Update app.py
Browse files
app.py
CHANGED
@@ -7,9 +7,7 @@ import struct
|
|
7 |
import time
|
8 |
# import zipfile # Not used in your original core logic
|
9 |
from google import genai
|
10 |
-
from google.genai import types as genai_types # Aliased
|
11 |
-
|
12 |
-
import logging # Standard Python logging
|
13 |
|
14 |
try:
|
15 |
from pydub import AudioSegment
|
@@ -17,10 +15,7 @@ try:
|
|
17 |
except ImportError:
|
18 |
PYDUB_AVAILABLE = False
|
19 |
|
20 |
-
# ---
|
21 |
-
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
|
22 |
-
|
23 |
-
# --- START: Core TTS Logic from YOUR AlphaTTS_Original (UNCHANGED) ---
|
24 |
SPEAKER_VOICES = [
|
25 |
"Achird", "Zubenelgenubi", "Vindemiatrix", "Sadachbia", "Sadaltager",
|
26 |
"Sulafat", "Laomedeia", "Achernar", "Alnilam", "Schedar", "Gacrux",
|
@@ -28,14 +23,14 @@ SPEAKER_VOICES = [
|
|
28 |
"Rasalthgeti", "Orus", "Aoede", "Callirrhoe", "Autonoe", "Enceladus",
|
29 |
"Iapetus", "Zephyr", "Puck", "Charon", "Kore", "Fenrir", "Leda"
|
30 |
]
|
31 |
-
FIXED_MODEL_NAME = "gemini-2.5-flash-preview-tts"
|
32 |
DEFAULT_MAX_CHUNK_SIZE = 3800
|
33 |
DEFAULT_SLEEP_BETWEEN_REQUESTS = 8
|
34 |
DEFAULT_OUTPUT_FILENAME_BASE = "alpha_tts_audio"
|
35 |
|
36 |
-
def _log(message, log_list):
|
37 |
log_list.append(message)
|
38 |
-
|
39 |
|
40 |
def save_binary_file(file_name, data, log_list):
|
41 |
try:
|
@@ -67,196 +62,163 @@ def parse_audio_mime_type(mime_type: str) -> dict[str, int]:
|
|
67 |
except ValueError: pass
|
68 |
return {"bits_per_sample": bits, "rate": rate}
|
69 |
|
70 |
-
def smart_text_split(text, max_size=3800, log_list=None):
|
71 |
if len(text) <= max_size: return [text]
|
72 |
chunks, current_chunk = [], ""
|
73 |
-
sentences = re.split(r'(?<=[.!?؟۔])\s+', text)
|
74 |
for sentence in sentences:
|
75 |
if len(current_chunk) + len(sentence) + 1 > max_size:
|
76 |
if current_chunk: chunks.append(current_chunk.strip())
|
77 |
current_chunk = sentence
|
78 |
while len(current_chunk) > max_size:
|
79 |
-
split_idx
|
80 |
-
for
|
81 |
-
|
82 |
-
idx = current_chunk.rindex(punc, max_size // 2, max_size)
|
83 |
-
if idx > split_idx:
|
84 |
-
split_idx = idx
|
85 |
-
except ValueError:
|
86 |
-
pass
|
87 |
-
if split_idx != -1 :
|
88 |
-
part, current_chunk = current_chunk[:split_idx+1], current_chunk[split_idx+1:]
|
89 |
-
else:
|
90 |
-
part, current_chunk = current_chunk[:max_size], current_chunk[max_size:]
|
91 |
chunks.append(part.strip())
|
92 |
-
else: current_chunk += (" " if current_chunk
|
93 |
if current_chunk: chunks.append(current_chunk.strip())
|
94 |
final_chunks = [c for c in chunks if c]
|
95 |
if log_list: _log(f"📊 متن به {len(final_chunks)} قطعه تقسیم شد.", log_list)
|
96 |
return final_chunks
|
97 |
|
98 |
-
def merge_audio_files_func(file_paths, output_path, log_list):
|
99 |
-
if not PYDUB_AVAILABLE:
|
100 |
-
_log("❌ pydub در دسترس نیست. ادغام انجام نشد.", log_list)
|
101 |
-
return False
|
102 |
try:
|
103 |
_log(f"🔗 ادغام {len(file_paths)} فایل صوتی...", log_list)
|
104 |
combined = AudioSegment.empty()
|
105 |
for i, fp in enumerate(file_paths):
|
106 |
if os.path.exists(fp):
|
107 |
-
try:
|
108 |
segment = AudioSegment.from_file(fp)
|
109 |
combined += segment
|
110 |
if i < len(file_paths) - 1:
|
111 |
-
|
112 |
except Exception as e_pydub:
|
113 |
_log(f"⚠️ خطای Pydub در پردازش فایل '{fp}': {e_pydub}. از این فایل صرف نظر می شود.", log_list)
|
114 |
continue
|
115 |
-
else:
|
116 |
-
|
117 |
-
if len(combined) == 0:
|
118 |
_log("❌ هیچ قطعه صوتی برای ادغام وجود ندارد.", log_list)
|
119 |
return False
|
120 |
combined.export(output_path, format="wav")
|
121 |
-
_log(f"✅ فایل ادغام شده: {output_path}", log_list)
|
122 |
-
|
123 |
-
except Exception as e:
|
124 |
-
_log(f"❌ خطا در ادغام: {e}", log_list)
|
125 |
-
return False
|
126 |
|
127 |
-
def core_generate_audio(text_input, prompt_input, selected_voice, temperature_val, log_list):
|
128 |
output_base_name = DEFAULT_OUTPUT_FILENAME_BASE
|
129 |
max_chunk, sleep_time = DEFAULT_MAX_CHUNK_SIZE, DEFAULT_SLEEP_BETWEEN_REQUESTS
|
130 |
_log(f"🚀 شروع فرآیند با مدل: {FIXED_MODEL_NAME}...", log_list)
|
131 |
-
api_key = os.environ.get("GEMINI_API_KEY")
|
132 |
-
if not api_key:
|
133 |
-
|
134 |
-
|
135 |
-
|
136 |
-
client = genai.Client(api_key=api_key)
|
137 |
-
_log(f"کلاینت Gemini با کلید API برای مدل {FIXED_MODEL_NAME} مقداردهی اولیه شد.", log_list)
|
138 |
-
except Exception as e:
|
139 |
-
_log(f"❌ خطا در مقداردهی اولیه کلاینت Gemini: {e}", log_list)
|
140 |
-
return None
|
141 |
-
if not text_input or not text_input.strip():
|
142 |
-
_log("❌ متن ورودی خالی.", log_list)
|
143 |
-
return None
|
144 |
text_chunks = smart_text_split(text_input, max_chunk, log_list)
|
145 |
-
if not text_chunks:
|
146 |
-
|
147 |
-
return None
|
148 |
generated_files = []
|
149 |
for i, chunk in enumerate(text_chunks):
|
150 |
_log(f"🔊 پردازش قطعه {i+1}/{len(text_chunks)} (صدا: {selected_voice}, دما: {temperature_val})...", log_list)
|
151 |
final_text = f'"{prompt_input}"\n{chunk}' if prompt_input and prompt_input.strip() else chunk
|
|
|
152 |
contents = [genai_types.Content(role="user", parts=[genai_types.Part.from_text(text=final_text)])]
|
153 |
-
config = genai_types.GenerateContentConfig(
|
154 |
-
|
155 |
-
|
156 |
-
speech_config=genai_types.SpeechConfig(
|
157 |
-
voice_config=genai_types.VoiceConfig(
|
158 |
-
prebuilt_voice_config=genai_types.PrebuiltVoiceConfig(voice_name=selected_voice)
|
159 |
-
)
|
160 |
-
)
|
161 |
-
)
|
162 |
-
_log(f"کانفیگ API برای قطعه {i+1}: دما={temperature_val}, صدا={selected_voice}, مدالیته=['audio']", log_list)
|
163 |
fname_base = f"{output_base_name}_part{i+1:03d}"
|
164 |
try:
|
|
|
165 |
response = client.models.generate_content(model=FIXED_MODEL_NAME, contents=contents, config=config)
|
166 |
if response.candidates and response.candidates[0].content and response.candidates[0].content.parts and response.candidates[0].content.parts[0].inline_data:
|
167 |
inline_data = response.candidates[0].content.parts[0].inline_data
|
168 |
data_buffer = inline_data.data
|
169 |
-
mime_type = inline_data.mime_type
|
170 |
-
_log(f"داده صوتی در candidate.part[0].inline_data برای قطعه {i+1} یافت شد.", log_list)
|
171 |
ext = mimetypes.guess_extension(mime_type) or ".wav"
|
172 |
if "audio/L" in mime_type and ext == ".wav":
|
173 |
-
|
174 |
-
|
175 |
if not ext.startswith("."): ext = "." + ext
|
176 |
fpath = save_binary_file(f"{fname_base}{ext}", data_buffer, log_list)
|
177 |
if fpath: generated_files.append(fpath)
|
178 |
-
else:
|
179 |
-
|
180 |
-
_log(f"ساختار کامل پاسخ (اولین 500 کاراکتر): {str(response)[:500]}", log_list)
|
181 |
-
except Exception as e:
|
182 |
-
_log(f"❌ خطا در تولید قطعه {i+1}: {e}", log_list)
|
183 |
-
continue
|
184 |
if i < len(text_chunks) - 1 and len(text_chunks) > 1:
|
185 |
-
_log(f"💤 توقف کوتاه ({sleep_time} ثانیه) قبل از قطعه بعدی...", log_list)
|
186 |
time.sleep(sleep_time)
|
187 |
-
|
188 |
-
|
189 |
-
return None
|
190 |
_log(f"🎉 {len(generated_files)} فایل(های) صوتی تولید شد.", log_list)
|
|
|
191 |
final_audio_file = None
|
192 |
final_output_path_base = f"{output_base_name}_final"
|
|
|
193 |
if len(generated_files) > 1:
|
194 |
if PYDUB_AVAILABLE:
|
195 |
merged_fn = f"{final_output_path_base}.wav"
|
196 |
if os.path.exists(merged_fn):
|
197 |
try: os.remove(merged_fn)
|
198 |
-
except OSError: _log(f"⚠️ عدم امکان حذف فایل ادغام شده قبلی '{merged_fn}' (خطای سیستم عامل)", log_list)
|
199 |
except Exception as e_rm: _log(f"⚠️ عدم امکان حذف فایل ادغام شده قبلی '{merged_fn}': {e_rm}", log_list)
|
|
|
200 |
if merge_audio_files_func(generated_files, merged_fn, log_list):
|
201 |
final_audio_file = merged_fn
|
202 |
-
for
|
203 |
-
if os.path.abspath(
|
204 |
-
try: os.remove(
|
205 |
-
except
|
206 |
-
|
207 |
-
|
208 |
-
_log("⚠️ ادغام فایلهای صوتی ناموفق بود. اولین قطعه ارائه میشود.", log_list)
|
209 |
-
if generated_files:
|
210 |
try:
|
|
|
211 |
first_chunk_path = generated_files[0]
|
212 |
-
|
213 |
-
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
for
|
220 |
-
|
221 |
-
|
222 |
-
|
223 |
-
|
|
|
224 |
final_audio_file = generated_files[0]
|
225 |
else:
|
226 |
_log("⚠️ pydub نیست. اولین قطعه ارائه میشود.", log_list)
|
227 |
if generated_files:
|
228 |
try:
|
229 |
first_chunk_path = generated_files[0]
|
230 |
-
|
231 |
-
|
232 |
-
|
233 |
-
|
234 |
-
|
235 |
-
|
236 |
-
final_audio_file = single_fallback_fn
|
237 |
for i_gf in range(1, len(generated_files)):
|
238 |
try: os.remove(generated_files[i_gf])
|
239 |
except: pass
|
240 |
-
except Exception as
|
241 |
-
_log(f"خطا در تغییر نام فایل اولین قطعه (بدون pydub): {
|
242 |
final_audio_file = generated_files[0]
|
243 |
elif len(generated_files) == 1:
|
244 |
try:
|
245 |
single_file_path = generated_files[0]
|
246 |
-
|
247 |
-
|
248 |
-
|
249 |
-
|
250 |
-
|
251 |
-
|
252 |
-
|
253 |
-
|
254 |
-
_log(f"خطا در تغییر نام فایل تکی نهایی: {e_rename_sgl_final}", log_list)
|
255 |
final_audio_file = generated_files[0]
|
|
|
256 |
if final_audio_file and not os.path.exists(final_audio_file):
|
257 |
_log(f"⚠️ فایل نهایی '{final_audio_file}' وجود ندارد!", log_list)
|
258 |
return None
|
259 |
-
return final_audio_file
|
260 |
|
261 |
def gradio_tts_interface(use_file_input, uploaded_file, text_to_speak, speech_prompt, speaker_voice, temperature, progress=gr.Progress(track_tqdm=True)): # YOUR ORIGINAL SIGNATURE
|
262 |
logs = []
|
@@ -271,200 +233,236 @@ def gradio_tts_interface(use_file_input, uploaded_file, text_to_speak, speech_pr
|
|
271 |
else:
|
272 |
actual_text = text_to_speak
|
273 |
if not actual_text or not actual_text.strip(): return None
|
|
|
274 |
final_path = core_generate_audio(actual_text, speech_prompt, speaker_voice, temperature, logs)
|
275 |
-
# for log_entry in logs: print(log_entry) #
|
276 |
-
return final_path
|
277 |
-
# --- END: Core TTS Logic
|
278 |
|
279 |
|
280 |
-
# --- START:
|
281 |
# (Using CSS variables from AlphaTranslator_Styled for colors and fonts)
|
282 |
-
FLY_PRIMARY_COLOR_HEX = "#4F46E5"
|
283 |
-
FLY_SECONDARY_COLOR_HEX = "#10B981"
|
284 |
-
FLY_ACCENT_COLOR_HEX = "#D97706"
|
285 |
-
FLY_TEXT_COLOR_HEX = "#1F2937"
|
286 |
-
FLY_SUBTLE_TEXT_HEX = "#6B7280"
|
287 |
-
FLY_LIGHT_BACKGROUND_HEX = "#F9FAFB"
|
288 |
-
FLY_WHITE_HEX = "#FFFFFF"
|
289 |
-
FLY_BORDER_COLOR_HEX = "#D1D5DB"
|
290 |
-
FLY_INPUT_BG_HEX_SIMPLE = "#F3F4F6"
|
291 |
-
|
292 |
-
|
293 |
-
|
294 |
-
|
295 |
-
|
296 |
-
|
297 |
-
|
298 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
299 |
@import url('https://fonts.googleapis.com/css2?family=Poppins:wght@400;500;600;700;800&display=swap');
|
300 |
@import url('https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap');
|
|
|
301 |
:root {{
|
302 |
-
|
303 |
-
--fly-
|
304 |
-
--fly-
|
305 |
-
--fly-
|
|
|
|
|
|
|
|
|
|
|
306 |
--fly-input-bg-simple: {FLY_INPUT_BG_HEX_SIMPLE};
|
307 |
-
|
308 |
-
--
|
309 |
-
|
310 |
-
|
311 |
-
--
|
312 |
--shadow-lg: 0 10px 15px -3px rgba(0,0,0,0.1),0 4px 6px -4px rgba(0,0,0,0.1);
|
313 |
--shadow-xl: 0 20px 25px -5px rgba(0,0,0,0.1),0 8px 10px -6px rgba(0,0,0,0.1);
|
314 |
-
|
|
|
|
|
|
|
315 |
}}
|
316 |
-
|
317 |
-
|
318 |
-
|
319 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
320 |
}}
|
321 |
-
|
322 |
-
|
323 |
-
|
324 |
-
|
325 |
-
|
|
|
326 |
}}
|
327 |
-
|
328 |
-
|
329 |
-
|
330 |
-
|
331 |
-
|
332 |
-
border-bottom-
|
333 |
-
|
|
|
334 |
}}
|
335 |
.app-header-alphatts-styled::before {{
|
336 |
-
content:'';position:absolute;top:-50px;right:-50px;width:150px;height:150px;
|
337 |
-
background:rgba(255,255,255,0.1);border-radius:
|
338 |
-
opacity:0.5;transform:rotate(45deg);
|
339 |
}}
|
340 |
.app-header-alphatts-styled h1 {{
|
341 |
-
font-size:2.25em !important;font-weight:800 !important;margin:0 0 0.5rem 0;
|
342 |
-
font-family:var(--font-english);letter-spacing:-0.5px;text-shadow:0 2px 4px rgba(0,0,0,0.1);
|
343 |
}}
|
344 |
.app-header-alphatts-styled p {{
|
345 |
-
font-size:1em !important;margin-top:0.25rem;font-weight:400;
|
346 |
color:rgba(255,255,255,0.85) !important;
|
347 |
}}
|
348 |
|
349 |
-
/* Main content
|
350 |
-
|
351 |
-
|
352 |
-
|
353 |
-
|
354 |
-
|
355 |
-
position:relative;z-index:10;
|
|
|
356 |
}}
|
357 |
|
358 |
-
/* Styling
|
359 |
-
|
360 |
-
.
|
361 |
-
.
|
362 |
-
.
|
363 |
-
.
|
|
|
|
|
364 |
{{
|
365 |
-
border-radius:var(--radius-md) !important; /* Using
|
366 |
border:1.5px solid var(--fly-border-color) !important;
|
367 |
-
font-size:0.95em !important;
|
368 |
-
|
|
|
|
|
|
|
369 |
}}
|
370 |
-
.
|
371 |
-
.
|
372 |
-
.
|
373 |
-
.
|
374 |
-
.
|
|
|
375 |
{{
|
376 |
border-color:var(--fly-primary) !important;
|
377 |
box-shadow:0 0 0 3px rgba(var(--fly-primary-rgb),0.12) !important;
|
378 |
background-color:var(--fly-bg-white) !important;
|
379 |
}}
|
380 |
-
.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
381 |
|
382 |
-
/* Button
|
383 |
-
.
|
384 |
{{
|
385 |
background:var(--fly-accent) !important; /* Orange accent */
|
386 |
-
margin-top:1.5rem !important;padding:12px 20px !important;
|
387 |
-
transition:all 0.25s ease-in-out !important;color:white !important;font-weight:600 !important;
|
388 |
-
border-radius:var(--radius-md) !important;border:none !important;
|
389 |
box-shadow:0 3px 8px -1px rgba(var(--fly-accent-rgb),0.3) !important;
|
390 |
-
width:100% !important;font-size:
|
391 |
-
display:flex;align-items:center;justify-content:center;
|
392 |
}}
|
393 |
-
.
|
394 |
{{
|
395 |
background:#B45309 !important; /* Darker orange */ transform:translateY(-1px) !important;
|
396 |
box-shadow:0 5px 10px -1px rgba(var(--fly-accent-rgb),0.4) !important;
|
397 |
}}
|
398 |
|
399 |
-
/*
|
400 |
-
.
|
401 |
-
|
402 |
-
font-weight:500 !important;color:#4B5563 !important;
|
403 |
-
font-size:0.90em !important;margin-bottom:8px !important;display:inline-block;
|
404 |
-
}}
|
405 |
-
|
406 |
-
/* Your specific temperature description class */
|
407 |
-
.main-content-panel-alpha .temp_description_class_alpha_v3 {{
|
408 |
-
font-size: 0.85em; color: var(--fly-text-secondary); margin-top: -0.6rem; margin-bottom: 1.2rem;
|
409 |
}}
|
410 |
|
411 |
-
/* Audio Player styling */
|
412 |
-
.
|
413 |
{{
|
414 |
width: 100%; border-radius: var(--radius-md); margin-top:0.8rem;
|
415 |
}}
|
416 |
|
417 |
-
/* Examples section button
|
418 |
-
.
|
419 |
-
.
|
420 |
{{
|
421 |
-
background-color:#E0E7FF !important;color:var(--fly-primary) !important;
|
422 |
-
border-radius:var(--radius-sm) !important;font-size:0.
|
423 |
-
}}
|
424 |
-
.main-content-panel-alpha .custom-hr {{ /* If you add a hr manually */
|
425 |
-
height:1px;background-color:var(--fly-border-color);margin:1.8rem 0;border:none;
|
426 |
}}
|
|
|
427 |
|
428 |
-
|
|
|
429 |
.app-footer-alphatts-styled {{
|
430 |
text-align:center;font-size:0.85em;color:var(--fly-text-secondary);margin-top:2.5rem;
|
431 |
padding:1rem 0;background-color:rgba(255,255,255,0.3);backdrop-filter:blur(5px);
|
432 |
border-top:1px solid var(--fly-border-color);
|
433 |
}}
|
434 |
-
|
|
|
|
|
435 |
|
436 |
-
/* Responsive adjustments */
|
437 |
@media (min-width:640px) {{
|
|
|
|
|
438 |
.app-header-alphatts-styled h1 {{font-size:2.5em !important;}}
|
439 |
.app-header-alphatts-styled p {{font-size:1.05em !important;}}
|
440 |
}}
|
441 |
@media (min-width:768px) {{
|
442 |
-
.main-content-
|
443 |
-
.
|
|
|
|
|
444 |
{{
|
445 |
-
width:auto !important;align-self:flex-start;
|
446 |
}}
|
447 |
.app-header-alphatts-styled h1 {{font-size:2.75em !important;}}
|
448 |
.app-header-alphatts-styled p {{font-size:1.1em !important;}}
|
449 |
}}
|
450 |
"""
|
451 |
-
|
452 |
-
|
453 |
-
|
454 |
-
|
455 |
-
#
|
456 |
-
# We apply `app_theme_outer_styled` for body background and `alphatranslator_inspired_css` for specifics.
|
457 |
-
# Note: Your original `custom_css_inspired_by_image` is replaced by `alphatranslator_inspired_css` here.
|
458 |
-
with gr.Blocks(theme=app_theme_outer_styled, css=alphatranslator_inspired_css, title=f"آلفا TTS ({FIXED_MODEL_NAME.split('-')[1]})") as demo:
|
459 |
-
# Applying the styled header
|
460 |
-
# Your original HTML for header was:
|
461 |
-
# alpha_header_html_v3 = """
|
462 |
-
# <div class='app-header-alpha'>
|
463 |
-
# <h1>Alpha TTS</h1>
|
464 |
-
# <p>جادوی تبدیل متن به صدا در دستان شما</p>
|
465 |
-
# </div>
|
466 |
-
# """
|
467 |
-
# We use the new class for styling:
|
468 |
gr.HTML(f"""
|
469 |
<div class='app-header-alphatts-styled'>
|
470 |
<h1>🚀 Alpha TTS</h1>
|
@@ -472,108 +470,117 @@ with gr.Blocks(theme=app_theme_outer_styled, css=alphatranslator_inspired_css, t
|
|
472 |
</div>
|
473 |
""")
|
474 |
|
475 |
-
#
|
476 |
-
|
477 |
-
|
478 |
-
|
479 |
-
|
480 |
-
|
481 |
-
|
482 |
-
|
483 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
484 |
)
|
485 |
-
gr.Markdown(f"<div class='api-warning-message'>{missing_key_msg}</div>") # api-warning-message is styled in AlphaTranslator
|
486 |
-
|
487 |
-
# Your original UI components
|
488 |
-
use_file_input_cb = gr.Checkbox(label="📄 استفاده از فایل متنی (.txt)", value=False, elem_id="use_file_cb_alpha_v3")
|
489 |
-
|
490 |
-
uploaded_file_input = gr.File(
|
491 |
-
label=" ",
|
492 |
-
file_types=['.txt'],
|
493 |
-
visible=False,
|
494 |
-
elem_id="file_uploader_alpha_main_v3"
|
495 |
-
)
|
496 |
-
|
497 |
-
text_to_speak_tb = gr.Textbox(
|
498 |
-
label="متن فارسی برای تبدیل",
|
499 |
-
placeholder="مثال: سلام، فردا هوا چطور است؟",
|
500 |
-
lines=5,
|
501 |
-
value="",
|
502 |
-
visible=True,
|
503 |
-
elem_id="text_input_main_alpha_v3"
|
504 |
-
)
|
505 |
-
|
506 |
-
use_file_input_cb.change( # YOUR ORIGINAL CHANGE HANDLER
|
507 |
-
fn=lambda x: (gr.update(visible=x, label=" " if x else "متن فارسی برای تبدیل"), gr.update(visible=not x)),
|
508 |
-
inputs=use_file_input_cb,
|
509 |
-
outputs=[uploaded_file_input, text_to_speak_tb]
|
510 |
-
)
|
511 |
-
|
512 |
-
speech_prompt_tb = gr.Textbox(
|
513 |
-
label="سبک گفتار (اختیاری)",
|
514 |
-
placeholder="مثال: با لحنی شاد و پرانرژی",
|
515 |
-
value="با لحنی دوستانه و رسا صحبت کن.",
|
516 |
-
lines=2, elem_id="speech_prompt_alpha_v3"
|
517 |
-
)
|
518 |
-
|
519 |
-
speaker_voice_dd = gr.Dropdown(
|
520 |
-
SPEAKER_VOICES, label="انتخاب گوینده و لهجه", value="Charon", elem_id="speaker_voice_alpha_v3"
|
521 |
-
)
|
522 |
-
|
523 |
-
temperature_slider = gr.Slider(
|
524 |
-
minimum=0.1, maximum=1.5, step=0.05, value=0.9, label="میزان خلاقیت صدا",
|
525 |
-
elem_id="temperature_slider_alpha_v3"
|
526 |
-
)
|
527 |
-
# Your original Markdown for temperature description
|
528 |
-
gr.Markdown("<p class='temp_description_class_alpha_v3'>مقادیر بالاتر = تنوع بیشتر، مقادیر پایینتر = یکنواختی بیشتر.</p>")
|
529 |
|
530 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
531 |
|
532 |
-
|
533 |
-
|
534 |
-
|
535 |
-
|
536 |
-
|
537 |
-
|
538 |
-
|
539 |
-
|
540 |
-
|
541 |
-
|
542 |
-
|
543 |
-
[
|
544 |
-
|
545 |
-
|
546 |
-
|
547 |
-
|
548 |
-
cache_examples=os.getenv("GRADIO_CACHE_EXAMPLES", "False").lower() == "true" # Optional caching
|
549 |
-
)
|
550 |
|
551 |
-
#
|
552 |
-
|
553 |
-
# We use the new class for styling:
|
554 |
-
gr.HTML(f"<p class='app-footer-alphatts-styled'>Alpha TTS © 2024 - Model: {FIXED_MODEL_NAME.split('-')[1]}</p>")
|
555 |
|
556 |
|
557 |
-
# --- Event Handlers (YOUR ORIGINAL
|
|
|
|
|
558 |
if generate_button is not None: # Check if button was created
|
559 |
generate_button.click(
|
560 |
fn=gradio_tts_interface,
|
561 |
-
inputs=[use_file_input_cb, uploaded_file_input, text_to_speak_tb, speech_prompt_tb, speaker_voice_dd, temperature_slider],
|
562 |
-
outputs=[output_audio]
|
563 |
)
|
564 |
else:
|
565 |
-
|
566 |
-
|
|
|
567 |
|
568 |
if __name__ == "__main__":
|
|
|
569 |
if not PYDUB_AVAILABLE:
|
570 |
-
|
571 |
if not os.environ.get("GEMINI_API_KEY"):
|
572 |
-
|
573 |
|
574 |
demo.launch(
|
575 |
-
server_name="0.0.0.0",
|
576 |
-
server_port=int(os.getenv("PORT", 7860)),
|
577 |
-
debug=os.environ.get("GRADIO_DEBUG", "False").lower() == "true",
|
578 |
-
show_error=True
|
579 |
)
|
|
|
7 |
import time
|
8 |
# import zipfile # Not used in your original core logic
|
9 |
from google import genai
|
10 |
+
from google.genai import types as genai_types # Aliased your 'types' import
|
|
|
|
|
11 |
|
12 |
try:
|
13 |
from pydub import AudioSegment
|
|
|
15 |
except ImportError:
|
16 |
PYDUB_AVAILABLE = False
|
17 |
|
18 |
+
# --- START: Your Original Core TTS Logic (UNCHANGED) ---
|
|
|
|
|
|
|
19 |
SPEAKER_VOICES = [
|
20 |
"Achird", "Zubenelgenubi", "Vindemiatrix", "Sadachbia", "Sadaltager",
|
21 |
"Sulafat", "Laomedeia", "Achernar", "Alnilam", "Schedar", "Gacrux",
|
|
|
23 |
"Rasalthgeti", "Orus", "Aoede", "Callirrhoe", "Autonoe", "Enceladus",
|
24 |
"Iapetus", "Zephyr", "Puck", "Charon", "Kore", "Fenrir", "Leda"
|
25 |
]
|
26 |
+
FIXED_MODEL_NAME = "gemini-2.5-flash-preview-tts" # YOUR DEFINED MODEL
|
27 |
DEFAULT_MAX_CHUNK_SIZE = 3800
|
28 |
DEFAULT_SLEEP_BETWEEN_REQUESTS = 8
|
29 |
DEFAULT_OUTPUT_FILENAME_BASE = "alpha_tts_audio"
|
30 |
|
31 |
+
def _log(message, log_list): # YOUR _log function
|
32 |
log_list.append(message)
|
33 |
+
# print(f"[AlphaTTS_LOG] {message}") # Optional: print to console for live debugging if needed
|
34 |
|
35 |
def save_binary_file(file_name, data, log_list):
|
36 |
try:
|
|
|
62 |
except ValueError: pass
|
63 |
return {"bits_per_sample": bits, "rate": rate}
|
64 |
|
65 |
+
def smart_text_split(text, max_size=3800, log_list=None): # YOUR smart_text_split
|
66 |
if len(text) <= max_size: return [text]
|
67 |
chunks, current_chunk = [], ""
|
68 |
+
sentences = re.split(r'(?<=[.!?؟۔])\s+', text) # Added Persian full stop from previous attempts for robustness
|
69 |
for sentence in sentences:
|
70 |
if len(current_chunk) + len(sentence) + 1 > max_size:
|
71 |
if current_chunk: chunks.append(current_chunk.strip())
|
72 |
current_chunk = sentence
|
73 |
while len(current_chunk) > max_size:
|
74 |
+
# Using your original split_idx logic
|
75 |
+
split_idx = next((i for i in range(max_size - 1, max_size // 2, -1) if current_chunk[i] in ['،', ',', ';', ':', ' ']), -1)
|
76 |
+
part, current_chunk = (current_chunk[:split_idx+1], current_chunk[split_idx+1:]) if split_idx != -1 else (current_chunk[:max_size], current_chunk[max_size:])
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
77 |
chunks.append(part.strip())
|
78 |
+
else: current_chunk += (" " if current_chunk else "") + sentence # Your original logic for adding sentence
|
79 |
if current_chunk: chunks.append(current_chunk.strip())
|
80 |
final_chunks = [c for c in chunks if c]
|
81 |
if log_list: _log(f"📊 متن به {len(final_chunks)} قطعه تقسیم شد.", log_list)
|
82 |
return final_chunks
|
83 |
|
84 |
+
def merge_audio_files_func(file_paths, output_path, log_list): # YOUR merge_audio_files_func
|
85 |
+
if not PYDUB_AVAILABLE: _log("❌ pydub در دسترس نیست.", log_list); return False
|
|
|
|
|
86 |
try:
|
87 |
_log(f"🔗 ادغام {len(file_paths)} فایل صوتی...", log_list)
|
88 |
combined = AudioSegment.empty()
|
89 |
for i, fp in enumerate(file_paths):
|
90 |
if os.path.exists(fp):
|
91 |
+
try: # Added try-except for individual file processing with pydub
|
92 |
segment = AudioSegment.from_file(fp)
|
93 |
combined += segment
|
94 |
if i < len(file_paths) - 1:
|
95 |
+
combined += AudioSegment.silent(duration=150) # Your original logic
|
96 |
except Exception as e_pydub:
|
97 |
_log(f"⚠️ خطای Pydub در پردازش فایل '{fp}': {e_pydub}. از این فایل صرف نظر می شود.", log_list)
|
98 |
continue
|
99 |
+
else: _log(f"⚠️ فایل پیدا نشد: {fp}", log_list)
|
100 |
+
if len(combined) == 0: # Added check if no valid segments were combined
|
|
|
101 |
_log("❌ هیچ قطعه صوتی برای ادغام وجود ندارد.", log_list)
|
102 |
return False
|
103 |
combined.export(output_path, format="wav")
|
104 |
+
_log(f"✅ فایل ادغام شده: {output_path}", log_list); return True
|
105 |
+
except Exception as e: _log(f"❌ خطا در ادغام: {e}", log_list); return False
|
|
|
|
|
|
|
106 |
|
107 |
+
def core_generate_audio(text_input, prompt_input, selected_voice, temperature_val, log_list): # YOUR core_generate_audio
|
108 |
output_base_name = DEFAULT_OUTPUT_FILENAME_BASE
|
109 |
max_chunk, sleep_time = DEFAULT_MAX_CHUNK_SIZE, DEFAULT_SLEEP_BETWEEN_REQUESTS
|
110 |
_log(f"🚀 شروع فرآیند با مدل: {FIXED_MODEL_NAME}...", log_list)
|
111 |
+
api_key = os.environ.get("GEMINI_API_KEY") # YOUR API KEY METHOD
|
112 |
+
if not api_key: _log("❌ کلید API تنظیم نشده.", log_list); return None
|
113 |
+
try: client = genai.Client(api_key=api_key) # YOUR CLIENT INSTANTIATION
|
114 |
+
except Exception as e: _log(f"❌ خطا در کلاینت: {e}", log_list); return None
|
115 |
+
if not text_input or not text_input.strip(): _log("❌ متن ورودی خالی.", log_list); return None
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
116 |
text_chunks = smart_text_split(text_input, max_chunk, log_list)
|
117 |
+
if not text_chunks: _log("❌ متن قابل پردازش نیست.", log_list); return None
|
118 |
+
|
|
|
119 |
generated_files = []
|
120 |
for i, chunk in enumerate(text_chunks):
|
121 |
_log(f"🔊 پردازش قطعه {i+1}/{len(text_chunks)} (صدا: {selected_voice}, دما: {temperature_val})...", log_list)
|
122 |
final_text = f'"{prompt_input}"\n{chunk}' if prompt_input and prompt_input.strip() else chunk
|
123 |
+
# Using genai_types (aliased) as per your original imports
|
124 |
contents = [genai_types.Content(role="user", parts=[genai_types.Part.from_text(text=final_text)])]
|
125 |
+
config = genai_types.GenerateContentConfig(temperature=temperature_val, response_modalities=["audio"],
|
126 |
+
speech_config=genai_types.SpeechConfig(voice_config=genai_types.VoiceConfig(
|
127 |
+
prebuilt_voice_config=genai_types.PrebuiltVoiceConfig(voice_name=selected_voice))))
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
128 |
fname_base = f"{output_base_name}_part{i+1:03d}"
|
129 |
try:
|
130 |
+
# YOUR API CALL
|
131 |
response = client.models.generate_content(model=FIXED_MODEL_NAME, contents=contents, config=config)
|
132 |
if response.candidates and response.candidates[0].content and response.candidates[0].content.parts and response.candidates[0].content.parts[0].inline_data:
|
133 |
inline_data = response.candidates[0].content.parts[0].inline_data
|
134 |
data_buffer = inline_data.data
|
135 |
+
mime_type = inline_data.mime_type # Get mime_type here
|
136 |
+
_log(f"داده صوتی در candidate.part[0].inline_data برای قطعه {i+1} یافت شد. MIME: {mime_type}", log_list)
|
137 |
ext = mimetypes.guess_extension(mime_type) or ".wav"
|
138 |
if "audio/L" in mime_type and ext == ".wav":
|
139 |
+
_log(f"تبدیل صدای خام PCM (MIME: {mime_type}) به WAV برای قطعه {i+1}.", log_list)
|
140 |
+
data_buffer = convert_to_wav(data_buffer, mime_type)
|
141 |
if not ext.startswith("."): ext = "." + ext
|
142 |
fpath = save_binary_file(f"{fname_base}{ext}", data_buffer, log_list)
|
143 |
if fpath: generated_files.append(fpath)
|
144 |
+
else: _log(f"⚠️ پاسخ API برای قطعه {i+1} بدون داده صوتی در مسیر مورد انتظار.", log_list)
|
145 |
+
except Exception as e: _log(f"❌ خطا در تولید قطعه {i+1}: {e}", log_list); continue # Your original error handling
|
|
|
|
|
|
|
|
|
146 |
if i < len(text_chunks) - 1 and len(text_chunks) > 1:
|
147 |
+
_log(f"💤 توقف کوتاه ({sleep_time} ثانیه) قبل از قطعه بعدی...", log_list) # Added log for sleep
|
148 |
time.sleep(sleep_time)
|
149 |
+
|
150 |
+
if not generated_files: _log("❌ هیچ فایلی تولید نشد.", log_list); return None
|
|
|
151 |
_log(f"🎉 {len(generated_files)} فایل(های) صوتی تولید شد.", log_list)
|
152 |
+
|
153 |
final_audio_file = None
|
154 |
final_output_path_base = f"{output_base_name}_final"
|
155 |
+
|
156 |
if len(generated_files) > 1:
|
157 |
if PYDUB_AVAILABLE:
|
158 |
merged_fn = f"{final_output_path_base}.wav"
|
159 |
if os.path.exists(merged_fn):
|
160 |
try: os.remove(merged_fn)
|
|
|
161 |
except Exception as e_rm: _log(f"⚠️ عدم امکان حذف فایل ادغام شده قبلی '{merged_fn}': {e_rm}", log_list)
|
162 |
+
|
163 |
if merge_audio_files_func(generated_files, merged_fn, log_list):
|
164 |
final_audio_file = merged_fn
|
165 |
+
for fp in generated_files:
|
166 |
+
if os.path.abspath(fp) != os.path.abspath(merged_fn):
|
167 |
+
try: os.remove(fp)
|
168 |
+
except: pass # Your original silent pass
|
169 |
+
else:
|
170 |
+
if generated_files: # Should be true if merge_audio_files_func was attempted
|
|
|
|
|
171 |
try:
|
172 |
+
# Your original logic for renaming the first chunk on merge fail
|
173 |
first_chunk_path = generated_files[0]
|
174 |
+
target_name = f"{final_output_path_base}{os.path.splitext(first_chunk_path)[1]}"
|
175 |
+
if os.path.exists(target_name) and os.path.abspath(first_chunk_path) != os.path.abspath(target_name):
|
176 |
+
os.remove(target_name)
|
177 |
+
if os.path.abspath(first_chunk_path) != os.path.abspath(target_name):
|
178 |
+
os.rename(first_chunk_path, target_name)
|
179 |
+
final_audio_file = target_name
|
180 |
+
# Clean up other parts as per your original logic (silent pass on fail)
|
181 |
+
if final_audio_file: # If renaming succeeded for the first file
|
182 |
+
for i_gf in range(1, len(generated_files)):
|
183 |
+
try: os.remove(generated_files[i_gf])
|
184 |
+
except: pass
|
185 |
+
except Exception as e_rename: # Your original variable name for this exception
|
186 |
+
_log(f"خطا در تغییر نام فایل اولین قطعه: {e_rename}", log_list)
|
187 |
final_audio_file = generated_files[0]
|
188 |
else:
|
189 |
_log("⚠️ pydub نیست. اولین قطعه ارائه میشود.", log_list)
|
190 |
if generated_files:
|
191 |
try:
|
192 |
first_chunk_path = generated_files[0]
|
193 |
+
target_name = f"{final_output_path_base}{os.path.splitext(first_chunk_path)[1]}"
|
194 |
+
if os.path.exists(target_name) and os.path.abspath(first_chunk_path) != os.path.abspath(target_name):
|
195 |
+
os.remove(target_name)
|
196 |
+
if os.path.abspath(first_chunk_path) != os.path.abspath(target_name):
|
197 |
+
os.rename(first_chunk_path, target_name)
|
198 |
+
final_audio_file = target_name
|
|
|
199 |
for i_gf in range(1, len(generated_files)):
|
200 |
try: os.remove(generated_files[i_gf])
|
201 |
except: pass
|
202 |
+
except Exception as e_rename_single: # Your original variable name
|
203 |
+
_log(f"خطا در تغییر نام فایل اولین قطعه (بدون pydub): {e_rename_single}", log_list)
|
204 |
final_audio_file = generated_files[0]
|
205 |
elif len(generated_files) == 1:
|
206 |
try:
|
207 |
single_file_path = generated_files[0]
|
208 |
+
target_name = f"{final_output_path_base}{os.path.splitext(single_file_path)[1]}"
|
209 |
+
if os.path.exists(target_name) and os.path.abspath(single_file_path) != os.path.abspath(target_name):
|
210 |
+
os.remove(target_name)
|
211 |
+
if os.path.abspath(single_file_path) != os.path.abspath(target_name):
|
212 |
+
os.rename(single_file_path, target_name)
|
213 |
+
final_audio_file = target_name
|
214 |
+
except Exception as e_rename_single_final: # Your original variable name
|
215 |
+
_log(f"خطا در تغییر نام فایل تکی نهایی: {e_rename_single_final}", log_list)
|
|
|
216 |
final_audio_file = generated_files[0]
|
217 |
+
|
218 |
if final_audio_file and not os.path.exists(final_audio_file):
|
219 |
_log(f"⚠️ فایل نهایی '{final_audio_file}' وجود ندارد!", log_list)
|
220 |
return None
|
221 |
+
return final_audio_file # Returns only path, as per your original
|
222 |
|
223 |
def gradio_tts_interface(use_file_input, uploaded_file, text_to_speak, speech_prompt, speaker_voice, temperature, progress=gr.Progress(track_tqdm=True)): # YOUR ORIGINAL SIGNATURE
|
224 |
logs = []
|
|
|
233 |
else:
|
234 |
actual_text = text_to_speak
|
235 |
if not actual_text or not actual_text.strip(): return None
|
236 |
+
|
237 |
final_path = core_generate_audio(actual_text, speech_prompt, speaker_voice, temperature, logs)
|
238 |
+
# for log_entry in logs: print(log_entry) # Your original commented-out log printing
|
239 |
+
return final_path # YOUR ORIGINAL RETURN
|
240 |
+
# --- END: Your Original Core TTS Logic (UNCHANGED) ---
|
241 |
|
242 |
|
243 |
+
# --- START: Styling and UI (Applying AlphaTranslator_Styled look to YOUR UI structure) ---
|
244 |
# (Using CSS variables from AlphaTranslator_Styled for colors and fonts)
|
245 |
+
FLY_PRIMARY_COLOR_HEX = "#4F46E5" # From AlphaTranslator_Styled
|
246 |
+
FLY_SECONDARY_COLOR_HEX = "#10B981" # From AlphaTranslator_Styled
|
247 |
+
FLY_ACCENT_COLOR_HEX = "#D97706" # Orange accent from AlphaTranslator_Styled (for buttons)
|
248 |
+
FLY_TEXT_COLOR_HEX = "#1F2937" # From AlphaTranslator_Styled
|
249 |
+
FLY_SUBTLE_TEXT_HEX = "#6B7280" # From AlphaTranslator_Styled
|
250 |
+
FLY_LIGHT_BACKGROUND_HEX = "#F9FAFB" # From AlphaTranslator_Styled (overall page bg)
|
251 |
+
FLY_WHITE_HEX = "#FFFFFF" # From AlphaTranslator_Styled (panel bg)
|
252 |
+
FLY_BORDER_COLOR_HEX = "#D1D5DB" # From AlphaTranslator_Styled
|
253 |
+
FLY_INPUT_BG_HEX_SIMPLE = "#F3F4F6" # From AlphaTranslator_Styled (input bg)
|
254 |
+
|
255 |
+
# Your original AlphaTTS CSS variables for comparison or specific overrides if needed
|
256 |
+
APP_HEADER_GRADIENT_START_IMG_ORIGINAL = "#2980b9" # Blue from your original AlphaTTS header
|
257 |
+
APP_HEADER_GRADIENT_END_IMG_ORIGINAL = "#2ecc71" # Green from your original AlphaTTS header
|
258 |
+
BUTTON_BG_IMG_ORIGINAL = "#2979FF" # Blue from your original AlphaTTS button (generate-button-final)
|
259 |
+
|
260 |
+
# Theme for Gradio Blocks (using a base font from AlphaTranslator_Styled)
|
261 |
+
# Your original `gr.Blocks` used `theme=gr.themes.Base(font=[gr.themes.GoogleFont("Vazirmatn")])`
|
262 |
+
# We'll use `app_theme_outer_styled` for body background and apply Vazirmatn in CSS.
|
263 |
+
app_theme_outer_styled = gr.themes.Base(
|
264 |
+
font=[gr.themes.GoogleFont("Inter"), "system-ui", "sans-serif"], # Base font from AlphaTranslator
|
265 |
+
).set(
|
266 |
+
body_background_fill=FLY_LIGHT_BACKGROUND_HEX, # Overall page background
|
267 |
+
)
|
268 |
+
|
269 |
+
# Combined and adapted CSS
|
270 |
+
# Prioritizing AlphaTranslator_Styled look, but using your specific elem_ids for targeting if possible
|
271 |
+
# The goal is to make YOUR UI components look like the AlphaTranslator_Styled components.
|
272 |
+
final_combined_css = f"""
|
273 |
+
@import url('https://fonts.googleapis.com/css2?family=Vazirmatn:wght@300;400;500;700;800&display=swap');
|
274 |
@import url('https://fonts.googleapis.com/css2?family=Poppins:wght@400;500;600;700;800&display=swap');
|
275 |
@import url('https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap');
|
276 |
+
|
277 |
:root {{
|
278 |
+
/* Colors from AlphaTranslator_Styled */
|
279 |
+
--fly-primary: {FLY_PRIMARY_COLOR_HEX};
|
280 |
+
--fly-secondary: {FLY_SECONDARY_COLOR_HEX};
|
281 |
+
--fly-accent: {FLY_ACCENT_COLOR_HEX}; /* Orange for primary actions */
|
282 |
+
--fly-text-primary: {FLY_TEXT_COLOR_HEX};
|
283 |
+
--fly-text-secondary: {FLY_SUBTLE_TEXT_HEX};
|
284 |
+
--fly-bg-light: {FLY_LIGHT_BACKGROUND_HEX};
|
285 |
+
--fly-bg-white: {FLY_WHITE_HEX};
|
286 |
+
--fly-border-color: {FLY_BORDER_COLOR_HEX};
|
287 |
--fly-input-bg-simple: {FLY_INPUT_BG_HEX_SIMPLE};
|
288 |
+
--fly-primary-rgb: 79,70,229;
|
289 |
+
--fly-accent-rgb: 217,119,6;
|
290 |
+
|
291 |
+
/* Radii and Shadows from AlphaTranslator_Styled */
|
292 |
+
--radius-sm: 0.375rem; --radius-md: 0.5rem; --radius-lg: 0.75rem; --radius-xl: 1rem;
|
293 |
--shadow-lg: 0 10px 15px -3px rgba(0,0,0,0.1),0 4px 6px -4px rgba(0,0,0,0.1);
|
294 |
--shadow-xl: 0 20px 25px -5px rgba(0,0,0,0.1),0 8px 10px -6px rgba(0,0,0,0.1);
|
295 |
+
|
296 |
+
/* Font stack prioritizing Vazirmatn (from your original theme) then AlphaTranslator's */
|
297 |
+
--font-global: 'Vazirmatn', 'Inter', 'Poppins', system-ui, sans-serif;
|
298 |
+
--font-english: 'Poppins', 'Inter', system-ui, sans-serif; /* For English text in header */
|
299 |
}}
|
300 |
+
|
301 |
+
body {{
|
302 |
+
font-family: var(--font-global);
|
303 |
+
direction: rtl;
|
304 |
+
background-color: var(--fly-bg-light);
|
305 |
+
color: var(--fly-text-primary);
|
306 |
+
line-height: 1.7;
|
307 |
+
font-size: 16px;
|
308 |
+
-webkit-font-smoothing: antialiased;
|
309 |
+
-moz-osx-font-smoothing: grayscale;
|
310 |
}}
|
311 |
+
|
312 |
+
.gradio-container {{ /* Overall container styling from AlphaTranslator_Styled */
|
313 |
+
max-width:100% !important; width:100% !important; min-height:100vh;
|
314 |
+
margin:0 auto !important; padding:0 !important; border-radius:0 !important;
|
315 |
+
box-shadow:none !important; background:linear-gradient(170deg, #E0F2FE 0%, #F3E8FF 100%);
|
316 |
+
display:flex; flex-direction:column;
|
317 |
}}
|
318 |
+
|
319 |
+
/* Header: Using AlphaTranslator_Styled header structure and classes */
|
320 |
+
.app-header-alphatts-styled {{
|
321 |
+
text-align:center; padding:2.5rem 1rem; margin:0;
|
322 |
+
background:linear-gradient(135deg, var(--fly-primary) 0%, var(--fly-secondary) 100%); /* Blue-Green gradient */
|
323 |
+
color:var(--fly-bg-white); border-bottom-left-radius:var(--radius-xl);
|
324 |
+
border-bottom-right-radius:var(--radius-xl); box-shadow:var(--shadow-lg);
|
325 |
+
position:relative; overflow:hidden;
|
326 |
}}
|
327 |
.app-header-alphatts-styled::before {{
|
328 |
+
content:''; position:absolute; top:-50px; right:-50px; width:150px; height:150px;
|
329 |
+
background:rgba(255,255,255,0.1); border-radius:9999px; /* full */
|
330 |
+
opacity:0.5; transform:rotate(45deg);
|
331 |
}}
|
332 |
.app-header-alphatts-styled h1 {{
|
333 |
+
font-size:2.25em !important; font-weight:800 !important; margin:0 0 0.5rem 0;
|
334 |
+
font-family:var(--font-english); letter-spacing:-0.5px; text-shadow:0 2px 4px rgba(0,0,0,0.1);
|
335 |
}}
|
336 |
.app-header-alphatts-styled p {{
|
337 |
+
font-size:1em !important; margin-top:0.25rem; font-weight:400;
|
338 |
color:rgba(255,255,255,0.85) !important;
|
339 |
}}
|
340 |
|
341 |
+
/* Main content panel that will wrap YOUR UI elements */
|
342 |
+
.main-content-area-alphatts-styled {{
|
343 |
+
flex-grow:1; padding:0.75rem; width:100%; margin:0 auto; box-sizing:border-box;
|
344 |
+
}}
|
345 |
+
.content-panel-alphatts-styled {{ /* This will wrap your gr.Column */
|
346 |
+
background-color:var(--fly-bg-white); padding:1rem; border-radius:var(--radius-xl);
|
347 |
+
box-shadow:var(--shadow-xl); margin-top:-2rem; position:relative; z-index:10;
|
348 |
+
margin-bottom:2rem; width:100%; box-sizing:border-box;
|
349 |
}}
|
350 |
|
351 |
+
/* Styling YOUR UI elements to match AlphaTranslator_Styled look */
|
352 |
+
/* Inputs (Textbox, Dropdown, File, Slider) */
|
353 |
+
.content-panel-alphatts-styled .gr-input > label + div > textarea,
|
354 |
+
.content-panel-alphatts-styled .gr-dropdown > label + div > div > input, /* For searchable dropdown */
|
355 |
+
.content-panel-alphatts-styled .gr-dropdown > label + div > div > select, /* For standard dropdown */
|
356 |
+
.content-panel-alphatts-styled .gr-textbox > label + div > textarea,
|
357 |
+
.content-panel-alphatts-styled .gr-file > label + div,
|
358 |
+
.content-panel-alphatts-styled .gr-slider input[type="range"] /* Basic slider styling */
|
359 |
{{
|
360 |
+
border-radius:var(--radius-md) !important; /* Using AlphaTranslator's medium radius */
|
361 |
border:1.5px solid var(--fly-border-color) !important;
|
362 |
+
font-size:0.95em !important;
|
363 |
+
background-color:var(--fly-input-bg-simple) !important;
|
364 |
+
padding:10px 12px !important;
|
365 |
+
color:var(--fly-text-primary) !important;
|
366 |
+
box-shadow: none !important; /* Remove default Gradio shadows if any */
|
367 |
}}
|
368 |
+
.content-panel-alphatts-styled .gr-input > label + div > textarea:focus,
|
369 |
+
.content-panel-alphatts-styled .gr-dropdown > label + div > div > input:focus,
|
370 |
+
.content-panel-alphatts-styled .gr-dropdown > label + div > div > select:focus,
|
371 |
+
.content-panel-alphatts-styled .gr-textbox > label + div > textarea:focus,
|
372 |
+
.content-panel-alphatts-styled .gr-file > label + div:focus-within,
|
373 |
+
.content-panel-alphatts-styled .gr-slider input[type="range"]:focus
|
374 |
{{
|
375 |
border-color:var(--fly-primary) !important;
|
376 |
box-shadow:0 0 0 3px rgba(var(--fly-primary-rgb),0.12) !important;
|
377 |
background-color:var(--fly-bg-white) !important;
|
378 |
}}
|
379 |
+
.content-panel-alphatts-styled .gr-file > label + div {{ text-align:center; border-style: dashed !important; }}
|
380 |
+
.content-panel-alphatts-styled .gr-dropdown select {{ cursor:pointer; }}
|
381 |
+
|
382 |
+
/* Labels: General style from AlphaTranslator_Styled */
|
383 |
+
.content-panel-alphatts-styled .gr-form > div > .gr-block > label > .label-text, /* Common path for labels */
|
384 |
+
.content-panel-alphatts-styled .gr-form > .gr-block.gr-input-label-ただ > label > .label-text, /* Another common path */
|
385 |
+
.content-panel-alphatts-styled label > span.label-text /* General target */
|
386 |
+
{{
|
387 |
+
font-weight:500 !important; color: var(--fly-text-secondary) !important;
|
388 |
+
font-size:0.88em !important; margin-bottom:6px !important; display:inline-block;
|
389 |
+
}}
|
390 |
+
/* Your specific elem_id label styling from AlphaTTS_Original (icon part) - this is tricky with Gradio's default label structure */
|
391 |
+
/* The ::before icons might not apply correctly without JS or if Gradio doesn't use `for` attributes on labels properly. */
|
392 |
+
/* For now, the general label style above will apply. If icons are crucial, it needs more specific handling. */
|
393 |
|
394 |
+
/* Button: Targeting your button by its elem_id for the AlphaTranslator_Styled primary button look */
|
395 |
+
.content-panel-alphatts-styled .gr-button[elem_id="generate_button_alpha_v3"]
|
396 |
{{
|
397 |
background:var(--fly-accent) !important; /* Orange accent */
|
398 |
+
margin-top:1.5rem !important; padding:12px 20px !important;
|
399 |
+
transition:all 0.25s ease-in-out !important; color:white !important; font-weight:600 !important;
|
400 |
+
border-radius:var(--radius-md) !important; border:none !important;
|
401 |
box-shadow:0 3px 8px -1px rgba(var(--fly-accent-rgb),0.3) !important;
|
402 |
+
width:100% !important; font-size:1.05em !important; /* Your original font size */
|
403 |
+
display:flex; align-items:center; justify-content:center;
|
404 |
}}
|
405 |
+
.content-panel-alphatts-styled .gr-button[elem_id="generate_button_alpha_v3"]:hover
|
406 |
{{
|
407 |
background:#B45309 !important; /* Darker orange */ transform:translateY(-1px) !important;
|
408 |
box-shadow:0 5px 10px -1px rgba(var(--fly-accent-rgb),0.4) !important;
|
409 |
}}
|
410 |
|
411 |
+
/* Temperature description (from your original AlphaTTS CSS) */
|
412 |
+
.content-panel-alphatts-styled .temp_description_class_alpha_v3 {{
|
413 |
+
font-size: 0.85em; color: var(--fly-text-secondary); margin-top: -0.4rem; margin-bottom: 1rem;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
414 |
}}
|
415 |
|
416 |
+
/* Audio Player styling (targeting your specific elem_id) */
|
417 |
+
.content-panel-alphatts-styled #output_audio_player_alpha_v3 audio
|
418 |
{{
|
419 |
width: 100%; border-radius: var(--radius-md); margin-top:0.8rem;
|
420 |
}}
|
421 |
|
422 |
+
/* Examples section (using AlphaTranslator_Styled examples button style) */
|
423 |
+
.content-panel-alphatts-styled .gr-examples .gr-button.gr-button-tool,
|
424 |
+
.content-panel-alphatts-styled .gr-examples .gr-sample-button
|
425 |
{{
|
426 |
+
background-color:#E0E7FF !important; color:var(--fly-primary) !important;
|
427 |
+
border-radius:var(--radius-sm) !important; font-size:0.78em !important; padding:4px 8px !important;
|
|
|
|
|
|
|
428 |
}}
|
429 |
+
.content-panel-alphatts-styled .custom-hr {{height:1px;background-color:var(--fly-border-color);margin:1.5rem 0;border:none;}}
|
430 |
|
431 |
+
|
432 |
+
/* Footer: Using AlphaTranslator_Styled footer */
|
433 |
.app-footer-alphatts-styled {{
|
434 |
text-align:center;font-size:0.85em;color:var(--fly-text-secondary);margin-top:2.5rem;
|
435 |
padding:1rem 0;background-color:rgba(255,255,255,0.3);backdrop-filter:blur(5px);
|
436 |
border-top:1px solid var(--fly-border-color);
|
437 |
}}
|
438 |
+
/* Hide default Gradio watermarks/footers */
|
439 |
+
footer.svelte-1gfkn6j, .gradio-footer, .flagging-container, .footer-utils {{display:none !important; visibility:hidden !important;}}
|
440 |
+
|
441 |
|
442 |
+
/* Responsive adjustments from AlphaTranslator_Styled */
|
443 |
@media (min-width:640px) {{
|
444 |
+
.main-content-area-alphatts-styled {{padding:1.5rem;max-width:700px;}}
|
445 |
+
.content-panel-alphatts-styled {{padding:1.5rem;}}
|
446 |
.app-header-alphatts-styled h1 {{font-size:2.5em !important;}}
|
447 |
.app-header-alphatts-styled p {{font-size:1.05em !important;}}
|
448 |
}}
|
449 |
@media (min-width:768px) {{
|
450 |
+
.main-content-area-alphatts-styled {{max-width:780px;}}
|
451 |
+
.content-panel-alphatts-styled {{padding:2rem;}}
|
452 |
+
/* Make button not full width on larger screens */
|
453 |
+
.content-panel-alphatts-styled .gr-button[elem_id="generate_button_alpha_v3"]
|
454 |
{{
|
455 |
+
width:auto !important; align-self:flex-start;
|
456 |
}}
|
457 |
.app-header-alphatts-styled h1 {{font-size:2.75em !important;}}
|
458 |
.app-header-alphatts-styled p {{font-size:1.1em !important;}}
|
459 |
}}
|
460 |
"""
|
461 |
+
|
462 |
+
# --- Gradio UI Definition (YOUR UI structure, with new CSS applied) ---
|
463 |
+
# Using app_theme_outer_styled for base theme, and final_combined_css for specifics
|
464 |
+
with gr.Blocks(theme=app_theme_outer_styled, css=final_combined_css, title=f"آلفا TTS ({FIXED_MODEL_NAME.split('-')[1]})") as demo:
|
465 |
+
# Header from AlphaTranslator_Styled structure
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
466 |
gr.HTML(f"""
|
467 |
<div class='app-header-alphatts-styled'>
|
468 |
<h1>🚀 Alpha TTS</h1>
|
|
|
470 |
</div>
|
471 |
""")
|
472 |
|
473 |
+
# Main content area wrapper from AlphaTranslator_Styled structure
|
474 |
+
with gr.Column(elem_classes=["main-content-area-alphatts-styled"]):
|
475 |
+
# Content panel wrapper from AlphaTranslator_Styled structure
|
476 |
+
# Your original UI was inside a gr.Column, which is fine. We wrap it with another for styling.
|
477 |
+
with gr.Column(elem_classes=["content-panel-alphatts-styled"]):
|
478 |
+
|
479 |
+
# Your original UI layout starts here
|
480 |
+
# `elem_id`s are from your original AlphaTTS code.
|
481 |
+
|
482 |
+
# Optional: Warning if GEMINI_API_KEY is not set
|
483 |
+
if not os.environ.get("GEMINI_API_KEY"):
|
484 |
+
missing_key_msg = (
|
485 |
+
"⚠️ هشدار: متغیر محیطی GEMINI_API_KEY تنظیم نشده است. "
|
486 |
+
"این برنامه برای کار کردن به آن نیاز دارد."
|
487 |
+
)
|
488 |
+
gr.Markdown(f"<div style='background-color:#FFFBEB; color:#92400E; padding:10px; border-radius:8px; border:1px solid #FDE68A; text-align:center; margin-bottom:1rem;'>{missing_key_msg}</div>")
|
489 |
+
|
490 |
+
use_file_input_cb = gr.Checkbox(label="📄 استفاده از فایل متنی (.txt)", value=False, elem_id="use_file_cb_alpha_v3")
|
491 |
+
|
492 |
+
uploaded_file_input = gr.File(
|
493 |
+
label=" ",
|
494 |
+
file_types=['.txt'],
|
495 |
+
visible=False,
|
496 |
+
elem_id="file_uploader_alpha_main_v3"
|
497 |
+
)
|
498 |
+
|
499 |
+
text_to_speak_tb = gr.Textbox(
|
500 |
+
label="متن فارسی برای تبدیل",
|
501 |
+
placeholder="مثال: سلام، فردا هوا چطور است؟",
|
502 |
+
lines=5,
|
503 |
+
value="",
|
504 |
+
visible=True,
|
505 |
+
elem_id="text_input_main_alpha_v3"
|
506 |
+
)
|
507 |
+
|
508 |
+
# Your original change function for checkbox
|
509 |
+
use_file_input_cb.change(
|
510 |
+
fn=lambda x: (gr.update(visible=x, label=" " if x else "متن فارسی برای تبدیل"), gr.update(visible=not x)),
|
511 |
+
inputs=use_file_input_cb,
|
512 |
+
outputs=[uploaded_file_input, text_to_speak_tb]
|
513 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
514 |
|
515 |
+
speech_prompt_tb = gr.Textbox(
|
516 |
+
label="سبک گفتار (اختیاری)",
|
517 |
+
placeholder="مثال: با لحنی شاد و پرانرژی",
|
518 |
+
value="با لحنی دوستانه و رسا صحبت کن.",
|
519 |
+
lines=2, elem_id="speech_prompt_alpha_v3"
|
520 |
+
)
|
521 |
+
|
522 |
+
speaker_voice_dd = gr.Dropdown(
|
523 |
+
SPEAKER_VOICES, label="انتخاب گوینده و لهجه", value="Charon", elem_id="speaker_voice_alpha_v3"
|
524 |
+
)
|
525 |
+
|
526 |
+
temperature_slider = gr.Slider(
|
527 |
+
minimum=0.1, maximum=1.5, step=0.05, value=0.9, label="میزان خلاقیت صدا",
|
528 |
+
elem_id="temperature_slider_alpha_v3"
|
529 |
+
)
|
530 |
+
# Your original temperature description
|
531 |
+
gr.Markdown("<p class='temp_description_class_alpha_v3'>مقادیر بالاتر = تنوع بیشتر، مقادیر پایینتر = یکنواختی بیشتر.</p>")
|
532 |
+
|
533 |
+
# Your original button, CSS will target its elem_id
|
534 |
+
generate_button = gr.Button("🚀 تولید و پخش صدا", elem_id="generate_button_alpha_v3")
|
535 |
+
|
536 |
+
# Your original audio output, CSS will target its elem_id
|
537 |
+
output_audio = gr.Audio(label=" ", type="filepath", elem_id="output_audio_player_alpha_v3")
|
538 |
|
539 |
+
# Your original Examples section
|
540 |
+
gr.HTML("<hr class='custom-hr'>") # Using the styled HR
|
541 |
+
gr.Markdown(
|
542 |
+
"<h3 style='text-align:center; font-weight:500; color:var(--fly-text-secondary); margin-top:1.5rem; margin-bottom:1rem;'>نمونههای کاربردی</h3>",
|
543 |
+
# elem_id="examples_section_title_v3" # Your original elem_id for the examples title
|
544 |
+
)
|
545 |
+
gr.Examples(
|
546 |
+
examples=[
|
547 |
+
[False, None, "سلام بر شما، امیدوارم روز خوبی داشته باشید.", "با لحنی گرم و صمیمی.", "Zephyr", 0.85],
|
548 |
+
[False, None, "این یک آزمایش برای بررسی کیفیت صدای تولید شده توسط هوش مصنوعی آلفا است.", "با صدایی طبیعی و روان.", "Charon", 0.9],
|
549 |
+
],
|
550 |
+
inputs=[use_file_input_cb, uploaded_file_input, text_to_speak_tb, speech_prompt_tb, speaker_voice_dd, temperature_slider],
|
551 |
+
outputs=[output_audio], # YOUR ORIGINAL OUTPUTS FOR EXAMPLES
|
552 |
+
fn=gradio_tts_interface,
|
553 |
+
cache_examples=os.getenv("GRADIO_CACHE_EXAMPLES", "False").lower() == "true" # Optional caching
|
554 |
+
)
|
|
|
|
|
555 |
|
556 |
+
# Footer from AlphaTranslator_Styled structure
|
557 |
+
gr.Markdown(f"<p class='app-footer-alphatts-styled'>Alpha TTS © 2024 - Model: {FIXED_MODEL_NAME.split('-')[0].upper()} {FIXED_MODEL_NAME.split('-')[1]}</p>")
|
|
|
|
|
558 |
|
559 |
|
560 |
+
# --- Event Handlers (YOUR ORIGINAL EVENT HANDLERS) ---
|
561 |
+
# Note: The `generate_button.click` from your original code did not include `progress` in inputs.
|
562 |
+
# I'm keeping it that way to match your original exactly.
|
563 |
if generate_button is not None: # Check if button was created
|
564 |
generate_button.click(
|
565 |
fn=gradio_tts_interface,
|
566 |
+
inputs=[use_file_input_cb, uploaded_file_input, text_to_speak_tb, speech_prompt_tb, speaker_voice_dd, temperature_slider], # NO progress
|
567 |
+
outputs=[output_audio]
|
568 |
)
|
569 |
else:
|
570 |
+
# This should not happen if the UI is defined correctly
|
571 |
+
logging.error("دکمه تولید صدا (generate_button_alpha_v3) در UI یافت نشد.")
|
572 |
+
|
573 |
|
574 |
if __name__ == "__main__":
|
575 |
+
# Basic check for PYDUB at launch
|
576 |
if not PYDUB_AVAILABLE:
|
577 |
+
print("WARNING: Pydub (for audio merging) not found. Please install with `pip install pydub`. Merging will be disabled.")
|
578 |
if not os.environ.get("GEMINI_API_KEY"):
|
579 |
+
print("WARNING: GEMINI_API_KEY environment variable not set. TTS functionality will likely fail.")
|
580 |
|
581 |
demo.launch(
|
582 |
+
server_name="0.0.0.0", # Make accessible on network
|
583 |
+
server_port=int(os.getenv("PORT", 7860)), # Use PORT from env or default
|
584 |
+
debug=os.environ.get("GRADIO_DEBUG", "False").lower() == "true", # Optional debug mode
|
585 |
+
show_error=True # Show errors in the browser
|
586 |
)
|