#121 avoiding forced-uppercase generation when identical dictionary-uppercase exists

[Kevin Brubeck Unhammer]

Originally discussed in thread
avoiding forced-uppercase generation when identical dictionary-uppercase exists Kevin Brubeck Unhammer Tue, 28 Feb 2017 01:34:30 -0800 (nabble)

---

If I have this in foo.dix:

<e><p><l>PC-ane</l><r>PC<s n="n"/><s n="m"/><s n="pl"/><s n="def"/></r></p></e>

then I get

$ echo '^PC<n><m><pl><def>$' | lt-proc -d foo.autogen.bin
PC-ane

which is what I want. However, sometimes people write compounds in
lowercase, so I try including both upper and lower-case lemmas (I only
care about generation now):

<e><p><l>pc-ane</l><r>pc<s n="n"/><s n="m"/><s n="pl"/><s n="def"/></r></p></e>
<e><p><l>PC-ane</l><r>PC<s n="n"/><s n="m"/><s n="pl"/><s n="def"/></r></p></e>

but that gives

$ echo '^PC<n><m><pl><def>$' | lt-proc -d foo.autogen.bin
PC-ane/PC-ANE

Would it make sense for lt-proc to not output forced-uppercase analyses
if there's an otherwise identical dictionary-uppercase analysis? (Would
it be easily implementable?)

The only other solution I can think of is to put a tag on anything that
has an uppercase lemma, which sounds like more testvoc.

[Kevin Brubeck Unhammer]

https://github.com/jimregan/lttoolbox/issues/31#issuecomment-292018281 – relevant?

[Jim O'Regan]

My changes there won't solve this particular issue (yet :), it just handles a somewhat similar problem with mixed case.

[Kevin Brubeck Unhammer]

Actually, it seems to:

$ cat tests/data/carefulcase-mono.dix
<?xml version="1.0" encoding="UTF-8"?>
<dictionary>
 <alphabet/>
 <sdefs>
   <sdef n="n"/>
   <sdef n="m"/>
   <sdef n="pl"/>
   <sdef n="def"/>
 </sdefs>
 <section id="main" type="standard">

<e><p><l>pc-ane</l><r>pc<s n="n"/><s n="m"/><s n="pl"/><s n="def"/></r></p></e>
<e><p><l>PC-ane</l><r>PC<s n="n"/><s n="m"/><s n="pl"/><s n="def"/></r></p></e>

 </section>
</dictionary>

$ lttoolbox/lt-comp rl tests/data/carefulcase-mono.dix tests/data/carefulcase-mono.bin
main@standard 8 8

$ echo '^PC<n><m><pl><def>$ ^pc<n><m><pl><def>$' | lttoolbox/lt-proc -C tests/data/carefulcase-mono.bin
PC-ane pc-ane

vs old

$ echo '^PC<n><m><pl><def>$ ^pc<n><m><pl><def>$' | lt-proc -g tests/data/carefulcase-mono.bin
PC-ane/PC-ANE pc-ane

[Jim O'Regan]

Ok, committed

[Kevin Brubeck Unhammer]

Nice! Btw, would it be possible to make the option compatible with -g or -d or -n? So we can do lt-proc -g -C or lt-proc -n -C, since modes.xml typically say lt-proc $1 in order to let the driver script optionally remove unknown word marks.

[Jim O'Regan]

On 18 May 2017 at 09:41, Kevin Brubeck Unhammer unhammer@users.sf.net wrote:

Nice! Btw, would it be possible to make the option compatible with -g or -d
or -n? So we can do lt-proc -g -C or lt-proc -n -C, since modes.xml
typically say lt-proc $1 in order to let the driver script optionally remove
unknown word marks.

Give it about two weeks, I'll probably need -d by then.