Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import json
|
| 3 |
+
from rapidfuzz import process, fuzz
|
| 4 |
+
import phonetics
|
| 5 |
+
from indic_transliteration import sanscript
|
| 6 |
+
from indic_transliteration.sanscript import transliterate
|
| 7 |
+
from collections import defaultdict
|
| 8 |
+
|
| 9 |
+
data = []
|
| 10 |
+
# Load and parse the JSON data efficiently
|
| 11 |
+
with open("hin_test.json", "r", encoding="utf-8") as file:
|
| 12 |
+
for line in file:
|
| 13 |
+
try:
|
| 14 |
+
# Parse each line as a JSON object
|
| 15 |
+
data.append(json.loads(line))
|
| 16 |
+
except json.JSONDecodeError as e:
|
| 17 |
+
print(f"Error parsing line: {line}. Error: {e}")
|
| 18 |
+
|
| 19 |
+
# Extract names from the dataset
|
| 20 |
+
names_database = [entry['english word'] for entry in data if 'english word' in entry]
|
| 21 |
+
|
| 22 |
+
# Precompute metaphone codes for the database names
|
| 23 |
+
phonetic_dict = defaultdict(list)
|
| 24 |
+
for name in names_database:
|
| 25 |
+
code = phonetics.metaphone(name)
|
| 26 |
+
phonetic_dict[code].append(name)
|
| 27 |
+
|
| 28 |
+
# Function to perform fuzzy matching
|
| 29 |
+
def fuzzy_match(name, threshold=70):
|
| 30 |
+
matches = process.extract(name, names_database, scorer=fuzz.ratio, limit=5)
|
| 31 |
+
matched_names = [
|
| 32 |
+
f"{match_name} - {score:.2f}" for match_name, score, _ in matches if score >= threshold
|
| 33 |
+
]
|
| 34 |
+
return ", ".join(matched_names) if matched_names else "No matches found."
|
| 35 |
+
|
| 36 |
+
# Function to perform phonetic matching
|
| 37 |
+
def phonetic_match(name):
|
| 38 |
+
input_phonetic = phonetics.metaphone(name)
|
| 39 |
+
matches = phonetic_dict.get(input_phonetic, [])
|
| 40 |
+
return ", ".join(matches) if matches else "No phonetic matches found."
|
| 41 |
+
|
| 42 |
+
# Function to transliterate Devanagari to Roman script
|
| 43 |
+
def devanagari_to_roman(hindi_name):
|
| 44 |
+
roman_name_itrans = transliterate(hindi_name, sanscript.DEVANAGARI, sanscript.ITRANS)
|
| 45 |
+
return roman_name_itrans
|
| 46 |
+
|
| 47 |
+
# Function to handle different types of matching/transliteration
|
| 48 |
+
def match_name(name, method):
|
| 49 |
+
if method == "Fuzzy Match":
|
| 50 |
+
return fuzzy_match(name)
|
| 51 |
+
elif method == "Phonetic Match":
|
| 52 |
+
return phonetic_match(name)
|
| 53 |
+
elif method == "Devanagari to Roman Transliteration":
|
| 54 |
+
return devanagari_to_roman(name)
|
| 55 |
+
else:
|
| 56 |
+
return "Select a valid method."
|
| 57 |
+
|
| 58 |
+
# Gradio Interface with updated components and parameters
|
| 59 |
+
interface = gr.Interface(
|
| 60 |
+
fn=match_name,
|
| 61 |
+
inputs=[
|
| 62 |
+
gr.Textbox(label="Enter Name"),
|
| 63 |
+
gr.Radio(
|
| 64 |
+
["Fuzzy Match", "Phonetic Match", "Devanagari to Roman Transliteration"],
|
| 65 |
+
label="Matching Method"
|
| 66 |
+
)
|
| 67 |
+
],
|
| 68 |
+
outputs=gr.Textbox(label="Result"),
|
| 69 |
+
title="नाम-सूचक (Naam-Suchak)",
|
| 70 |
+
allow_flagging='never',
|
| 71 |
+
description=(
|
| 72 |
+
"Welcome to **नाम-सूचक**, an innovative solution developed for the SIH1787 problem statement. "
|
| 73 |
+
"This tool offers a robust and intelligent approach for **fuzzy matching**, **phonetic matching**, "
|
| 74 |
+
"and **transliteration** of Hindi names, ensuring seamless handling of variations and phonetic similarities. "
|
| 75 |
+
"Effortlessly search, match, and transliterate names with high accuracy and efficiency."
|
| 76 |
+
),
|
| 77 |
+
examples=[
|
| 78 |
+
["Suresh", "Fuzzy Match"],
|
| 79 |
+
["Shiv", "Phonetic Match"],
|
| 80 |
+
["राहुल", "Devanagari to Roman Transliteration"]
|
| 81 |
+
],
|
| 82 |
+
cache_examples=True
|
| 83 |
+
)
|
| 84 |
+
|
| 85 |
+
# Launch the Gradio app
|
| 86 |
+
if __name__ == "__main__":
|
| 87 |
+
interface.launch()
|