projects/conversational-commerce-agent/data-ingestion/cosmetics_to_retail_search.py (196 lines of code) (raw):
# Copyright 2024 Google LLC
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""
Converting Cosmetic dataset to
Google Cloud Search for Retail data format.
Kaggle dataset:
https://www.kaggle.com/datasets/shivd24coder/cosmetic-brand-products-dataset
"""
import argparse
import json
import logging
import uuid
logging.basicConfig(level=logging.INFO)
def update_attributes(source_obj) -> dict:
"""
Update site level attribute controls
Args:
source_obj: source product attribute controls
Returns:
dict: updated attribute controls
"""
target_obj = {
"attributes" : {}
}
colors = source_obj.get("product_colors", [])
if colors is not None and colors != []:
c = {
"text": [
color["colour_name"] for color in colors
if "colour_name" in color and
color["colour_name"] is not None and
color["colour_name"] != ""
]
}
if c["text"] != []:
target_obj["attributes"]["Color"] = c
target_obj["attributes"]["Color"]["searchable"] = True
target_obj["attributes"]["Color"]["indexable"] = True
if ("tag_list" in source_obj and
source_obj["tag_list"] is not None and
source_obj["tag_list"] != []):
target_obj["attributes"]["Tags"] = {
"text": source_obj.get("tag_list", [])
}
target_obj["attributes"]["Tags"]["searchable"] = True
target_obj["Tags"] = {
"text": source_obj.get("tag_list", [])
}
return target_obj["attributes"]
def convert_flipkart_to_retail_search_product(
input_file:str,
output_file:str,
project_number:str,
branch:str="1") -> str:
"""
Transforms a Flipkart JSONL file to
Google Cloud Retail Search Product Schema.
Args:
input_file: Path to the input Flipkart JSONL file.
output_file: Path to the output JSONL file.
project_number: Google Cloud Project number.
branch: Retail Search Branch Id. defaults to 1
Returns:
Path to the output JSONL file.
"""
processed_products = ""
with open(input_file, "r", encoding="utf-8") as infile:
with open(output_file, "w", encoding="utf-8") as outfile:
source_objs = json.load(infile)
for source_obj in source_objs:
try:
target_obj = {}
# Required fields
target_obj["title"] = source_obj.get(
"name", "Unknown Product"
)
if "product_link" not in source_obj:
logging.warning(
(
"[Warning]Product doed not"
"have a product url:%s"
),
target_obj["title"]
)
continue
source_obj_brand = source_obj.get("brand")
if source_obj_brand == "":
logging.warning(
"[Warning]Product doed not have a brand:%s",
target_obj["title"]
)
source_obj_brand = "Unknown"
target_obj["brands"] = [
source_obj_brand
]
if target_obj["title"] in processed_products:
continue
else:
processed_products += f"""|{target_obj["title"]}"""
subcatagory = source_obj.get("category", "")
if subcatagory is None or subcatagory == "":
subcatagory = source_obj.get("product_type")
catagory = (
f"""{source_obj.get("product_type")} >> """
f"{subcatagory}"
)
target_obj["categories"] = catagory
prod_id = uuid.uuid4()
target_obj["id"] = f"""{source_obj.get("id", prod_id)}"""
target_obj["name"] = (
f"projects/{project_number}/locations/global/catalogs/"
f"""default_catalog/branches/{branch}"""
f"""/products/{target_obj["id"]}"""
)
target_obj["primaryProductId"] = target_obj["id"]
target_obj["type"] = "PRIMARY" # Assuming all are primary
target_obj["description"] = source_obj.get(
"description", {"description": ""}
)
target_desc = target_obj.get("description", "")
if target_desc is not None and len(target_desc) >= 5000:
# Max description
target_obj["description"] = target_desc[:5000]
target_obj["languageCode"] = "en-us" # Default language
source_image = source_obj.get("image_link", None)
if source_image is not None:
target_obj["images"] = [
{"uri": source_image}
]
else:
logging.error(
"[Error]product does not have images:%s",
target_obj["title"])
continue
target_obj["uri"] = source_obj["product_link"]
# Price Information
item_price = 0
item_original_price = 0
if "prince" in source_obj:
item_price = float(source_obj.get("price", 0))
if source_obj.get("price") is not None:
item_original_price = float(
source_obj.get("price")
)
if item_price > 0 or item_original_price > 0:
target_obj["priceInfo"] = {
"currencyCode": "USD",
"price": item_price
if item_price > 0 else item_original_price,
"originalPrice": item_original_price,
"priceRange": {},
}
# Attributes
target_obj["attributes"] = update_attributes(source_obj)
# Availability
target_obj["availability"] = "IN_STOCK"
target_obj["availableQuantity"] = 0
target_obj["fulfillmentInfo"] = [
{
"type": "custom-type-1",
"placeIds": ["mobile", "www"]
}
]
target_obj["retrievableFields"] = (
"name,title,brands,uri,categories,"
"priceInfo,description,attributes.Tags"
)
# For Promotion flow
if source_obj.get("product_type", "") == "mascara":
if ("Tags" in target_obj["attributes"] and
target_obj["attributes"]["Tags"] is not None):
target_obj["attributes"]["Tags"]["text"].append(
"PromotionItem"
)
outfile.write(json.dumps(target_obj) + "\n")
# For Blend recommendation flow
if target_obj["id"] == "828":
target_obj["id"] = "99828"
target_obj["categories"] = "foundation >> blend"
target_obj["attributes"]["Tags"] = {
"text": ["Natural", "Gluten Free"],
"searchable": True
}
target_obj["primaryProductId"] = "99828"
outfile.write(json.dumps(target_obj) + "\n")
except json.JSONDecodeError as e:
logging.error("""
======
* Error decoding JSON object in line:
Exception:%s
Line:%s}
======
""", e, source_obj)
logging.info(
"Successfully transformed %s to %s",
input_file,
output_file
)
return output_file
def prepare_arguments() -> dict:
"""
Configure and parse commandline arguments.
Returns:
A Dict holds commandline arguments.
"""
parser = argparse.ArgumentParser(
description=("Converting Flipkart dataset "
"to Search for Retail data format.")
)
parser.add_argument("-i", "--input",
help="Flipkart JSON file path.",
required=True)
parser.add_argument("-o", "--output",
help="Search for Retail Jsonl file path.",
required=True)
parser.add_argument("-p", "--project-number",
help="Search for Retail Jsonl file path.",
required=True)
parser.add_argument("-b", "--branch",
help="Search for Retail Jsonl file path.",
required=True)
args = vars(parser.parse_args())
return {
"input_file": args["input"],
"output_file": args["output"],
"project_number": args["project_number"],
"branch": args["branch"]
}
if __name__ == "__main__":
params = prepare_arguments()
FLIPKART_JSON_FILE = params["input_file"]
RETAIL_SEARCH_JSON_FILE = params["output_file"]
PROJECT_NUMBER = params["project_number"]
BRANCH = params["branch"]
convert_flipkart_to_retail_search_product(
input_file=FLIPKART_JSON_FILE,
output_file=RETAIL_SEARCH_JSON_FILE,
project_number=PROJECT_NUMBER,
branch=BRANCH)