forked from kamilstanuch/Autocrop-vertical
-
Notifications
You must be signed in to change notification settings - Fork 425
Expand file tree
/
Copy pathapp.py
More file actions
2224 lines (1848 loc) · 87.2 KB
/
app.py
File metadata and controls
2224 lines (1848 loc) · 87.2 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
import os
import uuid
import subprocess
import threading
import json
import shutil
import glob
import time
import asyncio
from dotenv import load_dotenv
from typing import Dict, Optional, List
from contextlib import asynccontextmanager
from fastapi import FastAPI, UploadFile, File, Form, HTTPException, Request, Header, BackgroundTasks
from fastapi.middleware.cors import CORSMiddleware
from fastapi.staticfiles import StaticFiles
from fastapi.responses import HTMLResponse
from pydantic import BaseModel
from s3_uploader import upload_job_artifacts, list_all_clips, upload_actor_to_s3, list_actor_gallery, upload_video_to_gallery, list_video_gallery
load_dotenv()
# Constants
UPLOAD_DIR = "uploads"
OUTPUT_DIR = "output"
os.makedirs(UPLOAD_DIR, exist_ok=True)
os.makedirs(OUTPUT_DIR, exist_ok=True)
# Configuration
# Default to 1 if not set, but user can set higher for powerful servers
MAX_CONCURRENT_JOBS = int(os.environ.get("MAX_CONCURRENT_JOBS", "5"))
MAX_FILE_SIZE_MB = 2048 # 2GB limit
JOB_RETENTION_SECONDS = 3600 # 1 hour retention
# Application State
job_queue = asyncio.Queue()
jobs: Dict[str, Dict] = {}
thumbnail_sessions: Dict[str, Dict] = {}
publish_jobs: Dict[str, Dict] = {} # {publish_id: {status, result, error}}
# Semester to limit concurrency to MAX_CONCURRENT_JOBS
concurrency_semaphore = asyncio.Semaphore(MAX_CONCURRENT_JOBS)
def _relocate_root_job_artifacts(job_id: str, job_output_dir: str) -> bool:
"""
Backward-compat rescue:
If main.py accidentally wrote metadata/clips into OUTPUT_DIR root (e.g. output/<jobid>_...),
move them into output/<job_id>/ so the API can find and serve them.
"""
try:
os.makedirs(job_output_dir, exist_ok=True)
root = OUTPUT_DIR
pattern = os.path.join(root, f"{job_id}_*_metadata.json")
meta_candidates = sorted(glob.glob(pattern), key=lambda p: os.path.getmtime(p), reverse=True)
if not meta_candidates:
return False
# Move the newest metadata and its associated clips.
metadata_path = meta_candidates[0]
base_name = os.path.basename(metadata_path).replace("_metadata.json", "")
# Move metadata
dest_metadata = os.path.join(job_output_dir, os.path.basename(metadata_path))
if os.path.abspath(metadata_path) != os.path.abspath(dest_metadata):
shutil.move(metadata_path, dest_metadata)
# Move any clips that match the same base_name into the job folder
clip_pattern = os.path.join(root, f"{base_name}_clip_*.mp4")
for clip_path in glob.glob(clip_pattern):
dest_clip = os.path.join(job_output_dir, os.path.basename(clip_path))
if os.path.abspath(clip_path) != os.path.abspath(dest_clip):
shutil.move(clip_path, dest_clip)
# Also move any temp_ clips that might remain
temp_clip_pattern = os.path.join(root, f"temp_{base_name}_clip_*.mp4")
for clip_path in glob.glob(temp_clip_pattern):
dest_clip = os.path.join(job_output_dir, os.path.basename(clip_path))
if os.path.abspath(clip_path) != os.path.abspath(dest_clip):
shutil.move(clip_path, dest_clip)
return True
except Exception:
return False
async def cleanup_jobs():
"""Background task to remove old jobs and files."""
import time
print("🧹 Cleanup task started.")
while True:
try:
await asyncio.sleep(300) # Check every 5 minutes
now = time.time()
# Simple directory cleanup based on modification time
# Check OUTPUT_DIR
for job_id in os.listdir(OUTPUT_DIR):
job_path = os.path.join(OUTPUT_DIR, job_id)
if os.path.isdir(job_path):
if now - os.path.getmtime(job_path) > JOB_RETENTION_SECONDS:
print(f"🧹 Purging old job: {job_id}")
shutil.rmtree(job_path, ignore_errors=True)
if job_id in jobs:
del jobs[job_id]
# Cleanup SaaSShorts jobs from memory
try:
saas_expired = [
jid for jid, jdata in list(saas_jobs.items())
if jdata.get("status") in ("completed", "failed")
and jdata.get("output_dir")
and os.path.isdir(jdata["output_dir"])
and now - os.path.getmtime(jdata["output_dir"]) > JOB_RETENTION_SECONDS
]
for jid in saas_expired:
del saas_jobs[jid]
except NameError:
pass
# Cleanup Uploads
for filename in os.listdir(UPLOAD_DIR):
file_path = os.path.join(UPLOAD_DIR, filename)
try:
if now - os.path.getmtime(file_path) > JOB_RETENTION_SECONDS:
os.remove(file_path)
except Exception: pass
except Exception as e:
print(f"⚠️ Cleanup error: {e}")
async def process_queue():
"""Background worker to process jobs from the queue with concurrency limit."""
print(f"🚀 Job Queue Worker started with {MAX_CONCURRENT_JOBS} concurrent slots.")
while True:
try:
# Wait for a job
job_id = await job_queue.get()
# Acquire semaphore slot (waits if max jobs are running)
await concurrency_semaphore.acquire()
print(f"🔄 Acquired slot for job: {job_id}")
# Process in background task to not block the loop (allowing other slots to fill)
asyncio.create_task(run_job_wrapper(job_id))
except Exception as e:
print(f"❌ Queue dispatch error: {e}")
await asyncio.sleep(1)
async def run_job_wrapper(job_id):
"""Wrapper to run job and release semaphore"""
try:
job = jobs.get(job_id)
if job:
await run_job(job_id, job)
except Exception as e:
print(f"❌ Job wrapper error {job_id}: {e}")
finally:
# Always release semaphore and mark queue task done
concurrency_semaphore.release()
job_queue.task_done()
print(f"✅ Released slot for job: {job_id}")
@asynccontextmanager
async def lifespan(app: FastAPI):
# Start worker and cleanup
worker_task = asyncio.create_task(process_queue())
cleanup_task = asyncio.create_task(cleanup_jobs())
yield
# Cleanup (optional: cancel worker)
app = FastAPI(lifespan=lifespan)
# Enable CORS for frontend
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
# Mount static files for serving videos
app.mount("/videos", StaticFiles(directory=OUTPUT_DIR), name="videos")
# Mount static files for serving thumbnails
THUMBNAILS_DIR = os.path.join(OUTPUT_DIR, "thumbnails")
os.makedirs(THUMBNAILS_DIR, exist_ok=True)
app.mount("/thumbnails", StaticFiles(directory=THUMBNAILS_DIR), name="thumbnails")
class ProcessRequest(BaseModel):
url: str
def enqueue_output(out, job_id):
"""Reads output from a subprocess and appends it to jobs logs."""
try:
for line in iter(out.readline, b''):
decoded_line = line.decode('utf-8').strip()
if decoded_line:
print(f"📝 [Job Output] {decoded_line}")
if job_id in jobs:
jobs[job_id]['logs'].append(decoded_line)
except Exception as e:
print(f"Error reading output for job {job_id}: {e}")
finally:
out.close()
async def run_job(job_id, job_data):
"""Executes the subprocess for a specific job."""
cmd = job_data['cmd']
env = job_data['env']
output_dir = job_data['output_dir']
jobs[job_id]['status'] = 'processing'
jobs[job_id]['logs'].append("Job started by worker.")
print(f"🎬 [run_job] Executing command for {job_id}: {' '.join(cmd)}")
try:
process = subprocess.Popen(
cmd,
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT, # Merge stderr to stdout
env=env,
cwd=os.getcwd()
)
# We need to capture logs in a thread because Popen isn't async
t_log = threading.Thread(target=enqueue_output, args=(process.stdout, job_id))
t_log.daemon = True
t_log.start()
# Async wait for process with incremental updates
start_wait = time.time()
while process.poll() is None:
await asyncio.sleep(2)
# Check for partial results every 2 seconds
# Look for metadata file
try:
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if json_files:
target_json = json_files[0]
# Read metadata (it might be being written to, so simple try/except or just read)
# Use a lock or just robust read? json.load might fail if file is partial.
# Usually main.py writes it once at start (based on my review).
if os.path.getsize(target_json) > 0:
with open(target_json, 'r') as f:
data = json.load(f)
base_name = os.path.basename(target_json).replace('_metadata.json', '')
clips = data.get('shorts', [])
cost_analysis = data.get('cost_analysis')
# Check which clips actually exist on disk
ready_clips = []
for i, clip in enumerate(clips):
clip_filename = f"{base_name}_clip_{i+1}.mp4"
clip_path = os.path.join(output_dir, clip_filename)
if os.path.exists(clip_path) and os.path.getsize(clip_path) > 0:
# Checking if file is growing? For now assume if it exists and main.py moves it there, it's done.
# main.py writes to temp_... then moves to final name. So presence means ready!
clip['video_url'] = f"/videos/{job_id}/{clip_filename}"
ready_clips.append(clip)
if ready_clips:
jobs[job_id]['result'] = {'clips': ready_clips, 'cost_analysis': cost_analysis}
except Exception as e:
# Ignore read errors during processing
pass
returncode = process.returncode
if returncode == 0:
jobs[job_id]['status'] = 'completed'
jobs[job_id]['logs'].append("Process finished successfully.")
# Start S3 upload in background (silent, non-blocking)
loop = asyncio.get_event_loop()
loop.run_in_executor(None, upload_job_artifacts, output_dir, job_id)
# Find result JSON
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if not json_files:
# Backward-compat rescue if outputs were written to OUTPUT_DIR root
if _relocate_root_job_artifacts(job_id, output_dir):
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if json_files:
target_json = json_files[0]
with open(target_json, 'r') as f:
data = json.load(f)
# Enhance result with video URLs
base_name = os.path.basename(target_json).replace('_metadata.json', '')
clips = data.get('shorts', [])
cost_analysis = data.get('cost_analysis')
for i, clip in enumerate(clips):
clip_filename = f"{base_name}_clip_{i+1}.mp4"
clip['video_url'] = f"/videos/{job_id}/{clip_filename}"
jobs[job_id]['result'] = {'clips': clips, 'cost_analysis': cost_analysis}
else:
jobs[job_id]['status'] = 'failed'
jobs[job_id]['logs'].append("No metadata file generated.")
else:
jobs[job_id]['status'] = 'failed'
jobs[job_id]['logs'].append(f"Process failed with exit code {returncode}")
except Exception as e:
jobs[job_id]['status'] = 'failed'
jobs[job_id]['logs'].append(f"Execution error: {str(e)}")
@app.post("/api/process")
async def process_endpoint(
request: Request,
file: Optional[UploadFile] = File(None),
url: Optional[str] = Form(None)
):
api_key = request.headers.get("X-Gemini-Key")
if not api_key:
raise HTTPException(status_code=400, detail="Missing X-Gemini-Key header")
# Handle JSON body manually for URL payload
content_type = request.headers.get("content-type", "")
if "application/json" in content_type:
body = await request.json()
url = body.get("url")
if not url and not file:
raise HTTPException(status_code=400, detail="Must provide URL or File")
job_id = str(uuid.uuid4())
job_output_dir = os.path.join(OUTPUT_DIR, job_id)
os.makedirs(job_output_dir, exist_ok=True)
# Prepare Command
cmd = ["python", "-u", "main.py"] # -u for unbuffered
env = os.environ.copy()
env["GEMINI_API_KEY"] = api_key # Override with key from request
if url:
cmd.extend(["-u", url])
else:
# Save uploaded file with size limit check
input_path = os.path.join(UPLOAD_DIR, f"{job_id}_{file.filename}")
# Read file in chunks to check size
size = 0
limit_bytes = MAX_FILE_SIZE_MB * 1024 * 1024
with open(input_path, "wb") as buffer:
while content := await file.read(1024 * 1024): # Read 1MB chunks
size += len(content)
if size > limit_bytes:
os.remove(input_path)
shutil.rmtree(job_output_dir)
raise HTTPException(status_code=413, detail=f"File too large. Max size {MAX_FILE_SIZE_MB}MB")
buffer.write(content)
cmd.extend(["-i", input_path])
cmd.extend(["-o", job_output_dir])
# Enqueue Job
jobs[job_id] = {
'status': 'queued',
'logs': [f"Job {job_id} queued."],
'cmd': cmd,
'env': env,
'output_dir': job_output_dir
}
await job_queue.put(job_id)
return {"job_id": job_id, "status": "queued"}
@app.get("/api/status/{job_id}")
async def get_status(job_id: str):
if job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
job = jobs[job_id]
return {
"status": job['status'],
"logs": job['logs'],
"result": job.get('result')
}
from editor import VideoEditor
from subtitles import generate_srt, burn_subtitles, generate_srt_from_video
from hooks import add_hook_to_video
from translate import translate_video, get_supported_languages
from thumbnail import analyze_video_for_titles, refine_titles, generate_thumbnail, generate_youtube_description
class EditRequest(BaseModel):
job_id: str
clip_index: int
api_key: Optional[str] = None
input_filename: Optional[str] = None
@app.post("/api/edit")
async def edit_clip(
req: EditRequest,
x_gemini_key: Optional[str] = Header(None, alias="X-Gemini-Key")
):
# Determine API Key
final_api_key = req.api_key or x_gemini_key or os.environ.get("GEMINI_API_KEY")
if not final_api_key:
raise HTTPException(status_code=400, detail="Missing Gemini API Key (Header or Body)")
if req.job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
job = jobs[req.job_id]
if 'result' not in job or 'clips' not in job['result']:
raise HTTPException(status_code=400, detail="Job result not available")
try:
# Resolve Input Path: Prefer explict input_filename from frontend (chaining edits)
if req.input_filename:
# Security: Ensure just a filename, no paths
safe_name = os.path.basename(req.input_filename)
input_path = os.path.join(OUTPUT_DIR, req.job_id, safe_name)
filename = safe_name
else:
# Fallback to original clip
clip = job['result']['clips'][req.clip_index]
filename = clip['video_url'].split('/')[-1]
input_path = os.path.join(OUTPUT_DIR, req.job_id, filename)
if not os.path.exists(input_path):
raise HTTPException(status_code=404, detail=f"Video file not found: {input_path}")
# Define output path for edited video
edited_filename = f"edited_{filename}"
output_path = os.path.join(OUTPUT_DIR, req.job_id, edited_filename)
# Run editing in a thread to avoid blocking main loop
# Since VideoEditor uses blocking calls (subprocess, API wait)
def run_edit():
editor = VideoEditor(api_key=final_api_key)
# SAFE FILE RENAMING STRATEGY (Avoid UnicodeEncodeError in Docker)
# Create a safe ASCII filename in the same directory
safe_filename = f"temp_input_{req.job_id}.mp4"
safe_input_path = os.path.join(OUTPUT_DIR, req.job_id, safe_filename)
# Copy original file to safe path
# (Copy is safer than rename if something crashes, we keep original)
shutil.copy(input_path, safe_input_path)
try:
# 1. Upload (using safe path)
vid_file = editor.upload_video(safe_input_path)
# 2. Get duration
import cv2
cap = cv2.VideoCapture(safe_input_path)
fps = cap.get(cv2.CAP_PROP_FPS)
frame_count = int(cap.get(cv2.CAP_PROP_FRAME_COUNT))
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
duration = frame_count / fps if fps else 0
cap.release()
# Load transcript from metadata
transcript = None
try:
meta_files = glob.glob(os.path.join(OUTPUT_DIR, req.job_id, "*_metadata.json"))
if meta_files:
with open(meta_files[0], 'r') as f:
data = json.load(f)
transcript = data.get('transcript')
except Exception as e:
print(f"⚠️ Could not load transcript for editing context: {e}")
# 3. Get Plan (Filter String)
filter_data = editor.get_ffmpeg_filter(vid_file, duration, fps=fps, width=width, height=height, transcript=transcript)
# 4. Apply
# Use safe output name first
safe_output_path = os.path.join(OUTPUT_DIR, req.job_id, f"temp_output_{req.job_id}.mp4")
editor.apply_edits(safe_input_path, safe_output_path, filter_data)
# Move result to final destination (rename works even if dest name has unicode if filesystem supports it,
# but python might still struggle if locale is broken? No, os.rename usually handles it better than subprocess args)
# Actually, output_path is defined above: f"edited_{filename}"
# If filename has unicode, output_path has unicode.
# Let's hope shutil.move / os.rename works.
if os.path.exists(safe_output_path):
shutil.move(safe_output_path, output_path)
return filter_data
finally:
# Cleanup temp safe input
if os.path.exists(safe_input_path):
os.remove(safe_input_path)
# Run in thread pool
loop = asyncio.get_event_loop()
plan = await loop.run_in_executor(None, run_edit)
# Update clip URL in the job result?
# Or return new URL and let frontend handle it?
# Updating job result allows persistence if page refreshes.
new_video_url = f"/videos/{req.job_id}/{edited_filename}"
# Start a new "edited" clip entry or just update the current one?
# Let's update the current one's video_url but keep backup?
# Or return the new URL to the frontend to display.
return {
"success": True,
"new_video_url": new_video_url,
"edit_plan": plan
}
except Exception as e:
print(f"❌ Edit Error: {e}")
raise HTTPException(status_code=500, detail=str(e))
class SubtitleRequest(BaseModel):
job_id: str
clip_index: int
position: str = "bottom" # top, middle, bottom
font_size: int = 16
font_name: str = "Verdana"
font_color: str = "#FFFFFF"
border_color: str = "#000000"
border_width: int = 2
bg_color: str = "#000000"
bg_opacity: float = 0.0
input_filename: Optional[str] = None
@app.get("/api/clip/{job_id}/{clip_index}/transcript")
async def get_clip_transcript(job_id: str, clip_index: int):
"""Return word-level captions for a specific clip, formatted for Remotion."""
if job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
output_dir = os.path.join(OUTPUT_DIR, job_id)
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if not json_files:
raise HTTPException(status_code=404, detail="Metadata not found")
with open(json_files[0], 'r') as f:
data = json.load(f)
transcript = data.get('transcript')
if not transcript:
raise HTTPException(status_code=400, detail="Transcript not found in metadata")
clips = data.get('shorts', [])
if clip_index >= len(clips):
raise HTTPException(status_code=404, detail="Clip not found")
clip_data = clips[clip_index]
clip_start = clip_data.get('start', 0)
clip_end = clip_data.get('end', 0)
# Extract words within clip range and convert to CaptionWord format
captions = []
for segment in transcript.get('segments', []):
for word_info in segment.get('words', []):
if word_info['end'] > clip_start and word_info['start'] < clip_end:
captions.append({
"text": word_info.get('word', '').strip(),
"startMs": int((max(0, word_info['start'] - clip_start)) * 1000),
"endMs": int((max(0, word_info['end'] - clip_start)) * 1000),
})
duration_sec = clip_end - clip_start
return {
"captions": captions,
"durationSec": duration_sec,
"language": transcript.get('language', 'en'),
}
# --- Remotion Render Proxy ---
RENDER_SERVICE_URL = os.getenv("RENDER_SERVICE_URL", "http://renderer:3100")
@app.post("/api/render")
async def proxy_render(request: Request):
"""Proxy render requests to the Node.js Remotion render service."""
import httpx
body = await request.json()
try:
async with httpx.AsyncClient(timeout=30.0) as client:
resp = await client.post(f"{RENDER_SERVICE_URL}/render", json=body)
return resp.json()
except Exception as e:
raise HTTPException(status_code=502, detail=f"Render service unavailable: {e}")
@app.get("/api/render/{render_id}")
async def proxy_render_status(render_id: str):
"""Proxy render status polling to the Node.js Remotion render service."""
import httpx
try:
async with httpx.AsyncClient(timeout=10.0) as client:
resp = await client.get(f"{RENDER_SERVICE_URL}/render/{render_id}")
return resp.json()
except Exception as e:
raise HTTPException(status_code=502, detail=f"Render service unavailable: {e}")
class EffectsGenerateRequest(BaseModel):
job_id: str
clip_index: int
input_filename: Optional[str] = None
@app.post("/api/effects/generate")
async def generate_effects_config(
req: EffectsGenerateRequest,
x_gemini_key: Optional[str] = Header(None, alias="X-Gemini-Key")
):
"""Generate structured EffectsConfig JSON for Remotion rendering via Gemini AI."""
final_api_key = x_gemini_key or os.environ.get("GEMINI_API_KEY")
if not final_api_key:
raise HTTPException(status_code=400, detail="Missing Gemini API Key (Header)")
if req.job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
job = jobs[req.job_id]
if 'result' not in job or 'clips' not in job['result']:
raise HTTPException(status_code=400, detail="Job result not available")
try:
# Resolve input path
if req.input_filename:
safe_name = os.path.basename(req.input_filename)
input_path = os.path.join(OUTPUT_DIR, req.job_id, safe_name)
else:
clip = job['result']['clips'][req.clip_index]
filename = clip['video_url'].split('/')[-1]
input_path = os.path.join(OUTPUT_DIR, req.job_id, filename)
if not os.path.exists(input_path):
raise HTTPException(status_code=404, detail=f"Video file not found: {input_path}")
def run_effects_generation():
editor = VideoEditor(api_key=final_api_key)
# Create safe ASCII filename to avoid encoding issues
safe_filename = f"temp_effects_{req.job_id}.mp4"
safe_input_path = os.path.join(OUTPUT_DIR, req.job_id, safe_filename)
shutil.copy(input_path, safe_input_path)
try:
# Upload video to Gemini
vid_file = editor.upload_video(safe_input_path)
# Get video metadata via ffprobe
probe_cmd = [
'ffprobe', '-v', 'error',
'-select_streams', 'v:0',
'-show_entries', 'stream=width,height,r_frame_rate,duration',
'-show_entries', 'format=duration',
'-of', 'json',
safe_input_path
]
probe_result = subprocess.check_output(probe_cmd).decode().strip()
probe_data = json.loads(probe_result)
stream = probe_data.get('streams', [{}])[0]
width = int(stream.get('width', 1080))
height = int(stream.get('height', 1920))
# Parse fps from r_frame_rate (e.g. "30/1")
r_frame_rate = stream.get('r_frame_rate', '30/1')
num, den = r_frame_rate.split('/')
fps = round(int(num) / int(den), 2)
# Get duration from stream or format
duration = float(stream.get('duration', 0))
if duration == 0:
duration = float(probe_data.get('format', {}).get('duration', 0))
# Load transcript from metadata
transcript = None
try:
meta_files = glob.glob(os.path.join(OUTPUT_DIR, req.job_id, "*_metadata.json"))
if meta_files:
with open(meta_files[0], 'r') as f:
data = json.load(f)
transcript = data.get('transcript')
except Exception as e:
print(f"⚠️ Could not load transcript for effects config: {e}")
# Generate effects config
effects_config = editor.get_effects_config(
vid_file, duration, fps=fps, width=width, height=height, transcript=transcript
)
return effects_config
finally:
if os.path.exists(safe_input_path):
os.remove(safe_input_path)
loop = asyncio.get_event_loop()
effects_config = await loop.run_in_executor(None, run_effects_generation)
if effects_config is None:
raise HTTPException(status_code=500, detail="Failed to generate effects config from Gemini")
return {"effects": effects_config}
except HTTPException:
raise
except Exception as e:
print(f"❌ Effects Generation Error: {e}")
raise HTTPException(status_code=500, detail=str(e))
@app.post("/api/subtitle")
async def add_subtitles(req: SubtitleRequest):
if req.job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
# Reload job data from disk just in case metadata was updated
job = jobs[req.job_id]
# We need to access metadata.json to get the transcript
output_dir = os.path.join(OUTPUT_DIR, req.job_id)
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if not json_files:
raise HTTPException(status_code=404, detail="Metadata not found")
with open(json_files[0], 'r') as f:
data = json.load(f)
transcript = data.get('transcript')
if not transcript:
raise HTTPException(status_code=400, detail="Transcript not found in metadata. Please process a new video.")
clips = data.get('shorts', [])
if req.clip_index >= len(clips):
raise HTTPException(status_code=404, detail="Clip not found")
clip_data = clips[req.clip_index]
# Video Path
if req.input_filename:
# Use chained file
filename = os.path.basename(req.input_filename)
else:
# Fallback to standard naming
filename = clip_data.get('video_url', '').split('/')[-1]
if not filename:
base_name = os.path.basename(json_files[0]).replace('_metadata.json', '')
filename = f"{base_name}_clip_{req.clip_index+1}.mp4"
input_path = os.path.join(output_dir, filename)
if not os.path.exists(input_path):
# Try looking for edited version if url implied it?
# Just fail if not found.
raise HTTPException(status_code=404, detail=f"Video file not found: {input_path}")
# Define outputs
srt_filename = f"subs_{req.clip_index}_{int(time.time())}.srt"
srt_path = os.path.join(output_dir, srt_filename)
# Output video
# We create a new file "subtitled_..."
output_filename = f"subtitled_{filename}"
output_path = os.path.join(output_dir, output_filename)
try:
# 1. Generate SRT
# Check if this is a dubbed video - if so, transcribe it fresh
is_dubbed = filename.startswith("translated_")
if is_dubbed:
print(f"🎙️ Dubbed video detected, transcribing audio for subtitles...")
def run_transcribe_srt():
return generate_srt_from_video(input_path, srt_path)
loop = asyncio.get_event_loop()
success = await loop.run_in_executor(None, run_transcribe_srt)
else:
success = generate_srt(transcript, clip_data['start'], clip_data['end'], srt_path)
if not success:
raise HTTPException(status_code=400, detail="No words found for this clip range.")
# 2. Burn Subtitles
# Run in thread pool
def run_burn():
burn_subtitles(input_path, srt_path, output_path,
alignment=req.position, fontsize=req.font_size,
font_name=req.font_name, font_color=req.font_color,
border_color=req.border_color, border_width=req.border_width,
bg_color=req.bg_color, bg_opacity=req.bg_opacity)
loop = asyncio.get_event_loop()
await loop.run_in_executor(None, run_burn)
except Exception as e:
print(f"❌ Subtitle Error: {e}")
raise HTTPException(status_code=500, detail=str(e))
# 3. Update Result and Metadata
# Update InMemory Jobs
if req.clip_index < len(job['result']['clips']):
job['result']['clips'][req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
# Update Metadata on Disk (Persistence)
try:
if req.clip_index < len(clips):
clips[req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
# Update the main data structure
data['shorts'] = clips
# Write back
with open(json_files[0], 'w') as f:
json.dump(data, f, indent=4)
print(f"✅ Metadata updated with subtitled video for clip {req.clip_index}")
except Exception as e:
print(f"⚠️ Failed to update metadata.json: {e}")
# Non-critical, but good for persistence
return {
"success": True,
"new_video_url": f"/videos/{req.job_id}/{output_filename}"
}
class HookRequest(BaseModel):
job_id: str
clip_index: int
text: str
input_filename: Optional[str] = None
position: Optional[str] = "top" # top, center, bottom
size: Optional[str] = "M" # S, M, L
@app.post("/api/hook")
async def add_hook(req: HookRequest):
if req.job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
job = jobs[req.job_id]
output_dir = os.path.join(OUTPUT_DIR, req.job_id)
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if not json_files:
raise HTTPException(status_code=404, detail="Metadata not found")
with open(json_files[0], 'r') as f:
data = json.load(f)
clips = data.get('shorts', [])
if req.clip_index >= len(clips):
raise HTTPException(status_code=404, detail="Clip not found")
clip_data = clips[req.clip_index]
# Video Path
if req.input_filename:
filename = os.path.basename(req.input_filename)
else:
filename = clip_data.get('video_url', '').split('/')[-1]
if not filename:
base_name = os.path.basename(json_files[0]).replace('_metadata.json', '')
filename = f"{base_name}_clip_{req.clip_index+1}.mp4"
input_path = os.path.join(output_dir, filename)
if not os.path.exists(input_path):
raise HTTPException(status_code=404, detail=f"Video file not found: {input_path}")
# Output video
output_filename = f"hook_{filename}"
output_path = os.path.join(output_dir, output_filename)
# Map Size to Scale
size_map = {"S": 0.8, "M": 1.0, "L": 1.3}
font_scale = size_map.get(req.size, 1.0)
try:
# Run in thread pool
def run_hook():
add_hook_to_video(input_path, req.text, output_path, position=req.position, font_scale=font_scale)
loop = asyncio.get_event_loop()
await loop.run_in_executor(None, run_hook)
except Exception as e:
print(f"❌ Hook Error: {e}")
raise HTTPException(status_code=500, detail=str(e))
# Update Persistence (Same logic as subtitles)
# Update InMemory Jobs
if req.clip_index < len(job['result']['clips']):
job['result']['clips'][req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
# Update Metadata on Disk
try:
if req.clip_index < len(clips):
clips[req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
data['shorts'] = clips
with open(json_files[0], 'w') as f:
json.dump(data, f, indent=4)
print(f"✅ Metadata updated with hook video for clip {req.clip_index}")
except Exception as e:
print(f"⚠️ Failed to update metadata.json: {e}")
return {
"success": True,
"new_video_url": f"/videos/{req.job_id}/{output_filename}"
}
class TranslateRequest(BaseModel):
job_id: str
clip_index: int
target_language: str
source_language: Optional[str] = None
input_filename: Optional[str] = None
@app.get("/api/translate/languages")
async def get_languages():
"""Return supported languages for translation."""
return {"languages": get_supported_languages()}
@app.post("/api/translate")
async def translate_clip(
req: TranslateRequest,
x_elevenlabs_key: Optional[str] = Header(None, alias="X-ElevenLabs-Key")
):
"""Translate a video clip to a different language using ElevenLabs dubbing."""
if not x_elevenlabs_key:
raise HTTPException(status_code=400, detail="Missing X-ElevenLabs-Key header")
if req.job_id not in jobs:
raise HTTPException(status_code=404, detail="Job not found")
job = jobs[req.job_id]
output_dir = os.path.join(OUTPUT_DIR, req.job_id)
json_files = glob.glob(os.path.join(output_dir, "*_metadata.json"))
if not json_files:
raise HTTPException(status_code=404, detail="Metadata not found")
with open(json_files[0], 'r') as f:
data = json.load(f)
clips = data.get('shorts', [])
if req.clip_index >= len(clips):
raise HTTPException(status_code=404, detail="Clip not found")
clip_data = clips[req.clip_index]
# Video Path
if req.input_filename:
filename = os.path.basename(req.input_filename)
else:
filename = clip_data.get('video_url', '').split('/')[-1]
if not filename:
base_name = os.path.basename(json_files[0]).replace('_metadata.json', '')
filename = f"{base_name}_clip_{req.clip_index+1}.mp4"
input_path = os.path.join(output_dir, filename)
if not os.path.exists(input_path):
raise HTTPException(status_code=404, detail=f"Video file not found: {input_path}")
# Output video with language suffix
base, ext = os.path.splitext(filename)
output_filename = f"translated_{req.target_language}_{base}{ext}"
output_path = os.path.join(output_dir, output_filename)
try:
# Run translation in thread pool (blocking API calls)
def run_translate():
return translate_video(
video_path=input_path,
output_path=output_path,
target_language=req.target_language,
api_key=x_elevenlabs_key,
source_language=req.source_language,
)
loop = asyncio.get_event_loop()
await loop.run_in_executor(None, run_translate)
except Exception as e:
print(f"❌ Translation Error: {e}")
raise HTTPException(status_code=500, detail=str(e))
# Update InMemory Jobs
if req.clip_index < len(job['result']['clips']):
job['result']['clips'][req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
# Update Metadata on Disk
try:
if req.clip_index < len(clips):
clips[req.clip_index]['video_url'] = f"/videos/{req.job_id}/{output_filename}"
data['shorts'] = clips