test: consistently use BTC/kB and sat/B in wallet_basic test

[decryp2kanon]

I found some confused case sensitive in BTC/kB and sat/B. It doesn't change the test result.

before: 😕

bTc/kB
btc/kb
BTc/Kb

sat/B
sAT/b
SAT/b
SAT/B

after: 😄

BTC/kB
sat/B

reference:

bitcoin/src/policy/feerate.h

Lines 22 to 23 in 1769828

	BTC_KB, //!< Use explicit BTC/kB fee given in coin control
	SAT_B, //!< Use explicit sat/B fee given in coin control

test result:

$ ./test/functional/test_runner.py --jobs=4 --loglevel=DEBUG --previous-releases --coverage --extended wallet_basic.py

Temporary test directory at /tmp/test_runner_₿_🏃_20200929_224115
Running Unit Tests for Test Framework Modules
.......
----------------------------------------------------------------------
Ran 7 tests in 0.008s

OK
Initializing coverage directory at /tmp/coveragevbx7unjc
Remaining jobs: [wallet_basic.py]
1/1 - wallet_basic.py passed, Duration: 32 s                   

TEST            | STATUS    | DURATION

wallet_basic.py | ✓ Passed  | 32 s

ALL             | ✓ Passed  | 32 s (accumulated) 
Runtime: 32 s

[decryp2kanon]

i just changed the commit title as suggested by fanquake. thanks!

[sipa]

I assume this is intentional? To test that units are parsed in a case-insenstive manner?

[decryp2kanon]

I assume this is intentional? To test that units are parsed in a case-insenstive manner?

you are right. it was intentional by the author, but it doesn't work correctly IMHO. so i suggest to use consistently. 05227a3

bitcoin/src/policy/feerate.h

Lines 22 to 23 in 1769828

	BTC_KB, //!< Use explicit BTC/kB fee given in coin control
	SAT_B, //!< Use explicit sat/B fee given in coin control

what do you think? @kallewoof

[kallewoof]

It's intentional, and it should work, so if it doesn't work, then I would prefer fixing it over changing the values.

[decryp2kanon]

It's intentional, and it should work, so if it doesn't work, then I would prefer fixing it over changing the values.

thanks to confirm! however in my case, after changed its still passing. it should be failure?

[kallewoof]

Why should it be a failure? Any variant is allowed, including the ones you are using. You're just removing variants in the test cases.

[jonatack]

Agree with the other reviewers, these tests are intentional and the proposed change would reduce the coverage.

[decryp2kanon]

Why should it be a failure? Any variant is allowed, including the ones you are using. You're just removing variants in the test cases.

i understood. thanks to make it clear! PR closing

[murchandamus]

I assume this is intentional? To test that units are parsed in a case-insenstive manner?

I was just remarking on the variety in capitalization in this test for the review of #20220. I think the change proposed by @decryp2kanon here would be great, if supplemented by an explicit test for the case insensitivity of the parameter value.