Global lyric-line metadata

[adrianholovaty]

This branch adds a new "lyrics" section to "global". This contains top-level metadata about each lyric line — at the moment, a language and label.

See new docs here:
https://cdn.githubraw.com/w3c/mnx/lyric-line-metadata/docs/mnx-reference/objects/lyrics-global/

And a new example document here:
https://cdn.githubraw.com/w3c/mnx/lyric-line-metadata/docs/mnx-reference/examples/lyric-line-metadata/

As a reminder, in MNX, a "lyric line" is a span of lyrics that should be considered a single unit. Here we have four separate lyric lines:

This is encoded as the following in "global":

{
   "lyrics": {
      "lineOrder": ["1", "2", "3", "4"],
      "lineMetadata": {
         "1": {"lang": "en", "label": "English"},
         "2": {"lang": "nl", "label": "Nederlands"},
         "3": {"lang": "uk", "label": "Українська"},
         "4": {"lang": "es", "label": "Español"}
      }
   }
}

The "1", "2", "3" and "4" are the arbitrary lyric line IDs I've chosen. These correspond to the IDs that are used in any "event" that has lyrics. For example, later in the document:

{
   "lyrics": {
      "lines": {
         "1": {"text": "John!"},
         "2": {"text": "Jan!"},
         "3": {"text": "ван!", "type": "end"},
         "4": {"text": "Juan!"}
      }
   }
}

[mscuthbert]

We decided to keep lineOrder separate so that later it could be redefined at a "zone" level (like systems) to omit or reorder different lyrics.

[samuelbradshaw]

Is it intentional that you're escaping non-ASCII characters in the JSON example? JSON should allow Unicode without escaping.

[adrianholovaty]

@samuelbradshaw Thanks for bringing that up. I've just improved our docs system not to force ASCII in JSON strings (commit here), and I've updated this pull request accordingly (commit here).