rewrite so webdriver is only used for the login

download/src/download.rs

@@ -0,0 +1,52 @@
+//! Handle downloads of newspaper issues.
+
+use std::{
+    fs::{self},
+    io::{Cursor, Read},
+    path::Path,
+};
+
+use anyhow::Result;
+use tracing::{debug, info};
+
+use crate::{nzz::Issue, pdf};
+
+/// Download all pages of the provided `issues` and save them merged to the directory `output_dir`.
+///
+/// Create `output_dir` if it does not exist.
+pub async fn fetch(issues: Vec<Issue>, output_dir: &Path) -> Result<()> {
+    debug!("ensuring {output_dir:?} exists");
+    fs::create_dir_all(output_dir)?;
+
+    for issue in issues {
+        info!("saving issue {}", issue.publication_date);
+
+        let tmp_dir = tempfile::tempdir()?;
+        let mut pages = Vec::new();
+        for (i, page) in issue.pages.into_iter().enumerate() {
+            debug!(
+                "fetching issue {}, page {}: {page}",
+                issue.publication_date,
+                i + 1
+            );
+
+            let response = reqwest::Client::new().get(page).send().await?;
+            let mut content = Cursor::new(response.bytes().await?);
+            let mut page_data = Vec::new();
+            content.read_to_end(&mut page_data)?;
+
+            let tmp_page = tmp_dir.path().join(i.to_string());
+            fs::write(&tmp_page, page_data)?;
+            pages.push(tmp_page);
+        }
+
+        let issue_name = format!("nzz_{}.pdf", issue.publication_date);
+        let issue_path = output_dir.join(issue_name);
+        let issue_title = format!("NZZ {}", issue.publication_date);
+
+        pdf::merge(pages, &issue_path, &issue_title)?;
+        debug!("issue {} saved", issue.publication_date);
+    }
+
+    Ok(())
+}