Re: [Gnumed-devel] Better-structured narrative output

[Busser, Jim]

On 2011-12-01, at 11:23 AM, Karsten Hilbert wrote:

> - PostgreSQL should not (and does not) play a role in
>  presentation of data

No, but it might provide a built in function (or might accept the insertion / 
creation of a function) that could compare two strings and resolve

        for strings X, Y

where X is new and Y is old, a merge or an array of values that represent the 
diffing between X and Y

        start, length, string, mode 

where, beginning with position 1 in string X, you would get

        1, 27, X, match <--- X can only (either) match Y or inform an insertion
        28, 7, X, insertion  <--- this would identify an X insertion into the 
original
        28, 10, Y, deletion <--- Y is needed in order to identify anything that 
was 'deleted'
                                                from this, we can know that 
position 38 will (at some
                                                point in X) again match X
        35, 53, X, match <--- obligate match at X 35 (28 + 7 = 35)
                                                else the insertion must have 
been longer than 7
        91, 16, Y, deletion <--- Y28 + 10 (past the above deletion) identifies 
at Y38
                                                where X and Y would have again 
matched and given that
                                                they then matched for length 53 
then it is at Y91 that there
                                                is a difference which may or 
may not reflect a Y-deletion
                                                (the alternative could be X 
insertion with or without Y-deletion)

> - presentation of diffs is a known-hard problem
>  user interface wise

The simplest form might then be

        This is the part of the string that remains the same [+ here showing a 
new value or words] [- and thusly marking the part that was replaced] following 
which to display the next, preserved segment of  the original string [- but 
showing thusly this additionally-deleted bit] followed by more that remains 
true to the original [+ and supplemented by this additional insertion].

??

-- Jim