(Moved from terki)
An odd feature of Lojban is that you actually have to use zoi-quotes to talk about some rafsi. I can say zo xuk. rafsi zo xruki and zo li'u rafsi zo litru, but I have to say zoi .ly. cpi .ly. rafsi zo cipni, since cpi- isn't a valid Lojban word of any sort. I suppose the same is true of lan for lanme, but for less meaningful reasons. --mi'e mark.
See the Book, chapter 19, section 10, example 10.5.
- Yuck. I foresee one change that will happen in Lojban usage - zo cpi would become acceptable even if it doesn't parse. Otherwise how can you fill the x1 of rafsi in a reasonable way?
- Well, you can use zoi, like I did above. That's "reasonable"... but only barely, I'll admit. lan is a harder one to remember, since it fails legality for more subtle reasons (I personally believe we should have abolished the nonsense about no la/lai/doi in cmene and just required a mandatory pause before as well as after in all settings.) --mi'e mark.
- I don't think that zo .cpi., with pauses before and after, would ever be ambiguous. Could this be considered an extremely minor extension to the word resolution algorithm? --rab.spirµµµ
- I like using cy.py.ibu or la me cy.py.ibu to refer to rafsi. It solves the CCV problem and also treats all rafsi identically, thus solving the problem of call the word li'u a rafsi and not a cmavo. — Adam
- cy.py.ibu is a variable, a pro-sumti referring to something called CPI. la me cy.py.ibu is no better. You need me'o cy.py.ibu, believe it or not. I'm told that's the way to refer to actual letters. --mi'e mark
- I use zoi for all rafsi. zo li'u is the word that is used to close a quotation, not the fragment of a word that is the rafsi of litru. --phma
actually, there is still no word resolution algorithm. And current lexers can easily be modified to treat anything delimited by pauses after "zo" correctly. — raphael
There is non-debugged and non-baselined WordResolutionAlgorithm at http://www.lojban.org/files/software/BRKWORDS.TXT. It would reject cpi, though with a pause both before and after it couldn't be mistake as a part of another word. — Adam