Trait Engine 

pub trait Engine:
    Send
    + Sync
    + 'static {
    // Required method
    fn scrape<'a>(
        &'a self,
        url: &'a str,
        opts: &'a ScrapeOptions,
    ) -> BoxFuture<'a, Result<Document, ApiError>>;

    // Provided methods
    fn parse<'a>(
        &'a self,
        _file: Upload,
        _opts: &'a ScrapeOptions,
    ) -> BoxFuture<'a, Result<Document, ApiError>> { ... }
    fn map<'a>(
        &'a self,
        _req: &'a MapRequest,
    ) -> BoxFuture<'a, Result<Vec<MapLink>, ApiError>> { ... }
    fn crawl<'a>(
        &'a self,
        _req: &'a CrawlRequest,
        _sink: Arc<dyn CrawlSink>,
    ) -> BoxFuture<'a, Result<(), ApiError>> { ... }
    fn search<'a>(
        &'a self,
        _req: &'a SearchRequest,
    ) -> BoxFuture<'a, Result<Value, ApiError>> { ... }
    fn extract<'a>(
        &'a self,
        _req: &'a ExtractRequest,
    ) -> BoxFuture<'a, Result<Value, ApiError>> { ... }
    fn agent_check(&self, _req: &AgentRequest) -> Result<(), ApiError> { ... }
    fn agent<'a>(
        &'a self,
        _req: &'a AgentRequest,
        _sink: Arc<dyn AgentSink>,
    ) -> BoxFuture<'a, AgentOutcome> { ... }
    fn supported_formats(&self) -> &'static [&'static str] { ... }
    fn supports_actions(&self) -> bool { ... }
    fn actions_unavailable(&self) -> &'static str { ... }
    fn supports_interact(&self) -> bool { ... }
    fn scrape_session<'a>(
        &'a self,
        _url: &'a str,
        _opts: &'a ScrapeOptions,
        _keep: bool,
        _owner: Option<&'a str>,
    ) -> BoxFuture<'a, Result<Document, ApiError>> { ... }
    fn observe<'a>(
        &'a self,
        _url: &'a str,
        _instruction: &'a str,
    ) -> BoxFuture<'a, Result<Observed, ApiError>> { ... }
    fn find_similar<'a>(
        &'a self,
        _url: &'a str,
        _selector: &'a str,
        _opts: &'a ScrapeOptions,
    ) -> BoxFuture<'a, Result<SimilarData, ApiError>> { ... }
    fn interact<'a>(
        &'a self,
        _id: &'a str,
        _req: &'a InteractRequest,
        _owner: Option<&'a str>,
    ) -> BoxFuture<'a, Result<InteractResponse, ApiError>> { ... }
    fn stop_interact<'a>(
        &'a self,
        _id: &'a str,
        _owner: Option<&'a str>,
    ) -> BoxFuture<'a, Result<InteractStopped, ApiError>> { ... }
    fn admitted_concurrency(&self) -> Option<usize> { ... }
}

Required Methods

fn scrape<'a>( &'a self, url: &'a str, opts: &'a ScrapeOptions, ) -> BoxFuture<'a, Result<Document, ApiError>>

Provided Methods

fn parse<'a>( &'a self, _file: Upload, _opts: &'a ScrapeOptions, ) -> BoxFuture<'a, Result<Document, ApiError>>

/v2/parse：解析调用方上传的文件（PDF、HTML），返回和 scrape 相同的文档。 opts 已经过 crate::check_parse_options 校验。

fn map<'a>( &'a self, _req: &'a MapRequest, ) -> BoxFuture<'a, Result<Vec<MapLink>, ApiError>>

fn crawl<'a>( &'a self, _req: &'a CrawlRequest, _sink: Arc<dyn CrawlSink>, ) -> BoxFuture<'a, Result<(), ApiError>>

整站爬取：抓到一页、出错一页、发现新 URL 时都通过 sink 回报，直到爬完或被取消。

fn search<'a>( &'a self, _req: &'a SearchRequest, ) -> BoxFuture<'a, Result<Value, ApiError>>

/v2/search：返回 Firecrawl 的 data（{ "web": [...] }）。

fn extract<'a>( &'a self, _req: &'a ExtractRequest, ) -> BoxFuture<'a, Result<Value, ApiError>>

/v2/extract：返回提取出的 JSON。

fn agent_check(&self, _req: &AgentRequest) -> Result<(), ApiError>

/v2/agent 开跑前的检查：没配 LLM、没有 urls 又没配搜索时，提交时就报错。

fn agent<'a>( &'a self, _req: &'a AgentRequest, _sink: Arc<dyn AgentSink>, ) -> BoxFuture<'a, AgentOutcome>

/v2/agent：跑完一次（中途出错也回已经攒下的证据，错误在 AgentOutcome::error）。

fn supported_formats(&self) -> &'static [&'static str]

引擎支持的 formats；请求了不支持的格式时返回 UNSUPPORTED_OPTION，不悄悄忽略。

fn supports_actions(&self) -> bool

是否支持 actions（需要浏览器渲染）。

fn actions_unavailable(&self) -> &'static str

不支持 actions 时给调用方的说明。

fn supports_interact(&self) -> bool

是否支持 /v2/scrape/{id}/interact。支持时 /v2/scrape 走 Engine::scrape_session。

fn scrape_session<'a>( &'a self, _url: &'a str, _opts: &'a ScrapeOptions, _keep: bool, _owner: Option<&'a str>, ) -> BoxFuture<'a, Result<Document, ApiError>>

/v2/scrape：抓一页，结果带 metadata.scrapeId。keep 时保留浏览器页面（keepSession）， 否则只记下回放信息。owner 是调用方的账号，之后的 interact 只认同一个账号。

fn observe<'a>( &'a self, _url: &'a str, _instruction: &'a str, ) -> BoxFuture<'a, Result<Observed, ApiError>>

/v2/observe：页面上和指令对应的候选元素（选择器、角色、名称），返回候选和用量警告。

fn find_similar<'a>( &'a self, _url: &'a str, _selector: &'a str, _opts: &'a ScrapeOptions, ) -> BoxFuture<'a, Result<SimilarData, ApiError>>

/v2/similar：和例子结构相同的兄弟元素。

fn interact<'a>( &'a self, _id: &'a str, _req: &'a InteractRequest, _owner: Option<&'a str>, ) -> BoxFuture<'a, Result<InteractResponse, ApiError>>

fn stop_interact<'a>( &'a self, _id: &'a str, _owner: Option<&'a str>, ) -> BoxFuture<'a, Result<InteractStopped, ApiError>>

fn admitted_concurrency(&self) -> Option<usize>

统一准入现在允许的在途抓取数：批量任务的并发按它定，用户的 max_batch_concurrency 只作上限。引擎还没接准入时返回 None，API 用配置或先验。

Dyn Compatibility

This trait is dyn compatible.

In older versions of Rust, dyn compatibility was called "object safety".

Implementors