User:Secular mind~enwiki/Datatable

Both the current and the proposed syntax for tables try to allow everything, which can be done via html-tables. Any syntax which tries to do that necessarily inherits the full complexity of html-tables.

I suggest using a syntax which does not try to mirror a specific html-element but which is tailored for a specific purpose.

One important purpose tables are used for in wikipedia is the presentation of data. There should be a syntax for data which reflects the data itself and not its layout as a table.

I suggest the following syntax for data:
 * Following Wikimark2 a data starts with  and ends with  . (Alternatively   and  )
 * Data contains at least one datablock.
 * Every datablock starts with a line containing only  (tailing whitespace is ignored).
 * An entry of a datablock starts with an word called identifier followed by a colon. The rest of the line is the corresponding value.
 * Everything before the first datablock is called the head (details below).

Example: <.

Default layout
Every identifier which appears in the data corresponds to a column of the table with the name of the identifier as the table heading. They are sorted by their order of appearance. Every datablock corresponds to one row of the table. Each cell contains the value of the identifier corresponding to the column in the datablock corresponding to the row. If an identifier is missing from a datablock or the value of this identifier is empty, the corresponding cell is empty.

Example:

Further Examples

 * User:Secular mind/Datatable/Example1
 * User:Secular mind/Datatable/Example2
 * User:Secular mind/Datatable/Example3