root/lang/python/twopy/trunk/twopy/comment.py @ 34396

#!/usr/bin/env python
#-*- coding:utf-8 -*-

import re, datetime
import xml.sax.saxutils
import unicodedata

class Comment (object):
        """
        スレッド上のコメントを管理するクラスです.
        """
        
        __delete_tag = re.compile(r"<.+?>")
        __date_and_id = re.compile(r"(?P<date>.*) ID:(?P<id>\S*)")
        __be = re.compile(r"BE:(?P<be>.*)")
        __datetime = re.compile(r"(?P<year>\d{4})/(?P<month>\d{2})/(?P<day>\d{2})\(.*\) (?P<hour>\d{2}):(?P<min>\d{2}):(?P<sec>\d{2})(\.(?P<csec>\d+)|)")
        __urls = re.compile(r"(ttps?:\/\/[-_.!~*'()a-zA-Z0-9;/?:@&=+$,%#]+)")
        __response = re.compile(r">>(\d{1,4}|＞＞[０-９]{1,4})")
        
        def __init__(self, thread, line, number):
                """
                オブジェクトのコンストラクタです.
                
                thread : コメントが保存されている親スレッドのインスタンス
                line   : 初期化に用いられる、datファイルの一行.
                         この引数はunicode型でなければなりません.
                number : スレッドからの、コメント位置.
                """
                self.thread = thread
                self.number = number
                self.line   = line
                
                self.__datetime_cache = None
                self.__urls_cache = None
                if type(line) == unicode:
                        columns = line.split("<>")
                elif type(line) == str:
                        columns = unicode(line, "Shift-JIS", "replace").split("<>")
                else: raise TypeError("the type of the argument 'line' is not unicode or str.")
                if len(columns) < 5 : raise TypeError
                self.__name = Comment.__delete_tag.sub("", columns[0])
                self.__mailaddr = columns[1]
                raw_body = Comment.__delete_tag.sub("", columns[3].replace(" <br> ", "\n"))
                self.__body = xml.sax.saxutils.unescape(raw_body)[1:-1] # 余分なスペースの削除
                di_result = Comment.__date_and_id.search(columns[2])
                if di_result:
                        self.__date = di_result.group("date") or ""
                        self.__id   = di_result.group("id") or ""
                be_result = Comment.__be.search(columns[2])
                self.__be = None
                if be_result:   self.__be = be_result.group("be")
                
                self.__responses_cache = None
                
        def getName(self): return self.__name
        name = property(getName)
        
        def getMailAddr(self): return self.__mailaddr
        mailaddr = property(getMailAddr)
        
        def getDate(self): return self.__date
        date = property(getDate)
        
        def getBody(self): return self.__body
        body = property(getBody)
        
        def getDatetime(self):
                if self.__datetime_cache: return self.__datetime_cache
                
                result = Comment.__datetime.search(self.__date)
                if result:
                        year  = int(result.group("year"))
                        month = int(result.group("month"))
                        day   = int(result.group("day"))
                        hour  = int(result.group("hour"))
                        m     = int(result.group("min"))
                        sec   = int(result.group("sec"))
                        c     = result.group("csec")
                        csec  = c == None and 0 or int(c)
                        
                        d = datetime.datetime(year, month, day, hour, m, sec, csec*10000)
                        self.__datetime_cache = d
                        return d
                else: return None
        datetime = property(getDatetime)
        
        def getID(self): return self.__id
        ID = property(getID)
        
        def getBe(self): return self.__be
        be = property(getBe)
        
        def extractUrls(self):
                """
                コメントの内容から、URLの一覧を抽出して返します.
                """
                if self.__urls_cache: return self.__urls_cache
                result = Comment.__urls.finditer(self.body)
                l = ["".join( ("h", r.group(0)) ) for r in result]
                self.__urls_cache = l
                return l
        urls = property(extractUrls)
        
        def extractResponses(self, returnType="str"):
                """
                コメントの内容から、レスポンスの一覧を抽出して返します.
                """
                if self.__responses_cache == None:
                        result = Comment.__response.finditer(self.body)
                        l = [r.group(0) for r in result]
                        self.__responses_cache = l
                
                l = self.__responses_cache
                if returnType == "str": return l
                elif returnType == "int":
                        ld = [int(unicodedata.normalize("NFKC", m[2:])) for m in l]
                        return ld
                if returnType == "comment":
                        ld = [int(unicodedata.normalize("NFKC", m[2:])) for m in l]
                        lc = [self.thread[i] for i in ld if 0 < i <= self.thread.res]
                        return lc 
                else: raise TypeError
                        
        res = property(extractResponses)
        responses = property(extractResponses)
        
        def render(self):
                """
                取得したコメントから、整形された文章を返します.
                """
                if self.be:
                        header = u"%i 名前:%s [%s]: %s ID:%s BE:%s" % \
                                (self.number, self.name, self.mailaddr, self.date, self.ID, self.be)
                else:
                        header = u"%i 名前:%s [%s]: %s ID:%s" % \
                                (self.number, self.name, self.mailaddr, self.date, self.ID)
                return u"%s\n%s\n" % (header, self.body)
        
        def __unicode__(self): return self.render()