Eprints2archives

* Retry network operations one time if get HTTP code 400.
* Internal network code refactoring.

* Add missing `requirements.txt` dependency for [`h2`](https://pypi.org/project/h2) package.
* Make parsing of malformed id ranges slightly more robust.
* Fix incorrect pluralization of an info message.
* Remove accidentally left-in invocation of `pdb` upon errors even if debugging not enabled.
* Edit the README.md file slightly.

* In addition to the record pages, `eprints2archives` now also harvests general URLs from the server, including the top-level URL and `/view` and 2 levels of pages underneath it. However, if a subset of records is requested, only gets those particular `/view/X/N.html` pages rather than all pages under `/view/X/`.
* Internal changes allow it to use protocol HTTP/2, which was necessary to communicate with Archive.Today (because it appears to have stopped accepting save requests unless HTTP2 is used).
* Now tries to add `https://` or `http://` if the user forgets to provide it, and also removes `/eprint` and adds `/rest` if needed. This makes it possible for the user to just provide a host name and `eprints2archives` will figure out the rest.
* Minor improvements to some of the run-time status messages.
* More progress bars!
* Improvements to debug logging.
* Improvements to README.md.
* Internal code refactoring.

* Include the top-level server URL among the URLs sent to archives, as well as `/view` and two levels of pages under `/view`.
* Make sure the set of URLs sent to archives is unique.
* Improve debug logging from low-level network module.
* Clarify some things in the README file.

First working version. Supports sending EPrints pages to the [Internet Archive](https://archive.org/web/) and [Archive.Today](https://archive.today). Runs with parallel threads and handles rate limits automatically. Currently implements a command-line interface only.