quarto-dev / gridtables Goto Github PK

Zero-width joiners are treated like half-width chars, but it would be better to treat the combined chars as a single character.

The following table

+-----------+-------+
| a         | back  |
+===========+=======+
| ø̞         | o̞     |
+-----------+-------+

gets parsed as having rowspans:

ArrayTable {arrayTableCells = array ((1,1),(2,2)) [((1,1),ContentCell 2 1 [" a         ","==========="," \248\798         "]),((1,2),ContentCell 2 1 [" back  ","======="," o\798     "]),((2,1),ContinuationCell (1,1)),((2,2),ContinuationCell (1,2))], arrayTableHead = Nothing, arrayTableFoot = Nothing, arrayTableColSpecs = array (1,2) [(1,(AlignDefault,11)),(2,(AlignDefault,7))]

Replacing the accented characters (formed with combining accents) with regular ascii letters makes the problem go away, so I assume it has to do with character width calculations.

Originally jgm/pandoc#9145.

Column alignments are not calculated correctly in tables like

+:---------:+
| Col 1 & 2 |
+-----+-----+
| foo | bar |
+-----+-----+

All columns spanned by a column with alignment information should share that alignment. Currently only the first column gets the right alignment.

Describe:

• scope of this package
• algorithm used
• example use

This table should not crash the tracer.

+---+
|one|
+---+---+
|one|two|
+---+---+

A malformed table like

+---+--+
|one|two|
+---+---+
|one|two|
+---+---+

should still be parsed as a table with four cells, despite the top line being too short.