merge.py

'''
A script to merge two tables generated by Piskvork

Kai Sun
'''

import sys

def cmp_to_key(mycmp):
    class K:
        def __init__(self, obj, *args):
            self.obj = obj
        def __lt__(self, other):
            return mycmp(self.obj, other.obj) < 0
        def __gt__(self, other):
            return mycmp(self.obj, other.obj) > 0
        def __eq__(self, other):
            return mycmp(self.obj, other.obj) == 0
        def __le__(self, other):
            return mycmp(self.obj, other.obj) <= 0
        def __ge__(self, other):
            return mycmp(self.obj, other.obj) >= 0
        def __ne__(self, other):
            return mycmp(self.obj, other.obj) != 0
    return K


def table_merge_fill(A, result):
    head = A[1]
    body = A[2]
    for i in xrange(len(body)):
        line = body[i]
        x = line[0]
        for j in xrange(1, len(line)):
            y = head[j-1]
            score = line[j]
            if '-' not in score:
                win, loss = score.replace(' ','').split(':')
                win, loss = int(win), int(loss)
                result[x][y][0] += win
                result[x][y][1] += loss


def table_merge(A, B):
    title = A[0]
    date = A[3]
    col = []
    for x in A[1] + B[1]:
        if x not in col:
            col += [x]
    row = []
    for x in A[2] + B[2]:
        if x[0] not in row:
            row += [x[0]]
    result = {}
    for x in row:
        result[x] = {}
        for y in col:
            result[x][y] = [0, 0]

    table_merge_fill(A, result)
    table_merge_fill(B, result)
    for x in result:
        for y in result[x]:
            if y not in result: continue
            if x not in result[y]: continue
            if result[x][y][0] != 0 or result[x][y][1] != 0:
                if result[y][x][0] == 0 and result[y][x][1] == 0:
                    result[y][x][0] = result[x][y][1]
                    result[y][x][1] = result[x][y][0]

    order = {}
    i = 0
    for x in A[2]:
        if x[0] not in order:
            order[x[0]] = [i,0,0,0]
            i += 1
    for x in B[2]:
        if x[0] not in order:
            order[x[0]] = [i,0,0,0]
            i += 1
    for x in A[1]:
        if x not in order:
            order[x] = [i,0,0,0]
            i += 1
    for x in B[1]:
        if x not in order:
            order[x] = [i,0,0,0]
            i += 1
            
    for x in row:
        for y in col:
            if x == y: continue
            order[y][1] += result[x][y][0]
            order[y][2] += result[x][y][1]
            if result[x][y][0] > result[x][y][1]:
                order[y][3] += 3
            elif result[x][y][0] == result[x][y][1]:
                order[y][3] += 1

    ordercol = [x for x in col]
    orderrow = [x for x in row]
    cmpf = lambda x,y: \
           order[x][3] - order[y][3] if order[x][3] != order[y][3] else \
           order[x][1] * order[y][2] - order[y][1] * order[x][2] if order[x][1] * order[y][2] != order[y][1] * order[x][2] else \
           order[x][0] - order[y][0]
    ordercol.sort(key = cmp_to_key(cmpf), reverse=True)
    orderrow.sort(key = cmp_to_key(cmpf), reverse=True)

    head = [x for x in ordercol]
    body = []
    
    for x in orderrow:
        line = []
        line += [x]
        for y in ordercol:
            if x == y and result[x][y][0] == 0 and result[x][y][1] == 0:
                line += ['-']
            else:
                line += [str(result[x][y][0]) + ' : ' + str(result[x][y][1])]
        body += [line]

    return [title, head, body, date]
        

def table_load(fn):
    f = open(fn, "r")

    title = f.readline()

    head = f.readline().replace('\n', '').replace('\r', '')
    head = head.replace('<TR><TD class="dash">-</TD>', '')
    head = head.replace('<TH>', '')
    head = head.replace('</TH></TR>', '')
    head = head.split('</TH>')

    body = []
    line = f.readline()
    while line:
        if '</HTML>' in line:
            date = line
            break
        if '<TR><TH>Total</TH>' not in line and \
           '<TR><TH>Ratio</TH>' not in line and \
           '<TR><TH>Points</TH>' not in line:
            line = line.replace('\n', '').replace('\r', '')
            line = line.replace('<TR><TH>', '')
            line = line.replace('</TH>', '</TD>')
            line = line.replace('</TD></TR>', '')
            line = line.replace('<TD class="dash">', '')
            line = line.replace('<TD class="win">', '')
            line = line.replace('<TD class="loss">', '')
            line = line.replace('<TD class="draw">', '')
            line = line.split('</TD>')
            body += [line]
        line = f.readline()

    f.close()
    return [title, head, body, date]

def table_dump(fn, C):
    f = open(fn, "w")
    title, head, body, date = C
    f.write(title)
    stat = []
    f.write('<TR><TD class="dash">-</TD>')
    for i in xrange(len(head)):
        stat += [[0,0,0]]
        f.write('<TH>' + head[i] + '</TH>')
    f.write('</TR>\n')
    for i in xrange(len(body)):
        line = body[i]
        f.write('<TR><TH>' + line[0] + '</TH>')
        for j in xrange(1, len(line)):
            score = line[j]
            if '-' in score:
                f.write('<TD class="dash">' + score + '</TD>')
            else:
                win, loss = score.replace(' ','').split(':')
                win, loss = int(win), int(loss)
                if win > loss:
                    f.write('<TD class="win">' + score + '</TD>')
                    stat[j-1][2] += 3
                elif win < loss:
                    f.write('<TD class="loss">' + score + '</TD>')
                    stat[j-1][2] += 0
                else:
                    f.write('<TD class="draw">' + score + '</TD>')
                    stat[j-1][2] += 1
                stat[j-1][0] += win
                stat[j-1][1] += loss
        f.write('</TR>\n')
    f.write('<TR><TH>Total</TH>')
    for i in xrange(len(stat)):
        score = str(stat[i][0]) + ' : ' + str(stat[i][1])
        if stat[i][0] > stat[i][1]:
            f.write('<TD class="win">' + score + '</TD>')
        elif stat[i][0] < stat[i][1]:
            f.write('<TD class="loss">' + score + '</TD>')
        else:
            f.write('<TD class="draw">' + score + '</TD>')
    f.write('</TR>\n')
    f.write('<TR><TH>Ratio</TH>')
    for i in xrange(len(stat)):
        score = '-' if int(stat[i][1]) == 0 else "%.3f" % (float(stat[i][0]) / float(stat[i][1]))
        f.write('<TD>' + score + '</TD>')
    f.write('</TR>\n')
    f.write('<TR><TH>Points</TH>')
    for i in xrange(len(stat)):
        f.write('<TD>' + str(stat[i][2]) + '</TD>')
    f.write('</TR>\n')
    f.write(date)
    f.close()

def main():
    if len(sys.argv) != 4:
        print 'Merge file A and file B, and output the result to file C'
        print '    python merge.py A B C'
        return
    A = table_load(sys.argv[1])
    B = table_load(sys.argv[2])
    C = table_merge(A, B)
    table_dump(sys.argv[3], C)

if __name__ == '__main__':
    main()