11.0
-Requires-Dist: bitsandbytes ==0.43.0
-Requires-Dist: Cartopy ==0.22.0
-Requires-Dist: chromadb ==0.4.24
-Requires-Dist: contourpy ==1.2.0
-Requires-Dist: datasets ==2.18.0
-Requires-Dist: faiss-cpu ==1.8.0
-Requires-Dist: fastavro ==1.9.4
-Requires-Dist: GitPython ==3.1.42
-Requires-Dist: gunicorn ==21.2.0
-Requires-Dist: jq ==1.7.0
-Requires-Dist: rank-bm25 ==0.2.2
-Requires-Dist: matplotlib ==3.8.3
-Requires-Dist: numba ==0.59.0
-Requires-Dist: opentelemetry-sdk ==1.24.0
-Requires-Dist: rapidocr-onnxruntime ==1.3.15
-Requires-Dist: pytesseract ==0.3.10
-Requires-Dist: python-docx ==1.1.0
-Requires-Dist: python-pptx ==0.6.23
-Requires-Dist: docx2txt ==0.8
-Requires-Dist: pytube ==15.0.0
-Requires-Dist: pydub ==0.25.1
-Requires-Dist: markdownify ==0.12.1
-Requires-Dist: librosa ==0.10.1
-Requires-Dist: yt-dlp ==2024.4.9
-Requires-Dist: moviepy ==1.0.3
-Requires-Dist: safetensors ==0.4.2
-Requires-Dist: sentence-transformers ==2.6.1
-Requires-Dist: tabulate ==0.9.0
-Requires-Dist: tiktoken ==0.7.0
-Requires-Dist: tokenizers ==0.19.1
-Requires-Dist: unstructured ==0.14.3
-Requires-Dist: unstructured-client ==0.18.0
-Requires-Dist: uvloop ==0.19.0
-Requires-Dist: XlsxWriter ==3.2.0
-Requires-Dist: youtube-transcript-api ==0.6.2
-Requires-Dist: selenium ==4.18.1
-Requires-Dist: webdriver-manager ==4.0.1
-Requires-Dist: transitions ==0.9.0
-Requires-Dist: sentencepiece ==0.2.0
-Requires-Dist: duckduckgo-search ==5.3.0
-Requires-Dist: google-search-results ==2.4.2
-Requires-Dist: google-api-python-client >=2.86.0
-Requires-Dist: gdown ==5.1.0
-Requires-Dist: weasyprint ==61.2
-Requires-Dist: markdown2 ==2.4.13
-Requires-Dist: xformers ==0.0.25.post1
-Requires-Dist: fastembed ==0.3.4
-Requires-Dist: mammoth ==1.7.1
-Requires-Dist: accelerate ==0.29.3
-Requires-Dist: langchain >=0.2.6
-Requires-Dist: langchain-community >=0.2.6
-Requires-Dist: langchain-core ==0.2.32
-Requires-Dist: langchain-experimental ==0.0.62
-Requires-Dist: langchainhub ==0.1.15
-Requires-Dist: langchain-text-splitters ==0.2.2
-Requires-Dist: huggingface-hub ==0.23.5
-Requires-Dist: llama-index ==0.10.20
-Requires-Dist: llama-cpp-python ==0.2.56
-Requires-Dist: asyncdb[all] >=2.7.10
-Requires-Dist: querysource >=3.10.1
-Requires-Dist: yfinance ==0.2.40
-Requires-Dist: youtube-search ==2.1.2
-Requires-Dist: wikipedia ==1.4.0
-Requires-Dist: mediawikiapi ==1.2
-Requires-Dist: wikibase-rest-api-client ==0.2.0
-Requires-Dist: asknews ==0.7.30
-Requires-Dist: pyowm ==3.3.0
-Requires-Dist: O365 ==2.0.35
-Requires-Dist: langchain-huggingface ==0.0.3
-Requires-Dist: stackapi ==0.3.1
-Provides-Extra: all
-Requires-Dist: langchain-milvus ==0.1.1 ; extra == 'all'
-Requires-Dist: milvus ==2.3.5 ; extra == 'all'
-Requires-Dist: pymilvus ==2.4.4 ; extra == 'all'
-Requires-Dist: groq ==0.6.0 ; extra == 'all'
-Requires-Dist: langchain-groq ==0.1.4 ; extra == 'all'
-Requires-Dist: llama-index-llms-huggingface ==0.2.7 ; extra == 'all'
-Requires-Dist: langchain-google-vertexai ==1.0.8 ; extra == 'all'
-Requires-Dist: langchain-google-genai ==1.0.8 ; extra == 'all'
-Requires-Dist: google-generativeai ==0.7.2 ; extra == 'all'
-Requires-Dist: vertexai ==1.60.0 ; extra == 'all'
-Requires-Dist: google-cloud-aiplatform >=1.60.0 ; extra == 'all'
-Requires-Dist: grpc-google-iam-v1 ==0.13.0 ; extra == 'all'
-Requires-Dist: langchain-openai ==0.1.21 ; extra == 'all'
-Requires-Dist: openai ==1.40.8 ; extra == 'all'
-Requires-Dist: llama-index-llms-openai ==0.1.11 ; extra == 'all'
-Requires-Dist: langchain-anthropic ==0.1.23 ; extra == 'all'
-Requires-Dist: anthropic ==0.34.0 ; extra == 'all'
-Provides-Extra: analytics
-Requires-Dist: annoy ==1.17.3 ; extra == 'analytics'
-Requires-Dist: gradio-tools ==0.0.9 ; extra == 'analytics'
-Requires-Dist: gradio-client ==0.2.9 ; extra == 'analytics'
-Requires-Dist: streamlit ==1.37.1 ; extra == 'analytics'
-Requires-Dist: simsimd ==4.3.1 ; extra == 'analytics'
-Requires-Dist: opencv-python ==4.10.0.84 ; extra == 'analytics'
-Provides-Extra: anthropic
-Requires-Dist: langchain-anthropic ==0.1.11 ; extra == 'anthropic'
-Requires-Dist: anthropic ==0.25.2 ; extra == 'anthropic'
-Provides-Extra: crew
-Requires-Dist: colbert-ai ==0.2.19 ; extra == 'crew'
-Requires-Dist: vanna ==0.3.4 ; extra == 'crew'
-Requires-Dist: crewai[tools] ==0.28.8 ; extra == 'crew'
-Provides-Extra: google
-Requires-Dist: langchain-google-vertexai ==1.0.4 ; extra == 'google'
-Requires-Dist: langchain-google-genai ==1.0.4 ; extra == 'google'
-Requires-Dist: google-generativeai ==0.5.4 ; extra == 'google'
-Requires-Dist: vertexai ==1.49.0 ; extra == 'google'
-Requires-Dist: google-cloud-aiplatform ==1.49.0 ; extra == 'google'
-Requires-Dist: grpc-google-iam-v1 ==0.13.0 ; extra == 'google'
-Provides-Extra: groq
-Requires-Dist: groq ==0.6.0 ; extra == 'groq'
-Requires-Dist: langchain-groq ==0.1.4 ; extra == 'groq'
-Provides-Extra: hunggingfaces
-Requires-Dist: llama-index-llms-huggingface ==0.2.7 ; extra == 'hunggingfaces'
-Provides-Extra: milvus
-Requires-Dist: langchain-milvus ==0.1.1 ; extra == 'milvus'
-Requires-Dist: milvus ==2.3.5 ; extra == 'milvus'
-Requires-Dist: pymilvus ==2.4.4 ; extra == 'milvus'
-Provides-Extra: openai
-Requires-Dist: langchain-openai ==0.1.21 ; extra == 'openai'
-Requires-Dist: openai ==1.40.3 ; extra == 'openai'
-Requires-Dist: llama-index-llms-openai ==0.1.11 ; extra == 'openai'
-Requires-Dist: tiktoken ==0.7.0 ; extra == 'openai'
-Provides-Extra: qdrant
-Requires-Dist: qdrant-client ==1.8.0 ; extra == 'qdrant'
-
-# AI Parrot: Python package for creating Chatbots
-This is an open-source Python package for creating Chatbots based on Langchain and Navigator.
-This README provides instructions for installation, development, testing, and releasing Parrot.
-
-## Installation
-
-**Creating a virtual environment:**
-
-This is recommended for development and isolation from system-wide libraries.
-Run the following command in your terminal:
-
-Debian-based systems installation:
-   ```
-   sudo apt install gcc python3.11-venv python3.11-full python3.11-dev libmemcached-dev zlib1g-dev build-essential libffi-dev unixodbc unixodbc-dev libsqliteodbc libev4 libev-dev
-   ```
-
-   For Qdrant installation:
-   ```
-   docker pull qdrant/qdrant
-   docker run -d -p 6333:6333 -p 6334:6334 --name qdrant -v $(pwd)/qdrant_storage:/qdrant/storage:z qdrant/qdrant
-   ```
-
-For VertexAI, creates a folder on "env" called "google" and copy the JSON credentials file into it.
-
-   ```bash
-   make venv
-   ```
-
-   This will create a virtual environment named `.venv`. To activate it, run:
-
-   ```bash
-   source .venv/bin/activate  # Linux/macOS
-   ```
-
-   Once activated, install Parrot within the virtual environment:
-
-   ```bash
-   make install
-   ```
-   The output will remind you to activate the virtual environment before development.
-
-   **Optional** (for developers):
-   ```bash
-   pip install -e .
-   ```
-
-## Start http server
-```bash
-python run.py
-```
-
-## Development Setup
-
-This section explains how to set up your development environment:
-
-1. **Install development requirements:**
-
-   ```bash
-   make setup
-   ```
-
-   This installs development dependencies like linters and test runners mentioned in the `docs/requirements-dev.txt` file.
-
-2. **Install Parrot in editable mode:**
-
-   This allows you to make changes to the code and test them without reinstalling:
-
-   ```bash
-   make dev
-   ```
-
-   This uses `flit` to install Parrot in editable mode.
-
-
-### Usage (Replace with actual usage instructions)
-
-*Once you have set up your development environment, you can start using Parrot.*
-
-#### Test with Code ChatBOT
-* Set environment variables for:
- [google]
-   GOOGLE_API_KEY=apikey
-   GOOGLE_CREDENTIALS_FILE=.../credentials.json
-   VERTEX_PROJECT_ID=vertex-project
-   VERTEX_REGION=region
-
-* Run the chatbot:
-    ```bash
-    python examples/test_agent.py
-    ```
-
-### Testing
-
-To run the test suite:
-
-```bash
-make test
-```
-
-This will run tests using `coverage` to report on code coverage.
-
-
-### Code Formatting
-
-To format the code with black:
-
-```bash
-make format
-```
-
-
-### Linting
-
-To lint the code for style and potential errors:
-
-```bash
-make lint
-```
-
-This uses `pylint` and `black` to check for issues.
-
-
-### Releasing a New Version
-
-This section outlines the steps for releasing a new version of Parrot:
-
-1. **Ensure everything is clean and tested:**
-
-   ```bash
-   make release
-   ```
-
-   This runs `lint`, `test`, and `clean` tasks before proceeding.
-
-2. **Publish the package:**
-
-   ```bash
-   make release
-   ```
-
-   This uses `flit` to publish the package to a repository like PyPI. You'll need to have publishing credentials configured for `flit`.
-
-
-### Cleaning Up
-
-To remove the virtual environment:
-
-```bash
-make distclean
-```
-
-
-### Contributing
-
-We welcome contributions to Parrot! Please refer to the CONTRIBUTING.md file for guidelines on how to contribute.