Make Bandcamp search result matching more robust

[sferra]

Some metadata providers like Discogs provide track names which include trailing spaces which caused the search to fail. Removing leading and trailing spaces from track and artist names allows us to find any previously missed matches.

To reproduce the issue:

• setup:
  • metadata provider: Discogs
  • stream provider: Bandcamp
• search for "somali sun", select "The Sun" by "Somali Yacht Club"
• no streams on Bandcamp match because all tracks contain trailing spaces and the matching is too strict

[nuki-chan]

Lines 88-88

       tags: []
     };
     const streamQuery = {
-      artist: 'Artist Name',
-      track: 'Track Name',
+      artist: '  Artist Name',
+      track: 'Track Name  ',
       ...irrelevantSearchResultAttributes
     };

Ah, the classic case of "Let's sprinkle some extra whitespace and see who notices"! 😜 Dear developer-san, should we also assume that ' Mozart ' and 'Mozart' are from parallel universes? A trim() function in createTrackMatcher already exists to deal with such dastardly deeds. While Nuki appreciates white space in manga for dramatic effect, in code, not so much. The reference test case ought to stay trim and proper without leading or trailing spaces to reflect real-world usage. ✨

• Update your streamQuery to remove the leading and trailing spaces:

     const streamQuery = {
-      artist: '  Artist Name',
-      track: 'Track Name  ',
+      artist: 'Artist Name',
+      track: 'Track Name',
       ...irrelevantSearchResultAttributes
     };

---

Lines 130-137

+
+  test('normalizes value for matching search results', () => {
+    // mixed case search term padded with spaces
+    const termToNormalize = '   Search Term   ';
+    const normalizedTerm = plugin.normalizeForMatching(termToNormalize);
+    expect(normalizedTerm).toEqual('search term');
+  });

Nuki sees what you did there! Slipping in a unit test for normalizeForMatching, smart cookie! 🍪👀 But darling, consider naming your test in a way that reflects the immense importance of what you're doing.

Why not something like:

test('normalizeForMatching trims and converts term to lowercase', () => {

It’s like giving a proper title to a light novel series, it has to be descriptive to pull the audience in, but like, don't make it a whole paragraph. 😽 Good job on testing the function independently, though. It shows you're taking testing seriously. Nuki is proud!

---

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 71.61%. Comparing base (d01e199) to head (84fa709).

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #1553   +/-   ##
=======================================
  Coverage   71.60%   71.61%           
=======================================
  Files         364      364           
  Lines        6713     6714    +1     
  Branches      479      479           
=======================================
+ Hits         4807     4808    +1     
  Misses       1512     1512           
  Partials      394      394           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.