comments, llm schema

src/models/ingress_object.rs

@@ -3,6 +3,7 @@ use serde::{Deserialize, Serialize};
 
 use super::{ingress_content::IngressContentError, text_content::TextContent};
 
+/// Knowledge object type, containing the content or reference to it, as well as metadata
 #[derive(Debug, Serialize, Deserialize, Clone)]
 pub enum IngressObject {
     Url {
@@ -21,7 +22,15 @@ pub enum IngressObject {
         category: String,
     },
 }
+
 impl IngressObject {
+    /// Creates a new `TextContent` instance from a `IngressObject`.
+    ///
+    /// # Arguments
+    /// `&self` - A reference to the `IngressObject`.
+    ///
+    /// # Returns
+    /// `TextContent` - An object containing a text representation of the object, could be a scraped URL, parsed PDF, etc.
     pub async fn to_text_content(&self) -> Result<TextContent, IngressContentError> {
         match self {
             IngressObject::Url { url, instructions, category } => {
@@ -66,6 +75,11 @@ impl IngressObject {
                 let content = tokio::fs::read_to_string(&file_info.path).await?;
                 Ok(content)
             }
+            "text/markdown" => {
+                // Read the file and return its content
+                let content = tokio::fs::read_to_string(&file_info.path).await?;
+                Ok(content)
+            }
             "application/pdf" => {
                 // TODO: Implement PDF text extraction using a crate like `pdf-extract` or `lopdf`
                 Err(IngressContentError::UnsupportedMime(file_info.mime_type.clone()))