Markdown format for encoding relational data

[icanzilb]

The current encoder/decoder could easily support relational data (to some extent). A single text file could contain multiple "entities" (for lack of a better term) like so:

# People
| name | age | Employers.id |
| - | - | - |
| Peter | 27 | asdf34 |

# Employers

| id | company |
| - | - |
| asdf34 | Pepsi Co |
| weqf17 | Adobe Inc |

This kind of format would still preserve the high human-editability but introduce a more database-like format for normalizing data.