jsbueno
commented
5 years ago "before :"
+-----------------------+----------------------------------+
| nome | endereco |
+-----------------------+----------------------------------+
| José da Silva | Rua dos Bobos, 0
Cidade Fantasma |
| José Maria
(Zé Maria) | R. XPTO, 1 |
+-----------------------+----------------------------------+"after" :
+---------------+------------------+
| nome | endereco |
+---------------+------------------+
| José da Silva | Rua dos Bobos, 0 |
| | Cidade Fantasma |
| | |
| José Maria | R. XPTO, 1 |
| (Zé Maria) | |
+---------------+------------------+
This PR fixes the immediate issues exposed at https://gist.github.com/turicas/8bd86dd425d2212dbe214d725121be7c?fbclid=IwAR1VitQE02l7OiCjCKDsdYTJJ9-mC53xlFOD2yhWoxZyCPWrjqchl7H7CPo
That is - any cell containing a line-feed character breaks the text-plugin output.
Besides creating multiline cells in the text output, the code also includes a "blank" line bellow any row having a multi-line cell, so that contents are visually separateted.
Not covered by this code: Parsing back text-tables with multi-line cells. (I invite the discussion of what features would be better to enable that in this very same PR - although merging it as is will be mostly harmless - the output of the text plug-in is supposed to be pretty, but it is a dangerous road to ensure data integrity round-trip between its output and back - multi-line cells will complicate a lot the current logic)