[developers] Treebanking with FFTB

Woodley Packard sweaglesw at sweaglesw.org
Sun Jun 25 00:20:55 CEST 2017 

Glad to hear the happy ending :-)
Woodley



> On Jun 24, 2017, at 9:28 AM, David Moeljadi <davidmoeljadi at gmail.com> wrote:
> 
> Dear Woodley,
> 
> I am sorry, my mistake, I could not annotate the first sentence because some words are not in the lexicon.
> I have just added the new words and run the FFTB again, now it works!
> 
> Best regards,
> David
> 
>> On Sun, Jun 25, 2017 at 12:13 AM, David Moeljadi <davidmoeljadi at gmail.com> wrote:
>> Dear Woodley,
>> 
>> Thank you very much! Yes, I missed this
>> 
>> --webdir=$LOGONROOT/lingo/answer/fftb
>> 
>> Now it works!!
>> 
>> I have another question, I have made a new test-suite called "kbbi". This test-suite has only two sentences.
>> It is completely new, so it does not have a gold profile.
>> I followed the step-by-step instruction in my documentation: http://moin.delph-in.net/IndraTreebanking
>> Last time (last year) it worked well, but this time no.
>> =====
>> ~/logon$ ./parse --binary --ind --protocol 2 --best 1 --limit 0 --count 8 kbbi
>> International Allegro CL Enterprise Edition
>> 8.2 [64-bit Linux (x86-64)] (May 24, 2016 0:22)
>> Copyright (C) 1985-2010, Franz Inc., Oakland, CA, USA.  All Rights Reserved.
>> 
>> This standard runtime copy of Allegro CL was built by:
>>    [TC13152] Universitetet i Oslo (IFI)
>> 
>> ; Loading /home/david/logon/dot.tsdbrc
>> ; Loading /home/david/.tsdbrc
>> 
>> [changing package from "COMMON-LISP-USER" to "TSDB"]
>> TSNLP(1): NIL
>> TSNLP(2): 5
>> TSNLP(3): 2
>> TSNLP(4): 5
>> TSNLP(5): NIL
>> TSNLP(6): T
>> TSNLP(7): 
>> "/home/david/logon/lingo/lkb/src/tsdb/home/"
>> TSNLP(8): 0
>> TSNLP(9): NIL
>> TSNLP(10): 1
>> TSNLP(11): 2
>> TSNLP(12): ; Loading /home/david/logon/ntu/ind/lkb/script
>> ;   Loading /home/david/grammar/ind/Version.lsp
>> ;   Loading /home/david/grammar/ind/lkb/globals.lsp
>> set-coding-system(): activated UTF8.
>> ;   Loading /home/david/grammar/ind/lkb/user-fns.lsp
>> ;   Loading /home/david/grammar/ind/lkb/user-prefs.lsp
>> read-repp(): reading file `vanilla.rpp'.
>> 
>> Reading in type file matrix
>> Reading in type file head-types
>> Reading in type file indonesian
>> WARNING: Type `SIGN-MIN' redefined.
>> WARNING: Type `HEAD-MIN' redefined.
>> Reading in type file mtr
>> Reading in type file tmt
>> Checking type hierarchy
>> Checking for unique greatest lower bounds
>> Expanding constraints
>> Making constraints well formed
>> [23:57:11] gc-after-hook(): {L#30 N=14M O=0 E=95%} [S=2.1G R=172M].
>> 
>> Expanding defaults
>> Type file checked successfully
>> Computing display ordering
>> Reading in lexical entry file lexicon
>> [23:57:11] gc-after-hook(): {L#31 N=30M O=0 E=90%} [S=2.1G R=282M].
>> [23:57:16] gc-after-hook(): {L#32 N=35M O=0 E=98%} [S=2.1G R=282M].
>> 
>> Reading in rules file rules
>> Reading in lexical rules file lrules
>> Reading in lexical rules file irules
>> Reading in root file roots
>> Reading in parse node file labels
>> ;   Loading /home/david/grammar/ind/lkb/mrsglobals.lsp
>> 
>>  (recompiling semantic indices)
>> [23:57:33] gc-after-hook(): {L#33 N=39M O=6.1K E=99%} [S=2.1G R=283M].
>> [23:57:34] gc-after-hook(): {L#34 N=41M O=0 E=91%} [S=2.1G R=283M].
>> 
>> Warning: TELAH has no semantics and no filter rule
>> [23:57:34] gc-after-hook(): {L#35 N=43M O=0 E=91%} [S=2.1G R=283M].
>> ...(some other warnings)...
>> 
>> Indexing complete
>> read-vpm(): reading file `semi.vpm'.
>> ;   Loading /home/david/grammar/ind/lkb/mt.lsp
>> read-transfer-rules(): reading file `trigger.mtr'.
>> read-transfer-rules(): reading file `gen-fixup.mtr'.
>> 
>> Building rule filter
>> Building lr connections table
>> Constructing lr table for non-morphological rules
>> Grammar input complete
>> NIL
>> TSNLP(13): 
>>   14 skeletons in `/home/david/logon/lingo/lkb/src/tsdb/skeletons/ind/'.
>> NIL
>> TSNLP(14): 
>> [t4000b] BEGIN
>> [t4000c] BEGIN
>> [t4000d] BEGIN
>> [t4000e] BEGIN
>> [t4000f] BEGIN
>> [t40010] BEGIN
>> [t40011] BEGIN
>> [t40012] BEGIN
>> [t40011] EOF
>> [t40010] EOF
>> [t4000f] EOF
>> [t4000e] EOF
>> [t4000d] EOF
>> cat: /tmp/.parse.tmp.david.12811: No such file or directory
>> ~/logon$ 
>> =====
>> 
>> Also, I tried 
>> =====
>> fftb -g ind.dat --browser --webdir=$LOGONROOT/lingo/answer/fftb /tmp/kbbi-demo/
>> =====
>> This command line launched the FFTB page, I could see the two sentences in my test-suite there (screenshot attached),
>> but when I clicked sentence number one, a page "404 no stored forest found for this item" appeared (screenshot attached).
>> However, when I clicked sentence number two, I could go to the annotation page (screenshot attached)!!
>> Maybe this time I missed something again?
>> 
>> Best regards,
>> David
>> 
>>> On Sat, Jun 24, 2017 at 1:09 AM, Woodley Packard <sweaglesw at sweaglesw.org> wrote:
>>> Dear David,
>>> 
>>> I believe your problem is that FFTB wasn't able to find the stylesheet and javascript files that it uses for its browser interface.  In the examples on the CapitolHillTreebank page, the working directory when invoking fftb was the same directory where fftb expects to be run from.  If that is not the case, you need to add a command-line option like:
>>> 
>>> --webdir=/path/to/fftb-dir/assets
>>> 
>>> or if you are using the FFTB binary from the LOGON tree:
>>> 
>>> --webdir=$LOGONROOT/lingo/answer/fftb
>>> 
>>> Can you try that and see if it helps?  Good luck,
>>> Woodley
>>> 
>>> 
>>>> On 06/23/2017 12:06 AM, David Moeljadi wrote:
>>>> Dear all,
>>>> 
>>>> I tried to do treebanking with FFTB, using the command lines in
>>>> http://moin.delph-in.net/CapitolHillTreebank.
>>>> I could see the FFTB page in my Firefox browser (screenshot attached), but
>>>> when I clicked the sentences (for example 11 Hujan turun), I got a page
>>>> "404 Not Found" (screenshot attached), I could not see the tree or the
>>>> discriminants.
>>>> Did I miss something here?
>>>> 
>>>> ====
>>>> ~/grammar/ind$ fftb -g ind.dat /tmp/mrs-demo/ --browser --gold
>>>> /tsdb/gold/mrs
>>>> grammar image: ind.dat
>>>> Just one TSDB profile: /tmp/mrs-demo/
>>>> Would update from profile: ../ind/tsdb/gold/mrs
>>>> listening on http://127.0.0.1:53055/private/
>>>> should GET    /private/
>>>> should GET    /favicon.ico
>>>> ...
>>>> [1031] {0 decisions} UCSTAT: input 0 edges, output 0 edges
>>>> {0 edges} {0 trees} no gold writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> [1041] {1 decisions} UCSTAT: input 4 edges, output 3 edges
>>>> {3 edges} {1 / 1 trees active} different writing tsdb relation 'tree' with
>>>> 172 tuples
>>>> 
>>>> [1042] {0 decisions} UCSTAT: input 6 edges, output 5 edges
>>>> {5 edges} {1 trees} no gold writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> [1051] {1 decisions} UCSTAT: input 4 edges, output 3 edges
>>>> {3 edges} {1 / 1 trees active} different writing tsdb relation 'tree' with
>>>> 172 tuples
>>>> 
>>>> [1061] {0 decisions} found a stored forest, but couldn't use it.
>>>> -> no usable stored forest
>>>> unable to get parse forest for 1061
>>>> writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> [1062] {0 decisions} found a stored forest, but couldn't use it.
>>>> -> no usable stored forest
>>>> unable to get parse forest for 1062
>>>> writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> [1071] {2 decisions} UCSTAT: input 12 edges, output 10 edges
>>>> {10 edges} {0 / 2 trees active} writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> [1072] {2 decisions} UCSTAT: input 9 edges, output 7 edges
>>>> {7 edges} {0 / 2 trees active} writing tsdb relation 'tree' with 172 tuples
>>>> 
>>>> should GET    /private/parse?profile=/&id=11
>>>> item id 11 -> input 'Hujan turun.'
>>>> profile parse id 11
>>>> gold parse id 11
>>>> preferred gold result id 0
>>>> -> got gold preferred tree 0xf05fd0
>>>> -> loaded stored forest
>>>> found stored forest (4 edges connected to 1 roots).<br/>
>>>> UCSTAT: input 4 edges, output 3 edges
>>>> should GET    /private/session?0
>>>> should GET    /private/parse?profile=/&id=1041
>>>> ...
>>>> TIMERS (792 calls = ~ 108,5µs overhead):
>>>> get decisions                   1,4ms for 102 events = 13,5µs per event
>>>> load gold tree                   0,7ms for 53 events = 13,0µs per event
>>>> load forest                     8,9ms for 171 events = 51,9µs per event
>>>> unary closure                   2,5ms for 117 events = 21,2µs per event
>>>> count solutions                 122,6µs for 162 events = 0,8µs per event
>>>> ~/grammar/ind$
>>>> ====
>>>> 
>>>> Also, I tried calling it without --gold /dir/ (screenshot attached) but
>>>> still the same, I got "404 Not Found" when I clicked the sentences.
>>>> ===
>>>> ~/grammar/ind$ fftb -g ind.dat /tmp/mrs-demo/ --browser
>>>> grammar image: ind.dat
>>>> listening on http://127.0.0.1:38959/private/
>>>> should GET    /private/
>>>> should GET    /favicon.ico
>>>> ...
>>>> should GET    /private/parse?profile=/&id=142
>>>> item id 142 -> input 'Dia menggonggong.'
>>>> profile parse id 142
>>>> -> loaded stored forest
>>>> found stored forest (3 edges connected to 1 roots).<br/>
>>>> UCSTAT: input 3 edges, output 3 edges
>>>> should GET    /private/session?0
>>>> should GET    /private/parse?profile=/&id=11
>>>> item id 11 -> input 'Hujan turun.'
>>>> profile parse id 11
>>>> -> loaded stored forest
>>>> found stored forest (4 edges connected to 1 roots).<br/>
>>>> UCSTAT: input 4 edges, output 3 edges
>>>> should GET    /private/session?1
>>>> should GET    /private/parse?profile=/&id=21
>>>> item id 21 -> input 'Adi menggonggong.'
>>>> profile parse id 21
>>>> -> loaded stored forest
>>>> found stored forest (3 edges connected to 1 roots).<br/>
>>>> UCSTAT: input 3 edges, output 3 edges
>>>> should GET    /private/session?2
>>>> should GET    /private/exit
>>>> TIMERS (0 calls = ~ 0,0µs overhead):
>>>> ~/grammar/ind$
>>>> ===
>>>> 
>>>> Best regards,
>>>> David
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.delph-in.net/archives/developers/attachments/20170624/5b532d27/attachment.html>

More information about the developers mailing list