Ja. Fik svar fra The Bat!'s tekniske liste. Undskyld, at jeg glemte at lukke spørgsmålet.
Koden er:
%headers%-
%_TEXT=">>;{;%SETPATTREGEXP=""(?is)(-+\s*Oprindelig meddelelse.*~-~>.*-*_->$\s+|-+\s*Original Message.*~-~>.*-*_->$\s+)%-
{0,1}(-+\s*Oprindelig meddelelse.*?\n|-+\s*Original Message.*?\n(Hash:.*?\n)?\s*)?%-
(.*?)(-+\s*Oprindelig meddelelse|-+\s*Original Message|\z)""%REGEXPBLINDMATCH=""%text""%SUBPATT=""4"""%-
%QINCLUDE="replace"%-
%_TEXT%-
Den kode fandt jeg frem til efter at have fået dette svar:
You might want to look at the two "Search and Replace" quick templates
in The Bat! Macro and Solutions Library,
http://cgi.marck.free-online.co.uk/library.php and copy (and possibly
adapt) one of them for your requirement.
Then you could write a template that would include (%Qinclude) the
Search and Replace template three times.
The first time you would use the Search and Replace template
to change all occurrences of ">> " (without the quotes of
course) to something unlikely ever to be in the text, perhaps
"~~ ".
The second time you would delete all occurrences of ">" by
changing them to "" (no character at all).
The third time you would change all occupancies of "~~ " to
"> ".
Men jeg går ud fra, at dette er specifikt for The Bat! Selv om der bruges reg.expr. så er brugen vist en del forskellig fra php. Eller hvad?
I hvert fald kan jeg nu fjerne disse irriterende "oprindelige meddelelser" i bunden af e-mails fra MS-brugere. Jeg skal -- når jeg får tid -- have lavet filteret om, så det også virker på html-mails og ikke kun tekst-mails som nu.