sbt-idp/cope2n-ai-fi/modules/sdsvkie/scripts/common/rename_labels.py

import argparse
from pathlib import Path
from tqdm import tqdm
import json 

INVOICE_MAPPING = {
    'no_key': 'No_key',    # số hóa đơn
    'no_value': 'No_value', 
    'form_key': 'Form_key',    # mẫu số hóa đơn
    'form_value': 'Form_value', 
    'serial_key': 'Serial_key',     # số kí hiệu hoá đơn
    'serial_value': 'Serial_value', 
    'date': 'Date_value',

    # seller info
    'seller_company_name_key': 'Seller_company_name_key', 
    'seller_company_name_value': 'Seller_company_name_value', 
    'seller_tax_code_key': 'Seller_tax_code_key', 
    'seller_tax_code_value': 'Seller_tax_code_value', 
    'seller_address_value': 'Seller_address_value',
    'seller_address_key': 'Seller_address_key', 
    'seller_mobile_key': 'Seller_tel_key',
    'seller_mobile_value': 'Seller_tel_value', 
    
    # buyer info
    'buyer_name_key': 'Buyer_personal_name_key',
    'buyer_name_value': 'Buyer_personal_name_value', 
    'buyer_company_name_value': 'Buyer_company_name_value', 
    'buyer_company_name_key': 'Buyer_company_name_key', 
    'buyer_tax_code_key': 'Buyer_tax_code_key', 
    'buyer_tax_code_value': 'Buyer_tax_code_value', 
    'buyer_address_key': 'Buyer_address_key', 
    'buyer_address_value': 'Buyer_address_value', 
    'buyer_mobile_key': 'Buyer_tel_key',
    'buyer_mobile_value': 'Buyer_tel_value',

    # money info
    'VAT_amount_key': 'Tax_amount_key', 
    'VAT_amount_value': 'Tax_amount_value', 
    'total_key': 'Total_key', 
    'total_value': 'Total_value', 
    'total_in_words_key': 'Total_in_words_key', 
    'total_in_words_value': 'Total_in_words_value',
    
    'other': 'Other', 
}

RECEIPT_MAPPING = {
    "Store_name_value": "seller_company_name_value",
    "Seller_company_name_value": "seller_company_name_value", 
    "id": "no_value",
    "No_value": "no_value",

    "Date_value": "date_value",
    "Total_key": "total_key",
    "Total_value": "total_value",

    "Others": "other",
    "others": "other",
    "Other": "other",
}

def write_txt(txt, data, mode="w"):
    with open(txt, mode, encoding="utf8") as f:
        for line in data:
            f.write(line + "\n")


def read_txt(txt):
    with open(txt, "r", encoding="utf8") as f:
        data = [line.strip() for line in f]
    return data



def edit_file(in_txt, out_txt, mapping):
    data = read_txt(in_txt)
    new_items = []
    not_exits_label = False 
    not_edit = True
    for item in data:
        splited_item = item.split("\t")
        label = splited_item[-1]
        if label in mapping.keys():
            new_label = mapping[label]
            splited_item[-1] = new_label
            not_edit = False
        else:
            # print(label, "not in ", mapping.keys())
            not_exits_label = True
            splited_item[-1] = label.lower()
            
        splited_item[-1] = splited_item[-1].lower()
        new_item = "\t".join(splited_item)
        new_items.append(new_item)
        
    if not_exits_label:
        print("Not exists label: ", in_txt)
    
    if not not_edit:
        print("Not edit: ", in_txt)
    write_txt(out_txt, new_items)

def rename_labels(data_dir, out_dir, doc_type):
    data_dir = Path(data_dir)
    out_dir = Path(out_dir)
    if not out_dir.exists():
        out_dir.mkdir(parents=True, exist_ok=True)
    if doc_type == "receipt":
        mapping = RECEIPT_MAPPING
    elif doc_type == 'invoice':
        mapping = INVOICE_MAPPING
    else:
        raise NotImplementedError()
    txt_paths = data_dir.rglob("*.txt")
    for txt_path in tqdm(txt_paths):
        txt_dir = str(Path(str(txt_path).replace(str(data_dir), "")).parent)  # a/b/c/x.txt -> c/x.txt -> c 
        
        if txt_dir[0] == "/":
            txt_dir = txt_dir[1:]
        out_sub_dir = out_dir / Path(txt_dir)

        if not out_sub_dir.exists():
            out_sub_dir.mkdir(parents=True, exist_ok=True)
        
        out_txt = out_sub_dir / txt_path.name
        
        # if "failure" in str(out_txt):
        #     # print(out_txt)
        #     print(out_sub_dir)
        #     print(out_txt)
        # print(out_txt)
        edit_file(str(txt_path), out_txt=out_txt, mapping=mapping)
        
def write_json(json_path, data):
    with open(json_path, "w", encoding="utf8") as f:
        json.dump(data, f, ensure_ascii=False, sort_keys=True)


def read_json(json_path):
    with open(json_path, "r", encoding="utf8") as f:
        data = json.load(f)
    return data
def rename_label_in_json(json_in, json_out, doc_type):
    if doc_type == "invoice":
        mapping = INVOICE_MAPPING
    else:
        mapping = RECEIPT_MAPPING
    ori_data = read_json(json_in)
    new_data = {}
    for img_key, field_item in ori_data.items():
        new_field_item = {}
        for field_key, field_value in field_item.items():
            if field_key in mapping:
                new_field_key = mapping[field_key]
            else:
                new_field_key = field_key
            new_field_key = new_field_key.lower()
            new_field_item[new_field_key] = field_value
            
        new_data[img_key] = new_field_item
        
    write_json(json_out, new_data)
                
            
    


if __name__ == "__main__":
    parser = argparse.ArgumentParser(prog="Rename labels")
    parser.add_argument("--in_dir", type=str, required=True, help="dataset directory")
    parser.add_argument("--out_dir", type=str, required=False, help="output")
    parser.add_argument("--doc_type", type=str, required=True, help="document type: receipt / invoice")
    
    args = parser.parse_args()
    rename_labels(
        data_dir=args.in_dir,
        out_dir=args.out_dir,
        doc_type=args.doc_type
    )

    # rename_label_in_json(
    #     json_in=args.in_dir,
    #     json_out=args.out_dir,
    #     doc_type=args.doc_type
    # )
    
    
"""

"""