Handle orm-rt cookie issue

[elrob]

• orm-rt cookie has floating point max-age (which it shouldn't https://tools.ietf.org/html/rfc6265#section-5.2.2)
• requests library fails to include it in the cookie jar (parsing the string as an int fails and the cookie is discarded: https://github.com/python/cpython/blob/20a4f6cde65549fd0252eb8c879963e0e8b40390/Lib/http/cookiejar.py#L1448)
• in-addition, the 'redirect-uri' was not the same as that used in the browser login flow which uses the 'next' parameter.
• one more change is that there is now a check after the login (or cookie load), to ensure that the profile page is accessible. This helps by failing fast if there is an authentication issue. Previously, the script would continue to process the book but just download truncated chapters. The issues people have raised about truncated chapters seem to all be caused by authentication issues.

Fixes:

• We can't download full ebook #146
• Truncated ( ... ) download still happening even after #133 fix #150

[McPatate]

Hey @elrob , I am getting an "Unexpected error" after pulling your repo. The script is unable to perform auth login.

[elrob]

Hey @elrob , I am getting an "Unexpected error" after pulling your repo. The script is unable to perform auth login.

I'm not having any issue with logging in. I've just tested again myself. How far does the script get for you?

[McPatate]

I was using the wrong password ...

But I'm still getting cropped books.

[phamhoangtuan]

I tried your code with a book 9781491983638 but the issue #150 still happened

[elrob]

@McPatate @phamhoangtuan
I have investigated further and made some more changes that resolve another couple of issues with login.
Can you try again with a fresh version of this: https://github.com/elrob/safaribooks
Make sure to delete anything in the Books directory before trying.

[lorenzodifuccia]

Thank you @elrob, you made a really good analysis of the problem.
Is the #152 up-to-date with all the fixes that you applied on your fork?

[elrob]

@lorenzodifuccia Yes, thank you. This PR is up-to-date with all the fixes mentioned in the PR description above (I just updated the description again).

[lorenzodifuccia]

Hi Rob, I pushed an update with your changes. Thanks you for what you did, nice job man!
Few point that i changed:

1. HEADERS are needed because of the user-agent and because of the accept-enconding, there was an issue related to Brotli compression algorithm (Problem with CSS file #21).
2. Renamed update_cookie_jar_with_float_max_age_cookies to handle_cookie_update to be more general and let this function be used also for future handling.
3. Personally, I think you do not really need to process the cookie through SimpleCookie, I think you can just search if there is a float num and set the cookie manually.
4. I added a try/catch block for the next_parameter parsing
5. I really like them, but I personally discard the use of the F-strings because of compatibility. String interpolation needs Py3.6+.
6. I fixed and changed some other things...

Thank you!
Cheers 🍻