Add --no-pager flag

[RomanPro100]

In my tealdeer config I have use_pager set to true, but if there's a case I want it not to be paged, there is no easy way to do this. The only way I found is to change the use_pager option in the config to false.
So I decided it would be nice to have a --no-pager flag that disables paging whether it's enabled in the config or by the --pager flag.
I also added some tests and edited the shell completion scripts, I hope I made everything right.

[niklasmohrin]

Hey @RomanPro100,

this is definitely a good motivation to implement this flag - thank you for submitting the PR!

I would have thought that clap would be able to generate the --no- version of a flag on its own, but clap-rs/clap#815 is still open. For the time being, I think we should include some more tweaks to the interaction of --[no-]pager, in particular tldr --pager --no-pager and tldr --no-pager --pager should do what the latest argument says. I believe this is done using clap's overrides_with attribute.

A good example on how to do this cleanly is uv:

• https://github.com/astral-sh/uv/blob/44664022148e92c00a0b534c6d17aedc41ad0488/crates/uv/src/settings.rs#L190
• https://github.com/astral-sh/uv/blob/dcd96a83aa5254d6b5068d2c68b074a0376d4100/crates/uv-cli/src/lib.rs#L2525-L2542

In our case, we use flag(args.pager, args.no_pager).unwrap_or(config.display.pager), so that the command line parameters always take precedence. All of this "deciding whether or not we spawn a pager" business can be done in main instead of in print_page, because it really is more CLI parsing logic than printing logic.

The tests should cover these cases, i.e. test that CLI takes precedence and that using both --pager and --no-pager together still works as expected. This should optimally include an actual check on whether or not the pager was spawned, for example by using tee tealdeer-pager-log as the pager and then checking for the presence of the tealdeer-pager-log file (we can set the environment variable PAGER when we run the tests).

---

All of these tasks might be a bit overwhelming, in particular the testing part, so if you have any troubles or questions please reach out :)