Spaces:
Sleeping
Sleeping
Shakshi3104
commited on
Commit
•
133e909
1
Parent(s):
e319108
[add] Implement notion_db to load contents from Notion database
Browse files- model/data/notion_db.py +216 -0
model/data/notion_db.py
ADDED
@@ -0,0 +1,216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
from pathlib import Path
|
3 |
+
import abc
|
4 |
+
|
5 |
+
import pandas as pd
|
6 |
+
from dotenv import load_dotenv
|
7 |
+
|
8 |
+
import notion_client as nt
|
9 |
+
from notion2md.exporter.block import StringExporter
|
10 |
+
|
11 |
+
from loguru import logger
|
12 |
+
|
13 |
+
|
14 |
+
class BaseNotionDatabase:
|
15 |
+
"""
|
16 |
+
Notion DBからページのコンテンツを取り出すベースのクラス
|
17 |
+
"""
|
18 |
+
def __init__(self):
|
19 |
+
load_dotenv()
|
20 |
+
self.notion_database_id = os.getenv("NOTION_DATABASE_ID")
|
21 |
+
self.integration_token = os.getenv("INTEGRATION_TOKEN")
|
22 |
+
|
23 |
+
# notion2mdの環境変数
|
24 |
+
os.environ["NOTION_TOKEN"] = os.getenv("INTEGRATION_TOKEN")
|
25 |
+
|
26 |
+
self.notion_client = nt.Client(auth=self.integration_token)
|
27 |
+
|
28 |
+
def load_database(self) -> list[dict]:
|
29 |
+
"""
|
30 |
+
Notion DBのページ一覧を取得
|
31 |
+
|
32 |
+
Returns:
|
33 |
+
|
34 |
+
"""
|
35 |
+
results = []
|
36 |
+
has_more = True
|
37 |
+
start_cursor = None
|
38 |
+
|
39 |
+
while has_more:
|
40 |
+
db = self.notion_client.databases.query(
|
41 |
+
**{
|
42 |
+
"database_id": self.notion_database_id,
|
43 |
+
"start_cursor": start_cursor
|
44 |
+
}
|
45 |
+
)
|
46 |
+
# 100件までしか1回に取得できない
|
47 |
+
# 100件以上ある場合 has_more = True
|
48 |
+
has_more = db["has_more"]
|
49 |
+
# 次のカーソル
|
50 |
+
start_cursor = db["next_cursor"]
|
51 |
+
|
52 |
+
# 取得結果
|
53 |
+
results += db["results"]
|
54 |
+
|
55 |
+
return results
|
56 |
+
|
57 |
+
@abc.abstractmethod
|
58 |
+
def load_content(self) -> list[dict]:
|
59 |
+
"""
|
60 |
+
Notion DBのページの中身をdictで返す
|
61 |
+
Returns:
|
62 |
+
|
63 |
+
"""
|
64 |
+
raise NotImplementedError
|
65 |
+
|
66 |
+
|
67 |
+
class SakurapDB(BaseNotionDatabase):
|
68 |
+
def load_database(self) -> list[dict]:
|
69 |
+
"""
|
70 |
+
Notion DBのページ一覧を取得
|
71 |
+
|
72 |
+
Returns:
|
73 |
+
results:
|
74 |
+
list[dict]
|
75 |
+
|
76 |
+
"""
|
77 |
+
results = []
|
78 |
+
has_more = True
|
79 |
+
start_cursor = None
|
80 |
+
|
81 |
+
while has_more:
|
82 |
+
# "Rap詞 : 櫻井翔"がTrueのもののみ取得
|
83 |
+
db = self.notion_client.databases.query(
|
84 |
+
**{
|
85 |
+
"database_id": self.notion_database_id,
|
86 |
+
"filter": {
|
87 |
+
"property": "Rap詞 : 櫻井翔",
|
88 |
+
"checkbox": {
|
89 |
+
"equals": True
|
90 |
+
}
|
91 |
+
},
|
92 |
+
"start_cursor": start_cursor
|
93 |
+
}
|
94 |
+
)
|
95 |
+
# 100件までしか1回に取得できない
|
96 |
+
# 100件以上ある場合 has_more = True
|
97 |
+
has_more = db["has_more"]
|
98 |
+
# 次のカーソル
|
99 |
+
start_cursor = db["next_cursor"]
|
100 |
+
|
101 |
+
# 取得結果
|
102 |
+
results += db["results"]
|
103 |
+
|
104 |
+
return results
|
105 |
+
|
106 |
+
def __load_blocks(self, block_id: str) -> str:
|
107 |
+
"""
|
108 |
+
Notionのページをプレーンテキストで取得する (Notion Official API)
|
109 |
+
|
110 |
+
Parameters
|
111 |
+
----------
|
112 |
+
block_id:
|
113 |
+
str, Block ID
|
114 |
+
|
115 |
+
Returns
|
116 |
+
-------
|
117 |
+
texts:
|
118 |
+
str
|
119 |
+
"""
|
120 |
+
block = self.notion_client.blocks.children.list(
|
121 |
+
**{
|
122 |
+
"block_id": block_id
|
123 |
+
}
|
124 |
+
)
|
125 |
+
|
126 |
+
# プレーンテキストを繋げる
|
127 |
+
def join_plain_texts():
|
128 |
+
text = [blck["paragraph"]["rich_text"][0]["plain_text"] if len(blck["paragraph"]["rich_text"])
|
129 |
+
else "\n" for blck in block["results"]]
|
130 |
+
|
131 |
+
texts = "\n".join(text)
|
132 |
+
return texts
|
133 |
+
|
134 |
+
return join_plain_texts()
|
135 |
+
|
136 |
+
def load_content(self) -> list[dict]:
|
137 |
+
"""
|
138 |
+
Notion DBのページの中身をdictで返す
|
139 |
+
|
140 |
+
Returns:
|
141 |
+
lyrics:
|
142 |
+
list[dict]
|
143 |
+
"""
|
144 |
+
|
145 |
+
# DBのページ一覧を取得
|
146 |
+
db_results = self.load_database()
|
147 |
+
logger.info("🚦 [Notion] load database...")
|
148 |
+
|
149 |
+
# コンテンツ一覧
|
150 |
+
lyrics = []
|
151 |
+
|
152 |
+
logger.info("🚦 [Notion] start to load each page content ...")
|
153 |
+
# 各ページの処理
|
154 |
+
for result in db_results:
|
155 |
+
block_id = result["id"]
|
156 |
+
# rap_lyric = self.__load_blocks(block_id)
|
157 |
+
|
158 |
+
# Markdown形式でページを取得
|
159 |
+
rap_lyric = StringExporter(block_id=block_id).export()
|
160 |
+
# Markdownの修飾子を削除
|
161 |
+
rap_lyric = rap_lyric.replace("\n\n", "\n").replace("<br/>", "\n").replace("*", "")
|
162 |
+
|
163 |
+
lyrics.append(
|
164 |
+
{
|
165 |
+
"title": result["properties"]["名前"]["title"][0]["plain_text"],
|
166 |
+
"content": rap_lyric
|
167 |
+
}
|
168 |
+
)
|
169 |
+
|
170 |
+
logger.info("🚦 [Notion] Finish to load.")
|
171 |
+
|
172 |
+
return lyrics
|
173 |
+
|
174 |
+
|
175 |
+
def fetch_sakurap_corpus(filepath: str, refetch=False) -> pd.DataFrame:
|
176 |
+
"""
|
177 |
+
サクラップのコーパスを取得する
|
178 |
+
CSVファイルが存在しないときにNotionから取得する
|
179 |
+
|
180 |
+
Parameters
|
181 |
+
----------
|
182 |
+
filepath:
|
183 |
+
str
|
184 |
+
refetch:
|
185 |
+
bool
|
186 |
+
|
187 |
+
Returns
|
188 |
+
-------
|
189 |
+
|
190 |
+
"""
|
191 |
+
filepath = Path(filepath)
|
192 |
+
|
193 |
+
if not filepath.exists() or refetch:
|
194 |
+
# CSVファイルを保存するディレクトリが存在しなかったら作成する
|
195 |
+
if not filepath.parent.exists():
|
196 |
+
logger.info(f"🚦 [Notion] mkdir {str(filepath.parent)} ...")
|
197 |
+
filepath.parent.mkdir(parents=True, exist_ok=True)
|
198 |
+
|
199 |
+
logger.info("🚦 [Notion] fetch from Notion DB ...")
|
200 |
+
# dictを取得
|
201 |
+
rap_db = SakurapDB()
|
202 |
+
lyrics = rap_db.load_content()
|
203 |
+
|
204 |
+
lyrics_df = pd.DataFrame(lyrics)
|
205 |
+
lyrics_df.to_csv(filepath, index=False)
|
206 |
+
else:
|
207 |
+
logger.info("🚦 [Notion] load CSV file.")
|
208 |
+
|
209 |
+
lyrics_df = pd.read_csv(filepath)
|
210 |
+
|
211 |
+
return lyrics_df
|
212 |
+
|
213 |
+
|
214 |
+
if __name__ == "__main__":
|
215 |
+
sakurap_db = SakurapDB()
|
216 |
+
lyrics = sakurap_db.load_content()
|