Re: Lpeg 0.12: "grammar has too many rules"

[Petite Abeille <petite.abeille@...>]

On Sep 23, 2013, at 9:53 AM, Sean Conner <sean@conman.org> wrote:

>  Oh wow ... the corpus of email I used for testing did not include an email
> address of that format.  My bad.  It's been fixed.

Much excellent. Survived around 700K messages with flying colors :)

> -spc (Thanks)

You are welcome. Very nicely done.


P.S.

Now, what would be *really* useful would be a little utility along the lines of reformime to detect MIME part boundaries:

% reformime -i < /Volumes/Queens/Mail/pa/mail/1jw/h2/m5lsig.eml

$ bin/reformime -i < m5lsig.eml
section: 1
content-type: multipart/mixed
content-transfer-encoding: 8bit
charset: UTF-8
starting-pos: 0
starting-pos-body: 3085
ending-pos: 1090972
line-count: 17973
body-line-count: 17913

section: 1.1
content-type: text/plain
content-transfer-encoding: base64
charset: utf-8
starting-pos: 3126
starting-pos-body: 3202
ending-pos: 3202
line-count: 4
body-line-count: 0

section: 1.2
content-type: message/rfc822
content-transfer-encoding: 7bit
charset: UTF-8
starting-pos: 3243
starting-pos-body: 3305
ending-pos: 279095
line-count: 4497
body-line-count: 4494

…

[1] http://www.courier-mta.org/reformime.html