Transducer infrastructure


- Images


VASTA - Trond says that it works so well that we can add a link for it in our doc.pages. I have today made some improvements. When are they put to the Vasta on the internet? (sme-ped.cg3) - /opt/smi/sme/bin/

LEKSA - bug 742 (words with different id-s, when the user chooses ALL)

NUMRA - add ordinals - an automat  

We must discuss this. I do not know whether our solution is the best one: We have ordinals and cardinals in the same automaton. Should we instead have two: sme-num.fst and sme-ord.fst? Now, numbers without period trigger cardinals and number with trigger ordinals, but I do not know whether we can say to oahpa "give me 10 ordinals, from word to arabic" (since the words are not tagged as ord or card).

Also: I added sjd numeral generator. These are of course a bit like curiosita, but I have come to think that teachers interested in showing the similarities between different Sámi languages could explain that to the pupils / students via this machin.

yksi = 1
ensimmäinen = 1.

MORFA - feedback also in Morfa-cealkagat (now it is only in Morfa-sánit)
		add paradigm-button to the feedback - NO GOOD IDEA AFTERALL
		placenames should not be in the plural Morfa
		eastern dialect - it is ready
		
I have done the eastern grammar - and I will today make a feedback-version
according to it.
The dialect is to be choosen in Makefile - eastern: NOT-KJ - and  western: NOT-GG:
-Makefile: 
$(RESTRFILES): $(SRCS)
	@echo  
	@echo "*** Removing non-orthographic entries - restr ***" ;
	@echo
	@egrep -v '(SUB|\^NG\^|NOT-KJ)' $(patsubst
$(TARGET)/int/%.restr,$(TARGET)/src/%.txt,$@) > $@

you remove NOT-KJ lines => 

and in lexiconfiles: 
- eastern: generate no dialect class and dialect class="n" 
- western: generate no dialect class and dialect class="o"		

pedlex      		grammar      											lexicon
class="not-kj"   everything but not NOT-KJ = choose NOT-GG, remove NOT-KJ  
class="not-gg"   everything but not NOT-GG = choose NOT-KJ, remove NOT-GG  	 
			
----------------------------
class="not-gg" = NOT-GG NOT-KJ = kaaresuoando
class="not-kj"
class="not-kj" class="not-gs"

    <dialect class="not-gg"/>
	<dialect class="not-kj"/>

To make dialect x:
remove NG and SUB
remove all not-x (thus, it goes without saying that we choose not-y, not-z, ...)
	
gánda dialect class="n"  (boy)
bárdni dialect class="o"		
nieida (no dialect class) (girl)



SAHKA - Kjellaug has started drawing. I will extend the dialogues and make more when the first-meeting and visit are implemented so we see if the format is good.
I think there could be three fields on the screen - Left: drawings    Middle: dialog    Right: grammarfeedback and so on


Feedback:
---------
Morfa: make eastern version - it will be finished tomorrow
Marfa and Vasta: translate feedback to sámi (finnish, english) - the sami version will be done this week




Grammar-explanations
-------------------------------
I will make small grammar explanations - like What means agreement between adjective and noun, what is finite verb, when do we use illative and so on.
What would be the best way of implementing them? Make a special window for them?

grammar doc
on the oahpa doc pages
in different languages
parametrised for that
correct links to page this-and-that in the game

or

in xml in the database, an integrated part of the game

integrated a (much)
 as feedback buttons
 
integrated b (little)
 upper right corner


coherent grammar on manpages

deeplinks in the game to relevant paragraph

doc pages work ourselves with them