Bug: If a TemporarilyBanned error happens right away, the output .json file gets deleted

Bug Report Summary

If you run the script generating a .json output once, it works great until it gets TemporarilyBanned; then it quits as expected. If you run it again immediately, the output json file gets deleted/overwritten with a 2-byte string.

Repo

1. Run the script, like python3 -m facebook_scraper --filename output.json --pages 1000 --cookies cookiefile.txt -fmt json --use-youtube-dl --source --resume-file resume_file.txt --dump ./html_dump_folder --group 123123123
2. Wait until it gets a TemporarilyBanned error.
3. See that the output.json file contains a valid, large json file.
4. Immediately re-run the script as in Step 1.
5. See that it gets the same TemporarilyBanned error almost immediately (before loading a single page successfully).
6. See that output.json is now basically empty, containing only 2 bytes of invalid json content.

What I expect to happen

output.json should not be overwritten with garbage if a TemporarilyBanned error happens right away. Instead, the script should simply exit right away without changing the ouput.json file at all.

Notes

The script works as expected (adds onto the output.json file) when 1 or more pages are successfully loaded in subsequent runs (i.e., repo Step #4).

[neon-ninja]

The resume file just contains a URL to start scraping from. There’s no dependency between the resume file and the output file - a user could input their desired URL to start from, and start collecting posts, without having any prior output file.

[DeflateAwning]

Would be nice if it handled that concatenation itself, considering it accepts a --resume-file argument (which appears to be wholly misleading).