How kathirkarky/tamilenkal works
What this project is (type + purpose)
From the file tree + README, this is a Python library/package (not a service) intended to perform Tamil numeral conversions and basic numeric operations.
• Primary purpose: Convert between Arabic numerals (int/float) and Tamil numerals (string representation) and provide arithmetic/rounding helpers that *return results in Tamil numerals*.
• Packaging/distribution: setup.py indicates it’s built and distributed as a Python package (likely to PyPI).
• CI/CD: .github/workflows/python-publish.yml indicates an automated publish pipeline.
• Usage example surface: TamilEnkalDemo.ipynb is a consumer/demo artifact.
Top-level:
• TamilEnmam/ → the actual Python package
• TamilEnmam/TamilEnkal.py → the core implementation module
• TamilEnmam/__init__.py → package exports / public API surface
• setup.py → packaging metadata, install entry point
• .github/workflows/python-publish.yml → publishing automation
• TamilEnkalDemo.ipynb → example consumer
• README.md, LICENSE → documentation/legal
Architecturally, this splits naturally into (A) runtime library components and (B) build/release components.
External actor / consumer
• Python user code (any app/script/notebook)
• Jupyter Notebook (explicitly represented by TamilEnkalDemo.ipynb)
Public API layer (package boundary)
• TamilEnmam package (TamilEnmam/__init__.py)
• Treat this as the public façade: what users import.
• If __init__.py re-exports functions from TamilEnkal.py, show that explicitly.
Core module
• TamilEnkal module (TamilEnmam/TamilEnkal.py)
• This contains the functions listed in the README and likely internal helpers/constants.
Conversion: Arabic → Tamil
• TamilEn(arabic_number) -> tamil_string
Conversion: Tamil → Arabic
• ArabuEn(tamil_string) -> arabic_number
Rounding layer
• TamilSutru(arabic_number) -> tamil_string (round then convert)
• ArabuSutru(tamil_string) -> arabic_number (parse then round)
Arithmetic operations (compute in Arabic, return Tamil)
• Perumam(a,b) max
• Sirumam(a,b) min
• Koottal(a,b) add
• Kazhithal(a,b) subtract
• Perukkal(a,b) multiply
• Eevu(a,b) quotient/division
• Meedhi(a,b) remainder/mod
• Tamil digit/symbol mapping tables (e.g., Tamil digits 0–9, tens, hundreds, etc.)
• Normalization/validation (strip spaces, validate characters)
• Tokenization/parsing logic (Tamil numeral string → numeric)
• Formatting logic (numeric → Tamil numeral string)
Flow 1: Arabic → Tamil conversion
Consumer (int/float) → calls TamilEnmam.TamilEn() → uses mapping/formatting → returns Tamil string
Flow 2: Tamil → Arabic conversion
Consumer (Tamil string) → calls TamilEnmam.ArabuEn() → uses parsing/tokenization + mapping → returns int/float
Flow 3: Arithmetic helpers
Consumer (a,b) → calls Koottal/Perukkal/etc.
These likely do:
• compute using Python numeric ops (a+b, a*b, etc.)
• then call TamilEn(result)
Koottal() → TamilEn() (dependency)
Same for Kazhithal, Perukkal, Eevu, Meedhi, Perumam, Sirumam.
Flow 4: Rounding helpers
• TamilSutru(x) → round(x) / rounding logic → TamilEn(rounded)
• ArabuSutru(t) → ArabuEn(t) → round(result) (or rounding logic)
These dependencies matter: it shows there’s a “core conversion primitive” reused by higher-level utilities.
Components
• Developer pushes tags/commits to GitHub
• GitHub Actions workflow (.github/workflows/python-publish.yml)
• Build step (sdist/wheel via setuptools)
• Package index (PyPI or equivalent target)
• End user installation (pip install ...) pulling from PyPI
Relationships
GitHub repo → triggers GitHub Actions → builds artifacts → publishes to Package Registry → user installs → imports TamilEnmam
This is part of “system architecture” for a library project: the distribution path is a core system concern.
• User / consumer code
• TamilEnkal library
• PyPI (distribution), GitHub Actions (delivery)
• “Consumer environment” (Notebook/App)
• “TamilEnmam package (in-process)”
• “Python runtime”
• (Optional) “PyPI” as external dependency only for installation time
• Inside the package: __init__.py API façade, TamilEnkal.py module
• Inside the module: Conversion, Parsing, Formatting, Rounding, Arithmetic
• Function-level dependencies (especially arithmetic → TamilEn)