Add z3998:name-title semantic

[erinendrei]

Issue #477 has been open for a while. If Robin still wants to do it, that's of course fine, but I thought I would try it as unless I'm missing something it doesn't look complicated.

• In 6af0b22 the z3998:name-title semantic has been added to every abbreviation in the semanticate function that looked like a relevant title. Notes:

  • "Mt." was not given the semantic as it seems to abbreviate only "Mount"/"Mountain" even though it's been placed in the middle of the list of all the other titles.
  • "Bros." was not given the semantic as it doesn't seem like a title (unless, I guess, it's referring to a group of religious brothers, for whom "Br." is actually a title; but this seems like it would occur less frequently than other possible uses of "Bros.").
  • "M." for "Monsieur", mentioned by Robin initially, does not appear to be automatically given <abbr> by semanticate and was not added here as I figured that that omission was intentional, perhaps because "M." could be part of an initialism or given name.

  If I'm mistaken in any choices here, let me know and I'll rebase.

• In dac0b89 the test for semanticate has been updated to accord with the changes in 6af0b22. If you want more test cases for z3998:name-title specifically, then I can add some. It seems like the one for "Mr." at line 12 of tests/semanticate/in/semanticate.xhtml was previously thought to be enough.

• In 433327f the other tests containing "Mr." and "Dr." were updated with the z3998:name-title semantic. I didn't see other title abbreviations, and force-pushed because I missed "Dr." the first time.

• There is no lint check here. For abbreviations that need the z3998:name-title semantic but lack it, would the correct error code be s-045 or would a new error code be required? I could add a lint check and test in a later PR or update this one if you want.

Even if the commits here are acceptable I obviously can't update the ebook repos as mentioned originally in #477 so this addresses only part of the issue and doesn't close it. And I imagine the manual would have to be updated too if you still want to accept this semantic change.

[acabal]

Great, thanks. The trickier part is updating the corpus. Can you write a Bash script for updating a set of repos locally? For example, if we have all of the repos as folders on a local machine, loop through them and use sed or something to make these replacements. You can email me that, instead of putting it in this PR. Then we can accept this since we can update the corpus at the same time.

You can use the sync-ebooks script to download the corpus locally to test if you want.

[erinendrei]

Sure, I'll work on that. Thanks.

[erinendrei]

Did you receive the email last week? No rush if so, but I realised the attachment might have sent it to spam.

[acabal]

Yes, I am unbelievably busy so this will probably have to wait till next week

[acabal]

OK, I've had time to look this over and it looks great. Thanks!