[CL] Clarification of section A.2.3.11 in 24707

[Pat Hayes]

Hi Cameron

The intention was that cltext would be the top node for the grammar,  
so that (d) would be the natural starting rule..

Pat

On Dec 1, 2009, at 9:24 AM, Cameron Ross wrote:

> Hi Pat,
>
> One final question in-line below:
>
> Cameron.
>
>
> Pat Hayes wrote:
>>
>> On Nov 30, 2009, at 3:24 PM, Cameron Ross wrote:
>>
>>> The 24707 defect report indicates that the productions in Section  
>>> A.2.3.11 should be written as:
>>>
>>> (a) phrase = sentence | module | ( open, 'cl-imports',  
>>> interpretablename, close ) | ( open, 'cl-comment', enclosedname  
>>> [cltext], close );
>>
>> as I recall, that last entry should be
>>
>> (open, 'cl-comment', (enclosedname | quotedstring), [cltext], close)
>>
>> This is to allow the use of either double or single quotes to  
>> enclose the comment text.
>>
>>>
>>> (b) text = { phrase };
>>>
>>> (c) namedtext = open, 'cl-text', interpretablename, text, close;
>>>
>>> (d) cltext = module | namedtext | text;
>>>
>>>
>>>
>>> I have the following questions related to these productions:
>>>
>>> 1)  What is the purpose of the last alternative in production (a)?
>>
>> To allow comments to be attached to top-level phrases and entire  
>> subtexts. For example
>>
>>
>> (cl-text
>> ...
>> (cl-comment "these axioms are used to define algebraic properties"
>> ...
>> )
>> ...
>> )
>>
>> where each ... is a whole sequence of axioms (phrases). The use of  
>> enclosedname here is simply to allow the use of double quotes (as  
>> well as single quotes, c.f. earlier remark)  to enclose comment  
>> text. The comment is not intended to be interpreted as a name  
>> semantically.
>>
>>>
>>> 2)  Given last alternative in production (a), what is the purpose  
>>> of production (c) given ?
>>
>> To attach a 'real' name to a text, this name being then used to  
>> import the text elsewhere. In this production, the name is  
>> understood to refer to the text.
>>
> I'm not sure if this is an implementation detail for my parser, or  
> if its a detail for the CLIF grammar...
> I need to define a "starting rule" for my parser and production (b)  
> above seems a likely candidate.  However, this choice would preclude  
> production (c) from being a "free-standing" sentence.  That is, one  
> could not directly attach a name to a text to enable importation.  I  
> can easily define my own starting rule, but I'm wondering if this is  
> something that should be stated explicitly within the CLIF grammar.
>
>>>
>>> 3)  How do these relate to each other?
>>
>> Hardly at all.
>>
>>>
>>> 4)  Should [cltext] really be optional?
>>
>> Yes, to allow 'in-line' comments not enclosing any CL content,  
>> acting as 'headers' in the text.
>>
>> Pat
>>
>>
>>>
>>>
>>>
>>>
>>> _______________________________________________
>>> CL mailing list
>>> CL at philebus.tamu.edu
>>> http://philebus.tamu.edu/mailman/listinfo/cl
>>>
>>
>> ------------------------------------------------------------
>> IHMC                                     (850)434 8903 or (650)494  
>> 3973
>> 40 South Alcaniz St.           (850)202 4416   office
>> Pensacola                            (850)202 4440   fax
>> FL 32502                              (850)291 0667   mobile
>> phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes
>>
>>
>>
>>
>>
>
> _______________________________________________
> CL mailing list
> CL at philebus.tamu.edu
> http://philebus.tamu.edu/mailman/listinfo/cl
>

------------------------------------------------------------
IHMC                                     (850)434 8903 or (650)494 3973
40 South Alcaniz St.           (850)202 4416   office
Pensacola                            (850)202 4440   fax
FL 32502                              (850)291 0667   mobile
phayesAT-SIGNihmc.us       http://www.ihmc.us/users/phayes