File size: 6,740 Bytes
cdaeb89
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
# -*- coding: utf-8 -*-
import os, glob, json, re
import numpy as np

import librosa
from scenedetect import VideoManager, SceneManager
from scenedetect.detectors import ContentDetector
from moviepy.editor import (
    VideoFileClip, concatenate_videoclips, TextClip, CompositeVideoClip, 
    AudioFileClip, ImageClip
)

def parse_duration_to_seconds(value):
    if value is None:
        return None
    if isinstance(value, (int, float)):
        return float(value)
    s = str(value).strip().lower()
    if s.endswith('s'):
        try:
            return float(s[:-1])
        except:
            pass
    m = re.match(r'^(?:(\d+):)?(\d{1,2}):(\d{1,2})(?:\.(\d+))?$', s)
    if m:
        h = int(m.group(1) or 0)
        mm = int(m.group(2) or 0)
        ss = int(m.group(3) or 0)
        frac = m.group(4)
        total = h*3600 + mm*60 + ss
        if frac:
            total += float("0." + frac)
        return float(total)
    try:
        return float(s)
    except:
        return None

def detect_beats(music_path):
    y, sr = librosa.load(music_path, sr=None, mono=True)
    tempo, beats = librosa.beat.beat_track(y=y, sr=sr, units='frames')
    times = librosa.frames_to_time(beats, sr=sr)
    return tempo, times.tolist()

def detect_scenes(video_path, threshold=27.0, min_scene_len=12):
    vm = VideoManager([video_path])
    sm = SceneManager()
    sm.add_detector(ContentDetector(threshold=threshold, min_scene_len=min_scene_len))
    vm.start()
    sm.detect_scenes(frame_source=vm)
    scene_list = sm.get_scene_list()
    vm.release()
    return [(s[0].get_seconds(), s[1].get_seconds()) for s in scene_list]

def choose_segments_from_scenes(scene_ranges, beats, min_slice=0.7):
    segments = []
    for i in range(len(beats)-1):
        bw_start, bw_end = beats[i], beats[i+1]
        picked = None
        for (s,e) in scene_ranges:
            start = max(s, bw_start)
            end = min(e, bw_end)
            if end - start >= min_slice:
                picked = (start, end)
                break
        if picked is None and (bw_end - bw_start) >= min_slice:
            picked = (bw_start, bw_end)
        if picked:
            segments.append(picked)
    return segments

def resize_fit(clip, target_w, target_h):
    r_target = target_w / target_h
    w, h = clip.size
    r_src = w / h
    if r_src > r_target:
        new_h = target_h
        new_w = int(round(r_src * new_h))
        c = clip.resize(height=new_h).crop(x_center=new_w//2, width=target_w, height=target_h)
    else:
        new_w = target_w
        new_h = int(round(new_w / r_src))
        c = clip.resize(width=new_w).crop(y_center=new_h//2, width=target_w, height=target_h)
    return c

def build_video_single_aspect(segments, W, H, music_path, out_path, intro_text=None, logo_path=None, crossfade=0.0, fps=30):
    clips=[]
    for path,(s,e) in segments:
        c = VideoFileClip(path).subclip(s,e)
        c = resize_fit(c, W, H)
        if crossfade>0:
            c = c.crossfadein(crossfade) if clips else c
        clips.append(c)
    if not clips:
        raise RuntimeError("No segments to compile.")
    body = concatenate_videoclips(clips, method="compose", padding=-crossfade if crossfade>0 else 0)

    overlays = []
    if intro_text:
        try:
            txt = TextClip(intro_text, fontsize=90, font="Arial-Bold", color="white").set_duration(2).set_pos("center")
        except Exception:
            txt = TextClip(intro_text, fontsize=90, color="white").set_duration(2).set_pos("center")
        overlays.append(txt.set_start(0))

    if logo_path and os.path.exists(logo_path):
        logo = ImageClip(logo_path).set_duration(body.duration).resize(width=int(W*0.18)).set_pos(("right","bottom")).margin(right=40, bottom=40, opacity=0)
        overlays.append(logo)

    final = CompositeVideoClip([body] + overlays, size=(W,H))

    if os.path.exists(music_path):
        a = AudioFileClip(music_path).subclip(0, final.duration)
        final = final.set_audio(a)

    final.write_videofile(out_path, codec="libx264", audio_codec="aac", fps=fps, threads=4)

def run_job(job_dir):
    cfg_path = os.path.join(job_dir, "config.json")
    with open(cfg_path, "r", encoding="utf-8") as f:
        cfg = json.load(f)

    # Inputs
    footage_dir = os.path.join(job_dir, "footage")
    music_path = os.path.join(job_dir, "music")
    if not os.path.exists(music_path):
        # try known extensions
        for ext in (".mp3",".wav",".m4a"):
            p = music_path+ext
            if os.path.exists(p):
                music_path = p
                break
    logo_path = os.path.join(job_dir, "logo.png")
    if not os.path.exists(footage_dir):
        raise RuntimeError("footage/ missing")

    videos = sorted(glob.glob(os.path.join(footage_dir, "*.mp4")) + glob.glob(os.path.join(footage_dir, "*.mov")))
    if not videos:
        raise RuntimeError("No videos found in footage/. Upload .mp4 or .mov files.")
    if not os.path.exists(music_path):
        raise RuntimeError("Music file missing.")

    # Prompt-ish
    duration = parse_duration_to_seconds(cfg.get("duration"))
    intro_text = cfg.get("intro_text") or None
    crossfade = float(cfg.get("crossfade", 0.0))
    aspects = cfg.get("aspects", ["9:16"])  # list
    threshold = float(cfg.get("scene_threshold", 27.0))

    # analysis
    tempo, beats = detect_beats(music_path)

    scene_map = {}
    for p in videos:
        scene_map[p] = detect_scenes(p, threshold=threshold)

    # build candidate segments
    segs = []
    for p in videos:
        for (s,e) in choose_segments_from_scenes(scene_map[p], beats):
            segs.append((p,(s,e)))

    # trim to duration
    if duration:
        trimmed, acc = [], 0.0
        for p,(s,e) in segs:
            dur = e - s
            if acc + dur > duration:
                e = s + max(0.7, duration - acc)
                dur = e - s
            trimmed.append((p,(s,e)))
            acc += dur
            if acc >= duration:
                break
        segs = trimmed

    # map aspect to (W,H)
    def parse_aspect(a):
        if a in ("9:16","1080x1920"): return (1080,1920)
        if a in ("16:9","1920x1080"): return (1920,1080)
        if "x" in a:
            try:
                w,h = [int(x) for x in a.split("x")]
                return (w,h)
            except: pass
        return (1080,1920)

    outputs = []
    for a in aspects:
        W,H = parse_aspect(a)
        out_path = os.path.join(job_dir, f"output_{W}x{H}.mp4")
        build_video_single_aspect(segs, W,H, music_path, out_path, intro_text=intro_text, logo_path=(logo_path if os.path.exists(logo_path) else None), crossfade=crossfade, fps=30)
        outputs.append(out_path)

    return outputs