[SUGGESTION] Table parsing happens too easily

Completed

AmberV

May 30, 2019 01:35

I've run into a few cases now where the "Automatically format next line in tables" setting will trigger on lines that are not table lines. It almost seems to be considering any line with more than one pipe in it, but that's going to make a mess of any code samples, shell commands, regex lines and so forth. Speaking of regex, there is a convenient example that can be used to trigger the behaviour in the software data itself:

 

1. Open preferences, and from the General pane, copy the "Ignore files matching" text field.

2. Paste the regex into a note on its own empty line, and press Return.

 

   (Guess how I found this one!)

 

Now to be fair, that one pretty much does match a table row. :) It might be fun finding a way around regex examples. But the example I ran into myself was when using one of my own personal archival codes. I mark internal links with the following syntax: <|LINKID|>, which may point to a document ID, or to a matching >|ANCHORID|< typed into a document somewhere. Both examples might commonly appear in paragraphs much like this one---and like this one, I'll have to go and turn off table parsing in order to continue drafting this report.

 

I've checked, and MMD4 does *not* recognise either of these examples as looking like tables---so hopefully you already have a good algorithm floating around.

0

• 

  Fletcher Penney

  May 30, 2019 19:25 Edited

  The MMD Composer 4 "algorithm" simply requires that the first character of the line be `|`.  Which is not necessary for MMD tables, so it technically misses some.

  I'll look to see if there is a better approach, but if not can consider using the old behavior.

  0
• 

  Fletcher Penney

  May 30, 2019 21:35

  I'm not sure there is a better algorithm since anything with a `|` could technically be a table line.  So that newline behavior will only be triggered with a leading `|` before other content in the line.

  0

Please sign in to leave a comment.