starred/minne
1
0
Fork 0
mirror of https://github.com/perstarkse/minne.git synced 2026-04-27 03:08:41 +02:00
Code Issues Packages Projects Releases 10 Wiki Activity

refactoring: working macro and generics

Browse Source
This commit is contained in:
Per Stark
2024-11-20 22:44:30 +01:00
parent 7222223c31
commit 41134cfa49
11 changed files with 198 additions and 167 deletions
Download Patch File Download Diff File Expand all files Collapse all files

41
src/models/ingress_object.rs
View File

@@ -1,9 +1,9 @@
use super::ingress_content::IngressContentError;
use crate::models::file_info::FileInfo;
use crate::storage::types::text_content::TextContent;
use serde::{Deserialize, Serialize};
use uuid::Uuid;
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 {
@@ -34,7 +34,11 @@ impl IngressObject {
/// `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 } => {
IngressObject::Url {
url,
instructions,
category,
} => {
let text = Self::fetch_text_from_url(url).await?;
let id = Uuid::new_v4();
Ok(TextContent {
@@ -44,8 +48,12 @@ impl IngressObject {
category: category.clone(),
file_info: None,
})
},
IngressObject::Text { text, instructions, category } => {
}
IngressObject::Text {
text,
instructions,
category,
} => {
let id = Uuid::new_v4();
Ok(TextContent {
id: id.to_string(),
@@ -54,8 +62,12 @@ impl IngressObject {
category: category.clone(),
file_info: None,
})
},
IngressObject::File { file_info, instructions, category } => {
}
IngressObject::File {
file_info,
instructions,
category,
} => {
let id = Uuid::new_v4();
let text = Self::extract_text_from_file(file_info).await?;
Ok(TextContent {
@@ -65,7 +77,7 @@ impl IngressObject {
category: category.clone(),
file_info: Some(file_info.clone()),
})
},
}
}
}
@@ -89,11 +101,15 @@ impl IngressObject {
}
"application/pdf" => {
// TODO: Implement PDF text extraction using a crate like `pdf-extract` or `lopdf`
Err(IngressContentError::UnsupportedMime(file_info.mime_type.clone()))
Err(IngressContentError::UnsupportedMime(
file_info.mime_type.clone(),
))
}
"image/png" | "image/jpeg" => {
// TODO: Implement OCR on image using a crate like `tesseract`
Err(IngressContentError::UnsupportedMime(file_info.mime_type.clone()))
Err(IngressContentError::UnsupportedMime(
file_info.mime_type.clone(),
))
}
"application/octet-stream" => {
let content = tokio::fs::read_to_string(&file_info.path).await?;
@@ -104,8 +120,9 @@ impl IngressObject {
Ok(content)
}
// Handle other MIME types as needed
_ => Err(IngressContentError::UnsupportedMime(file_info.mime_type.clone())),
_ => Err(IngressContentError::UnsupportedMime(
file_info.mime_type.clone(),
)),
}
}
}