Functions — github.com/Forget-C/Jellyfish

hub / github.com/Forget-C/Jellyfish / functions

Functions2,103 in github.com/Forget-C/Jellyfish

⨍ Functions 2,103◇ Types & classes 497↳ Endpoints 138

↓ 1 callersFunction_build_display_name

(filename: str, name: str | None)

backend/app/services/studio/files.py:37

↓ 1 callersFunction_build_extraction_state

( *, shot: Shot, asset_candidate_total: int, dialogue_candidate_total: int, pending_asset_

backend/app/services/studio/shots.py:34

↓ 1 callersFunction_build_frame_specific_guidance

按首帧/关键帧/尾帧生成专项提示，拉开三类帧职责差异。

backend/app/services/film/shot_frame_prompt_tasks.py:301

↓ 1 callersFunction_build_guidance_suffix

生成一段稳定的镜头执行约束，供模板渲染结果补强使用。

backend/app/services/studio/shot_video_prompt_pack.py:103

↓ 1 callersFunction_build_image_body

(inp: ImageGenerationInput)

backend/app/core/integrations/volcengine/images.py:87

↓ 1 callersFunction_build_retry_guidance

(issues: list[str])

backend/app/services/film/shot_frame_prompt_tasks.py:604

↓ 1 callersFunction_build_screen_direction_guidance

构造视频镜头的人物朝向、视线与左右轴线建议。

backend/app/services/studio/shot_video_prompt_pack.py:299

↓ 1 callersFunction_build_screen_direction_guidance

生成人物朝向、视线与左右轴线建议。

backend/app/services/film/shot_frame_prompt_tasks.py:232

↓ 1 callersMethod_build_structured_chain

构建 structured output chain（优先 ToolStrategy；缺 langchain 时退回 with_structured_output）。

backend/app/chains/agents/base.py:282

↓ 1 callersFunction_build_subject_priority

( *, characters: list[Character], scenes: list[Scene], props: list[Prop], costumes: list[C

backend/app/services/film/shot_frame_prompt_tasks.py:105

↓ 1 callersFunction_build_sync_engine

()

backend/app/core/db_sync.py:24

↓ 1 callersFunction_count_candidates

统计镜头候选项总数和未处理数量。这里显式走 SQL 计数，避免在异步场景下访问 relationship 触发隐式懒加载。

backend/app/services/studio/shot_status.py:28

↓ 1 callersFunction_count_candidates_sync

(db: Session, *, shot_id: str)

backend/app/services/studio/shot_status.py:99

↓ 1 callersFunction_count_dialogue_candidates

统计镜头对白候选项总数和未处理数量。

backend/app/services/studio/shot_status.py:47

↓ 1 callersFunction_count_dialogue_candidates_sync

(db: Session, *, shot_id: str)

backend/app/services/studio/shot_status.py:113

↓ 1 callersFunction_count_pending_candidates

(db: AsyncSession, *, shot_id: str)

backend/app/services/studio/shot_video_readiness.py:50

↓ 1 callersMethod_create_task

发起供应商创建任务请求，并设置 self._provider_task_id。

backend/app/core/tasks/video_generation_tasks.py:52

↓ 1 callersMethod_create_task

发起供应商请求（或委托 adapter）。

backend/app/core/tasks/image_generation_tasks.py:56

↓ 1 callersFunction_default_model_id

(settings_row: ModelSettings | None, category: ModelCategoryKey)

backend/app/services/llm/runtime.py:15

↓ 1 callersFunction_detect_file_type

(filename: str)

backend/app/services/studio/files.py:28

↓ 1 callersFunction_dialogue_line_mode_str

(mode: Any)

backend/app/services/studio/shot_extraction_draft.py:45

↓ 1 callersFunction_ensure_chapter_optional

(db: AsyncSession, *, project_id: str, chapter_id: str | None)

backend/app/utils/project_links.py:21

↓ 1 callersFunction_ensure_project_exists

(db: AsyncSession, project_id: str)

backend/app/utils/project_links.py:16

↓ 1 callersFunction_ensure_shot_optional

(db: AsyncSession, *, project_id: str, chapter_id: str | None, shot_id: str | None)

backend/app/utils/project_links.py:31

↓ 1 callersMethod_execute_phase

(self, task_id: str, run_args: dict[str, Any])

backend/app/services/worker/task_executor.py:107

↓ 1 callersFunction_extract_context_names

(context_text: str | None)

backend/app/services/film/shot_frame_prompt_tasks.py:556

↓ 1 callersFunction_extract_first_json_object

尽量从文本中提取第一个 JSON 对象/数组片段。

backend/app/chains/agents/base.py:30

↓ 1 callersFunction_find_active_analysis_task

( db: AsyncSession, *, relation_type: str, relation_entity_id: str, )

backend/app/services/script_processing_tasks.py:235

↓ 1 callersFunction_find_character_id

(q: str)

backend/app/services/studio/entity_existence.py:51

↓ 1 callersFunction_find_linked_costume

(q: str)

backend/app/services/studio/entity_existence.py:91

↓ 1 callersFunction_find_linked_prop

(q: str)

backend/app/services/studio/entity_existence.py:65

↓ 1 callersFunction_find_linked_scene

(q: str)

backend/app/services/studio/entity_existence.py:78

↓ 1 callersFunction_format_action_beat_phase_summary

格式化动作拍点阶段摘要，供 agent 输入与调试预览复用。

backend/app/services/film/shot_frame_prompt_tasks.py:343

↓ 1 callersFunction_has_active_video_task

(db: AsyncSession, *, shot_id: str)

backend/app/services/studio/shot_video_readiness.py:64

↓ 1 callersFunction_has_sequential_reaction_chain

判断文本是否包含明显的连续反应链，供首帧时间切片加权使用。

backend/app/services/film/shot_frame_prompt_tasks.py:291

↓ 1 callersFunction_infer_file_type_from_content_type

(content_type: str | None)

backend/app/utils/files.py:39

↓ 1 callersFunction_load_json_like

尽量解析 LLM 返回的 JSON / JSON-like 文本。

backend/app/chains/agents/base.py:87

↓ 1 callersMethod_load_run_args

(self, task_id: str)

backend/app/services/worker/task_executor.py:238

↓ 1 callersFunction_main

初始化对象存储（例如创建 bucket）。

backend/init_storage.py:6

↓ 1 callersFunction_main

初始化数据库，创建所有表。

backend/init_db.py:6

↓ 1 callersMethod_mark_cancelled_if_requested

(self, task_id: str)

backend/app/services/worker/task_executor.py:278

↓ 1 callersMethod_mark_running_and_load_run_args

(self, task_id: str)

backend/app/services/worker/task_executor.py:94

↓ 1 callersMethod_normalize

规范化脚本分割结果。

backend/app/chains/agents/script_divider_agent.py:81

↓ 1 callersFunction_normalize_optional_text

归一化可选文本参数：空字符串视为未设置。

backend/app/services/film/generated_video.py:130

↓ 1 callersFunction_pack_variables

为 DB 模板暴露稳定变量名。同时保留扁平变量和 `pack`，让模板可以逐步迁移到更结构化的写法。

backend/app/services/studio/shot_video_prompt_pack.py:57

↓ 1 callersFunction_parse_openai_images_payload

(data: dict[str, Any])

backend/app/core/integrations/openai/images.py:130

↓ 1 callersFunction_parse_python_call_kwargs

解析 `Foo(a=1, b='x')` / `dict(a=1)` 这类 kwargs 风格输出。

backend/app/chains/agents/base.py:66

↓ 1 callersFunction_parse_volcengine_images_payload

(data: dict[str, Any])

backend/app/core/integrations/volcengine/images.py:111

↓ 1 callersFunction_persist_images_to_assets

将图片生成结果落库到 FileItem 与业务图片表。

backend/app/services/studio/image_task_runner.py:57

↓ 1 callersFunction_pick_override

(model: str | None)

backend/app/core/integrations/openai/video_capabilities.py:34

↓ 1 callersFunction_pick_override

(model: str | None)

backend/app/core/integrations/openai/image_capabilities.py:32

↓ 1 callersFunction_pick_override

(model: str | None)

backend/app/core/integrations/volcengine/video_capabilities.py:34

↓ 1 callersFunction_pick_override

(model: str | None)

backend/app/core/integrations/volcengine/image_capabilities.py:51

↓ 1 callersMethod_poll_and_get_result

轮询至终态并解析为 VideoGenerationResult。

backend/app/core/tasks/video_generation_tasks.py:56

↓ 1 callersMethod_poll_and_get_result

解析为统一结果。

backend/app/core/tasks/image_generation_tasks.py:60

↓ 1 callersFunction_quote_unquoted_object_keys

为常见 JSON-like 输出补齐未加引号的对象键名。

backend/app/chains/agents/base.py:42

↓ 1 callersFunction_record_executor_dispatch

(task_id: str, *, executor_type: str, executor_task_id: str | None)

backend/app/tasks/execute_task.py:23

↓ 1 callersFunction_reference_frames_ready

( db: AsyncSession, *, shot_id: str, reference_mode: str, )

backend/app/services/studio/shot_video_readiness.py:77

↓ 1 callersFunction_render_template

(content: str, variables: dict[str, Any])

backend/app/services/studio/shot_video_prompt_pack.py:97

↓ 1 callersFunction_repair_json_like

修复 LLM 常见 JSON 格式问题。

backend/app/chains/agents/base.py:56

↓ 1 callersFunction_replace_json_literals_for_python

将 JSON 字面量替换为 Python 字面量，便于 ast.literal_eval 解析。

backend/app/chains/agents/base.py:48

↓ 1 callersFunction_require_provider_and_model_sync

( db: Session, *, category: ModelCategoryKey, )

backend/app/services/llm/runtime.py:25

↓ 1 callersFunction_resolve_dialog_index

(db: AsyncSession, *, shot_id: str, preferred_index: int)

backend/app/services/studio/shot_extracted_dialogue_candidates.py:256

↓ 1 callersFunction_resolve_download_media_type

(filename: str)

backend/app/services/studio/files.py:44

↓ 1 callersFunction_resolve_navigation_targets

( db: AsyncSession, *, link_map: dict[str, GenerationTaskLink], )

backend/app/core/task_manager/stores.py:68

↓ 1 callersMethod_resolve_timeout_seconds

(self, run_args: dict[str, Any])

backend/app/services/worker/task_executor.py:268

↓ 1 callersFunction_resolve_video_prompt_template

( db: AsyncSession, *, template_id: str | None, )

backend/app/services/studio/shot_video_prompt_pack.py:363

↓ 1 callersMethod_run_async

(self, task_id: str, run_args: dict[str, Any], timeout_seconds: float | None)

backend/app/services/worker/task_executor.py:259

↓ 1 callersMethod_run_async_with_runtime

( self, task_id: str, run_args: dict[str, Any], timeout_seconds: float | None,

backend/app/services/worker/task_executor.py:245

↓ 1 callersFunction_scope_filters

返回与 FileUsage 行组合的 WHERE 条件（不含 project_id 基条件）。

backend/app/services/studio/file_usages.py:122

↓ 1 callersFunction_score_director_guidance_item

为 guidance 句子打分，优先保留更能稳定镜头连续性的约束。

backend/app/services/film/shot_frame_prompt_tasks.py:366

↓ 1 callersFunction_seed_fear_reaction_shot_graph

(db: AsyncSession)

backend/tests/test_shot_frame_prompt_tasks_service.py:166

↓ 1 callersFunction_seed_image_model

(db: _FakeLlmDB, *, provider_id: str = "p-image", model_id: str = "m-image")

backend/tests/test_llm_api_responses.py:97

↓ 1 callersFunction_seed_prompt

( db: _FakePromptDB, *, prompt_id: str = "tpl-1", is_default: bool = False, is_system: boo

backend/tests/test_api_response_envelopes.py:45

↓ 1 callersFunction_seed_shot

( db: _FakeStudioDB, *, shot_id: str = "shot-1", chapter_id: str = "ch-1", )

backend/tests/test_studio_api_responses.py:131

↓ 1 callersFunction_seed_sync_graph

同步版本的最小项目/章节/镜头图。

backend/tests/test_shot_semantic_defaults_service.py:49

↓ 1 callersFunction_seed_video_model

(db: _FakeLlmDB, *, provider_id: str = "p-video", model_id: str = "m-video")

backend/tests/test_llm_api_responses.py:85

↓ 1 callersFunction_settings_model_id

(settings_row: ModelSettings | None, category: ModelCategoryKey)

backend/app/services/llm/resolver.py:15

↓ 1 callersFunction_status_value

(value: ProviderStatus | str | None)

backend/app/services/llm/provider_resolver.py:25

↓ 1 callersFunction_to_app_mode

(mode: str | GenerationDeliveryMode)

backend/app/core/task_manager/stores.py:32

↓ 1 callersFunction_to_db_mode

(mode: DeliveryMode)

backend/app/core/task_manager/stores.py:40

↓ 1 callersFunction_validate_provider_status

(provider: Provider)

backend/app/services/llm/provider_resolver.py:31

↓ 1 callersFunction_video_model_and_provider_ready

(db: AsyncSession)

backend/app/services/studio/shot_video_readiness.py:101

↓ 1 callersMethodacceptExtractedDialogueCandidateApiV1StudioShotsExtractedDialogueCandidatesCandidateIdAcceptPatch

* 接受镜头提取对白候选项 * @returns ApiResponse_ShotPreparationMutationResultRead_ Successful Response * @throws ApiError

front/src/services/generated/services/StudioShotsService.ts:378

↓ 1 callersMethodanalyzeCharacterPortraitAsyncApiV1ScriptProcessingAnalyzeCharacterPortraitAsyncPost

* 异步分析人物画像缺失信息 * 创建人物画像分析任务并立即返回 task_id；前端可通过任务状态接口轮询。 * @returns ApiResponse_AsyncTaskCreateRead_ Successful Response * @throws ApiEr

front/src/services/generated/services/ScriptProcessingService.ts:206

↓ 1 callersMethodanalyzeCostumeInfoAsyncApiV1ScriptProcessingAnalyzeCostumeInfoAsyncPost

* 异步分析服装信息缺失项 * 创建服装信息分析任务并立即返回 task_id；前端可通过任务状态接口轮询。 * @returns ApiResponse_AsyncTaskCreateRead_ Successful Response * @throws ApiErr

front/src/services/generated/services/ScriptProcessingService.ts:332

↓ 1 callersMethodanalyzePropInfoAsyncApiV1ScriptProcessingAnalyzePropInfoAsyncPost

* 异步分析道具信息缺失项 * 创建道具信息分析任务并立即返回 task_id；前端可通过任务状态接口轮询。 * @returns ApiResponse_AsyncTaskCreateRead_ Successful Response * @throws ApiErr

front/src/services/generated/services/ScriptProcessingService.ts:248

↓ 1 callersMethodanalyzeSceneInfoAsyncApiV1ScriptProcessingAnalyzeSceneInfoAsyncPost

* 异步分析场景信息缺失项 * 创建场景信息分析任务并立即返回 task_id；前端可通过任务状态接口轮询。 * @returns ApiResponse_AsyncTaskCreateRead_ Successful Response * @throws ApiErr

front/src/services/generated/services/ScriptProcessingService.ts:290

↓ 1 callersFunctionappend

(key: string, value: any)

front/src/services/generated/core/request.ts:53

↓ 1 callersFunctionapplyCancelToRelationTaskState

( currentTask: RelationTaskState, data?: TaskCancelLike | null, )

front/src/pages/aiStudio/project/ProjectWorkbench/chapterDivisionTasks.ts:96

↓ 1 callersFunctionapplyCardImage

(frameType: PromptFrameType, fileId: string)

front/src/pages/aiStudio/chapter/ChapterStudio.tsx:4675

↓ 1 callersFunctionapplyDefaultBaseUrlForDisplayName

(displayName: string)

front/src/pages/aiStudio/models/ProvidersTab.tsx:122

↓ 1 callersFunctionapply_division_result

( db: Session, *, chapter_id: str, result: ScriptDivisionResult, )

backend/app/services/script_processing_worker.py:263

↓ 1 callersMethodapply_result

(self, ctx: WorkerTaskContext, run_args: dict[str, Any], result: Any)

backend/app/services/worker/task_executor.py:175

↓ 1 callersFunctionapply_shot_semantic_defaults_from_draft_sync

同步任务执行器版本：将镜头语言默认建议回写到 ShotDetail。

backend/app/services/studio/shot_semantic_defaults.py:103

↓ 1 callersFunctionassetDetailUrl

(kind: AssetKind, id: string, projectId: string)

front/src/pages/aiStudio/shots/components/ChapterShotAssetConfirmation.tsx:35

↓ 1 callersFunctionasset_prompt_category

( *, relation_type: str, is_front_view: bool, )

backend/app/services/studio/image_tasks.py:263

↓ 1 callersFunctionbase64

(str: string)

front/src/services/generated/core/request.ts:41

↓ 1 callersFunctionbatchOpenDownloads

(urls: string[])

front/src/pages/aiStudio/project/ProjectWorkbench/tabs/FilesTab.tsx:17

↓ 1 callersFunctionbootstrap_builtin_providers

()

backend/app/services/llm/provider_bootstrap.py:7

← previous next →601–700 of 2,103, ranked by callers