Fix escaped character handling in regex

[msm-code]

I've tried to fix issue #186.

This was suspiciously easy (especially compared to my previous PR: #93), so please check that I don't do something obviously wrong. FWIW this fixes the issue at hand. Using the POC from the linked issue:

import yaramod

y = yaramod.Yaramod(yaramod.Features.AllCurrent)
yara_file = y.parse_string(r'''
rule regex_test {
    strings:
        $a = /1\x32?3/
    condition:
        $a
}''')
for rule in yara_file.rules:
    for string in rule.strings:
        for unit in string.unit.units:
            print(type(unit), unit.text)

Previous results:

<class 'yaramod.RegexpText'> 1
<class 'yaramod.RegexpText'> \x
<class 'yaramod.RegexpText'> 3
<class 'yaramod.RegexpOptional'> 2?
<class 'yaramod.RegexpText'> 3

Current (correct) results:

<class 'yaramod.RegexpText'> 1
<class 'yaramod.RegexpOptional'> \x32?
<class 'yaramod.RegexpText'> 3

I've also added a test.

Fixes #186.

[msm-code]

Matches hex escaped characters, like "\x61", as a single token.

[nazywam]

Should we care about other encoding methods as well? It seems like \65? is a valid regex as well, not sure about the YARA side though 🤔

[msm-code]

I think yara doesn't support that:

rule kot {
    strings: $a = /kot\61kot/
    condition: $a
}

❯ yara -rs kot.yar /tmp/a.txt
kot.yar(2): error: invalid regular expression "$a": backreferences are not allowed

But sure, we should check for any other alternative encodings (if they exist). By the way I've checked, and \X61 is not a valid encoding (x must be lowercase).

[TadeasKucera]

Thank you very much for this very nicely written PR. It is good, I will merge it.