def _today_ticks(symbol, tdate, pageNo, retry_count, pause): ct._write_console() for _ in range(retry_count): time.sleep(pause) try: html = lxml.html.parse(ct.TODAY_TICKS_URL % (ct.P_TYPE['http'], ct.DOMAINS['vsf'], ct.PAGES['t_ticks'], symbol, tdate, pageNo )) res = html.xpath('//table[@id=\"datatbl\"]/tbody/tr') if ct.PY3: sarr = [etree.tostring(node).decode('utf-8') for node in res] else: sarr = [etree.tostring(node) for node in res] sarr = ''.join(sarr) sarr = '<table>%s</table>'%sarr sarr = sarr.replace('--', '0') df = pd.read_html(StringIO(sarr), parse_dates=False)[0] df.columns = ct.TODAY_TICK_COLUMNS df['pchange'] = df['pchange'].map(lambda x : x.replace('%', '')) except Exception as e: print(e) else: return df raise IOError(ct.NETWORK_URL_ERROR_MSG)
def _newstocks(data, pageNo, retry_count, pause): for _ in range(retry_count): time.sleep(pause) ct._write_console() try: html = lxml.html.parse(rv.NEW_STOCKS_URL%(ct.P_TYPE['http'],ct.DOMAINS['vsf'], ct.PAGES['newstock'], pageNo)) res = html.xpath('//table[@id=\"NewStockTable\"]/tr') if ct.PY3: sarr = [etree.tostring(node).decode('utf-8') for node in res] else: sarr = [etree.tostring(node) for node in res] sarr = ''.join(sarr) sarr = sarr.replace('<font color="red">*</font>', '') sarr = '<table>%s</table>'%sarr df = pd.read_html(StringIO(sarr), skiprows=[0, 1])[0] df = df.drop([df.columns[idx] for idx in [1, 12, 13, 14]], axis=1) df.columns = rv.NEW_STOCKS_COLS df['code'] = df['code'].map(lambda x : str(x).zfill(6)) res = html.xpath('//table[@class=\"table2\"]/tr[1]/td[1]/a/text()') tag = '下一页' if ct.PY3 else unicode('下一页', 'utf-8') hasNext = True if tag in res else False data = data.append(df, ignore_index=True) pageNo += 1 if hasNext: data = _newstocks(data, pageNo, retry_count, pause) except Exception as ex: print(ex) else: return data
def _parsing_dayprice_json(pageNum=1): """ 处理当日行情分页数据,格式为json Parameters ------ pageNum:页码 return ------- DataFrame 当日所有股票交易数据(DataFrame) """ ct._write_console() request = Request(ct.SINA_DAY_PRICE_URL%(ct.P_TYPE['http'], ct.DOMAINS['vsf'], ct.PAGES['jv'], pageNum)) text = urlopen(request, timeout=10).read() if text == 'null': return None reg = re.compile(r'\,(.*?)\:') text = reg.sub(r',"\1":', text.decode('gbk') if ct.PY3 else text) text = text.replace('"{symbol', '{"symbol') text = text.replace('{symbol', '{"symbol"') if ct.PY3: jstr = json.dumps(text) else: jstr = json.dumps(text, encoding='GBK') js = json.loads(jstr) df = pd.DataFrame(pd.read_json(js, dtype={'code':object}), columns=ct.DAY_TRADING_COLUMNS) df = df.drop('symbol', axis=1) # df = df.ix[df.volume > 0] return df
def _get_report_data(year, quarter, pageNo, dataArr): ct._write_console() try: request = Request( ct.REPORT_URL % (ct.P_TYPE['http'], ct.DOMAINS['vsf'], ct.PAGES['fd'], year, quarter, pageNo, ct.PAGE_NUM[1])) text = urlopen(request, timeout=10).read() text = text.decode('GBK') text = text.replace('--', '') html = lxml.html.parse(StringIO(text)) res = html.xpath("//table[@class=\"list_table\"]/tr") if ct.PY3: sarr = [etree.tostring(node).decode('utf-8') for node in res] else: sarr = [etree.tostring(node) for node in res] sarr = ''.join(sarr) sarr = '<table>%s</table>' % sarr df = pd.read_html(sarr)[0] df = df.drop(11, axis=1) df.columns = ct.REPORT_COLS dataArr = dataArr.append(df, ignore_index=True) nextPage = html.xpath('//div[@class=\"pages\"]/a[last()]/@onclick') if len(nextPage) > 0: pageNo = re.findall(r'\d+', nextPage[0])[0] return _get_report_data(year, quarter, pageNo, dataArr) else: return dataArr except Exception as e: print(e)
def _inst_detail(pageNo=1, retry_count=3, pause=0.001, dataArr=pd.DataFrame()): ct._write_console() for _ in range(retry_count): time.sleep(pause) try: request = Request(rv.LHB_SINA_URL % (ct.P_TYPE['http'], ct.DOMAINS['vsf'], rv.LHB_KINDS[3], ct.PAGES['fd'], '', pageNo)) text = urlopen(request, timeout=10).read() text = text.decode('GBK') html = lxml.html.parse(StringIO(text)) res = html.xpath("//table[@id=\"dataTable\"]/tr") if ct.PY3: sarr = [etree.tostring(node).decode('utf-8') for node in res] else: sarr = [etree.tostring(node) for node in res] sarr = ''.join(sarr) sarr = '<table>%s</table>' % sarr df = pd.read_html(sarr)[0] df.columns = rv.LHB_JGMX_COLS dataArr = dataArr.append(df, ignore_index=True) nextPage = html.xpath('//div[@class=\"pages\"]/a[last()]/@onclick') if len(nextPage) > 0: pageNo = re.findall(r'\d+', nextPage[0])[0] return _inst_detail(pageNo, retry_count, pause, dataArr) else: return dataArr except Exception as e: print(e)
def _sh_mx(data, date='', start='', end='', symbol='', pageNo='', beginPage='', endPage='', retry_count=3, pause=0.001): for _ in range(retry_count): time.sleep(pause) ct._write_console() try: tail = '&pageHelp.pageNo=%s&pageHelp.beginPage=%s&pageHelp.endPage=%s'%(pageNo, beginPage, endPage) if pageNo == '': pageNo = 6 tail = '' else: pageNo += 5 beginPage = pageNo endPage = pageNo + 4 ref = rv.MAR_SH_HZ_REF_URL%(ct.P_TYPE['http'], ct.DOMAINS['sse']) clt = Client(rv.MAR_SH_MX_URL%(ct.P_TYPE['http'], ct.DOMAINS['sseq'], ct.PAGES['qmd'], _random(5), date, symbol, start, end, tail, _random()), ref=ref, cookie=rv.MAR_SH_COOKIESTR) lines = clt.gvalue() lines = lines.decode('utf-8') if ct.PY3 else lines lines = lines[19:-1] lines = json.loads(lines) pagecount = int(lines['pageHelp'].get('pageCount')) datapage = int(pagecount/5+1 if pagecount%5>0 else pagecount/5) if pagecount == 0: return data if pageNo == 6: ct._write_tips(lines['pageHelp'].get('total')) df = pd.DataFrame(lines['result'], columns=rv.MAR_SH_MX_COLS) df['opDate'] = df['opDate'].map(lambda x: '%s-%s-%s'%(x[0:4], x[4:6], x[6:8])) data = data.append(df, ignore_index=True) if beginPage < datapage*5: data = _sh_mx(data, start=start, end=end, pageNo=pageNo, beginPage=beginPage, endPage=endPage, retry_count=retry_count, pause=pause) except Exception as e: print(e) else: return data raise IOError(ct.NETWORK_URL_ERROR_MSG)
def _day_cinema(date=None, pNo=1, retry_count=3, pause=0.001): ct._write_console() for _ in range(retry_count): time.sleep(pause) try: request = Request( ct.BOXOFFICE_CBD % (ct.P_TYPE['http'], ct.DOMAINS['mbox'], ct.BOX, pNo, date)) lines = urlopen(request, timeout=10).read() if len(lines) < 15: #no data return None except Exception as e: print(e) else: js = json.loads(lines.decode('utf-8') if ct.PY3 else lines) df = pd.DataFrame(js['data1']) df = df.drop(['CinemaID'], axis=1) return df
def _sz_hz(date='', retry_count=3, pause=0.001): for _ in range(retry_count): time.sleep(pause) ct._write_console() try: request = Request(rv.MAR_SZ_HZ_URL%(ct.P_TYPE['http'], ct.DOMAINS['szse'], ct.PAGES['szsefc'], date)) lines = urlopen(request, timeout = 10).read() if len(lines) <= 200: return pd.DataFrame() df = pd.read_html(lines, skiprows=[0])[0] df.columns = rv.MAR_SZ_HZ_COLS df['opDate'] = date except Exception as e: print(e) else: return df raise IOError(ct.NETWORK_URL_ERROR_MSG)
def _holding_cotent(start, end, pageNo, retry_count, pause): for _ in range(retry_count): time.sleep(pause) if pageNo>0: ct._write_console() try: request = Request(rv.FUND_HOLDS_URL%(ct.P_TYPE['http'], ct.DOMAINS['163'], ct.PAGES['163fh'], ct.PAGES['163fh'], pageNo, start, end, _random(5))) lines = urlopen(request, timeout = 10).read() lines = lines.decode('utf-8') if ct.PY3 else lines lines = lines.replace('--', '0') lines = json.loads(lines) data = lines['list'] df = pd.DataFrame(data) df = df.drop(['CODE', 'ESYMBOL', 'EXCHANGE', 'NAME', 'RN', 'SHANGQIGUSHU', 'SHANGQISHIZHI', 'SHANGQISHULIANG'], axis=1) for col in ['GUSHU', 'GUSHUBIJIAO', 'SHIZHI', 'SCSTC27']: df[col] = df[col].astype(float) df['SCSTC27'] = df['SCSTC27']*100 df['GUSHU'] = df['GUSHU']/10000 df['GUSHUBIJIAO'] = df['GUSHUBIJIAO']/10000 df['SHIZHI'] = df['SHIZHI']/10000 df['GUSHU'] = df['GUSHU'].map(ct.FORMAT) df['GUSHUBIJIAO'] = df['GUSHUBIJIAO'].map(ct.FORMAT) df['SHIZHI'] = df['SHIZHI'].map(ct.FORMAT) df['SCSTC27'] = df['SCSTC27'].map(ct.FORMAT) df.columns = rv.FUND_HOLDS_COLS df = df[['code', 'name', 'date', 'nums', 'nlast', 'count', 'clast', 'amount', 'ratio']] except Exception as e: print(e) else: if pageNo == 0: return df, int(lines['pagecount']) else: return df raise IOError(ct.NETWORK_URL_ERROR_MSG)
def _get_detail(tag, retry_count=3, pause=0.001): for _ in range(retry_count): time.sleep(pause) try: ct._write_console() request = Request( ct.SINA_DATA_DETAIL_URL % (ct.P_TYPE['http'], ct.DOMAINS['vsf'], ct.PAGES['jv'], tag)) text = urlopen(request, timeout=10).read() text = text.decode('gbk') except _network_error_classes: pass else: reg = re.compile(r'\,(.*?)\:') text = reg.sub(r',"\1":', text) text = text.replace('"{symbol', '{"symbol') text = text.replace('{symbol', '{"symbol"') jstr = json.dumps(text) js = json.loads(jstr) df = pd.DataFrame(pd.read_json(js, dtype={'code': object}), columns=ct.THE_FIELDS) df = df[ct.FOR_CLASSIFY_B_COLS] return df
def _dist_cotent(year, pageNo, retry_count, pause): for _ in range(retry_count): time.sleep(pause) try: if pageNo > 0: ct._write_console() html = lxml.html.parse(rv.DP_163_URL%(ct.P_TYPE['http'], ct.DOMAINS['163'], ct.PAGES['163dp'], year, pageNo)) res = html.xpath('//div[@class=\"fn_rp_list\"]/table') if ct.PY3: sarr = [etree.tostring(node).decode('utf-8') for node in res] else: sarr = [etree.tostring(node) for node in res] sarr = ''.join(sarr) df = pd.read_html(sarr, skiprows=[0])[0] df = df.drop(df.columns[0], axis=1) df.columns = rv.DP_163_COLS df['divi'] = df['plan'].map(_fun_divi) df['shares'] = df['plan'].map(_fun_into) df = df.drop('plan', axis=1) df['code'] = df['code'].astype(object) df['code'] = df['code'].map(lambda x : str(x).zfill(6)) pages = [] if pageNo == 0: page = html.xpath('//div[@class=\"mod_pages\"]/a') if len(page)>1: asr = page[len(page)-2] pages = asr.xpath('text()') except Exception as e: print(e) else: if pageNo == 0: return df, pages[0] if len(pages)>0 else 0 else: return df raise IOError(ct.NETWORK_URL_ERROR_MSG)
def get_h_data(code, start=None, end=None, autype='qfq', index=False, retry_count=3, pause=0.001, drop_factor=True): ''' 获取历史复权数据 Parameters ------ code:string 股票代码 e.g. 600848 start:string 开始日期 format:YYYY-MM-DD 为空时取当前日期 end:string 结束日期 format:YYYY-MM-DD 为空时取去年今日 autype:string 复权类型,qfq-前复权 hfq-后复权 None-不复权,默认为qfq retry_count : int, 默认 3 如遇网络等问题重复执行的次数 pause : int, 默认 0 重复请求数据过程中暂停的秒数,防止请求间隔时间太短出现的问题 drop_factor : bool, 默认 True 是否移除复权因子,在分析过程中可能复权因子意义不大,但是如需要先储存到数据库之后再分析的话,有该项目会更加灵活 return ------- DataFrame date 交易日期 (index) open 开盘价 high 最高价 close 收盘价 low 最低价 volume 成交量 amount 成交金额 ''' start = du.today_last_year() if start is None else start end = du.today() if end is None else end qs = du.get_quarts(start, end) qt = qs[0] ct._write_head() data = _parse_fq_data(_get_index_url(index, code, qt), index, retry_count, pause) if len(qs)>1: for d in range(1, len(qs)): qt = qs[d] ct._write_console() df = _parse_fq_data(_get_index_url(index, code, qt), index, retry_count, pause) if df is None: # 可能df为空,退出循环 break else: data = data.append(df, ignore_index=True) if len(data) == 0 or len(data[(data.date>=start)&(data.date<=end)]) == 0: return None data = data.drop_duplicates('date') if index: data = data[(data.date>=start) & (data.date<=end)] data = data.set_index('date') data = data.sort_index(ascending=False) return data if autype == 'hfq': if drop_factor: data = data.drop('factor', axis=1) data = data[(data.date>=start) & (data.date<=end)] for label in ['open', 'high', 'close', 'low']: data[label] = data[label].map(ct.FORMAT) data[label] = data[label].astype(float) data = data.set_index('date') data = data.sort_index(ascending = False) return data else: if autype == 'qfq': if drop_factor: data = data.drop('factor', axis=1) df = _parase_fq_factor(code, start, end) df = df.drop_duplicates('date') df = df.sort('date', ascending=False) firstDate = data.head(1)['date'] frow = df[df.date == firstDate[0]] rt = get_realtime_quotes(code) if rt is None: return None if ((float(rt['high']) == 0) & (float(rt['low']) == 0)): preClose = float(rt['pre_close']) else: if du.is_holiday(du.today()): preClose = float(rt['price']) else: if (du.get_hour() > 9) & (du.get_hour() < 18): preClose = float(rt['pre_close']) else: preClose = float(rt['price']) rate = float(frow['factor']) / preClose data = data[(data.date >= start) & (data.date <= end)] for label in ['open', 'high', 'low', 'close']: data[label] = data[label] / rate data[label] = data[label].map(ct.FORMAT) data[label] = data[label].astype(float) data = data.set_index('date') data = data.sort_index(ascending = False) return data else: for label in ['open', 'high', 'close', 'low']: data[label] = data[label] / data['factor'] if drop_factor: data = data.drop('factor', axis=1) data = data[(data.date>=start) & (data.date<=end)] for label in ['open', 'high', 'close', 'low']: data[label] = data[label].map(ct.FORMAT) data = data.set_index('date') data = data.sort_index(ascending = False) data = data.astype(float) return data