unmungeabe.py

#!/usr/bin/python
import mysql.connector as db
import re

dbhost = 'xxx'
dbuser = 'xxx'
dbpassword = 'xxx'
releases_regex = '^(?:u4a-)?[b-df-hj-np-tv-z0-9\.\-]+$' # extra prefix on u4a posts, vowels removed
releases_groupid = 1000040

nzedb_connection = db.connect(host=dbhost, user=dbuser, password=dbpassword, database='nzedb')
nzedb_cursor = nzedb_connection.cursor()
nzedb_cursor.execute("select `id`,`searchname`,`postdate` from `releases` where `groups_id`=%s and `searchname` REGEXP %s order by `postdate` desc;", (releases_groupid, releases_regex,))
releases_matches = nzedb_cursor.fetchall()
print 'processing ' + str(len(releases_matches)) + ' releases'
for row in releases_matches:
    id = row[0]
    searchname = row[1]
    postdate = row[2]
    cleansearchname = searchname

    # irregular site names
    if re.search('^mmgs\.', cleansearchname):
        cleansearchname = re.sub('^mmgs\.', 'mmysgrl.', cleansearchname)
    elif re.search('^1kf\.', cleansearchname):
        cleansearchname = re.sub('^1kf\.', '1000fcls.', cleansearchname)
    elif re.search('^3rd\.', cleansearchname):
        cleansearchname = re.sub('^3rd\.', 'thrdmvs.', cleansearchname)
    elif re.search('^ztd\.', cleansearchname):
        cleansearchname = re.sub('^ztd\.', 'zrtlrnc.', cleansearchname)
    elif re.search('^pssn-hd\.', cleansearchname):
        cleansearchname = re.sub('^pssn-hd\.', 'pssnhd.', cleansearchname) # only so it isn't confused for a group name below

    # check for resolution
    resolution = None
    if re.search('\.(720p|hd)$', cleansearchname):
        resolution = '720p'
        cleansearchname = re.sub('\.(720p|hd)$', '', cleansearchname)
    elif re.search('\.(1080p|fllhd)$', cleansearchname):
        resolution = '1080p'
        cleansearchname = re.sub('\.(1080p|fllhd)$', '', cleansearchname)
    elif re.search('\.4k$', cleansearchname):
        resolution = '2160p'
        cleansearchname = re.sub('\.4k$', '', cleansearchname)
    else:
        resolution = '1080p'

    # ignore 2d
    if re.search('^tmwvrnt\.', cleansearchname) and re.search('\.2d$', cleansearchname):
        cleansearchname = re.sub('\.2d$', '', cleansearchname)

    # check for group prefix
    grpname = None
    grpmatch = re.search('^(gsh)\.', cleansearchname) # these guys use gush. not gush- for some reason
    if grpmatch:
        cleansearchname = re.sub('^gsh\.', 'gsh-', cleansearchname)
    if not grpmatch:
        grpmatch = re.search('^(u4a)-', cleansearchname)
    if not grpmatch:
        grpmatch = re.search('^([a-z]+)-', cleansearchname)
    if grpmatch:
        # remove group name (replaced at end later)
        cleansearchname = re.sub('^[a-z0-9]+-', '', cleansearchname)
        # normalize separators
        cleansearchname = re.sub('-', '.', cleansearchname)
        # attempt to fix disc numbers
        cleansearchname = re.sub('cd1$', 'disc1', cleansearchname)
        cleansearchname = re.sub('cd2$', 'disc2', cleansearchname)
        cleansearchname = re.sub('d1$', 'disc1', cleansearchname)
        cleansearchname = re.sub('d2$', 'disc2', cleansearchname)
        grpname = grpmatch.group(1)

    # begin pattern
    # this wildcard is notionally complete for scene releases but just harder to read
    #wc = '[a-z0-9\.\-_]*'
    wc = '.*'
    pattern = wc.join(list(cleansearchname))

    # add .resolution. (if known)
    if resolution:
        pattern += wc + resolution

    # add -group (if known)
    if grpname:
        pattern += wc + '-' + wc + wc.join(list(grpname))

    # complete pattern
    print 'searchname     : ' + searchname
    print 'postdate       : ' + str(postdate)
    print 'cleansearchname: ' + cleansearchname
    print 'resolution     : ' + (resolution if resolution else 'unknown')
    print 'release group  : ' + (grpname if grpname else 'unknown')
    print 'pattern        : ' + pattern
    nzedb_cursor.execute("select `title`,`source` from `predb` where `created` >= date_sub(%s, interval 1 day) and `created` <= date_add(%s, interval 1 day) and `title` like '%.XXX.%' and `title` REGEXP %s;", (postdate, postdate, pattern,))
    predb_matches = nzedb_cursor.fetchall()
    if predb_matches:
        if len(predb_matches) == 1:
            match = predb_matches[0]
            title = match[0]
            source = match[1]
            # if the only match was 4k but 4k wasn't detected as a resolution, change title resolution to unknown
            if resolution != '2160p' and re.search('\.2160p\.', title):
                title = re.sub('\.2160p\.', '.UNKNOWN.', title)
            #print str(id) + ':' + searchname + ':' + title
            #query = 'update `releases` set `searchname`=' + title + ' where `id`=' + str(id) + ';';
            #print query
            print 'rename         : ' + title
            print 'source         : ' + source
            nzedb_cursor.execute("update `releases` set `searchname`=%s where `id`=%s;", (title, id))
            nzedb_connection.commit()
            #break
        else:
            print 'rename         : [skipping, multiple matches]'
            for match in predb_matches:
                title = match[0]
                print '                 ' + title
    else:
        print 'rename         : [no match]'
    print ''