import MySQLdb
from ConfigParser import ConfigParser
import os, re, time

WEBDIR = '/home/mark/web/diveintomark.org/'
BASEDIR = os.path.join(WEBDIR, 'archives', 'citations')
TEMPLATEDIR = '/home/mark/templates/citations/'
CITATION_TEMPLATE = os.path.join(TEMPLATEDIR, 'citation')
CITATION_INDEX = os.path.join(TEMPLATEDIR, 'citationindex')
MINIMUM_NUMBER_OF_CITATIONS = 2
config = ConfigParser()
config.read([os.path.expanduser('~/db.conf')])
connectionparams = {}
for o in config.options('weblog'):
    connectionparams[o] = config.get('weblog', o)
dbhandle = apply(MySQLdb.connect, (), connectionparams)
cursor = dbhandle.cursor()

def read(filename):
    fsock = open(filename, 'rb')
    data = fsock.read()
    fsock.close()
    return data

def write(filename, data):
    fsock = open(filename, 'wb')
    fsock.write(data)
    fsock.close()
    
def stripTags(htmlSource):
  """strip all HTML tags"""
  return "".join([e.split(">", 1).pop() for e in htmlSource.split("<")]).strip()

def q(sql):
    cursor.execute(sql)
    def car(t): return t[0]
    columnNames = map(car, cursor.description)
    rows = []
    for rowdata in cursor.fetchall():
        thisrow = {}
        for name, value in map(None, columnNames, rowdata):
            thisrow[name] = value
        rows.append(thisrow)
    return rows

def x(sql):
    cursor.execute(sql)

def getCitationPosts():
    return q("""select entry_id, entry_title, entry_text, entry_created_on, year(entry_created_on) as entry_year, lpad(concat(month(entry_created_on), ''), 2, '0') as entry_month, lpad(concat(dayofmonth(entry_created_on), ''), 2, '0') as entry_day from mt_entry where entry_blog_id = 3 and entry_status = 2 and entry_text regexp '<cite( title=".*")?>'""")

dirmap = (('%', ''),
          ('$', ''),
          (')', ''),
          ('.', ''),
          (';', ''),
          ('=', ''),
          ('(', ''),
          ('-', ''),
          ('/', ''),
          (':', ''),
          ('+', ''),
          ('#', ''),
          ('!', ''),
          ('?', ''),
          (',', ''),
          ("'", ''),
          (' ', '_'),
          ('__', '_'))
def dirify(s):
    s = s.lower()
    for source, target in dirmap:
        s = s.replace(source, target)
    return s

def getPostURL(year, month, day, title):
    return 'http://diveintomark.org/archives/%s/%s/%s/%s' % (year, month, day, dirify(title))

def getAllPostURLs():
    return [getPostURL(row['entry_year'], row['entry_month'], row['entry_day'], row['entry_title']) for row in q("""select year(entry_created_on) as entry_year, lpad(concat(month(entry_created_on), ''), 2, '0') as entry_month, lpad(concat(dayofmonth(entry_created_on), ''), 2, '0') as entry_day, entry_title from mt_entry where entry_blog_id = 3 and entry_status = 2""")]

citeMatch = re.compile(r'<cite( title="(?P<title>.*?)")?>(?P<name>.*?)</cite>', re.DOTALL)
def aggregateCitations(rows):
    d = {}
    for row in rows:
        row['entry_url'] = getPostURL(row['entry_year'], row['entry_month'], row['entry_day'], row['entry_title'])
        entry_text = row['entry_text']
        del row['entry_text']
        for unused1, unused2, name in citeMatch.findall(entry_text):
            name = stripTags(name)
            if name:
                if not d.has_key(name):
                    d[name] = []
                entry_date = row['entry_created_on'].strftime('%Y%m%dT%H%M%S')
                if not [onedate for onedate, onerow in d[name] if onedate==entry_date]:
                    d[name].append((entry_date, row))
    items = []
    for k, v in d.items():
        if len(v) >= MINIMUM_NUMBER_OF_CITATIONS:
            v.sort()
            v.reverse()
            items.append((len(v), k, v))
    items.sort()
    items.reverse()
    return items

def outputCitation(name, cites):
    fullpath = os.path.join(BASEDIR, dirify(name))
    if not os.path.exists(fullpath):
        os.makedirs(fullpath)
    fullpath = os.path.join(fullpath, 'index')
    citelist = "\n".join(["""<li><p>[%s] <a href="%s">%s</a></p></li>""" % (params['entry_created_on'].strftime('%m/%d/%Y'), params['entry_url'], params['entry_title']) for dummy, params in cites])
    write(fullpath, read(CITATION_TEMPLATE) % vars())
        
def output(agg):
    for c, name, cites in agg:
        outputCitation(name, cites)

def outputIndex(agg):
    fullpath = os.path.join(BASEDIR, 'index')
    citelist = "\n".join(["""<li><a href="%s/">%s (%s)</a></li>""" % (dirify(name), name, c) for c, name, cites in agg])
    write(fullpath, read(CITATION_INDEX) % vars())

def zapIndividualCites():
    all = getAllPostURLs()
    for url in all:
        fullpath = url.replace('http://diveintomark.org/', WEBDIR)
        olddata = read(fullpath)
        newdata = olddata.split('<!-- begin citations -->', 1)[0] + '<!-- begin citations -->\r\n<!-- end citations -->' + olddata.split('<!-- end citations -->', 1)[1]
        if olddata != newdata:
            write(fullpath, newdata)
            
def outputIndividual(agg):
    zapIndividualCites()
    byurl = {}
    for c, name, cites in agg:
        for dummy, params in cites:
            url = params['entry_url']
            if not byurl.has_key(url):
                byurl[url] = []
            byurl[url].append(name)
    for url, names in byurl.items():
        names.sort()
        output = []
        for n in names:
            output.append(r'''<li><a href="/">Home</a> &rarr; <a href="/archives/">Archives</a> &rarr; <a href="/archives/citations/">Citations</a> &rarr; <a href="/archives/citations/%s/">%s</a></li>''' % (dirify(n), n))
        fullpath = url.replace('http://diveintomark.org/', WEBDIR)
        olddata = read(fullpath)
        newdata = olddata.split('<!-- begin citations -->', 1)[0] + '<!-- begin citations -->' + '\n'.join(output) + '<!-- end citations -->' + olddata.split('<!-- end citations -->', 1)[1]
        if olddata != newdata:
            write(fullpath, newdata)
    
def main():
    posts = getCitationPosts()
    agg = aggregateCitations(posts)
    output(agg)
    outputIndex(agg)
    outputIndividual(agg)

if __name__ == '__main__':
    main()