content-find-agent-new/content_agent/schemas.py

from __future__ import annotations

from typing import Any, Literal

from pydantic import BaseModel, ConfigDict, Field, model_validator

from content_agent.constants import DEFAULT_STRATEGY_VERSION


class RunStartRequest(BaseModel):
    model_config = ConfigDict(extra="forbid")

    run_id: str | None = Field(
        default=None,
        description="Optional fixed run_id (V3-M5: deterministic replay for concurrency-consistency tests).",
    )
    source: str | None = Field(
        default=None,
        description="Optional source_context.json or DemandAgent demand_content.json path.",
    )
    demand_content_id: int | None = Field(
        default=None,
        description="Optional content-deconstruction-supply.demand_content id.",
    )
    run_label: str | None = Field(
        default=None,
        description="Optional ext_data.run_label selector in demand_content.",
    )
    platform: str = Field(default="douyin", description="Requested platform.")
    platform_mode: Literal["mock", "real"] = Field(
        default="real",
        description="mock uses deterministic fixtures; real calls the configured platform API.",
    )
    strategy_version: str = Field(default=DEFAULT_STRATEGY_VERSION)

    @model_validator(mode="after")
    def validate_source_selector(self) -> "RunStartRequest":
        selectors = [
            bool(self.source),
            self.demand_content_id is not None,
            bool(self.run_label),
        ]
        if sum(selectors) > 1:
            raise ValueError("source, demand_content_id, and run_label are mutually exclusive")
        return self


class RunStartResponse(BaseModel):
    run_id: str
    policy_run_id: str
    status: str
    policy_bundle_id: str
    strategy_version: str
    platform: str
    platform_mode: str
    output_dir: str


class RunSummaryResponse(BaseModel):
    run_id: str
    policy_run_id: str | None = None
    status: str
    current_step: str
    output_dir: str
    files: dict[str, bool]
    validation_status: str
    errors: list[str] = Field(default_factory=list)


class RecordsResponse(BaseModel):
    run_id: str
    records: list[dict[str, Any]]


class JsonFileResponse(BaseModel):
    run_id: str
    data: dict[str, Any]


class ValidationResponse(BaseModel):
    run_id: str
    status: str
    findings: list[dict[str, Any]]


class RunListItem(BaseModel):
    run_id: str
    policy_run_id: str | None = None
    status: str | None = None
    current_step: str | None = None
    platform: str | None = None
    platform_mode: str | None = None
    content_format: str | None = None
    strategy_version: str | None = None
    demand_content_id: int | None = None
    demand_name: str | None = None
    demand_desc: str | None = None
    validation_status: str | None = None
    error_code: str | None = None
    started_at: str | None = None
    completed_at: str | None = None


class RunListResponse(BaseModel):
    items: list[RunListItem]
    page: int
    page_size: int
    total: int
    data_origin: str


class DashboardResponse(BaseModel):
    run_id: str
    summary: dict[str, Any]
    counts: dict[str, int]
    files: dict[str, bool]
    runtime_files: list[dict[str, Any]]
    validation: dict[str, Any]
    final_output_summary: dict[str, Any]
    strategy_review_status: str
    business_summary: dict[str, Any]
    stage_conclusions: list[dict[str, Any]]
    rule_application_summary: list[dict[str, Any]]
    walk_graph: dict[str, Any]
    primary_failure_reason: dict[str, Any] | None = None
    technical_refs: dict[str, Any]
    data_origin: str
    links: dict[str, str]


class RuntimeFilesResponse(BaseModel):
    run_id: str
    files: list[dict[str, Any]]
    data_origin: str


class RuntimeFileResponse(BaseModel):
    run_id: str
    filename: str
    data_origin: str
    data: dict[str, Any] | None = None
    records: list[dict[str, Any]] | None = None
    offset: int | None = None
    limit: int | None = None
    total: int | None = None


class QueryListResponse(BaseModel):
    run_id: str
    items: list[dict[str, Any]]
    total: int
    data_origin: str


class ConfigFileResponse(BaseModel):
    source_file: str
    data: dict[str, Any]


class PlatformDescriptor(BaseModel):
    # 平台展示元数据(从 platform_profiles 派生,前端按此渲染,不再写死抖音)
    platform: str
    label: str
    status: str | None = None
    # V3 历史展示字段:来自 profile.heat.signals,保留兼容,不等同于 V4 平台表现分。
    heat_fields: list[str] = []
    # V4 可观测字段:来自 profile.observable_fields,用于解释平台可观测表现。
    observable_fields: list[str] = []


class PlatformCatalogResponse(BaseModel):
    platforms: dict[str, PlatformDescriptor]


class TimelineResponse(BaseModel):
    run_id: str
    items: list[dict[str, Any]]
    total: int
    data_origin: str
    summary: dict[str, Any]


class ContentItemsResponse(BaseModel):
    run_id: str
    items: list[dict[str, Any]]
    total: int
    data_origin: str


class FlowLedgerResponse(BaseModel):
    run_id: str
    data_origin: str
    data_origin_label: str
    summary: dict[str, Any]
    demand_summary: dict[str, Any] | None = None
    rows: list[dict[str, Any]]
    extension_queries: list[dict[str, Any]]
    debug: dict[str, Any] | None = None


class FlowLedgerVideosResponse(BaseModel):
    run_id: str
    query: dict[str, Any] | None = None
    videos: list[dict[str, Any]]
    summary: dict[str, Any]
    data_origin: str
    data_origin_label: str


class FlowLedgerWalkResponse(BaseModel):
    run_id: str
    query: dict[str, Any] | None = None
    actions: list[dict[str, Any]]
    lanes: dict[str, list[dict[str, Any]]]
    extension_queries: list[dict[str, Any]]
    summary: dict[str, Any]
    data_origin: str
    data_origin_label: str


class FlowLedgerVideoResponse(BaseModel):
    run_id: str
    video: dict[str, Any] | None = None
    source_paths: list[dict[str, Any]] = Field(default_factory=list)
    decision: dict[str, Any] | None = None
    data_origin: str
    data_origin_label: str


class FlowLedgerVideoWalkResponse(BaseModel):
    run_id: str
    root_video: dict[str, Any] | None = None
    actions: list[dict[str, Any]]
    summary: dict[str, Any]
    data_origin: str
    data_origin_label: str


class AigcPushRecordListResponse(BaseModel):
    items: list[dict[str, Any]]
    total: int
    data_origin: str