fix: strip trailing slash from indexer URL

[sdbondi]

Summary

• Fix 404 errors in the indexer sync loop caused by a double-slash in the URL (https://host//templates/catalogue)
• The configured indexer URL ends with /, and the format string adds another, producing an invalid path
• Trim trailing slash from indexer_url in both sync_once and fetch_definition

Test plan

• Run with --network esmeralda and verify no 404 errors in logs
• Verify sync loop successfully fetches catalogue entries

🤖 Generated with Claude Code

[gemini-code-assist]

Code Review

This pull request refactors URL construction in server/src/sync/indexer.rs to ensure the base indexer URL is correctly formatted by trimming trailing slashes. A review comment suggests optimizing this by moving the trimming logic outside of the pagination loop to avoid redundant operations in each iteration.

[gemini-code-assist]

The indexer_url is trimmed inside the loop, which is inefficient as the URL base remains constant across all iterations of the pagination loop. It is better to normalize the URL once before entering the loop. Shadowing the indexer_url variable also ensures that subsequent calls (like fetch_definition later in the function) benefit from the normalized URL without needing to re-trim.

Suggested change

	let mut total_new = 0usize;
	loop {
	let mut url = format!("{}/templates/catalogue?limit={}", indexer_url, limit);
	let base = indexer_url.trim_end_matches('/');
	let mut url = format!("{base}/templates/catalogue?limit={limit}");
	let mut total_new = 0usize;
	let indexer_url = indexer_url.trim_end_matches('/');
	loop {
	let mut url = format!("{indexer_url}/templates/catalogue?limit={limit}");