[Date Prev][Date Next][Subject Prev][Subject Next][ Date Index][ Subject Index]

Re: uniqcomp.pm - Working for others?



Hi Patricia,
I tried you text and it got as far as the last sentence before it starts putting an endless loop of nmin's

Hope this helps,

Russ
On Nov 11, 2005, at 1:02 PM, Patricia M. Godfrey wrote:
Robert Holmgren wrote:
Here's my version (frame COMP):<<

OK, I decoded that, added to U2. Still the same problem, but I think
I've got a clue. It appears to run OK through the end of the files, then
goes into the endless loop. So I called each of the files in turn in
List in hex format, but the last three chars in both were 0D 0A 1A
(13,10, 26--a CR/LF and EOf, no?).

As these files were pretty heavily formatted, and with Use Styles at
that, I then tried getting rid of the US command. Still no luck. So then I cut out the last paragraph, on which comp appeared to choke. I enclose it here, XplNcoded. I suspect there is some tipping point in the amount of MD commands before it chokes. Anyone else want to try with this file?
Save it, then CI possessive to genitive and save that as another file,
then Comp them.
XPLeNCODE v2.0
b-gin [UNTITLED]
{<}MD+BO{>}{<}C3{>}.{tab}Formation of the possessive of words
 ending in sibilants other than ess:{<}MD-BO{>} Do we say {<}
MD+IT{>}Xerox[267]{<}MD-IT{>} or {<}MD+IT{>}Xerox[267]s, Berl
ioz[267] or Berlioz[267]s?{<}MD-IT{>} Again, house style must
 decide. But the rule should note that foreign names ending i
n silent sibilants (Descartes) or sibilants not pronounced as
 such (Vel{160}squez in Castilian, where the {<}MD+IT{>}z{<}M
D-IT{>} is pronounced {<}MD+IT{>}th{<}MD-IT{>}) should logica
lly not be classed as sibilants.[cr|lf]
-nd
XPLeNCODE

Patricia M. Godfrey