From 67d1e0ee2f869e37b886988c50c62894b707bd6b Mon Sep 17 00:00:00 2001
From: perf3ct <jonfuller2012@gmail.com>
Date: Mon, 23 Jun 2025 19:39:39 +0000
Subject: [PATCH] feat(webdav): move etag parser to own function, create
 required migration

---
 ...0250623000001_normalize_existing_etags.sql | 11 ++++++
 src/webdav_xml_parser.rs                      | 35 +++++++++++++++----
 tests/webdav_enhanced_unit_tests.rs           |  6 ++--
 3 files changed, 42 insertions(+), 10 deletions(-)
 create mode 100644 migrations/20250623000001_normalize_existing_etags.sql

diff --git a/migrations/20250623000001_normalize_existing_etags.sql b/migrations/20250623000001_normalize_existing_etags.sql
new file mode 100644
index 0000000..3198d35
--- /dev/null
+++ b/migrations/20250623000001_normalize_existing_etags.sql
@@ -0,0 +1,11 @@
+-- Normalize existing ETags in webdav_files table to match new normalization format
+-- This migration ensures that existing ETag values are normalized to prevent
+-- unnecessary re-downloads of unchanged files after the ETag normalization fix
+
+-- Update ETags to remove quotes and W/ prefixes
+UPDATE webdav_files 
+SET etag = TRIM(BOTH '"' FROM TRIM(LEADING 'W/' FROM etag))
+WHERE etag LIKE '"%"' OR etag LIKE 'W/%';
+
+-- Add a comment to document this normalization
+COMMENT ON COLUMN webdav_files.etag IS 'Normalized ETag without quotes or W/ prefix (since migration 20250623000001)';
\ No newline at end of file
diff --git a/src/webdav_xml_parser.rs b/src/webdav_xml_parser.rs
index ac6a737..328a09c 100644
--- a/src/webdav_xml_parser.rs
+++ b/src/webdav_xml_parser.rs
@@ -83,13 +83,7 @@ pub fn parse_propfind_response(xml_text: &str) -> Result<Vec<FileInfo>> {
                                 resp.content_type = Some(text.trim().to_string());
                             }
                             "getetag" => {
-                                // Normalize ETag by removing quotes and weak ETag prefix
-                                let etag = text.trim();
-                                let normalized = etag
-                                    .trim_start_matches("W/")
-                                    .trim_matches('"')
-                                    .to_string();
-                                resp.etag = Some(normalized);
+                                resp.etag = Some(normalize_etag(&text));
                             }
                             "status" if in_propstat => {
                                 // Check if status is 200 OK
@@ -206,6 +200,20 @@ fn parse_http_date(date_str: &str) -> Option<DateTime<Utc>> {
         })
 }
 
+/// Normalize ETag by removing quotes and weak ETag prefix
+/// This ensures consistent ETag comparison across different WebDAV servers
+/// 
+/// Examples:
+/// - `"abc123"` → `abc123`
+/// - `W/"abc123"` → `abc123`
+/// - `abc123` → `abc123`
+fn normalize_etag(etag: &str) -> String {
+    etag.trim()
+        .trim_start_matches("W/")
+        .trim_matches('"')
+        .to_string()
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -344,4 +352,17 @@ mod tests {
         let files = parse_propfind_response(xml).unwrap();
         assert_eq!(files.len(), 0);
     }
+
+    #[test]
+    fn test_normalize_etag() {
+        // Test various ETag formats that WebDAV servers might return
+        assert_eq!(normalize_etag("abc123"), "abc123");
+        assert_eq!(normalize_etag("\"abc123\""), "abc123");
+        assert_eq!(normalize_etag("W/\"abc123\""), "abc123");
+        assert_eq!(normalize_etag("  \"abc123\"  "), "abc123");
+        assert_eq!(normalize_etag("W/\"abc-123-def\""), "abc-123-def");
+        assert_eq!(normalize_etag(""), "");
+        assert_eq!(normalize_etag("\"\""), "");
+        assert_eq!(normalize_etag("W/\"\""), "");
+    }
 }
\ No newline at end of file
diff --git a/tests/webdav_enhanced_unit_tests.rs b/tests/webdav_enhanced_unit_tests.rs
index 112c0f6..d229205 100644
--- a/tests/webdav_enhanced_unit_tests.rs
+++ b/tests/webdav_enhanced_unit_tests.rs
@@ -201,21 +201,21 @@ fn test_webdav_response_parsing_comprehensive() {
     let pdf_file = files.iter().find(|f| f.name == "report.pdf").unwrap();
     assert_eq!(pdf_file.size, 2048000);
     assert_eq!(pdf_file.mime_type, "application/pdf");
-    assert_eq!(pdf_file.etag, "\"pdf123\"");
+    assert_eq!(pdf_file.etag, "pdf123"); // ETag should be normalized (quotes removed)
     assert!(!pdf_file.is_directory);
 
     // Verify second file (photo.png)
     let png_file = files.iter().find(|f| f.name == "photo.png").unwrap();
     assert_eq!(png_file.size, 768000);
     assert_eq!(png_file.mime_type, "image/png");
-    assert_eq!(png_file.etag, "\"png456\"");
+    assert_eq!(png_file.etag, "png456"); // ETag should be normalized (quotes removed)
     assert!(!png_file.is_directory);
 
     // Verify third file (unsupported.docx)
     let docx_file = files.iter().find(|f| f.name == "unsupported.docx").unwrap();
     assert_eq!(docx_file.size, 102400);
     assert_eq!(docx_file.mime_type, "application/vnd.openxmlformats-officedocument.wordprocessingml.document");
-    assert_eq!(docx_file.etag, "\"docx789\"");
+    assert_eq!(docx_file.etag, "docx789"); // ETag should be normalized (quotes removed)
     assert!(!docx_file.is_directory);
 }