DenyTranDFW/edgar_xbrl_companyfacts

SCRIPT

from IPython.display import clear_output, display, HTML

import os, time, shutil, sys, json, io, zipfile, requests, shutil, re
import pandas as pd
from os import makedirs as mk, remove as rm, getcwd as cwd, listdir as ls
from os.path import join as osj, isdir as osd, isfile as osf, basename as osb

url = 'https://www.sec.gov/Archives/edgar/daily-index/xbrl/companyfacts.zip'

headers = { 'User-Agent': f'{name} {email}',
'Accept-Encoding': 'text/plain',
'Host': 'www.sec.gov'
}

response = requests.get(url, headers=headers)

if response.status_code == 200:
zip_file = zipfile.ZipFile(io.BytesIO(response.content))
zip_file.extractall()
else: print(f"STATUS CODE: {response.status_code}")

def process_json_file(json_data, api, token, repo_id, work):
cik = json_data.split('.')[0]

with open(json_data, 'r') as file:  
    data = json.load(file)  

data_keys = data.keys()  

if not data or not 'facts' in data_keys:  
    return  
    
fact_keys = data['facts'].keys()  

entity_name = data['entityName']  

cik_folder = re.sub(r'[^a-zA-Z0-9]', '_', osj(f'{cik}_{entity_name}')).replace('__', '_').strip('_')  

mk(cik_folder, exist_ok=True)  

fact_keys = data['facts'].keys()  

def process_key(key, units_data):  
    if key in data['facts']['dei']:  
        label = data['facts']['dei'][key]['label']  
        desc = data['facts']['dei'][key]['description']  
        data_list = []  
        for unit_key, unit_values in units_data.items():  
            for unit in unit_values:  
                data_list.append({**unit, 'label': label, 'description': desc, 'unit': unit_key})  
        return pd.DataFrame(data_list)  

df_dei_EntityCommonStockSharesOutstanding = None  
df_dei_EntityListingParValuePerShare = None  
df_dei_EntityPublicFloat = None  

if 'dei' in fact_keys:  
    dei_items = data['facts']['dei']  
    
    if 'EntityCommonStockSharesOutstanding' in dei_items:  
        df_dei_EntityCommonStockSharesOutstanding = process_key('EntityCommonStockSharesOutstanding', dei_items['EntityCommonStockSharesOutstanding']['units'])  
    
    if 'EntityListingParValuePerShare' in dei_items:  
        df_dei_EntityListingParValuePerShare = process_key('EntityListingParValuePerShare', dei_items['EntityListingParValuePerShare']['units'])  

    if 'EntityPublicFloat' in dei_items:  
        df_dei_EntityPublicFloat = process_key('EntityPublicFloat', dei_items['EntityPublicFloat']['units'])  

    for df, name in [  
        [df_dei_EntityCommonStockSharesOutstanding, 'EntityCommonStockSharesOutstanding'],   
        [df_dei_EntityListingParValuePerShare, 'EntityListingParValuePerShare'],   
        [df_dei_EntityPublicFloat, 'EntityPublicFloat']  
    ]:
        if df is not None:  
            filename = osj(cik_folder, name) + '.parquet'  
            df.to_parquet(filename, index=False)  

def process_fact_section(section_key, section_data):  
    fact_list = []
    for k, v in section_data.items():  
        base_item = {'item': k, 'label': v['label'], 'description': v['description']}  
        for unit_type, unit_values in v['units'].items():  
            for unit in unit_values:  
                fact_list.append({**unit, **base_item, 'unit_type': unit_type})  
    return pd.DataFrame(fact_list) if fact_list else None  

df_facts_invest = process_fact_section('invest', data['facts'].get('invest', {}))  
if df_facts_invest is not None:  
    filename = osj(cik_folder, 'Facts_Invest.parquet')  
    df_facts_invest.to_parquet(filename, index=False)  

df_facts_srt = process_fact_section('srt', data['facts'].get('srt', {}))  
if df_facts_srt is not None:  
    filename = osj(cik_folder, 'Facts_Srt.parquet')  
    df_facts_srt.to_parquet(filename, index=False)  

df_facts_usgaap = process_fact_section('us-gaap', data['facts'].get('us-gaap', {}))  
if df_facts_usgaap is not None:  
    filename = osj(cik_folder, 'Facts_UsGaap.parquet')  
    df_facts_usgaap.to_parquet(filename, index=False)  

df_facts_ifrs = process_fact_section('ifrs-full', data['facts'].get('ifrs-full', {}))  
if df_facts_ifrs is not None:  
    filename = osj(cik_folder, 'Facts_IfrsFull.parquet')  
    df_facts_ifrs.to_parquet(filename, index=False)  


rm(json_data)  

work = cwd()
json_files = [i for i in ls(work) if i.endswith('.json')]

for json_data in json_files:
process_json_file(json_data, api, token, repo_id, work)
clear_output()