From 78452c2eb59cf997f54d4f3a95b1c8233e7e04ab Mon Sep 17 00:00:00 2001
From: SixtyGrit <nick.steinwachs@gmail.com>
Date: Fri, 12 Jun 2026 12:13:50 -0700
Subject: [PATCH] Resolve CSV file references with spaces against nested
 attachments

CsvParser#file_paths transliterated every file reference's spaces to
underscores before checking existence, while
remove_spaces_from_filenames only renames the top level of files/
(nested entries are deliberately left as-is, per the unzip_spec added
with the guided importer). A CSV referencing a nested attachment whose
name contains spaces (files/subdir/photo name.jpg) could therefore
never resolve: the lookup demanded underscores, the disk kept spaces.

Try the reference as written first, then fall back to the underscored
variant for top-level files that remove_spaces_from_filenames renamed.
---
 app/parsers/bulkrax/csv_parser.rb       | 17 +++++++++++------
 spec/parsers/bulkrax/csv_parser_spec.rb | 24 ++++++++++++++++++++++++
 2 files changed, 35 insertions(+), 6 deletions(-)

diff --git a/app/parsers/bulkrax/csv_parser.rb b/app/parsers/bulkrax/csv_parser.rb
index 5d66277e..65d2222c 100644
--- a/app/parsers/bulkrax/csv_parser.rb
+++ b/app/parsers/bulkrax/csv_parser.rb
@@ -372,12 +372,17 @@ def file_paths
         raise StandardError, "Record references local files but no files directory could be resolved from the import path" if files_dir.nil?
 
         r[file_mapping].split(split_pattern).map do |f|
-          file = File.join(files_dir, f.strip.tr(' ', '_'))
-          if File.exist?(file) # rubocop:disable Style/GuardClause
-            file
-          else
-            raise "File #{file} does not exist"
-          end
+          reference = f.strip
+          # Try the reference as written first, then the space->underscore
+          # variant produced by #remove_spaces_from_filenames. The rename only
+          # rewrites the top level of files/ (see unzip_spec), so nested
+          # attachments keep their spaces on disk and only match as written.
+          file = [reference, reference.tr(' ', '_')].uniq
+                                                    .map { |candidate| File.join(files_dir, candidate) }
+                                                    .find { |candidate| File.exist?(candidate) }
+          raise "File #{File.join(files_dir, reference)} does not exist" if file.nil?
+
+          file
         end
       end.flatten.compact.uniq
     end
diff --git a/spec/parsers/bulkrax/csv_parser_spec.rb b/spec/parsers/bulkrax/csv_parser_spec.rb
index 712f20cf..f12c1983 100644
--- a/spec/parsers/bulkrax/csv_parser_spec.rb
+++ b/spec/parsers/bulkrax/csv_parser_spec.rb
@@ -622,6 +622,30 @@ module Bulkrax
         end
       end
 
+      context 'when a record references file names containing spaces' do
+        before do
+          allow(subject).to receive(:records).and_return([{ file: 'subdir/has space.jpg|top space.jpg' }])
+          allow(subject).to receive(:path_to_files).and_return('spec/fixtures/csv/files')
+        end
+
+        it 'resolves nested files as written and top-level files via the underscored variant' do
+          allow(File).to receive(:exist?).and_return(false)
+          allow(File).to receive(:exist?).with('spec/fixtures/csv/files/subdir/has space.jpg').and_return(true)
+          allow(File).to receive(:exist?).with('spec/fixtures/csv/files/top_space.jpg').and_return(true)
+
+          expect(subject.file_paths).to contain_exactly(
+            'spec/fixtures/csv/files/subdir/has space.jpg',
+            'spec/fixtures/csv/files/top_space.jpg'
+          )
+        end
+
+        it 'raises naming the as-written reference when neither variant exists' do
+          allow(File).to receive(:exist?).and_return(false)
+
+          expect { subject.file_paths }.to raise_error(RuntimeError, %r{subdir/has space\.jpg does not exist})
+        end
+      end
+
       context 'when a record file value is blank' do
         before do
           allow(subject).to receive(:records).and_return([{ file: '' }, { file: nil }])