btcscript contains an amount of debug logging that is very useful to
have. We have agreed that testing this isn't realy practical or indeed
really useful and thus is rather unlikely to grow test coverage any time
soon.
Turns out that there are some signatures in the bitcoin blockchain that have
trailing 0s, for example
12a1b29fd6c295075b6a66f5fd90f0126ceb1fda4f15e4b44d92518bd52a5cdf has a signature
length of 0x45 where there are 0x47 bytes following that length check (one is
hashtype and is supposed to be trimmed out prior to calling the function). We
relax the paranoid length check to permit traling data, but not to permit
buffers that are too short. Change the test to passing with a big comment
stating why this is now considered a valid case.
To be usd for validation. Most of the codepaths teste, a few tests
missing for cases needed tests in the validation codepaths too. To be
worked on in tree.
The only time we need to zero out scripts is for calcScriptHash which operates
on a deep copy anyway. This should make the tx passed to us unmodified now.
Although you can technically get at this value via the MaxPayloadLength
function on a block, it is less overhead for any consumers that need to
know the value to simply export it directly.
We were counting the number of push ops instead of the number of non
push ops. Add tests that found this (checking tha the max operations
check fires).
Use this to test the pubkey paths in checksig which return btcec errors
which we don't define. all of the other active tests know the return
code we need.
Use it to reduce code dpulication in the bip16 case.
In addition we export it so that that a user could run:
for !done && err == nil {
done, err = s.Step()
}
err = s.CheckErrorCondition()
manually instead of having to run Execute().
Since we already have the specific username and password we want to strip
from errors, use a specific search string rather than a generic regular
expression. This is quite a bit more efficient than using regular
expressions and also has the benefit of being more accurate.
Also, rather than using the added overhead of fmt to convert the error to
a string, just call Error() directly on it to get the string.
Finally, instead of just stripping it, replace it with the literal
string "<username>:<password>" to avoid any possible confusion in the
error messages where it might otherwise appear the url was being
constructed incorrectly.
ok jcv@
