Re: xml workspace

[Dr . Jürgen Sauermann]

    Hi Bill,
    
    it looks like one of your functions (I can't tell which because I am
    not
    familiar with your functions) keeps increasing the depth of the
    string(s)
    involved.
    
    To see this, add 2 lines to utl∆split:
    
    [0]   v←delim utl∆split string;b
    [1]    ⍝ Split a string at delim.  No recursive algorithm
    [1.1] 'utl∆split INPUT (string):' ◊ 4 ⎕CR string
    [2]    b←(string∊delim)/⍳⍴string←,string
    [3]    b←(0,b),[1.1]-/(b,1+⍴string),[1.1] 1,b+1
    [4]    v←,(⊂string) utl∆sph ¨ ⊂[2]b
    [5.1] 'utl∆split OUTPUT (v):' ◊ 4 ⎕CR v
        ∇
    
    
    Then:
    
    
          t1←xml∆parseStylesheet 'body { background-color: blue; }'
    utl∆split INPUT (string):
    ┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
    ┃body { background-color: blue; }┃
    ┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
    utl∆split OUTPUT (v):
    ┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
    ┃┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓ ┏⊖┓┃
    ┃┃body { background-color: blue; ┃ ┃ ┃┃
    ┃┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛ ┗━┛┃
    ┗∊━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
    utl∆split INPUT (string):
    ┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
    ┃┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓┃
    ┃┃body { background-color: blue; ┃┃
    ┃┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛┃
    ┗∊━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
    utl∆split OUTPUT (v):
    ┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
    ┃┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓┃
    ┃┃┏→━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓┃┃
    ┃┃┃body { background-color: blue; ┃┃┃
    ┃┃┗━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛┃┃
    ┃┗∊━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛┃
    ┗∊∊━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┛
    
    
    I can't tell where this happens (line [4] maybe?), but please note
    the
    following: if the function used with the EACH operator returns a
    non-scalar
    (like the first output above), then the non-scalar will be enclosed
    automatically
    (so that  the shape of the result of EACH matches the shape of the
    argument(s)
    of EACH. This is often being overlooked.
    
    Maybe v←⊃... instead of v←,... in line 4?
    
    Best Regards,
    Jürgen

On 1/5/21 4:44 PM, Bill Daly wrote:

> Can
>       you take a look at this log file? My effort to parse a cascading
>       style sheet is failing apparently because the string passed by
>       xml∆parseStylesheet to utl∆split does not disclose the string to
>       be split.
>       
>       
>       xml∆parseStylesheet  works as long as there is more than one rule
>       to parse.
>       
>       
>       w
>       
>       
>