pipe jBin to sed to remove trailing empty columns#37
pipe jBin to sed to remove trailing empty columns#37davidworkman9 wants to merge 4 commits intodbashford:masterfrom
Conversation
d163120 to
4b90114
Compare
|
Bump. |
|
Not familiar enough with windows to know if sed is installed by default? I imagine it isn't. I struggle with this because of the added external dependency. Can you provide an example problematic xls for me to play around with? |
|
I can understand not wanting to add an external dependency but I think there are other things in this library that aren't windows compatible. I'll work on getting a document that demonstrates this issue. Would you merge a request that didn't use sed unless it was not running on windows? |
|
I've been trying to reduce the number of external dependencies, not grow them. I had more, like unzip, but systematically pushed through nuking those. Some, the ones that just outright do the extraction, are hard to avoid. Ideally the solution is one performed in node. I wonder if the |
Makes sense.
That's probably a workable solution. I would still see value in getting rid of the trailing empty columns, but at least that could be done in JS then. I'll see if I can implement that. |
Yep! That's the idea. |
I had a problem where XLS and XLSX files had reference to a column number way off in the distance with nothing in it (probably from someone applying formatting an entire row). I started hitting max buffer limits, and I couldn't even find a number that would handle it without hitting it! So instead I pipe the output of jBin to sed and remove all the trailing empty columns in a row.
I also opened an issue with J here, and it sounds like there might eventually be a switch to have it to this itself, or even be the default behaviour, so this could be removed when that becomes available.