djstevenson
commented
4 years ago Current schema:
| Column | Embedded links? | List links? | Both |
|---|---|---|---|
| id | SAME | ||
| song_id | SAME | ||
| identifier | YES | ||
| class | YES | ||
| url | DIFF | ||
| description | DIFF | ||
| priority | YES | ||
| extras | YES | ||
| description | DIFF | ||
| title | Not used | ||
| css | YES |
"SAME" means used by embedded and list links, the same value used for both.
"DIFF" means used by embedded and list links, but we may want different values for each.
So, proposed new schema is:
| Column | Meaning |
|---|---|
| id | Primary key |
| song_id | belongs_to song |
| embed_identifier | ^^identifier^^ for referring to this link in the article. Empty string if no embedded link required. |
| embed_class | Perl class for rendering embedded links. Ignored if embed_identifier is empty |
| embed_url | URL for embedded links. Ignored if embed_identifier is empty |
| embed_description | Link text for embedded links. Ignored if embed_identifier is empty |
| list_priority | Sort order for list links. Zero means "not in list" |
| list_description | Link text for list links. Ignored if list_priority is 0 |
| list_url | URL for list links. Ignored if list_priority is 0 |
| list_css | CSS class / icon name for list links. Ignored if list_priority is 0 |
title, extras are dropped until we have a use-case for them.
We end up with similar links repeated, cos
Should allow one link record to serve multiple purposes.
e.g. Single YT link:
So, a single link can have TWO descriptions (inline and list).
If embedded, then we mangle the URL.
Single identifier...
e.g. ^^yt^^ uses 'yt' identifier inline, with the 'inline_description' If there is a list description, It appears in list with the list_description If you do ^^yt-embedded^^ it mangles the URL and embeds it