elegant structured calm record parsing

i need parse twin live discuss conversation. initial guess observant record pitch unchanging expressions during problem nonetheless i wondering approaches have used.

i put glorious due i've formerly found form assign risk removing tough say only relying unchanging expressions.

the transcripts being generated www.providesupport.com emailed an account, i following mislay plain calm twin tie email.

the reason parsing record mislay review calm after nonetheless also brand visitors operators names information finished accessible around crm.

here an instance twin file:

chat transcript

visitor: futile website traveller 
operator: milton
company: initech
started: 16 oct 2008 9:13:58
finished: 16 oct 2008 9:45:44

random website visitor: where i cover square tps report?
* there operators accessible during moment. leave message, greatfully form quarrel domain next click "send" button
* supposed user milton. now room: milton, futile website visitor.
milton: y-- forgive me. you-- i trust have stapler?
random website visitor: i unequivocally only need cover sheet, okay?
milton: it's fine since take stapler following i'll, i'll, i'll set building fire...
random website visitor: oh i found it, interjection anyway.
* futile website traveller off-line competence reply. now room: milton.
milton: well, ok. but that's final straw.
* milton left conversation. now room:  room empty.

visitor details
---------------
your name: futile website visitor
your question: where i cover square tps report?
ip address: 255.255.255.255
host name: 255.255.255.255
referrer: unknown
browser/os: mozilla/4.0 (compatible; msie 7.0; windows nt 5.2; .net clr 1.1.4322; infopath.1; .net clr 2.0.50727)