Sorting Smileys

And so I started to make copy-paste. As I was at the tenth entried, I though "How much is this" and scrolled down. And what to say: it is way to much. So I made a small python script to let me do this. And that script improved nearly by itself, because I wasn't happy with it after every step. Here are the steps of the process:

1. copy all with all strings
2. escape the strings (with CDDATA)
3. sort the result
4. better finding of the best suiting smiley
5. improve finding
6. and again
7. add list of already known smileys-codes (because they should not be doubled)
8. add a sections-ignoring filter

Well, anyway. Here is the finished script. Feel free to copy it as you like. I put it under the MIT. 

ignore_sections = ('[QQ]')

writer = open('output.xml', 'w')

start = '<messaging-emoticon-map>\n'

writer.write(start)

read = False

printer_dict = {}
already_found = []

def strange_sort(in_list):
    """
    This is a strange sort-algorithm. it returns a sorted list. Strange means,
    that it is meant for smileys  and does not make really sense to use
    it...
    """
    in_list.sort()
    # normally sorting works quiet good. the bad thing is, that it prefer /
    # before : and that it prefers uppercase from lowercase. To have the most
    # simple and easiest smiley as the first one, we have this algorithm


    def simple_first(first, new):
        if len(new) < len(first):
            return True
        if first < new:
            return True

        return False

    def has_eyes(item):
        if item.find(':') != -1:
            return True
        if item.find(';') != -1:
            return True
        if item.find('8') != -1:
            return True

        return False

    next_list = []
    for item in in_list:
        if len(next_list) == 0:
            # at least one item!
            next_list.append(item)
            continue

        if has_eyes(item):
            # we prefer items with eyes
            if '-' not in item:
                # if we also don't have a nose, we are really easy, so lets
                # use this one as the first one
                next_list.insert(0, item)
            else:
                # if we have a nose, we should check if the 'first' item is
                # easier than we are
                first = next_list[0]
                if not has_eyes(first):
                    next_list.insert(0, item)
                elif '-' not in first:
                    next_list.insert(1, item)
                
                else:
                    # smiley counter: both easy. use the one that is lowercase
                    if simple_first(first, item):
                        next_list.insert(0, item)
                    else:
                        next_list.insert(1, item)
        elif '-' in item:
            first = next_list[0]
            if has_eyes(first):
                # it has eyes!
                next_list.insert(1, item)
            elif '-' not in first:
                # it has no nose and no eyes, but we have a nose!
                next_list.insert(0, item)
            else:
                # smiley counter: both easy. use the one that is lowercase
                if simple_first(first, item):
                    next_list.insert(0, item)
                else:
                    next_list.insert(1, item)
        else:
            # no eyes, we did not identify the smiley. hopefully that sort was
            # doing it thing good. just place it at the end again
            next_list.append(item)

    return next_list



for line in open('theme'):
    if line.startswith('['):
        line = line.rstrip(' \n')
        print line
        if line not in ignore_sections:
            read = True
        else:
            read = False
        continue
    if not read:
        continue
    parts = line.split()
    if len(parts) < 2 or parts[0][0] in ('#', '!'):
        # this one should be ignored
        continue

    file = parts.pop(0)
    if file not in printer_dict.keys():
        printer_dict[file] = []
    
    while(len(parts) > 0):
        string = parts.pop(0)
        if string not in printer_dict[file] and string not in already_found:
            already_found.append(string)
            printer_dict[file].append(string)


for file, chars in printer_dict.iteritems():

    writer.write('        <emoticon file="%s">\n' % file)
    chars = strange_sort(chars)
    for char in chars:
        writer.write('                    <string><![CDATA[%s]]></string>\n' % char)
    writer.write('        </emoticon>\n') 

writer.write('</messaging-emoticon-map>')
print "done!"

Anyway, you can it here. And as I might think, most of you are more intrested in the output.xml, which you have to rename to emoticons.xml and put in the directory of the smileys. After that put it under ~/.kde/emoticons/. So that youn have a directory there and in that directory is the emoticons.xml file. Now Kopete should find it. If you want to have a finished packages, you might write this here as a comment!