Implements #1552: Change how the title in the head is defined. #1832
Conversation
There was a problem hiding this comment.
This currently breaks the integration tests:
(Information, , 15: 2025-09-05T08:38:25.8100000Z info ::e.d.t.f.topwatchFilter:: repo build-all :: Finished in '00:00:09.0448737')
(Information, , 16: 2025-09-05T08:38:25.8210000Z error::Program :: System.IndexOutOfRangeException: Index was outside the bounds of the array.)
(Information, , 17: 2025-09-05T08:38:25.8210000Z at System.Collections.Generic.Dictionary`2.TryInsert(TKey key, TValue value, InsertionBehavior behavior))
(Information, , 18: 2025-09-05T08:38:25.8210000Z at System.Collections.Generic.Dictionary`2.set_Item(TKey key, TValue value))
I think a different extension point might be better (see other comment).
| { | ||
| // Create a dictionary where keys are the titles | ||
| // and values are files with that title | ||
| var titleMap = new Dictionary<string, List<MarkdownFile>>(StringComparer.OrdinalIgnoreCase); |
There was a problem hiding this comment.
HashSet<string> will suffice we care about something being duplicate not necessary how many times since the diagnostics reporting will take care of that.
I think it makes more sense to implement this as IMarkdownExporter implementations (maybe ValidationMarkdownExporter).
That way it works for both docs-builder and docs-assembler. Here the duplicate file checking is per DocumentationSet not overall.
There was a problem hiding this comment.
If a title is flagged as a duplicate, it is good to know the set of pages with the same title. Then the author can coordinate with the author(s) of the other pages for a solution. There is no harm in maintaining the full context.
| var titleMap = new Dictionary<string, List<MarkdownFile>>(StringComparer.OrdinalIgnoreCase); | ||
| foreach (var file in files) | ||
| { | ||
| if (string.IsNullOrWhiteSpace(file.Title)) |
There was a problem hiding this comment.
This should be reported as an error.
No description provided.