Raw Input

Each non-blank line of the datafile either consists of

a tag only, or
a tag, followed by a space, minus sign, a space, and some data

The following table describes each tag:

BOF begin of file

EOF end of file

AU author

TI title

A sample of the datafile is:

BOF
BOR
AU - Daniel P. Bovet
AU - Macro Cesati
TI - Understanding the Linux Kernel
TI - From I/O Ports
TI - to Process Management
LA - English
EOR
BOR  
AU - John E. Hopcroft
AU - Jeffery D. Ullman
TI - Introduction to Automata Theory
TI - Languages, and Computation
EOR
EOF

Desired Output

We find that in some record (block delimited by a pair of BOR and EOR), there exist consecutive lines which begin with the same tag. What we want is joining these lines together and remove the duplicate tags. For example:

BOF
BOR
AU - Daniel P. Bovet, Macro Cesati
TI - Understanding the Linux Kernel, From I/O Ports, to Process Management
LA - English
EOR
BOR  
AU - John E. Hopcroft, Jeffery D. Ullman
TI - Introduction to Automata Theory, Languages, and Computation
EOR
EOF

Script and Comments

Script₁

[ 1] :loop
[ 2] N
[ 3] /^\([^ ]*\) .*\n\1 /{
[ 4] s/\n\([^-]*\)-/,/
[ 5] b loop
[ 6] }
[ 7] P
[ 8] D