Module http 

HTTP scraping adapter with anti-bot capabilities HTTP scraping adapter with anti-bot features

Implements the ScrapingService port using reqwest with:

• Realistic browser headers and User-Agent rotation
• Cookie jar persistence across requests in a session
• Exponential backoff retry (up to 3 attempts)
• Configurable timeouts
• Optional proxy support
• T112 catalogue-fingerprint rejection (default-on): outbound User-Agents are matched against a deny-list of known library banners (requests, urllib3, httpx, Scrapy, axios, node-fetch, curl <8.4). A catalogue hit is rejected before hitting the wire unless the caller opts in via HttpConfig::allow_plain_http. This is the safe default — see the module-level docs for HttpAdapterError::PlainJa4Rejected.

Example

use stygian_graph::adapters::http::{HttpAdapter, HttpConfig};
use stygian_graph::ports::{ScrapingService, ServiceInput};
use serde_json::json;

let adapter = HttpAdapter::with_config(HttpConfig::default());
let input = ServiceInput {
    url: "https://httpbin.org/get".to_string(),
    params: json!({}),
};
// let result = adapter.execute(input).await.unwrap();

Structs

HttpAdapter

HTTP client adapter with anti-bot features.

HttpConfig

Configuration for the HTTP adapter

Enums

CatalogueFingerprint

Known catalogue fingerprints that HttpAdapter refuses to send by default.

HttpAdapterError

Errors specific to catalogue-rejection (T112).

StealthProfile

Stealth-profile reference — a named UA/TLS bundle.