From: ldo@nz.invalid   
      
   On Fri, 16 Jan 2026 21:14:33 -0000 (UTC), Lawrence D’Oliveiro wrote:   
      
   >> On Fri, 16 Jan 2026 11:31:29 -0500, Evelyn C. Leeper wrote:   
   >>   
   >>> In 2010, the Academia Real Española declared that 'ch' and 'll'   
   >>> were no longer letters in their own right, but digraphs (like 'ph'   
   >>> in English). As such words with 'ch' would be alphabetized after   
   >>> 'cg' and before 'ci', and those with 'll' would have that between   
   >>> 'lk' and 'lm'.   
   >>>   
   >>> I personally think this was because computers could not handle   
   >>> them as single letters, and sort algorithms in particular would   
   >>> just break.   
   >>   
   >> It’s not the computers at fault, it’s the programmers who don’t   
   >> know to write text-processing code that caters for different   
   >> languages.   
   >   
   > Not the programmers fault. It's a problem using ASCII (or EBCDIC)   
   > rather than unicode.   
      
   It’s the programmer’s fault. As proof, here is some code that   
   implements the old rule. (Sorry I don’t know any actual Spanish words,   
   so I just made up some dummy text using those character sequences.)   
      
       def custom_sort_key(s) :   
           "generates a sort key for the string s, by splitting it into" \   
           " segments at instances of the character sequences to be treated" \   
           " specially."   
           key = []   
           while True :   
               positions = list \   
                 (   
                   (c1, c2, pos)   
                   for c1, c2 in (("c", "h"), ("l", "l"))   
                   for pos in (s.find(c1),)   
                   if pos >= 0   
                 )   
               if len(positions) == 0 :   
                   key.append(s)   
                   break   
               #end if   
               positions.sort(key = lambda e : e[1])   
               c1, c2, pos = positions[0]   
                 # whichever occurs earliest   
               if pos > 0 :   
                   key.append(s[:pos])   
               #end if   
               s = s[pos + 1:]   
               if len(s) != 0 and s[0] == c2 :   
                   key.append(c1 + chr(0))   
                     # ensure it sorts before other instances of c1   
                   s = s[1:]   
               else :   
                   key.append(c1 + chr(127))   
                     # ensure it sorts after other instances of c1   
               #end if   
               if len(s) == 0 :   
                   break   
           #end while   
           return key   
       #end custom_sort_key   
      
       text = ["xxlkxx", "xxcgxx", "xxllxx", "xxchxx", "xxlmxx", "xxcixx"]   
       print("input", text)   
       for t in text :   
           print("sort %s => %s" % (repr(t), repr(custom_sort_key(t))))   
       #end for   
       text.sort(key = custom_sort_key)   
       print("sorted", text)   
      
   Output:   
      
       input ['xxlkxx', 'xxcgxx', 'xxllxx', 'xxchxx', 'xxlmxx', 'xxcixx']   
       sort 'xxlkxx' => ['xx', 'l\x7f', 'kxx']   
       sort 'xxcgxx' => ['xx', 'c\x7f', 'gxx']   
       sort 'xxllxx' => ['xx', 'l\x00', 'xx']   
       sort 'xxchxx' => ['xx', 'c\x00', 'xx']   
       sort 'xxlmxx' => ['xx', 'l\x7f', 'mxx']   
       sort 'xxcixx' => ['xx', 'c\x7f', 'ixx']   
       sorted ['xxchxx', 'xxcgxx', 'xxcixx', 'xxllxx', 'xxlkxx', 'xxlmxx']   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)