Changed extraction of data from sync to async and include a get endpoint that handles all files

This commit is contained in:
krut_ni 2025-03-31 19:21:22 +02:00
parent ef8d1051ae
commit 0e4c49d3c0
3 changed files with 224 additions and 280 deletions

View File

@ -15,7 +15,7 @@ serde = { version = "1.0.217", features = ["derive"] }
serde_json = "1.0.134"
keyring = { version = "3.6.1", features = ["apple-native", "windows-native", "sync-secret-service"] }
arboard = "3.4.1"
tokio = { version = "1.42", features = ["rt", "rt-multi-thread", "macros"] }
tokio = { version = "1.42", features = ["rt", "rt-multi-thread", "macros", "process"] }
flexi_logger = "0.29.8"
log = { version = "0.4", features = ["kv"] }
once_cell = "1.20"
@ -31,12 +31,14 @@ hmac = "0.12.1"
sha2 = "0.10.8"
rcgen = { version = "0.13.2", features = ["pem"] }
file-format = "0.26.0"
calamine = { version = "0.26.1", features = ["dates"] }
chrono = "0.4.39"
calamine = "0.22"
pdfium-render = "0.8.27"
# Fixes security vulnerability downstream, where the upstream is not fixed yet:
url = "2.5"
async-stream = "0.3"
futures = "0.3"
tokio-stream = "0.1"
[target.'cfg(target_os = "linux")'.dependencies]
# See issue https://github.com/tauri-apps/tauri/issues/4470

View File

@ -1,30 +1,28 @@
use std::path::Path;
use std::io::{BufRead, BufReader};
use std::pin::Pin;
use async_stream::stream;
use base64::{engine::general_purpose, Engine as _};
use calamine::{open_workbook_auto, Reader};
use file_format::{FileFormat, Kind};
use futures::{Stream, StreamExt};
use pdfium_render::prelude::Pdfium;
use std::error::Error;
use std::fs::File;
use std::io::Read;
use std::process::Command;
use rocket::post;
use rocket::response::stream::{Event, EventStream};
use tokio::io::AsyncBufReadExt;
use tokio::process::Command;
use tokio::sync::mpsc;
use tokio_stream::wrappers::ReceiverStream;
use rocket::{State, Shutdown};
use rocket::fs::{relative, FileServer};
use rocket::form::Form;
use rocket::serde::{Serialize, Deserialize};
use rocket::tokio::sync::broadcast::{channel, Sender, error::RecvError};
use rocket::response::stream::{EventStream, Event};
use rocket::tokio::select;
use rocket::serde::Serialize;
use rocket::get;
#[derive(Debug)]
#[derive(Debug, Serialize)]
pub struct Chunk {
pub content: String,
pub metadata: Metadata,
}
#[derive(Debug)]
#[derive(Debug, Serialize)]
pub enum Metadata {
Text { line_number: usize },
Pdf { page_number: usize },
@ -37,289 +35,233 @@ const TO_MARKDOWN: &str = "markdown";
const DOCX: &str = "docx";
const ODT: &str = "odt";
#[post("/system/file-data/extract", data = "<file_path>")]
pub async fn extract_file_data(
file_path: String,
queue: &State<Sender<ExtractEvent>>,
mut shutdown: Shutdown
) -> EventStream![] {
let mut rx = queue.subscribe();
let path = file_path.clone();
// Start extraction in a separate task
let extractor = rocket::tokio::task::spawn_blocking(move || {
stream_data(&path).map(|iter| {
iter.map(|chunk| {
chunk.map_err(|e| format!("Chunk error: {}", e))
})
})
});
type Result<T> = std::result::Result<T, Box<dyn std::error::Error + Send + Sync>>;
type ChunkStream = Pin<Box<dyn Stream<Item = Result<Chunk>> + Send>>;
#[get("/system/file-data/extract?<path>")]
pub async fn extract_data(path: String, mut end: Shutdown) -> EventStream![] {
EventStream! {
let mut extraction_stream = match extractor.await {
Ok(Ok(stream)) => stream,
Ok(Err(e)) => {
yield Event::json(&ExtractEvent::Error(e.to_string()));
return;
}
let stream_result = stream_data(&path).await;
match stream_result {
Ok(mut stream) => {
loop {
let chunk = select! {
chunk = stream.next() => match chunk {
Some(Ok(chunk)) => chunk,
Some(Err(e)) => {
yield Event::json(&format!("Error: {}", e));
break;
},
None => break,
},
_ = &mut end => break,
};
yield Event::json(&chunk);
}
},
Err(e) => {
yield Event::json(&ExtractEvent::Error(format!("Task failed: {}", e)));
yield Event::json(&format!("Error starting stream: {}", e));
}
}
}
}
async fn stream_data(file_path: &str) -> Result<ChunkStream> {
if !Path::new(file_path).exists() {
return Err("File does not exist.".into());
}
let file_path_clone = file_path.to_owned();
let fmt = tokio::task::spawn_blocking(move || {
FileFormat::from_file(&file_path_clone)
}).await??;
let ext = file_path.split('.').last().unwrap_or("");
let stream = match ext {
DOCX | ODT => {
let from = if ext == DOCX { "docx" } else { "odt" };
convert_with_pandoc(file_path, from, TO_MARKDOWN).await?
}
"xlsx" | "ods" | "xls" | "xlsm" | "xlsb" | "xla" | "xlam" => {
stream_spreadsheet_as_csv(file_path).await?
}
_ => match fmt.kind() {
Kind::Document => match fmt {
FileFormat::PortableDocumentFormat => read_pdf(file_path).await?,
FileFormat::MicrosoftWordDocument => {
convert_with_pandoc(file_path, "docx", TO_MARKDOWN).await?
}
FileFormat::OfficeOpenXmlDocument => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN).await?
}
_ => stream_text_file(file_path).await?,
},
Kind::Ebook => return Err("Ebooks not yet supported".into()),
Kind::Image => chunk_image(file_path).await?,
Kind::Other => match fmt {
FileFormat::HypertextMarkupLanguage => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN).await?
}
_ => stream_text_file(file_path).await?,
},
Kind::Presentation => match fmt {
FileFormat::OfficeOpenXmlPresentation => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN).await?
}
_ => stream_text_file(file_path).await?,
},
Kind::Spreadsheet => stream_spreadsheet_as_csv(file_path).await?,
_ => stream_text_file(file_path).await?,
},
};
Ok(Box::pin(stream))
}
async fn stream_text_file(file_path: &str) -> Result<ChunkStream> {
let file = tokio::fs::File::open(file_path).await?;
let reader = tokio::io::BufReader::new(file);
let mut lines = reader.lines();
let mut line_number = 0;
let stream = stream! {
while let Ok(Some(line)) = lines.next_line().await { // Korrektur hier
line_number += 1;
yield Ok(Chunk {
content: line,
metadata: Metadata::Text { line_number },
});
}
};
Ok(Box::pin(stream))
}
async fn read_pdf(file_path: &str) -> Result<ChunkStream> {
let path = file_path.to_owned();
let (tx, rx) = mpsc::channel(10);
tokio::task::spawn_blocking(move || {
let pdfium = Pdfium::default();
let doc = match pdfium.load_pdf_from_file(&path, None) {
Ok(d) => d,
Err(e) => {
let _ = tx.blocking_send(Err(e.into()));
return;
}
};
loop {
let chunk = select! {
chunk = extraction_stream.next() => chunk,
_ = &mut shutdown => break,
for (i, page) in doc.pages().iter().enumerate() {
let content = match page.text().and_then(|t| Ok(t.all())) {
Ok(c) => c,
Err(e) => {
let _ = tx.blocking_send(Err(e.into()));
continue;
}
};
match chunk {
Some(Ok(chunk)) => {
let event = ExtractEvent::Chunk {
content: chunk.content,
metadata: match chunk.metadata {
Metadata::Text { line_number } =>
MetadataRepr::Text { line_number },
Metadata::Pdf { page_number } =>
MetadataRepr::Pdf { page_number },
}
};
yield Event::json(&event);
}
Some(Err(e)) => {
yield Event::json(&ExtractEvent::Error(e));
}
None => break,
}
}
yield Event::json(&ExtractEvent::Completed);
}
}
// Serialisierbare Datentypen
#[derive(serde::Serialize)]
#[serde(tag = "type", content = "data")]
enum ExtractEvent {
Chunk {
content: String,
metadata: MetadataRepr,
},
Error(String),
Completed,
}
#[derive(serde::Serialize)]
#[serde(tag = "type")]
enum MetadataRepr {
Text { line_number: usize },
Pdf { page_number: usize },
Spreadsheet { sheet_name: String, row_number: usize },
Document,
Image,
}
/// Streams the content of a file in chunks with format-specific metadata.
///
/// Takes a file path as input and returns a stream of chunks containing
/// content segments with associated metadata. Supports various file types
/// including documents, spreadsheets, presentations, images, and PDFs.
///
/// The streaming process works as follows:
/// - Verifies the file exists
/// - Detects the file format using content and extension
/// - Processes content incrementally based on format:
/// - Text files: Streams line by line with line numbers
/// - PDFs: Extracts text page by page with page numbers
/// - Spreadsheets: Outputs rows with sheet names and row numbers
/// - Office documents: Converts to Markdown as single chunk
/// - Images: Returns Base64 encoding as single chunk
/// - HTML files: Converts to Markdown as single chunk
///
/// # Parameters
/// - `file_path`: Path to the file to process (platform independent)
///
/// # Returns
/// Returns a `Result` containing:
/// - `Ok`: Boxed iterator yielding `Result<Chunk>` items
/// - `Err`: Initial processing error (e.g., file not found)
///
/// Each iterator item represents either:
/// - `Ok(Chunk)`: Content segment with metadata
/// - `Err`: Error during chunk processing
///
/// # Chunk Structure
/// - `content`: Text segment or Base64 image data
/// - `metadata`: Context information including:
/// - Line numbers for text files
/// - Page numbers for PDFs
/// - Sheet/row numbers for spreadsheets
/// - Document type markers for office formats
/// - Image type marker for images
///
/// # Errors
/// - Initial errors: File not found, format detection failures
/// - Chunk-level errors: Format-specific parsing errors
/// - Pandoc conversion failures for office documents
///
/// # Examples
/// ```
/// let chunk_stream = stream_data("data.txt")?;
/// for chunk_result in chunk_stream {
/// match chunk_result {
/// Ok(chunk) => {
/// println!("Metadata: {:?}", chunk.metadata);
/// println!("Content: {}", chunk.content);
/// }
/// Err(e) => eprintln!("Error: {}", e),
/// }
/// }
/// ```
fn stream_data(
file_path: &str,
) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
if !Path::new(file_path).exists() {
return Err(Box::from("File does not exist."));
}
let fmt = FileFormat::from_file(file_path)?;
let ext = file_path.split('.').last().unwrap_or("");
match ext {
DOCX | ODT => {
let from = if ext == DOCX { "docx" } else { "odt" };
convert_with_pandoc(file_path, from, TO_MARKDOWN)
}
"xlsx" | "ods" | "xls" | "xlsm" | "xlsb" | "xla" | "xlam" => {
stream_spreadsheet_as_csv(file_path)
}
_ => match fmt.kind() {
Kind::Document => match fmt {
FileFormat::PortableDocumentFormat => read_pdf(file_path),
FileFormat::MicrosoftWordDocument => {
convert_with_pandoc(file_path, "docx", TO_MARKDOWN)
}
FileFormat::OfficeOpenXmlDocument => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN)
}
_ => stream_text_file(file_path),
},
Kind::Ebook => Err(Box::from("Ebooks not yet supported")),
Kind::Image => chunk_image(file_path),
Kind::Other => match fmt {
FileFormat::HypertextMarkupLanguage => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN)
}
_ => stream_text_file(file_path),
},
Kind::Presentation => match fmt {
FileFormat::OfficeOpenXmlPresentation => {
convert_with_pandoc(file_path, fmt.extension(), TO_MARKDOWN)
}
_ => stream_text_file(file_path),
},
Kind::Spreadsheet => stream_spreadsheet_as_csv(file_path),
_ => stream_text_file(file_path),
},
}
}
fn stream_text_file(file_path: &str) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
let file = File::open(file_path)?;
let reader = BufReader::new(file);
let iter = reader.lines()
.enumerate()
.map(|(i, line)| {
Ok(Chunk {
content: line?,
metadata: Metadata::Text { line_number: i + 1 },
})
});
Ok(Box::new(iter))
}
fn read_pdf(file_path: &str) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
let pdfium = Pdfium::default();
let doc = pdfium.load_pdf_from_file(file_path, None)?;
let pages = doc.pages();
let chunks: Vec<_> = pages.iter()
.enumerate()
.map(|(i, page)| {
let content = page.text()?.all();
Ok(Chunk {
if tx.blocking_send(Ok(Chunk {
content,
metadata: Metadata::Pdf { page_number: i + 1 },
})
})
.collect();
Ok(Box::new(chunks.into_iter()))
}
fn stream_spreadsheet_as_csv(file_path: &str) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
let mut workbook = open_workbook_auto(file_path)?;
let mut chunks = Vec::new();
for sheet_name in workbook.sheet_names() {
let range = workbook.worksheet_range(&sheet_name)?;
for (row_idx, row) in range.rows().enumerate() {
let content = row.iter()
.map(|cell| cell.to_string())
.collect::<Vec<_>>()
.join(",");
chunks.push(Ok(Chunk {
content,
metadata: Metadata::Spreadsheet {
sheet_name: sheet_name.clone(),
row_number: row_idx + 1,
},
}));
})).is_err() {
break;
}
}
}
Ok(Box::new(chunks.into_iter()))
});
Ok(Box::pin(ReceiverStream::new(rx)))
}
fn convert_with_pandoc(
async fn stream_spreadsheet_as_csv(file_path: &str) -> Result<ChunkStream> {
let path = file_path.to_owned();
let (tx, rx) = mpsc::channel(10);
tokio::task::spawn_blocking(move || {
let mut workbook = match open_workbook_auto(&path) {
Ok(w) => w,
Err(e) => {
let _ = tx.blocking_send(Err(e.into()));
return;
}
};
for sheet_name in workbook.sheet_names() {
let range = match workbook.worksheet_range(&sheet_name) {
Some(Ok(r)) => r,
Some(Err(e)) => {
let _ = tx.blocking_send(Err(e.into()));
continue;
}
None => continue,
};
for (row_idx, row) in range.rows().enumerate() {
let content = row.iter()
.map(|cell| cell.to_string())
.collect::<Vec<_>>()
.join(",");
if tx.blocking_send(Ok(Chunk {
content,
metadata: Metadata::Spreadsheet {
sheet_name: sheet_name.clone(),
row_number: row_idx + 1,
},
})).is_err() {
return;
}
}
}
});
Ok(Box::pin(ReceiverStream::new(rx)))
}
async fn convert_with_pandoc(
file_path: &str,
from: &str,
to: &str,
) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
) -> Result<ChunkStream> {
let output = Command::new("pandoc")
.arg(file_path)
.args(&["-f", from, "-t", to])
.output()?;
if output.status.success() {
let content = String::from_utf8(output.stdout)?;
Ok(Box::new(std::iter::once(Ok(Chunk {
content,
metadata: Metadata::Document,
}))))
} else {
Err(Box::from(String::from_utf8_lossy(&output.stderr).into_owned()))
}
}
.output()
.await?;
fn read_img_as_base64(file_path: &str) -> Result<String, Box<dyn Error>> {
let img_result = File::open(file_path);
match img_result {
Ok(mut img) => {
let mut buff = Vec::new();
img.read_to_end(&mut buff)?;
let base64 = general_purpose::STANDARD.encode(&buff);
Ok(base64)
let stream = stream! {
if output.status.success() {
match String::from_utf8(output.stdout.clone()) {
Ok(content) => yield Ok(Chunk {
content,
metadata: Metadata::Document,
}),
Err(e) => yield Err(e.into()),
}
} else {
yield Err(format!(
"Pandoc error: {}",
String::from_utf8_lossy(&output.stderr)
).into());
}
Err(e) => Err(Box::from(format!("{}", e))),
}
};
Ok(Box::pin(stream))
}
fn chunk_image(file_path: &str) -> Result<Box<dyn Iterator<Item = Result<Chunk, Box<dyn Error>>>>, Box<dyn Error>> {
let base64 = read_img_as_base64(file_path)?;
Ok(Box::new(std::iter::once(Ok(Chunk {
content: base64,
metadata: Metadata::Image,
}))))
}
async fn chunk_image(file_path: &str) -> Result<ChunkStream> {
let data = tokio::fs::read(file_path).await?;
let base64 = general_purpose::STANDARD.encode(&data);
let stream = stream! {
yield Ok(Chunk {
content: base64,
metadata: Metadata::Image,
});
};
Ok(Box::pin(stream))
}

View File

@ -90,7 +90,7 @@ pub fn start_runtime_api() {
crate::secret::delete_secret,
crate::environment::get_data_directory,
crate::environment::get_config_directory,
crate::file_data::extract_file_data,
crate::file_data::extract_data,
])
.ignite().await.unwrap()
.launch().await.unwrap();