🐛 Fix rerun for entries that came from zips (#1013)

Previously, when rerunning an entry that came from a zip file, the rerun
would fail because it would look for the CSV in an assumed location
which is based on it's last importer ID.  Since this was a rerun, it
does not do an unzip into the assumed location so the directory does not
exist.

This commit will first check if the assumed location exists, and if not,
it will look for the location of the last unizpped files and use that
for the rerun.

Ref:
- notch8/palni_palci_knapsack#210

app/models/bulkrax/importer.rb

@@ -239,6 +239,14 @@ def import_metadata_format
     # If the import data is zipped, unzip it to this path
     def importer_unzip_path
       @importer_unzip_path ||= File.join(parser.base_path, "import_#{path_string}")
+      return @importer_unzip_path if Dir.exist?(@importer_unzip_path)
+
+      # turns "tmp/imports/tenant/import_1_20250122035229_1" to "tmp/imports/tenant/import_1_20250122035229"
+      base_importer_unzip_path = @importer_unzip_path.split('_')[0...-1].join('_')
+
+      # If we don't have an existing unzip path, we'll try and find it.
+      # Just in case there are multiple paths, we sort by the number at the end of the path and get the last one
+      @importer_unzip_path = Dir.glob(base_importer_unzip_path + '*').sort_by { |path| path.split(base_importer_unzip_path).last[1..-1].to_i }.last
     end
 
     def errored_entries_csv_path

app/parsers/bulkrax/csv_parser.rb

@@ -360,6 +360,10 @@ def path_to_files(**args)
       @path_to_files = File.join(
           zip? ? importer_unzip_path : File.dirname(import_file_path), 'files', filename
         )
+
+      return @path_to_files if File.exist?(@path_to_files)
+
+      File.join(importer_unzip_path, 'files', filename) if file? && zip?
     end
 
     private