Skip to content

Commit

Permalink
[KAN-63] 음식점 조회 API를 위한 네이밍 변경 (#12)
Browse files Browse the repository at this point in the history
  • Loading branch information
@sinkyoungdeok
sinkyoungdeok authored May 14, 2024
1 parent 27375b5 commit 3340610
Show file tree
Hide file tree
Showing 2 changed files with 387 additions and 365 deletions.
190 changes: 106 additions & 84 deletions csv-to-es.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -14,97 +14,119 @@

# 새 인덱스 생성 및 매핑 설정
if not es.indices.exists(index=index_name):
es.indices.create(index=index_name, body={
"settings": {
"analysis": {
"analyzer": {
"korean": {
"type": "custom",
"tokenizer": "nori_tokenizer",
"filter": ["nori_readingform"]
}
es.indices.create(index=index_name, body={
"settings": {
"analysis": {
"analyzer": {
"korean": {
"type": "custom",
"tokenizer": "nori_tokenizer",
"filter": ["nori_readingform"]
}
}
}
},
"mappings": {
"properties": {
"id": {"type": "long"},
"name": {"type": "text", "analyzer": "korean"},
"original_category": {"type": "text", "analyzer": "korean"},
"naver_review_count": {"type": "long"},
"address": {"type": "text", "analyzer": "korean"},
"naver_rating": {"type": "float"},
"number": {"type": "text"},
"image_url": {"type": "text"},
"category": {"type": "text", "analyzer": "korean"},
"discount_content": {"type": "text", "analyzer": "korean"},
"menus": {
"type": "nested",
"properties": {
"menu_name": {"type": "text", "analyzer": "korean"},
"price": {"type": "integer"},
"description": {"type": "text", "analyzer": "korean"},
"is_representative": {"type": "text"},
"image_url": {"type": "text"}
}
}
}
}
}
},
"mappings": {
"properties": {
"name": {"type": "text", "analyzer": "korean"},
"original_category": {"type": "text", "analyzer": "korean"},
"naver_review_count": {"type": "text", "analyzer": "korean"},
"address": {"type": "text", "analyzer": "korean"},
"naver_rating": {"type": "float"},
"number": {"type": "text"},
"image_url": {"type": "text"},
"category": {"type": "text", "analyzer": "korean"},
"menus": {
"type": "nested",
"properties": {
"menu_name": {"type": "text", "analyzer": "korean"},
"price": {"type": "integer"},
"description": {"type": "text", "analyzer": "korean"},
"is_representative": {"type": "text"},
"image_url": {"type": "text"}
}
}
}
}
})
})

# 데이터 인덱싱
for _, row in restaurant_df.iterrows():
menus = menu_df[menu_df['restaurant_id'] == row['id']].to_dict('records')

for menu in menus:
if pd.isna(menu['image_url']):
menu.pop('image_url') # image_url 필드가 NaN이면 제거

menu['price'] = int(menu['price'].replace(',', '')) # 가격에서 쉼표 제거 및 정수 변환
menu['is_representative'] = menu['is_representative'] == '대표' # 대표 여부를 Boolean 값으로 변환

if pd.isna(row['image_url']):
restaurant_image_url = None # NaN 값을 None으로 설정
else:
restaurant_image_url = row['image_url']

if pd.notna(row['rating']):
rating = float(row['rating'])
else:
rating = None

if pd.notna(row['number']):
number = row['number']
else:
number = None

print(row['name'], row['category'], row['review_count'], row['address'], rating, number, restaurant_image_url, menus)
response = es.index(index=index_name, id=row['name'], document={
"name": row['name'],
"original_category": row['category'],
"naver_review_count": row['review_count'],
"address": row['address'],
"naver_rating": rating,
"number": number,
"image_url": restaurant_image_url,
"category": row['custom_category'],
"menus": menus,
})
print(f"Indexed document ID: {response['_id']}, Result: {response['result']}")
menus = menu_df[menu_df['restaurant_id'] == row['id']].to_dict('records')

for menu in menus:
if pd.isna(menu['image_url']):
menu.pop('image_url') # image_url 필드가 NaN이면 제거

menu['price'] = int(menu['price'].replace(',', '')) # 가격에서 쉼표 제거 및 정수 변환
menu['is_representative'] = menu['is_representative'] == '대표' # 대표 여부를 Boolean 값으로 변환

if pd.isna(row['image_url']):
restaurant_image_url = None # NaN 값을 None으로 설정
else:
restaurant_image_url = row['image_url']

if pd.notna(row['rating']):
rating = float(row['rating'])
else:
rating = None

if pd.notna(row['number']):
number = row['number']
else:
number = None

if pd.notna(row['discount_content']):
discount_content = row['discount_content']
else:
discount_content = None

print(row['name'], row['category'], row['review_count'], row['address'], rating, number, restaurant_image_url,
menus)
data = {
"id": row['id'],
"name": row['name'],
"original_category": row['category'],
"naver_review_count": row['review_count'].replace('+', ''),
"address": row['address'],
"naver_rating": rating,
"number": number,
"image_url": restaurant_image_url,
"category": row['custom_category'],
"discount_content": discount_content,
"menus": menus,
}
if data.get("discount_content") is None:
data.pop("discount_content")
if data.get("naver_review_count") is None:
data.pop("naver_review_count")
if data.get("naver_rating") is None:
data.pop("naver_rating")
if data.get("number") is None:
data.pop("number")
if data.get("image_url") is None:
data.pop("image_url")

response = es.index(index=index_name, id=row['name'], document=data)
print(f"Indexed document ID: {response['_id']}, Result: {response['result']}")

# 앨리어스 확인 및 설정
if not es.indices.exists_alias(name="restaurant"):
# 앨리어스가 없으면 새 인덱스에 앨리어스 생성
es.indices.put_alias(index=index_name, name="restaurant")
# 앨리어스가 없으면 새 인덱스에 앨리어스 생성
es.indices.put_alias(index=index_name, name="restaurant")
else:
# 기존에 앨리어스가 있다면, 앨리어스를 새 인덱스로 업데이트하고, 기존 인덱스 삭제
old_indices = list(es.indices.get_alias(name="restaurant").keys())
es.indices.update_aliases(body={
"actions": [
{"remove": {"index": "*", "alias": "restaurant"}},
{"add": {"index": index_name, "alias": "restaurant"}}
]
})
for idx in old_indices:
if idx != index_name:
es.indices.delete(index=idx)
# 기존에 앨리어스가 있다면, 앨리어스를 새 인덱스로 업데이트하고, 기존 인덱스 삭제
old_indices = list(es.indices.get_alias(name="restaurant").keys())
es.indices.update_aliases(body={
"actions": [
{"remove": {"index": "*", "alias": "restaurant"}},
{"add": {"index": index_name, "alias": "restaurant"}}
]
})
for idx in old_indices:
if idx != index_name:
es.indices.delete(index=idx)

print("Indexing complete, and alias updated.")
Loading

0 comments on commit 3340610

Please sign in to comment.