Struct ExtractOptions 

pub struct ExtractOptions<'a> {

    pub raw_html: &'a str,
    pub source_url: &'a str,
    pub status_code: u16,
    pub rendered_with: Option<String>,
    pub elapsed_ms: u64,
    pub render_decision: Option<RenderDecision>,
    pub credit_cost: u32,
    pub warnings: Vec<String>,
    pub formats: &'a [OutputFormat],
    pub only_main_content: bool,
    pub include_tags: &'a [String],
    pub exclude_tags: &'a [String],
    pub css_selector: Option<&'a str>,
    pub xpath: Option<&'a str>,
    pub chunk_strategy: Option<&'a ChunkStrategy>,
    pub query: Option<&'a str>,
    pub filter_mode: Option<&'a FilterMode>,
    pub top_k: Option<usize>,
    pub domain_selectors: Option<&'a HashMap<String, String>>,
    pub captured_responses: &'a [CapturedNetworkResponse],
    pub llm_fallback: Option<LlmFallbackParams<'a>>,
    pub debug: bool,
    pub debug_sink: Option<Arc<Mutex<DebugCollector>>>,
}

Options for the high-level extraction pipeline.

Fields

raw_html: &'a str

source_url: &'a str

status_code: u16

rendered_with: Option<String>

elapsed_ms: u64

render_decision: Option<RenderDecision>

Routing decision metadata to surface to API consumers.

credit_cost: u32

Credit cost attributed to this fetch.

warnings: Vec<String>

Soft-failure warnings collected through the render chain.

formats: &'a [OutputFormat]

only_main_content: bool

include_tags: &'a [String]

exclude_tags: &'a [String]

css_selector: Option<&'a str>

CSS selector to narrow content before readability extraction.

xpath: Option<&'a str>

XPath expression to narrow content before readability extraction.

chunk_strategy: Option<&'a ChunkStrategy>

Strategy for chunking the extracted markdown.

query: Option<&'a str>

Query for chunk filtering (requires filter_mode).

filter_mode: Option<&'a FilterMode>

Filtering algorithm for chunk ranking.

top_k: Option<usize>

Number of top chunks to return (default: 5).

domain_selectors: Option<&'a HashMap<String, String>>

Per-host CSS selector overrides. Used only when the request did not supply an explicit css_selector / xpath. The selector for the source URL’s host is applied before readability narrowing.

captured_responses: &'a [CapturedNetworkResponse]

XHR/fetch responses captured during navigation. Used as a fallback content source when DOM-based extraction is low quality.

llm_fallback: Option<LlmFallbackParams<'a>>

LLM-assisted extraction fallback configuration. When the chosen candidate’s quality score is below quality_threshold and enable is true, the raw HTML (truncated to max_html_bytes) is sent to the configured LLM provider for re-extraction.

debug: bool

Opt-in extraction debug trace. When true, the extractor populates debug_sink with one DebugAttempt per extract() invocation.

debug_sink: Option<Arc<Mutex<DebugCollector>>>

Sink for debug attempts. Shared across the multi-attempt JS-escalation loop so that all attempts land in one trace.