CODEWITHSUNDEEP
pythonstreamlit
Mahdi Amrollahi
Mahdi Amrollahi

Reputation: 3021

Prevent another process from writing to the same file

I am using streamlit to show some results of my Python app. In this app, I need to update some files every 30 minutes. In the process of updating, if two or more users load the page, the updating files will be corrupted because they may be overwritten. So, how can I lock the process of updating files in python so that corruption never happens?

Here is part of the file that I use it to update:

config = f_getConfigs()
str_ftime = "%Y-%m-%d %H:%M:%S%Z:%z"

dt_lastup = datetime.now(timezone.utc)) 
if (dt_lastup - datetime.strptime(config["last_update"], str_ftime)).total_seconds() < 30 * 60:
    return #Use the available files
else:
    ... #Get update files from API

Upvotes: 0

Views: 263

Answers (1)

tripleee
tripleee

Reputation: 189317

I'm guessing you are really trying to ask "if I have two or more processes which run this code, how can I prevent them from trampling each other?"

One common approach is to use a lock directory:

def fetch_newer(datadir, new_data_file):
    lockdir = os.path.join(datadir, "fetch_newer")
    try:
        os.mkdir(lockdir)
    except:
        # maybe check if the directory is older than, say, 20 minutes
        # and if so, just remove it? or raise an error?
        logging.info("fetch already in progress; not starting another")
        return
    r = requests.get("https://location.example/data/" + new_data_file)
    with open(os.path.join(datadir, new_data_file), "rw") as dest:
        dest.write(r.text)
    os.rmdir(lockdir)

Upvotes: 1

Related Questions