Functions — github.com/GVCLab/CutClaw

hub / github.com/GVCLab/CutClaw / functions

Functions377 in github.com/GVCLab/CutClaw

⨍ Functions 377◇ Types & classes 16

↓ 1 callersFunctionto_rgb

(pil_image: Image.Image)

src/utils/video_tools.py:89

↓ 1 callersFunctiontranscribe_audio

Unified ASR entry point. backend: "whisper_cpp" | "litellm" Defaults to config.ASR_BACKEND or "whisper_cpp".

src/video/preprocess/asr.py:525

↓ 1 callersFunctionvalidate_section_durations

Validate that all sections have duration within the specified range. Args: sections: List of section dicts with Start_Time and End_T

src/audio/audio_caption_madmom.py:335

↓ 1 callersFunctionvalidate_sections_within_duration

Validate that all sections are within the audio duration. Args: sections: List of section dicts with Start_Time and End_Time

src/audio/audio_caption_madmom.py:301

↓ 1 callersFunctionwrite_character_info

Write character information to a JSON file. Args: character_info: Character identification results output_path: Path to outp

src/video/deconstruction/get_character.py:296

↓ 1 callersFunctionwrite_srt

Write subtitles to an SRT file. Args: subtitles: List of subtitle entries output_path: Path to output SRT file

src/video/deconstruction/get_character.py:275

FunctionReview_audio_video_alignment

Review Execution by raising this exception (Signal that the task is Finished).

src/Reviewer.py:52

FunctionReview_timeline

Review Execution by raising this exception (Signal that the task is Finished).

src/Reviewer.py:47

Method__init__

(self, video_caption_path, video_scene_path, audio_caption_path, output_path, max_iterations, video_path=None,

src/core.py:804

Method__init__

(self, video_caption_path, video_scene_path, audio_caption_path, output_path, max_iterations, video_path=None,

src/core.py:1551

Method__init__

(self, video_scene_path, audio_caption_path, output_path, video_path=None, subtitle_path=None, main_character=

src/Screenwriter_scene_short.py:1015

Method__init__

Initialize ReviewerAgent. Args: frame_folder_path: Path to extracted video frames for review. video_path: Pa

src/Reviewer.py:251

Method__init__

音频感官关键点检测器 - 支持多种检测方法 Args: detection_method: 检测方法 ("downbeat", "pitch", "mel_energy")

src/audio/audio_Madmom.py:394

Method__init__

(self)

src/audio/interactive/structure.py:19

Method__init__

(self)

src/audio/interactive/detectors.py:40

Method__init__

(self)

src/audio/interactive/caption.py:12

Method__init__

Args: vr: Dict returned by decode_video_to_frames, containing 'video_reader' (decord VideoReader) and 'frame_indi

src/video/deconstruction/scene_analysis_video.py:109

Method__init__

(self, model_name='all-MiniLM-L6-v2')

src/video/deconstruction/scene_merge.py:13

Function_audio_to_base64_mp3

Convert audio file to base64-encoded MP3 for cloud API submission.

src/audio/litellm_client.py:60

Method_caption_one

(in_path, out_path, scene_data, frames, semaphore)

src/video/deconstruction/scene_analysis_video.py:354

Method_detect_protagonist_in_frames_vlm

Call VLM once for multiple frames and return per-frame results in order.

src/Reviewer.py:545

Function_find_split_points_near_midpoints

在均分点附近搜索最接近的关键点作为分割点（参考 interactive 逻辑）。 Args: start_time: 片段开始时间 end_time: 片段结束时间 num_parts: 需要分成的部分数 all_k

src/audio/audio_caption_madmom.py:398

Function_patched_dbn_process

修复 NumPy 2.x 兼容性的 DBNDownBeatTrackingProcessor.process 方法

src/audio/audio_Madmom.py:77

Function_preprocess_allow_none

(self, payload)

src/audio/interactive/utils.py:25

Function_read_stdout

Read process stdout line-by-line into queue. Sentinel None signals EOF.

app.py:451

Function_read_video_decord

read video using decord.VideoReader Args: ele (dict): a dict contains the configuration of video. support keys: - vid

src/utils/video_tools.py:292

Function_read_video_torchcodec

read video using torchcodec.decoders.VideoDecoder Args: ele (dict): a dict contains the configuration of video. support keys:

src/utils/video_tools.py:339

Function_read_video_torchvision

read video using torchvision.io.read_video Args: ele (dict): a dict contains the configuration of video. support keys:

src/utils/video_tools.py:189

Method_run_all

(task_list, pbar)

src/video/deconstruction/scene_analysis_video.py:361

Function_run_scenedetect_segment

Worker function for parallel scenedetect. Returns scene list for one segment.

src/video/preprocess/video_utils.py:187

Method_run_worker

(self, shot, sec_idx, shot_idx, guidance_text=None, forbidden_time_ranges=None)

src/core.py:1639

Function_scene_number

(filename: str)

src/utils/media_utils.py:237

Method_to_sec

(t)

src/Screenwriter_scene_short.py:1081

Function_update_method_groups

(selected_methods)

src/audio/interactive/interface.py:223

Functionadd_filtered_keypoint

Add a keypoint to filtered list

src/audio/interactive/interface.py:793

Functionapply_filtered_type_filter

Apply type filter to filtered keypoints table

src/audio/interactive/interface.py:913

Functionapply_raw_type_filter

Apply type filter to raw keypoints table

src/audio/interactive/interface.py:892

Functionapply_type_filter

Apply type filter - update info table and return script to filter markers.

src/audio/interactive/logic.py:218

Methodcheck_aesthetic_quality

Check aesthetic quality of a video clip using VLM analysis. Analyzes visual appeal, lighting, composition, colors, and cinematography

Functionclear_structure_cache

Clear structure analysis cache (memory and disk)

src/audio/interactive/interface.py:681

Functioncommit

Call this function to finalize the shot selection and save the result. NOTE: You MUST call commit first to validate the shot before calling t

src/core.py:124

Functiondelete_filtered_keypoint

Delete a keypoint from filtered list

src/audio/interactive/interface.py:830

Functionderive_shot_plan_path

(video_path: str, audio_path: str, instruction: str)

app.py:392

Methoddetect

Detect keypoints using Madmom

src/audio/interactive/detectors.py:7

Functiondetect_keypoints_madmom_from_params

Convenience wrapper: accept a params dict (e.g. loaded from UI settings).

src/audio/madmom_api.py:242

Functiondoc

Annotate a variable with a description.

src/func_call_shema.py:37

Functionedit_keypoint

Unified handler for Adding/Deleting keypoints

src/audio/interactive/logic.py:135

Functionescape_drawtext

Escape text for ffmpeg drawtext filter.

render/render_video.py:108

Methodfield_title_should_be_set

(self, schema)

src/func_call_shema.py:117

Functionfilter_by_type

按关键点类型进行过滤或增强关键点类型包括: - "Downbeat" (重拍): 节奏上的强拍 Args: keypoints: 原始关键点列表 preferred_types: 优先类型列表，支持部分匹配

src/audio/audio_Madmom.py:915

Functionfine_grained_shot_trimming

Analyze a video clip time range and return detailed scene information and usability assessment. Returns: A JSON string with

src/core.py:449

Functiongather_clip_frames

( video_frame_folder, clip_secs: int, subtitle_file_path: str = None )

src/video/deconstruction/video_caption.py:38

Functiongather_clip_frames_from_long_shots

Gather frames based on physical shot boundaries (long_shots). Reads frames directly from a decord VideoReader (in-memory, no disk I/O). I

src/video/deconstruction/video_caption.py:92

Methodgenerate

( self, schema, mode="validation" )

src/func_call_shema.py:101

Functiongenerate_caption_callback

Generate captions and save to JSON

src/audio/interactive/interface.py:935

Methodget_downbeat_activation

获取时间t处的downbeat激活值

src/audio/audio_Madmom.py:640

Methodget_protagonist_frame_data

Get frame-level protagonist detection data for a time range. Returns structured data instead of a summary string. Args:

src/Reviewer.py:609

Functionget_returns_schema

(func: Callable)

src/func_call_shema.py:69

Methodget_schema_from_definitions

( self, json_ref )

src/func_call_shema.py:109

Methodhhmmss_to_sec

(t)

src/core.py:1358

Functionload_audio_no_librosa

(file, sr=16000)

src/audio/interactive/utils.py:44

Functionnatural_sort_key

Key function for natural (human) sort order (e.g. clip_2 before clip_10).

src/utils/media_utils.py:215

Functionpreview_segments

Preview how segments will be merged/split based on current keypoints and parameters. Returns a markdown table showing the resulting segments

src/audio/interactive/logic.py:254

Functionprocess_mm_info_no_librosa

Process multimodal information without using librosa. This is a replacement for qwen_omni_utils.process_mm_info that doesn't depend

src/audio/audio_utils.py:247

Functionreset_all_params

()

src/audio/interactive/interface.py:1087

Functionreview_clip

Check if the proposed time range overlaps with any previously used clips. You MUST call this tool BEFORE calling finish to ensure no duplicat

src/core.py:395

Functionrun_analysis

Unified handler for running analysis

src/audio/interactive/logic.py:12

Functionrun_asr_and_character_id

Thread A: ASR + Character ID (film only). Sets asr_done_event when complete.

local_run.py:180

Functionrun_audio_analysis

Thread C: Audio Analysis — fully independent of video pipeline.

local_run.py:338

Functionrun_filter_keypoints

Filter keypoints by sections

src/audio/interactive/interface.py:709

Functionrun_madmom_detection

Run Madmom detection and return raw keypoints

src/audio/interactive/interface.py:431

Functionrun_structure_analysis

Run Omni structure analysis

src/audio/interactive/interface.py:618

Functionrun_video_captioning

Thread B: Video Captioning → Scene Merge → Scene Analysis. For film: waits for ASR to complete first. For vlog: starts immediately.

local_run.py:254

Functionsave_current_params

( methods, bpb, dbn_th, correct_beats, min_bpm, max_bpm, num_tempi, transition_lambda,

src/audio/interactive/interface.py:1014

Functionsemantic_neighborhood_retrieval

Retrieves shot information from specified scenes. You can optionally specify which scenes to search by passing a 'related_scenes' list.

src/core.py:274

Functiontimecode_to_seconds

Convert timecode HH:MM:SS.mmm to seconds.

src/utils/time_format_convert.py:1

← previous301–377 of 377, ranked by callers