[osis-core] Use of "+" in grain of a reference.

[Steve DeRose]

At 01:49 PM -0400 06/14/02, Patrick Durusau wrote:
>Todd,
>
>The general structure of the grain for character references is:
>
>grain ::= 'char:' offset length? content
>offset ::= integer
>length ::= '+' integer
>content ::= '(' [^0]+ ')'
>integer ::= [0-9]+
>
>The '+' sign is just a separator character. (Steve, didn't XPointer 
>use a comma?)

We used a comma since that's the general parameter syntax in the language.
The reason for the content and length is that the content piece need 
not be the whole thing linked -- for example, you might want to link 
to a very long or very short portion, but use a moderate size content 
string for the checking function (which is mainly what the content is 
there for -- checking and reattachment).

>
>Look at Steve's post on 9 May 2002, subject line Pointer syntax.
>
>I have discussed this with Steve and the intent is to be able to point to:
>
>1. a particular location, i.e., just the offset

We need to state explicitly what 'just the offset' means. I recommend 
that it mean the point immediately preceding the nth character -- and 
that this be defined as the same as a length of 0. No negative 
offsets or lengths allowed, and the first text character within the 
unit that the (pre-grain) reference identifies, is counted as 
character number 1.

>
>2. a particular range of characters, offset+length
>
>3. a particular range of characters, offset+lenth and then to a 
>particular word in that range (at least that is how I read it)

I was thinking of the (string) part as merely a check -- that's what 
you should find at the stated offset, and if you don't, you can issue 
a warning (and maybe offer to reattach to the nearest place that 
string *does* occur).

>
>(Don't know that I want to complicate the regex any more but I 
>suspect that the text string would be more useful if it were match 
>on this ref, then find this string and not have string as part of 
>offset and length. If I know that much detail, why do I need the 
>string? (Steve, can you help out here?)
>
>Patrick
>
>
>Todd Tillinghast wrote:
>
>>And I did not use the "+" behavior?  It seems to indicate a range or
>>characters rather than a point.  When referring to a segment of a verse
>>it would seem the only appropriate value following the "+" would be the
>>number of characters in the segment not the number of characters in the
>>first word in the segment. 
>>Have I missed the boat on the "+"? Is it to help parse the word in 
>>parentheses?  If this is the purpose
>>would it be more "third normal" to just count the characters of the word
>>in parentheses?
>>
>>Is there a reason to not remove the "+"?
>>
>>Todd
>>
>
>--
>Patrick Durusau
>Director of Research and Development
>Society of Biblical Literature
>pdurusau@emory.edu


-- 

Steve DeRose -- http://www.stg.brown.edu/~sjd
Chair, Bible Technologies Group -- http://www.bibletechnologies.net
Email: sderose@speakeasy.net
Backup email: sderose@mac.com, sjd@stg.brown.edu