Add ListExtractor

crates/pile-dataset/src/extract/pdf/pdf_meta.rs

@@ -4,7 +4,8 @@ use pile_config::Label;
 use std::{collections::HashMap, io::BufReader, sync::OnceLock};
 use tracing::trace;
 
-use crate::{Item, PileValue, SyncReadBridge, extract::Extractor};
+use crate::extract::ObjectExtractor;
+use crate::{Item, PileValue, SyncReadBridge};
 
 pub struct PdfMetaExtractor<'a> {
 	item: &'a Item,
@@ -40,6 +41,8 @@ impl<'a> PdfMetaExtractor<'a> {
 				}
 			};
 
+			let page_count = file.num_pages();
+
 			let mut meta: Vec<(&'static str, Option<String>)> = Vec::new();
 
 			if let Some(info) = &file.trailer.info_dict {
@@ -64,12 +67,12 @@ impl<'a> PdfMetaExtractor<'a> {
 				meta.push(("mod_date", info.mod_date.as_ref().map(format_date)));
 			}
 
-			Ok::<_, std::io::Error>(meta)
+			Ok::<_, std::io::Error>((page_count, meta))
 		})
 		.await
 		.map_err(std::io::Error::other)?;
 
-		let raw_meta = match raw_meta {
+		let (page_count, raw_meta) = match raw_meta {
 			Ok(x) => x,
 			Err(error) => {
 				trace!(message = "Could not process pdf", ?error, key = ?self.item.key());
@@ -79,6 +82,12 @@ impl<'a> PdfMetaExtractor<'a> {
 
 		let mut output: HashMap<Label, PileValue<'a>> = HashMap::new();
 
+		#[expect(clippy::unwrap_used)]
+		output.insert(
+			Label::new("pages").unwrap(),
+			PileValue::U64(page_count as u64),
+		);
+
 		#[expect(clippy::unwrap_used)]
 		for (key, val) in raw_meta {
 			let label = Label::new(key).unwrap();
@@ -106,7 +115,7 @@ fn format_date(d: &Date) -> String {
 }
 
 #[async_trait::async_trait]
-impl Extractor for PdfMetaExtractor<'_> {
+impl ObjectExtractor for PdfMetaExtractor<'_> {
 	async fn field<'a>(
 		&'a self,
 		name: &Label,