krutiextract 1.0.0__tar.gz

krutiextract-1.0.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 KrutiExtract Contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

krutiextract-1.0.0/PKG-INFO

@@ -0,0 +1,119 @@
+Metadata-Version: 2.4
+Name: krutiextract
+Version: 1.0.0
+Summary: A universal toolkit for extracting legacy Hindi PDFs (KrutiDev/Devlys/Chanakya) to Markdown
+Author-email: "Bharat Choudhary (Lecturer, Department of Secondary Education, Govt. of Rajasthan)" <epicbharat@gmail.com>
+Project-URL: Homepage, https://github.com/epicbharat
+Project-URL: Bug Tracker, https://github.com/epicbharat/krutiextract/issues
+Classifier: Programming Language :: Python :: 3
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Text Processing :: Linguistic
+Classifier: Natural Language :: Hindi
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: pymupdf
+Requires-Dist: pymupdf4llm
+Requires-Dist: nltk
+Requires-Dist: opencv-python-headless
+Requires-Dist: numpy
+Dynamic: license-file
+
+<div align="center">
+  <img src="assets/logo.png" alt="KrutiExtract Logo" width="200"/>
+  <h1>KrutiExtract</h1>
+  <p><strong>A Universal Legacy Hindi PDF to Markdown Extractor</strong></p>
+
+  [![Python Version](https://img.shields.io/badge/python-3.8%2B-blue)](https://www.python.org/)
+  [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+  [![Maintenance](https://img.shields.io/badge/Maintained%3F-yes-green.svg)](https://github.com/epicbharat)
+</div>
+
+<hr/>
+
+Extracting legacy Hindi fonts (like **KrutiDev**, **Devlys**, or **Chanakya**) from PDF documents has traditionally been a nightmare. Standard OCR and PDF text extractors read characters based on visual coordinates. Because Hindi matras (vowel marks like `ि` or `े`) overlap with consonants, this visual sorting scrambles the text (e.g., typing `vkSj` for 'और' gets extracted as `vkjS`, breaking the conversion into the nonsense word 'आरै'). 
+
+**KrutiExtract** solves this problem by bypassing visual coordinate sorting and extracting the **logical character stream** directly from the PDF.
+
+## ✨ Core Features
+
+- **Perfect Character Order:** Preserves the original typist's keystrokes, completely avoiding spatial scrambling of matras.
+- **Smart Auto-Detect Engine:** Automatically mathematically calculates text frequencies to detect whether the typist used **KrutiDev / Devlys** or **Chanakya**, instantly routing the text to the correct translation engine!
+- **English / Hinglish Bypass Mode:** If the engine detects that the document contains modern Unicode Hindi, standard English, or Hinglish, it automatically bypasses legacy conversion entirely—perfectly preserving your modern texts!
+- **Dynamic English Preservation:** Uses NLTK's English corpus to proactively detect and protect English words, numbers, acronyms, and HTML tags interspersed within mixed-font Hindi texts.
+- **Collision-Free Safe-Tokens:** Markdown syntax elements (like `#` or `*`) clash with KrutiDev maps. KrutiExtract shields these structural elements using a unique `$$$INDEX$$$` safe-token system during font conversion.
+
+## 👁️ Advanced OCR & Image Cleaning
+
+When the toolkit encounters text trapped inside a scanned image, it seamlessly hands it over to **Tesseract OCR** in the background.
+
+- **Native Hindi OCR:** The toolkit hardcodes `ocr_language="hin"` into the internal pipeline. This permanently stops Tesseract from trying to hallucinate English letters onto Hindi image shapes, ensuring natively output Unicode Devanagari.
+- **OpenCV Image Cleaning (`--enhance-ocr`):** Pass this flag to activate our built-in computer vision pipeline. The toolkit will automatically intercept blurry images in the PDF, upscale the resolution, apply unsharp masking, and run adaptive thresholding to perfectly reconstruct the text shapes *before* OCR scans it!
+
+## 🚀 Installation
+
+Ensure you have Python 3.8+ installed.
+
+```bash
+# Clone the repository
+git clone https://github.com/epicbharat/krutiextract.git
+cd krutiextract
+
+# Create and activate a virtual environment (Recommended)
+python -m venv venv
+venv\Scripts\activate  # On Windows
+# source venv/bin/activate  # On macOS/Linux
+
+# Install dependencies
+pip install -r requirements.txt
+
+# Download the required NLTK English word corpus
+python -c "import nltk; nltk.download('words')"
+```
+
+### Note on Tesseract (Required for OCR)
+If you are parsing scanned PDFs or PDFs with embedded images, you must have [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) installed on your system.
+- **Windows:** Download the installer and ensure you check the **"Hindi"** language pack during installation. Add Tesseract to your PATH.
+- **Linux:** `sudo apt-get install tesseract-ocr tesseract-ocr-hin`
+
+## 💻 CLI Usage Guide
+
+KrutiExtract provides a powerful Command Line Interface (CLI).
+
+### 1. Process a Single File
+```bash
+python -m src.cli --input "path/to/document.pdf" --output "path/to/output.md"
+```
+
+### 2. Batch Process a Directory
+Recursively process an entire folder of PDFs:
+```bash
+python -m src.cli --dir "path/to/pdf/folder" --out-dir "path/to/markdown/folder"
+```
+
+### 3. Force a Specific Font Encoding
+By default, the font is set to `auto`. You can force a specific bypass or mapping:
+```bash
+# Available options: krutidev, chanakya, auto, unicode, english
+python -m src.cli --input "doc.pdf" --font chanakya
+```
+
+### 4. Enable OpenCV OCR Enhancement
+Use this flag for old, blurry, or low-resolution scanned PDFs to artificially sharpen the text before extraction:
+```bash
+python -m src.cli --input "blurry_scan.pdf" --enhance-ocr
+```
+
+## 🤝 Contributing
+
+Contributions are welcome! If you find a PDF that breaks the extraction logic or a rare character combination we missed, please open an issue and attach the sample PDF.
+
+## 📄 License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+---
+**Author:** Bharat Choudhary (Lecturer, Department of Secondary Education, Govt. of Rajasthan)  
+**Email:** epicbharat@gmail.com  
+**GitHub:** [https://github.com/epicbharat](https://github.com/epicbharat)

krutiextract-1.0.0/README.md

@@ -0,0 +1,97 @@
+<div align="center">
+  <img src="assets/logo.png" alt="KrutiExtract Logo" width="200"/>
+  <h1>KrutiExtract</h1>
+  <p><strong>A Universal Legacy Hindi PDF to Markdown Extractor</strong></p>
+
+  [![Python Version](https://img.shields.io/badge/python-3.8%2B-blue)](https://www.python.org/)
+  [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+  [![Maintenance](https://img.shields.io/badge/Maintained%3F-yes-green.svg)](https://github.com/epicbharat)
+</div>
+
+<hr/>
+
+Extracting legacy Hindi fonts (like **KrutiDev**, **Devlys**, or **Chanakya**) from PDF documents has traditionally been a nightmare. Standard OCR and PDF text extractors read characters based on visual coordinates. Because Hindi matras (vowel marks like `ि` or `े`) overlap with consonants, this visual sorting scrambles the text (e.g., typing `vkSj` for 'और' gets extracted as `vkjS`, breaking the conversion into the nonsense word 'आरै'). 
+
+**KrutiExtract** solves this problem by bypassing visual coordinate sorting and extracting the **logical character stream** directly from the PDF.
+
+## ✨ Core Features
+
+- **Perfect Character Order:** Preserves the original typist's keystrokes, completely avoiding spatial scrambling of matras.
+- **Smart Auto-Detect Engine:** Automatically mathematically calculates text frequencies to detect whether the typist used **KrutiDev / Devlys** or **Chanakya**, instantly routing the text to the correct translation engine!
+- **English / Hinglish Bypass Mode:** If the engine detects that the document contains modern Unicode Hindi, standard English, or Hinglish, it automatically bypasses legacy conversion entirely—perfectly preserving your modern texts!
+- **Dynamic English Preservation:** Uses NLTK's English corpus to proactively detect and protect English words, numbers, acronyms, and HTML tags interspersed within mixed-font Hindi texts.
+- **Collision-Free Safe-Tokens:** Markdown syntax elements (like `#` or `*`) clash with KrutiDev maps. KrutiExtract shields these structural elements using a unique `$$$INDEX$$$` safe-token system during font conversion.
+
+## 👁️ Advanced OCR & Image Cleaning
+
+When the toolkit encounters text trapped inside a scanned image, it seamlessly hands it over to **Tesseract OCR** in the background.
+
+- **Native Hindi OCR:** The toolkit hardcodes `ocr_language="hin"` into the internal pipeline. This permanently stops Tesseract from trying to hallucinate English letters onto Hindi image shapes, ensuring natively output Unicode Devanagari.
+- **OpenCV Image Cleaning (`--enhance-ocr`):** Pass this flag to activate our built-in computer vision pipeline. The toolkit will automatically intercept blurry images in the PDF, upscale the resolution, apply unsharp masking, and run adaptive thresholding to perfectly reconstruct the text shapes *before* OCR scans it!
+
+## 🚀 Installation
+
+Ensure you have Python 3.8+ installed.
+
+```bash
+# Clone the repository
+git clone https://github.com/epicbharat/krutiextract.git
+cd krutiextract
+
+# Create and activate a virtual environment (Recommended)
+python -m venv venv
+venv\Scripts\activate  # On Windows
+# source venv/bin/activate  # On macOS/Linux
+
+# Install dependencies
+pip install -r requirements.txt
+
+# Download the required NLTK English word corpus
+python -c "import nltk; nltk.download('words')"
+```
+
+### Note on Tesseract (Required for OCR)
+If you are parsing scanned PDFs or PDFs with embedded images, you must have [Tesseract OCR](https://github.com/tesseract-ocr/tesseract) installed on your system.
+- **Windows:** Download the installer and ensure you check the **"Hindi"** language pack during installation. Add Tesseract to your PATH.
+- **Linux:** `sudo apt-get install tesseract-ocr tesseract-ocr-hin`
+
+## 💻 CLI Usage Guide
+
+KrutiExtract provides a powerful Command Line Interface (CLI).
+
+### 1. Process a Single File
+```bash
+python -m src.cli --input "path/to/document.pdf" --output "path/to/output.md"
+```
+
+### 2. Batch Process a Directory
+Recursively process an entire folder of PDFs:
+```bash
+python -m src.cli --dir "path/to/pdf/folder" --out-dir "path/to/markdown/folder"
+```
+
+### 3. Force a Specific Font Encoding
+By default, the font is set to `auto`. You can force a specific bypass or mapping:
+```bash
+# Available options: krutidev, chanakya, auto, unicode, english
+python -m src.cli --input "doc.pdf" --font chanakya
+```
+
+### 4. Enable OpenCV OCR Enhancement
+Use this flag for old, blurry, or low-resolution scanned PDFs to artificially sharpen the text before extraction:
+```bash
+python -m src.cli --input "blurry_scan.pdf" --enhance-ocr
+```
+
+## 🤝 Contributing
+
+Contributions are welcome! If you find a PDF that breaks the extraction logic or a rare character combination we missed, please open an issue and attach the sample PDF.
+
+## 📄 License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
+
+---
+**Author:** Bharat Choudhary (Lecturer, Department of Secondary Education, Govt. of Rajasthan)  
+**Email:** epicbharat@gmail.com  
+**GitHub:** [https://github.com/epicbharat](https://github.com/epicbharat)

krutiextract-1.0.0/pyproject.toml

@@ -0,0 +1,34 @@
+[build-system]
+requires = ["setuptools>=61.0"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "krutiextract"
+version = "1.0.0"
+authors = [
+  { name="Bharat Choudhary (Lecturer, Department of Secondary Education, Govt. of Rajasthan)", email="epicbharat@gmail.com" },
+]
+description = "A universal toolkit for extracting legacy Hindi PDFs (KrutiDev/Devlys/Chanakya) to Markdown"
+readme = "README.md"
+requires-python = ">=3.8"
+classifiers = [
+    "Programming Language :: Python :: 3",
+    "License :: OSI Approved :: MIT License",
+    "Operating System :: OS Independent",
+    "Topic :: Text Processing :: Linguistic",
+    "Natural Language :: Hindi"
+]
+dependencies = [
+    "pymupdf",
+    "pymupdf4llm",
+    "nltk",
+    "opencv-python-headless",
+    "numpy"
+]
+
+[project.scripts]
+krutiextract = "hindi_pdf_extractor.src.cli:main"
+
+[project.urls]
+Homepage = "https://github.com/epicbharat"
+"Bug Tracker" = "https://github.com/epicbharat/krutiextract/issues"

krutiextract-1.0.0/setup.cfg

@@ -0,0 +1,4 @@
+[egg_info]
+tag_build = 
+tag_date = 0
+

krutiextract-1.0.0/src/__init__.py

@@ -0,0 +1 @@
+"""Source module for hindi_pdf_extractor."""

krutiextract-1.0.0/src/chanakya_converter.py

@@ -0,0 +1,69 @@
+import re
+from .chanakya_map import array_one, array_two
+
+def escape_regex(s):
+    return re.escape(s)
+
+def chanakya_to_unicode(text: str) -> str:
+    """
+    Converts a Chanakya encoded string into standard Devanagari Unicode.
+    """
+    if not text:
+        return ""
+    
+    modified_substring = text
+
+    # Apply base array replacements
+    for i in range(len(array_one)):
+        pattern = re.compile(escape_regex(array_one[i]))
+        modified_substring = pattern.sub(array_two[i], modified_substring)
+
+    # Chanakya special rules
+    modified_substring = re.sub(r'Z', "üं", modified_substring)
+
+    # Handle short 'i' matra (ç)
+    position_of_f = modified_substring.find("ç")
+    while position_of_f != -1:
+        if position_of_f + 1 < len(modified_substring):
+            character_right_to_f = modified_substring[position_of_f + 1]
+            modified_substring = modified_substring.replace(
+                "ç" + character_right_to_f,
+                character_right_to_f + "ि"
+            )
+            # Handle half consonants after 'i' matra
+            pos_temp = position_of_f + 1
+            while pos_temp + 1 < len(modified_substring) and modified_substring[pos_temp + 1] == "्":
+                if pos_temp + 2 < len(modified_substring):
+                    string_to_be_replaced = modified_substring[pos_temp + 1] + modified_substring[pos_temp + 2]
+                    modified_substring = modified_substring.replace("ि" + string_to_be_replaced, string_to_be_replaced + "ि")
+                    pos_temp += 2
+                else:
+                    break
+        position_of_f = modified_substring.find("ç", position_of_f + 1)
+
+    # Reorder 'R' matras (ü -> र्)
+    set_of_matras = "ा ि ी ु ू ृ े ै ो ौ ं ः ँ ॅ"
+    position_of_Z = modified_substring.find("ü")
+    
+    while position_of_Z > 0:
+        probable_position_of_half_r = position_of_Z - 1
+        character_at_probable_position_of_half_r = modified_substring[probable_position_of_half_r]
+
+        while character_at_probable_position_of_half_r in set_of_matras and probable_position_of_half_r >= 0:
+            probable_position_of_half_r -= 1
+            if probable_position_of_half_r >= 0:
+                character_at_probable_position_of_half_r = modified_substring[probable_position_of_half_r]
+
+        if probable_position_of_half_r >= 0:
+            substring_to_be_replaced = modified_substring[probable_position_of_half_r:position_of_Z + 1]
+            replace_with = "र्" + modified_substring[probable_position_of_half_r:position_of_Z]
+            modified_substring = modified_substring.replace(substring_to_be_replaced, replace_with)
+
+        position_of_Z = modified_substring.find("ü", position_of_Z + 1)
+
+    # Final cleanup
+    modified_substring = modified_substring.replace("ंे", "ें")
+    modified_substring = modified_substring.replace("ंो", "ों")
+    modified_substring = modified_substring.replace("ाे", "ो")
+
+    return modified_substring