Perhaps you want to split on /\n\b/. That will require new records to begin with a word char (because \n is not a word char).
I am not sure how this helps me. In the example I have provided above, I have added the newlines. In reality, the email message is provided to me as a scalar, with all the newlines and all in it already. I just use Mail::IMAPClient to grab the message.
As I said, the code I have above works fine even with splitting on newlines. Where it fails is its inability to "lookbehind" when it encounters a line without a label -- something that looks like so --
: sometext
I believe that when I encounter a line like above, I have to lookbehind and say, "Ha, this line doesn't begin with a label, hence it is just a continuation of the value of the previous label."
This is where I am lost. | [reply]
[d/l] |
| [reply] |
And you want to split on those newlines, except where the next line doesn't begin with a letter. You want those lines that don't begin with a letter not to be split from the previous lines. Maybe you should just try it and see what you get.
I tried. And... Sweet. This works just swell. I wish I could vote more than once for your tip for what I learned from it.
Many thanks.
| [reply] |