Can't write wget snapshots of URL with query-string

Describe the bug

Most of Linux filesystems allow to use ? and other special characters as part of folder and filenames. But if you run this using ExFAT as the storage, it will throw the following error:

Got wget response code 3:
Cannot write to ‘www.youtube.com/watch?v=BkW1xQgrSPQ.html’ (No such file or directory).

ExFAT are common filesystems for portable USB drives, specially if there is the intention to use it with Linux, Windows and MacOS. It is not the best archival filesystem (eg: zfs), but it is the most portable filesystem.

I’m not sure how to do that yet, but it would be nice to convert special characters into safer ones when writing it down to the disk.

Steps to reproduce

1. Create a ExFAT partition, cd into it
2. Execute archivebox using a YouTube URL: echo 'www.youtube.com/watch?v=BkW1xQgrSPQ' | ./archive

Software versions

• OS: Debian
• ArchiveBox version: docker image: 296aa767078f

[pirate]

Done https://github.com/pirate/ArchiveBox/commit/4f599c0b0b07c842b1a2d0ec31f229d8fa0d6294

[bltavares]

I’ve noticed that the Dockerfile clones the latest commit on Github, instead of copying the project with their changes into the image. This could cause some confusion on why a change is not being built, as it is being cached by Docker automated build layers, and given the text don’t change it will not update the image.

I’ll send a PR soon (unless its intentional) 😃