Rp2

* refactored to_string() method to be top-level, so that it's easier to use in external projects.
* added [RP2 Ecosystem](https://github.com/eprbell/rp2/tree/main/README.md#rp2-ecosystem) paragraph to README.md
* added a few more FAQs to user frequently asked questions document
* improved CLI description
* minor fixes

* fixed [issue 6](https://github.com/eprbell/rp2/issues/6): OUT transaction fee is now a deduction as per https://www.irs.gov/publications/p544 and https://taxbit.com/cryptocurrency-tax-guide (previously it was considered part of proceeds and therefore taxed)
* restructured User FAQ document
* added several questions to User FAQ document (many of them DeFi-related)
* various edits and improvements throughout entire documentation

* add DONATE and GIFT-typed InTransactions
* changed Legend sheet to include time filter information
* updated documentation to reflect changes in this version

* added hyperlinks to rp2_full_report output:
* taxable events and acquired lots in tax sheet are now hyperlinked to their definition line in the in-out sheet. In LibreOffice, CTRL-click on them to jump to the target (on Mac Command-click)
* summary lines in the Summary sheet are now hyperlinked to the first line of the given year in the tax sheet
* modified time filter CLI options (-f and -t) to accept a full date (YYYY-MM-DD): so it's now possible to filter transactions before and after a given day
* major refactoring: renamed "from lot" and "in lot" to "acquired lot" throughout identifiers, strings, templates, documentation, etc.
* updated documentation to reflect changes in this version

* added profiler instrumentation
* added large input test and refactored ODS output test
* removed rmtree("output") from test_ods_output_diff.py: it modifies the file system, which can cause problems when running tests in parallel (currently not enabled, but that may change in the future).
* minor fixes

* fixed a bug in timestamp check: in and out lots were not allowed to have the same timestamp, but in certain scenarios it could happen (e.g. high-frequency trading)
* fixed a subtle corner-case bug in LIFO, which caused the first gain/loss pair to be incorrect: the minimal input that reproduced the bug is now captured in one of the tests (test_data2.ods, sheet B1). This ensures there will not be a regression.
* disable a link check in user faq: the link was failing the check (incorrectly) on Github actions