Module multimodal 

Multi-modal content extraction (CSV, JSON, XML, images, PDFs) Multi-modal content extraction adapter

Routes non-HTML content (PDFs, images, CSV, JSON, XML) through appropriate parsers or AI vision providers depending on content type.

Content Routing

Content type	Strategy
text/csv	Parse in-process via CSV iterator
application/json	Parse + re-format via serde_json
text/xml / application/xml	Lightweight attribute extraction
image/*	Delegate to vision-capable AIProvider
application/pdf	Text extraction (requires pdf feature)
Unknown	Return raw bytes as UTF-8 string

Example

use stygian_graph::adapters::multimodal::{MultiModalAdapter, MultiModalConfig};
use stygian_graph::ports::{ScrapingService, ServiceInput};
use serde_json::json;

let adapter = MultiModalAdapter::new(MultiModalConfig::default(), None);
let input = ServiceInput {
    url: "data:text/csv,name,age\nalice,30\nbob,25".to_string(),
    params: json!({ "content_type": "text/csv" }),
};
// let output = adapter.execute(input).await.unwrap();

Structs

MultiModalAdapter

Multi-modal content extraction adapter

MultiModalConfig

Configuration for multi-modal extraction

Enums

ContentType

Detected or declared content type for multi-modal routing