How to parse a wiki style table in a multi-text field

[#1]

So my friend @Mourad Marzouk asked in the Adaptavist forum:

• How to use script runner to parse a wiki style table in a multi-text field 

But because Groovy makes my head hurt, I went ahead and posted an answer on how to do it with Automation, which well... probably belongs here.

Also, I think this is a simpler solution for this (if I'm interpreting the original question correctly):

• Parse table in Description column using Jira automation 

The table we're splitting looks like this:

Old Part #	Old Rev	New Part #	New Rev	Part Description	Procurement/End Item
Part number 1	Old Revision 1	Part number 1	Revision 1	Description 1	End Item 1
Part number 2	Old Revision 2	Part number 2	Revision 2	Description 2	End Item 2

If you look at the wiki source for this table, it looks like this:

||*Old Part #*||*Old Rev*||*New Part #*||*New Rev*||*Part Description*||*Procurement/End Item*||
|Part number 1|Old Revision 1|Part number 1|Revision 1|Description 1|End Item 1|
|Part number 2|Old Revision 2|Part number 2|Revision 2|Description 2|End Item 2|

First you split the description by newlines: {{issue.description.split("\n")}}, and iterate through each line with the For each advanced branching.

Then for each line ({{row}}), if it doesn't contain || (header), but does start with a |, then create a new subtask.

You then can split the row to get each "column", and those could be put in fields or whatever. I just threw them all in the Description.

First you split the description by newlines: {{issue.description.split("\n")}}, and iterate.

First column (Old part #) is: {{row.split("\|").get(1)}}

(Numbering starts with 0, but because the tables _start_ with a pipe character, we actually are getting the second column). Example:

0|1|2|3|4|5|6|

|Part number 1|Old Revision 1|Part number 1|Revision 1|Description 1|End Item 1|

So you'll get sub-tasks for each of the content rows in the table, and if you wanted, you could populate fields with different column data.

Now this assumes that you only have one table per Description, and also that there aren't any other random lines that start with a | (pipe) character.

It might be safer to change the second comparison to exactly matches regular expression and use this:

 ^\|.*\|$ 

That will ensure that the line both starts and ends with a pipe character.

[Dave Liao]

@Darryl Lee - the images aren't loading in this article FYI!

[Darryl Lee]

Gah. Thanks. This is what happens when I try to copy-paste images here. Will fix in a bit. 

[Bill Sheboy]

Good stuff, @Darryl Lee !

Another good pairing of techniques with that newline split is to add regEx and smart value, list filtering.  Now...if smart value, list filters also worked with smart value params and smart branches could produce objects...we'd have a lot more options  :^)

Kind regards,
Bill