correctly handle NFD content in media DB from older Anki versions

ankitects · Mar 19, 2020 · 857c0dd · 857c0dd
1 parent 0ecc189
commit 857c0dd
Showing 1 changed file with 11 additions and 0 deletions.
diff --git a/rslib/src/media/sync.rs b/rslib/src/media/sync.rs
@@ -20,6 +20,7 @@ use std::collections::HashMap;
 use std::io::{Read, Write};
 use std::path::Path;
 use std::{io, time};
+use unicode_normalization::is_nfc;
 
 static SYNC_MAX_FILES: usize = 25;
 static SYNC_MAX_BYTES: usize = (2.5 * 1024.0 * 1024.0) as usize;
@@ -717,6 +718,16 @@ fn zip_files<'a>(
             break;
         }
 
+        #[cfg(target_vendor = "apple")]
+        {
+            if !is_nfc(&file.fname) {
+                // older Anki versions stored non-normalized filenames in the DB; clean them up
+                debug!(log, "clean up non-nfc entry"; "fname"=>&file.fname);
+                invalid_entries.push(&file.fname);
+                continue;
+            }
+        }
+
         let file_data = if file.sha1.is_some() {
             match data_for_file(media_folder, &file.fname) {
                 Ok(data) => data,