Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,331 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
import re
|
| 3 |
+
from typing import Dict, List
|
| 4 |
+
|
| 5 |
+
class PersianVowelizer:
|
| 6 |
+
"""
|
| 7 |
+
A rule-based Persian vowel mark (diacritic) system.
|
| 8 |
+
This is a simplified version - for production use, consider using
|
| 9 |
+
pre-trained models like BERT-based Persian vocalization models.
|
| 10 |
+
"""
|
| 11 |
+
|
| 12 |
+
def __init__(self):
|
| 13 |
+
# Persian vowel marks (diacritics)
|
| 14 |
+
self.vowel_marks = {
|
| 15 |
+
'fatha': '\u064E', # َ (a)
|
| 16 |
+
'kasra': '\u0650', # ِ (e/i)
|
| 17 |
+
'damma': '\u064F', # ُ (o/u)
|
| 18 |
+
'sukun': '\u0652', # ْ (silence)
|
| 19 |
+
'shadda': '\u0651', # ّ (doubling)
|
| 20 |
+
'tanwin_fath': '\u064B', # ً
|
| 21 |
+
'tanwin_kasr': '\u064D', # ٍ
|
| 22 |
+
'tanwin_damm': '\u064C', # ٌ
|
| 23 |
+
}
|
| 24 |
+
|
| 25 |
+
# Common Persian words with their vocalized forms
|
| 26 |
+
self.common_words = {
|
| 27 |
+
'سلام': 'سَلام',
|
| 28 |
+
'خوب': 'خوب',
|
| 29 |
+
'بد': 'بَد',
|
| 30 |
+
'خانه': 'خانِه',
|
| 31 |
+
'مدرسه': 'مَدرِسِه',
|
| 32 |
+
'کتاب': 'کِتاب',
|
| 33 |
+
'آب': 'آب',
|
| 34 |
+
'نان': 'نان',
|
| 35 |
+
'دوست': 'دوست',
|
| 36 |
+
'عشق': 'عِشق',
|
| 37 |
+
'زندگی': 'زِندِگی',
|
| 38 |
+
'کار': 'کار',
|
| 39 |
+
'خواهر': 'خواهَر',
|
| 40 |
+
'برادر': 'بَرادَر',
|
| 41 |
+
'مادر': 'مادَر',
|
| 42 |
+
'پدر': 'پِدَر',
|
| 43 |
+
'فرد': 'فَرد',
|
| 44 |
+
'شب': 'شَب',
|
| 45 |
+
'روز': 'روز',
|
| 46 |
+
'صبح': 'صُبح',
|
| 47 |
+
'عصر': 'عَصر',
|
| 48 |
+
'شام': 'شام',
|
| 49 |
+
'درس': 'دَرس',
|
| 50 |
+
'استاد': 'اُستاد',
|
| 51 |
+
'دانشجو': 'دانِشجو',
|
| 52 |
+
'دانشگاه': 'دانِشگاه',
|
| 53 |
+
'شهر': 'شَهر',
|
| 54 |
+
'روستا': 'روستا',
|
| 55 |
+
'کشور': 'کِشوَر',
|
| 56 |
+
'ایران': 'ایران',
|
| 57 |
+
'فارسی': 'فارسی',
|
| 58 |
+
'زبان': 'زَبان',
|
| 59 |
+
'کلام': 'کَلام',
|
| 60 |
+
'سخن': 'سُخن',
|
| 61 |
+
'گفتن': 'گُفتَن',
|
| 62 |
+
'رفتن': 'رَفتَن',
|
| 63 |
+
'آمدن': 'آمَدَن',
|
| 64 |
+
'خوردن': 'خوردَن',
|
| 65 |
+
'نوشتن': 'نِویسیدَن',
|
| 66 |
+
'خواندن': 'خواندَن',
|
| 67 |
+
'دیدن': 'دیدَن',
|
| 68 |
+
'شنیدن': 'شِنیدَن',
|
| 69 |
+
'فهمیدن': 'فَهمیدَن',
|
| 70 |
+
'دانستن': 'دانِستَن',
|
| 71 |
+
'بودن': 'بودَن',
|
| 72 |
+
'کردن': 'کَردَن',
|
| 73 |
+
'داشتن': 'داشتَن',
|
| 74 |
+
'گذاشتن': 'گُذاشتَن',
|
| 75 |
+
'نشستن': 'نِشَستَن',
|
| 76 |
+
'ایستادن': 'ایستادَن',
|
| 77 |
+
'خوابیدن': 'خوابیدَن',
|
| 78 |
+
'بیدار': 'بیدار',
|
| 79 |
+
'خسته': 'خَستِه',
|
| 80 |
+
'شاد': 'شاد',
|
| 81 |
+
'غمگین': 'غَمگین',
|
| 82 |
+
'خوشحال': 'خوشحال',
|
| 83 |
+
'عاشق': 'عاشِق',
|
| 84 |
+
'دوستدار': 'دوستدار',
|
| 85 |
+
'محبت': 'مُحِبَّت',
|
| 86 |
+
'عشق': 'عِشق',
|
| 87 |
+
'نفرت': 'نَفرَت',
|
| 88 |
+
'ترس': 'تَرس',
|
| 89 |
+
'امید': 'اُمید',
|
| 90 |
+
'آرزو': 'آرِزو',
|
| 91 |
+
'خواب': 'خواب',
|
| 92 |
+
'رویا': 'رؤیا'
|
| 93 |
+
}
|
| 94 |
+
|
| 95 |
+
# Pattern-based rules for common Persian structures
|
| 96 |
+
self.patterns = [
|
| 97 |
+
# Past tense patterns
|
| 98 |
+
(r'(\w+)ید', r'\1ید'), # -id ending
|
| 99 |
+
(r'(\w+)د', r'\1َد'), # -ad ending for past tense
|
| 100 |
+
|
| 101 |
+
# Present tense patterns
|
| 102 |
+
(r'می(\w+)', r'می\1'), # mi- prefix
|
| 103 |
+
(r'(\w+)م', r'\1َم'), # -am ending
|
| 104 |
+
(r'(\w+)ی', r'\1ی'), # -i ending
|
| 105 |
+
(r'(\w+)د', r'\1َد'), # -ad ending
|
| 106 |
+
(r'(\w+)یم', r'\1یم'), # -im ending
|
| 107 |
+
(r'(\w+)ید', r'\1ید'), # -id ending
|
| 108 |
+
(r'(\w+)ند', r'\1َند'), # -and ending
|
| 109 |
+
]
|
| 110 |
+
|
| 111 |
+
def add_vowel_marks(self, text: str) -> str:
|
| 112 |
+
"""
|
| 113 |
+
Add vowel marks to Persian text using dictionary lookup and pattern matching.
|
| 114 |
+
"""
|
| 115 |
+
if not text.strip():
|
| 116 |
+
return text
|
| 117 |
+
|
| 118 |
+
# Split text into words while preserving spaces and punctuation
|
| 119 |
+
words = re.findall(r'\S+|\s+', text)
|
| 120 |
+
vocalized_words = []
|
| 121 |
+
|
| 122 |
+
for word in words:
|
| 123 |
+
if word.isspace() or not re.search(r'[\u0600-\u06FF]', word):
|
| 124 |
+
# Skip whitespace and non-Persian characters
|
| 125 |
+
vocalized_words.append(word)
|
| 126 |
+
continue
|
| 127 |
+
|
| 128 |
+
# Clean word for lookup (remove punctuation)
|
| 129 |
+
clean_word = re.sub(r'[^\u0600-\u06FF]', '', word)
|
| 130 |
+
|
| 131 |
+
# Try dictionary lookup first
|
| 132 |
+
if clean_word in self.common_words:
|
| 133 |
+
# Replace the Persian part with vocalized version
|
| 134 |
+
vocalized_word = word.replace(clean_word, self.common_words[clean_word])
|
| 135 |
+
vocalized_words.append(vocalized_word)
|
| 136 |
+
else:
|
| 137 |
+
# Apply pattern-based rules
|
| 138 |
+
vocalized_word = word
|
| 139 |
+
for pattern, replacement in self.patterns:
|
| 140 |
+
vocalized_word = re.sub(pattern, replacement, vocalized_word)
|
| 141 |
+
|
| 142 |
+
# Add basic vowel marks based on common patterns
|
| 143 |
+
vocalized_word = self._apply_basic_rules(vocalized_word)
|
| 144 |
+
vocalized_words.append(vocalized_word)
|
| 145 |
+
|
| 146 |
+
return ''.join(vocalized_words)
|
| 147 |
+
|
| 148 |
+
def _apply_basic_rules(self, word: str) -> str:
|
| 149 |
+
"""
|
| 150 |
+
Apply basic vowel marking rules to Persian words.
|
| 151 |
+
"""
|
| 152 |
+
# This is a simplified approach - real vocalization needs more sophisticated rules
|
| 153 |
+
result = word
|
| 154 |
+
|
| 155 |
+
# Add fatha to common consonant patterns
|
| 156 |
+
result = re.sub(r'([بپتثجچحخدذرزژسشصضطظعغفقکگلمنوهی])([بپتثجچحخدذرزژسشصضطظعغفقکگلمنوهی])',
|
| 157 |
+
r'\1َ\2', result)
|
| 158 |
+
|
| 159 |
+
return result
|
| 160 |
+
|
| 161 |
+
def vocalize_persian_text(text: str) -> str:
|
| 162 |
+
"""
|
| 163 |
+
Main function to add vowel marks to Persian text.
|
| 164 |
+
"""
|
| 165 |
+
if not text.strip():
|
| 166 |
+
return "لطفاً متنی وارد کنید."
|
| 167 |
+
|
| 168 |
+
vowelizer = PersianVowelizer()
|
| 169 |
+
vocalized_text = vowelizer.add_vowel_marks(text)
|
| 170 |
+
return vocalized_text
|
| 171 |
+
|
| 172 |
+
# Custom CSS for RTL support and Persian fonts
|
| 173 |
+
custom_css = """
|
| 174 |
+
.rtl-text {
|
| 175 |
+
direction: rtl !important;
|
| 176 |
+
text-align: right !important;
|
| 177 |
+
font-family: 'Vazir', 'Tahoma', 'Arial Unicode MS', sans-serif !important;
|
| 178 |
+
font-size: 16px !important;
|
| 179 |
+
line-height: 1.6 !important;
|
| 180 |
+
}
|
| 181 |
+
|
| 182 |
+
.rtl-text textarea {
|
| 183 |
+
direction: rtl !important;
|
| 184 |
+
text-align: right !important;
|
| 185 |
+
font-family: 'Vazir', 'Tahoma', 'Arial Unicode MS', sans-serif !important;
|
| 186 |
+
font-size: 16px !important;
|
| 187 |
+
}
|
| 188 |
+
|
| 189 |
+
.rtl-text .output-text {
|
| 190 |
+
direction: rtl !important;
|
| 191 |
+
text-align: right !important;
|
| 192 |
+
font-family: 'Vazir', 'Tahoma', 'Arial Unicode MS', sans-serif !important;
|
| 193 |
+
font-size: 18px !important;
|
| 194 |
+
background-color: #f8f9fa !important;
|
| 195 |
+
padding: 15px !important;
|
| 196 |
+
border-radius: 8px !important;
|
| 197 |
+
border: 1px solid #dee2e6 !important;
|
| 198 |
+
min-height: 100px !important;
|
| 199 |
+
}
|
| 200 |
+
|
| 201 |
+
/* Import Persian fonts */
|
| 202 |
+
@import url('https://cdn.jsdelivr.net/gh/rastikerdar/vazir-font@v30.1.0/dist/font-face.css');
|
| 203 |
+
|
| 204 |
+
.gradio-container {
|
| 205 |
+
max-width: 900px !important;
|
| 206 |
+
margin: auto !important;
|
| 207 |
+
}
|
| 208 |
+
|
| 209 |
+
.title-text {
|
| 210 |
+
text-align: center !important;
|
| 211 |
+
color: #2c3e50 !important;
|
| 212 |
+
margin-bottom: 20px !important;
|
| 213 |
+
}
|
| 214 |
+
|
| 215 |
+
.description-text {
|
| 216 |
+
text-align: center !important;
|
| 217 |
+
color: #7f8c8d !important;
|
| 218 |
+
margin-bottom: 30px !important;
|
| 219 |
+
}
|
| 220 |
+
"""
|
| 221 |
+
|
| 222 |
+
# Sample Persian texts for examples
|
| 223 |
+
sample_texts = [
|
| 224 |
+
"سلام دوست عزیز چطوری؟",
|
| 225 |
+
"من دانشجوی دانشگاه تهران هستم",
|
| 226 |
+
"فردا به کتابخانه خواهم رفت",
|
| 227 |
+
"مادرم برای من غذای خوشمزه درست کرد",
|
| 228 |
+
"زبان فارسی زیبا و شیرین است"
|
| 229 |
+
]
|
| 230 |
+
|
| 231 |
+
# Create Gradio interface
|
| 232 |
+
with gr.Blocks(css=custom_css, title="Persian Vowel Marks Generator") as demo:
|
| 233 |
+
gr.HTML("""
|
| 234 |
+
<div class="title-text">
|
| 235 |
+
<h1>🎯 افزودن اعراب به متن فارسی</h1>
|
| 236 |
+
</div>
|
| 237 |
+
<div class="description-text">
|
| 238 |
+
<p>این ابزار به شما کمک میکند تا اعراب (حرکات) را به متن فارسی اضافه کنید</p>
|
| 239 |
+
<p>Add vowel marks (diacritics) to Persian text</p>
|
| 240 |
+
</div>
|
| 241 |
+
""")
|
| 242 |
+
|
| 243 |
+
with gr.Row():
|
| 244 |
+
with gr.Column(scale=1):
|
| 245 |
+
input_text = gr.Textbox(
|
| 246 |
+
label="📝 متن فارسی (Persian Text Input)",
|
| 247 |
+
placeholder="متن فارسی خود را اینجا وارد کنید...",
|
| 248 |
+
lines=5,
|
| 249 |
+
rtl=True,
|
| 250 |
+
elem_classes=["rtl-text"]
|
| 251 |
+
)
|
| 252 |
+
|
| 253 |
+
with gr.Row():
|
| 254 |
+
submit_btn = gr.Button("🎯 افزودن اعراب", variant="primary", size="lg")
|
| 255 |
+
clear_btn = gr.Button("🗑️ پاک کردن", variant="secondary")
|
| 256 |
+
|
| 257 |
+
with gr.Column(scale=1):
|
| 258 |
+
output_text = gr.Textbox(
|
| 259 |
+
label="✨ متن با اعراب (Text with Vowel Marks)",
|
| 260 |
+
lines=5,
|
| 261 |
+
rtl=True,
|
| 262 |
+
elem_classes=["rtl-text"],
|
| 263 |
+
interactive=False
|
| 264 |
+
)
|
| 265 |
+
|
| 266 |
+
# Examples section
|
| 267 |
+
gr.HTML("<h3 style='text-align: center; margin-top: 30px;'>🔸 نمونه متنها (Sample Texts)</h3>")
|
| 268 |
+
|
| 269 |
+
examples = gr.Examples(
|
| 270 |
+
examples=sample_texts,
|
| 271 |
+
inputs=input_text,
|
| 272 |
+
outputs=output_text,
|
| 273 |
+
fn=vocalize_persian_text,
|
| 274 |
+
cache_examples=False,
|
| 275 |
+
label="نمونههای آماده"
|
| 276 |
+
)
|
| 277 |
+
|
| 278 |
+
# Information section
|
| 279 |
+
with gr.Accordion("📚 راهنما و اطلاعات (Guide & Information)", open=False):
|
| 280 |
+
gr.HTML("""
|
| 281 |
+
<div style="direction: rtl; text-align: right; font-family: Vazir, Tahoma;">
|
| 282 |
+
<h4>🔹 نحوه استفاده:</h4>
|
| 283 |
+
<ul>
|
| 284 |
+
<li>متن فارسی خود را در کادر بالا وارد کنید</li>
|
| 285 |
+
<li>روی دکمه "افزودن اعراب" کلیک کنید</li>
|
| 286 |
+
<li>نتیجه در کادر سمت راست نمایش داده میشود</li>
|
| 287 |
+
</ul>
|
| 288 |
+
|
| 289 |
+
<h4>🔹 ویژگیها:</h4>
|
| 290 |
+
<ul>
|
| 291 |
+
<li>پشتیبانی از متن راست به چپ</li>
|
| 292 |
+
<li>استفاده از فونت فارسی مناسب</li>
|
| 293 |
+
<li>افزودن اعراب بر اساس کلمات رایج</li>
|
| 294 |
+
<li>پشتیبانی از الگوهای گرامری فارسی</li>
|
| 295 |
+
</ul>
|
| 296 |
+
|
| 297 |
+
<h4>🔹 نکات مهم:</h4>
|
| 298 |
+
<ul>
|
| 299 |
+
<li>این سیستم بر اساس قوانین و واژگان رایج کار میکند</li>
|
| 300 |
+
<li>برای دقت بیشتر، از مدلهای یادگیری ماشین پیشرفته استفاده کنید</li>
|
| 301 |
+
<li>اعراب اضافه شده ممکن است نیاز به ویرایش دستی داشته باشد</li>
|
| 302 |
+
</ul>
|
| 303 |
+
</div>
|
| 304 |
+
""")
|
| 305 |
+
|
| 306 |
+
# Event handlers
|
| 307 |
+
submit_btn.click(
|
| 308 |
+
fn=vocalize_persian_text,
|
| 309 |
+
inputs=input_text,
|
| 310 |
+
outputs=output_text
|
| 311 |
+
)
|
| 312 |
+
|
| 313 |
+
clear_btn.click(
|
| 314 |
+
fn=lambda: ("", ""),
|
| 315 |
+
outputs=[input_text, output_text]
|
| 316 |
+
)
|
| 317 |
+
|
| 318 |
+
input_text.submit(
|
| 319 |
+
fn=vocalize_persian_text,
|
| 320 |
+
inputs=input_text,
|
| 321 |
+
outputs=output_text
|
| 322 |
+
)
|
| 323 |
+
|
| 324 |
+
# Launch configuration
|
| 325 |
+
if __name__ == "__main__":
|
| 326 |
+
demo.launch(
|
| 327 |
+
server_name="0.0.0.0",
|
| 328 |
+
server_port=7860,
|
| 329 |
+
share=True,
|
| 330 |
+
debug=True
|
| 331 |
+
)
|