LT-22261: Problems with copied text when importing from flextext

[jtmaxwell3]

This fixes several bugs in the flextext importer:

1. The "pos" item was being ignored by the importer.
2. FindOrCreateWfiAnalysis wasn't adding affix markers when matching morphemes.
3. CreateWordformWithWfiAnalysis erroneously assumed that the "cf" item referred to citation forms, when in fact it refers to lexeme forms. This means that homograph numbers are not unique. (This may be a bug in the code for displaying interlinear text.)
4. CreateWordformWithWfiAnalysis was setting bundle.MorphRA to the lexeme form when it should have been setting it to the allomorph matching the "txt" item.

There is also a bug where an extra § is inserted.

I'm going to create JIRA issues for the bugs.

---

This change is 

[papeh]

LG through BIRD line 1000; I'll look at the rest after lunch.

@papeh reviewed all commit messages.
Reviewable status: 0 of 2 files reviewed, 1 unresolved discussion

---

Src/LexText/Interlinear/BIRDInterlinearImporter.cs line 865 at r1 (raw file):

				foreach (var item in word.Items)
				{
					if (wordForm.Analysis.CategoryRA != null)

What besides line 880 wordForm.Analysis.CategoryRA = cat; might set CategoryRA in this loop?

[papeh]

@papeh reviewed 2 of 2 files at r1, 1 of 1 files at r2, all commit messages.
Reviewable status: all files reviewed, 2 unresolved discussions (waiting on @jtmaxwell3)

---

Src/LexText/Interlinear/ITextDllTests/BIRDFormatImportTests.cs line 827 at r1 (raw file):

					ISegment segment = imported.ContentsOA[0].SegmentsOS[0];
					// Verify that we found the category.
					Assert.True(segment.AnalysesRS[0].Analysis.CategoryRA.Equals(cat));

nit: you could get more helpful error reports from

Assert.That(segment.AnalysesRS[0].Analysis.CategoryRA).Is.Equalto(cat);