#!/usr/bin/env python
# -*- coding: utf-8 -*-
import codecs
import cjson
import datetime
import sys

#saldo_file="/home/markus/sblex_update/saldo/saldo.txt"
#saldo_morpho_file="/home/markus/sblex_update/saldo/saldo.lex"
saldo_file="saldo.txt"
saldo_morpho_file="saldo.lex"

(saldo,saldom,saldo_sem)=([],[],set([]))

def fathers(d):
    if d[2] == 'PRIM..1':
        return []
    return d[2].split(' ')
    
def number_of_entries():
    return len(saldo)

def number_of_senses():
    return len(set([d[0] for d in saldo_sem]))

def descriptor_count(d):
    if d[1] == 'PRIM..1':
        return 0
    fs = fathers(d)
    return 1+len(fs)

def number_of_mothers():
    return len(set([d[1] for d in saldo_sem]))

def number_of_fathers():
    return len(set([f for d in saldo_sem for f in d[2].split(' ')]))

def number_of_lemgrams():
    return len(set([d[3] for d in saldo]))

def number_of_gf():
    return len(set([d[4] for d in saldo]))

def number_of_pos():
    return len(set([d[5] for d in saldo]))

def number_of_paradigms():
    return len(set([d[6] for d in saldo]))

def number_of_wordforms():
    return len(saldom)

def pref(pos,lang):
    if lang == 'swe':
        return '<a href="http://spraakbanken.gu.se/swe/forskning/saldo/taggmangd#%s">%s</a>' % (pos,pos)
    else:
        return '<a href="http://spraakbanken.gu.se/eng/research/saldo/tagset#%s">%s</a>' % (pos,pos)

def pos_table(lang):
    pos  = [d[5] for d in saldo]
    table = sorted([(pos.count(p),p) for p in set(pos)],reverse=True)
    return "\n".join(['<table border="1"><tr>',
                      "\n".join(['<th>%s</th>' % (pref(p,lang)) for (_,p) in table[:15]]),
                      "</tr><tr>",
                      "\n".join(['<td>%s</td>' % (pr_number(n,lang)) for (n,_) in table[:15]]),
                      "</tr>",
                      '<tr>',
                      "\n".join(['<th>%s</th>' % (pref(p,lang)) for (_,p) in table[15:30]]),
                      "</tr><tr>",
                      "\n".join(['<td>%s</td>' % (pr_number(n,lang)) for (n,_) in table[15:30]]),
                      "</tr>",
                      '<tr>',
                      "\n".join(['<th>%s</th>' % (pref(p,lang)) for (_,p) in table[30:]]),
                      "%s</tr><tr>" %("".join(['<td></td>' for _ in range(14-len(table[30:]))])),
                      "\n".join(['<td>%s</td>' % (pr_number(n,lang)) for (n,_) in table[30:]]),
                      "%s</td><td>" %("".join(['<td></td>' for _ in range(14-len(table[30:]))])),
                      "</tr>",
                      '</table>'
                      ]).encode('UTF-8')

def number_of_descriptors(lang):
    count = {}
    for d in saldo_sem:
        c = descriptor_count(d)
        if c in count:
            count[c] += 1
        else:
            count[c] = 1
    res = sorted(count.items())
    return "\n".join(['<table border="1"><tr>',
                      '<th>%s</th>' % (locale['desc_count'][lang]),
                      "".join(['<th>%s</th>' % (pr_number(n,lang)) for (n,_) in res]),
                      "</tr>",
                      '<th>%s</th>' % (locale['senses'][lang]),
                      "".join(['<td>%s</td>' % (pr_number(n,lang)) for (_,n) in res]),
                      '</tr></table>'
                      ])

def pr_number(n,lang):
    n_str = str(n)
    res = ''
    if lang == 'swe':
        dot = '.'
    else:
        dot = ','
    count = 1
    for x in reversed(n_str):
        res = x + res
        if (count % 3 == 0 and count != len(n_str)):
            res = dot + res
        count += 1
    return res

def read_data():
    #count = 0
    for l in codecs.open(saldo_file, encoding='UTF-8'):
        lt = l[:-1].split('\t')
        if len(lt) == 7:
            saldo.append(lt)
            saldo_sem.add((lt[0],lt[1],lt[2]))
    #    count += 1
    #    if count > 10000:
    #        break
    for l in codecs.open(saldo_morpho_file, encoding='UTF-8'):
        try:
            saldom.append(cjson.decode(l))
        except:
            pass
    #    count += 1
    #    if count > 20000:
    #        break

locale = {
    'paradigm_list':{'swe':'paradigmlista','eng':'paradigm list'},
    'compact_paradigm_list':{'swe':'kompakt paradigmlista','eng':'compact paradigm list'},
    'paradigm_distribution':{'swe':'paradigmdistribution','eng':'paradigm distribution'},
    'entries':{'swe':'ingångar','eng':'entries'},
    'wordforms':{'swe':'ordformer','eng':'wordforms'},
    'senses':{'swe':'betydelser','eng':'senses'},
    'primaries':{'swe':'primärer','eng':'primaries'},
    'secondaries':{'swe':'sekundärer','eng':'secondaries'},
    'lemgrams':{'swe':'lemgram','eng':'lemgrams'},
    'lemmas':{'swe':'grundformer','eng':'lemmas'},
    'paradigms':{'swe':'paradigm','eng':'paradigms'},
    'word_classes':{'swe':'ordklasser','eng':'word classes'},
    'desc_count':{'swe':'antal deskriptorer','eng':'descriptor count'}
    }

def produce_report(lang):
    tdate = str(datetime.date.today())
    date_and_time = str(datetime.datetime.now()).split('.')[0]
    return "\n".join([
        "<html>",
        " <head>",
        '  <meta content="text/html; charset=utf-8" http-equiv="Content-Type"/>',
        '  <link rel="stylesheet" type="text/css" href="http://demo.spraakdata.gu.se/markus/saldo.css" />',
        "  <title>Statistik: SALDO (%s)</title>" % (date_and_time),
        " </head>",
        " <body>",
        "  <center>",
        "  <p>",
        '   <a href="http://spraakbanken.gu.se/%s/saldo"><img src="http://spraakbanken.gu.se/sites/spraakbanken.gu.se/files/img/saldo/saldo.gif" align="top" alt="SALDO" /></a>' %(lang),
        "<br/>%s" % (date_and_time),
        "  </p>",
        "  </center>",
        '<table border="1">',
        "<tr><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th><th>%s</th></tr>" % (
            locale['entries'][lang],
            locale['wordforms'][lang],
            locale['senses'][lang],
            locale['primaries'][lang],
            locale['secondaries'][lang],
            locale['lemgrams'][lang],
            locale['lemmas'][lang],
            locale['paradigms'][lang],
            locale['word_classes'][lang]),
        "<tr><td>%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td><td>%s</td></tr>" %
         (
            pr_number(number_of_entries(),lang),
            pr_number(number_of_wordforms(),lang),
            pr_number(number_of_senses(),lang),
            pr_number(number_of_mothers(),lang),
            pr_number(number_of_fathers(),lang),
            pr_number(number_of_lemgrams(),lang),
            pr_number(number_of_gf(),lang),
            pr_number(number_of_paradigms(),lang),
            pr_number(number_of_pos(),lang)),
        "</table>",
        number_of_descriptors(lang),
        '<table border="1">',
        '<tr><th colspan="4">%s</th></tr>' % (locale['paradigms'][lang]),
        "<tr>",
        " <td><a href=\"https://svn.spraakdata.gu.se/repos/sblex/pub/saldo/stat/paradigm.pdf\">%s (PDF)</a></td>" % (locale['paradigm_list'][lang]),
        " <td><a href=\"https://svn.spraakdata.gu.se/repos/sblex/pub/saldo/stat/paradigm.txt\">%s (TXT)</a></td>" % (locale['paradigm_list'][lang]),
        " <td><a href=\"https://svn.spraakdata.gu.se/repos/sblex/pub/saldo/stat/paradigm_kompakt.txt\">%s (TXT)</a></td>" % (locale['compact_paradigm_list'][lang]),
        " <td><a href=\"https://svn.spraakdata.gu.se/repos/sblex/pub/saldo/stat/paradigm_frekvens.txt\">%s (TXT)</a></td>" % (locale['paradigm_distribution'][lang]),
        "</tr>",
        "</table>",
        pos_table(lang),
        " </body>",
        "</html>"])

read_data()

if __name__ == '__main__':
    lang = sys.argv[1]
    print produce_report(lang)