Table of Contents
前言
2023 鐵人賽 DAY-08 Shioaji
昨天我們取得了基本資料
今天會進入到去券商那邊取得歷史資料
一個下單軟體，還是要能夠有一些資料
不然就有點盲目下單的感覺了
包裝
繼續重複著相同的事情
將官方提供的查詢歷史資料方法
包裝一層到數層不等，來符合我們的架構
最終需要變成 gRPC 的接口
傳輸的媒介應該要是 Protobuf
歷史資料有很多種類型可以選擇
今天會挑最常看到的 K 線來做演示
K 線
先看文件
看起來相當簡單
傳入開始日期、結束日期
以及股票號碼
就可以做第一層包裝，如下
def stock_kbars(self, num, date):
try:
return self.__api.kbars(
contract=self.get_contract_by_stock_num(num),
start=date,
end=date,
)
except TimeoutError:
return self.stock_kbars(num, date)可以很輕易注意到，我的 date，同時拿來當作 start, end
一次只處理一天，這蠻看人
我不太喜歡一次拿了好多天，又要再一個陣列裡面去做切割
那就可以來試跑一下
沒意外你會看到一大坨資料
當然，這不會是我們最終需要的型態
先來定義一下 proto
// HistoryKbarMessage
message HistoryKbarMessage {
int64 ts = 1;
double close = 2;
double open = 3;
double high = 4;
double low = 5;
int64 volume = 6;
string code = 7;
}編譯過程我就省略了
可以看得出來
我把券商 API 給的資料
拆散，重新整合成傳統意義上的 OHLC
美國線（英語：Open-High-Low-Close chart，OHLC chart），以豎立的線條表現股票價格的變化，可以呈現「開盤價、最高價、最低價、收盤價」，豎線呈現最高價和最低價間的價差間距，左側橫線代表開盤價，右側橫線代表收盤價，繪製上較K線簡單。另有一種美國線僅呈現「最高價、最低價、收盤價」（HLC）三項訊息。
另外，在線圖的繪製上，也有人採用單色製圖，並不影響「開盤價、最高價、最低價、收盤價」四項訊息的呈現。
引述自維基百科
至少我認為這樣是我比較好閱讀的
在包裝一次
這次加入 gRPC
service HistoryDataInterface {
rpc GetStockHistoryKbar(StockNumArrWithDate) returns (HistoryKbarResponse) {}
}下面是節錄這次比較重要的部分
class RPCHistory(history_pb2_grpc.HistoryDataInterfaceServicer):
def __init__(self, worker: Shioaji):
self.worker = worker
def GetStockHistoryKbar(self, request, _):
response = history_pb2.HistoryKbarResponse()
threads = []
for num in request.stock_num_arr:
thread = threading.Thread(
target=self.fill_stock_history_kbar_response,
args=(
num,
request.date,
response,
self.worker,
),
daemon=True,
)
threads.append(thread)
thread.start()
for thread in threads:
thread.join()
return response
def fill_stock_history_kbar_response(self, num, date, response, worker: Shioaji):
kbar = worker.stock_kbars(num, date)
total_count = len(kbar.ts)
tmp_length = [
len(kbar.Close),
len(kbar.Open),
len(kbar.High),
len(kbar.Low),
len(kbar.Volume),
]
for length in tmp_length:
if length - total_count != 0:
return
for pos in range(total_count):
response.data.append(
history_pb2.HistoryKbarMessage(
code=num,
ts=kbar.ts[pos],
close=kbar.Close[pos],
open=kbar.Open[pos],
high=kbar.High[pos],
low=kbar.Low[pos],
volume=kbar.Volume[pos],
)
)
# Run the gRPC server
def run_server():
server = grpc.server(futures.ThreadPoolExecutor(max_workers=10))
basic = Basic()
basic_pb2_grpc.add_BasicDataInterfaceServicer_to_server(basic, server)
history_pb2_grpc.add_HistoryDataInterfaceServicer_to_server(RPCHistory(basic.get_shioaji()), server)
server.add_insecure_port("[::]:50051")
server.start()
print("Server started. Listening on port 50051.")
server.wait_for_termination()
也用一個比較簡單的 Golang 來試著串串
package main
import (
"context"
"fmt"
"log"
"gop/pb"
"google.golang.org/grpc"
"google.golang.org/protobuf/types/known/emptypb"
)
const (
address = "localhost:50051"
)
func main() {
conn, err := grpc.Dial(address, grpc.WithInsecure())
if err != nil {
log.Fatalf("did not connect: %v", err)
}
defer conn.Close()
c := pb.NewBasicDataInterfaceClient(conn)
_, err = c.Login(context.Background(), &emptypb.Empty{})
if err != nil {
log.Fatalf("could not greet: %v", err)
}
h := pb.NewHistoryDataInterfaceClient(conn)
kbarArr, err := h.GetStockHistoryKbar(context.Background(), &pb.StockNumArrWithDate{
StockNumArr: []string{"2330"},
Date: "2023-09-22",
})
if err != nil {
log.Fatalf("could not greet: %v", err)
}
for _, v := range kbarArr.GetData() {
fmt.Println(v)
}
}
可以看到資料已經被整理的漂漂亮亮
而且是跨 Python, Golang 哦
ts:1695386880000000000 close:524 open:524 high:524 low:523 volume:39 code:"2330"
ts:1695386940000000000 close:523 open:524 high:524 low:523 volume:46 code:"2330"
ts:1695387000000000000 close:523 open:524 high:524 low:523 volume:36 code:"2330"
ts:1695387060000000000 close:523 open:523 high:524 low:523 volume:31 code:"2330"
ts:1695387120000000000 close:523 open:524 high:524 low:523 volume:85 code:"2330"
ts:1695387180000000000 close:523 open:523 high:524 low:523 volume:32 code:"2330"
ts:1695387240000000000 close:523 open:523 high:524 low:523 volume:27 code:"2330"
ts:1695387300000000000 close:523 open:524 high:524 low:523 volume:62 code:"2330"
ts:1695387360000000000 close:523 open:524 high:524 low:523 volume:41 code:"2330"
ts:1695387420000000000 close:524 open:523 high:524 low:523 volume:92 code:"2330"
ts:1695387480000000000 close:524 open:523 high:524 low:523 volume:47 code:"2330"
ts:1695387540000000000 close:523 open:523 high:524 low:523 volume:32 code:"2330"
ts:1695387600000000000 close:523 open:524 high:524 low:523 volume:54 code:"2330"
ts:1695387660000000000 close:524 open:524 high:524 low:523 volume:89 code:"2330"
ts:1695387720000000000 close:524 open:524 high:524 low:523 volume:43 code:"2330"
ts:1695387780000000000 close:524 open:523 high:524 low:523 volume:38 code:"2330"
ts:1695387840000000000 close:523 open:524 high:524 low:523 volume:57 code:"2330"
ts:1695387900000000000 close:524 open:524 high:524 low:523 volume:51 code:"2330"
ts:1695387960000000000 close:524 open:524 high:524 low:523 volume:102 code:"2330"
ts:1695388020000000000 close:523 open:524 high:524 low:523 volume:49 code:"2330"
ts:1695388080000000000 close:524 open:524 high:524 low:523 volume:134 code:"2330"
ts:1695388140000000000 close:524 open:524 high:524 low:523 volume:67 code:"2330"
ts:1695388200000000000 close:523 open:523 high:524 low:523 volume:74 code:"2330"
ts:1695388260000000000 close:524 open:523 high:524 low:523 volume:56 code:"2330"
ts:1695388320000000000 close:524 open:524 high:524 low:523 volume:91 code:"2330"
ts:1695388380000000000 close:523 open:524 high:524 low:523 volume:87 code:"2330"
ts:1695388440000000000 close:524 open:524 high:524 low:523 volume:51 code:"2330"
ts:1695388500000000000 close:524 open:524 high:524 low:523 volume:79 code:"2330"
ts:1695388560000000000 close:523 open:524 high:524 low:523 volume:90 code:"2330"
ts:1695388620000000000 close:524 open:524 high:524 low:523 volume:66 code:"2330"
ts:1695388680000000000 close:523 open:524 high:524 low:523 volume:127 code:"2330"
ts:1695388740000000000 close:524 open:524 high:524 low:523 volume:84 code:"2330"
ts:1695388800000000000 close:523 open:524 high:524 low:523 volume:79 code:"2330"
ts:1695388860000000000 close:524 open:524 high:524 low:523 volume:114 code:"2330"
ts:1695388920000000000 close:524 open:524 high:524 low:523 volume:90 code:"2330"
ts:1695388980000000000 close:524 open:524 high:524 low:523 volume:316 code:"2330"
ts:1695389040000000000 close:524 open:524 high:524 low:523 volume:347 code:"2330"
ts:1695389100000000000 close:524 open:524 high:524 low:523 volume:252 code:"2330"
ts:1695389400000000000 close:522 open:522 high:522 low:522 volume:6622 code:"2330"總結
這兩天進入了基本資料、歷史資料
但都還沒進入到真正的邏輯
以及介面，開始懷疑三十天是否夠用了🤣
