test(#3807): Add disabled test for XMIR bytes with anonymous abstract object

[volodya-lombrozo]

This pull request introduces a new test case in EoSyntaxTest to verify the correct compilation of bytes to XMIR with an anonymous abstract object. The test is currently disabled due to an existing issue with the XMIR format, which is missing an anonymous abstract object for bytes.

Related to #3807

[github-actions]

🚀 Performance Analysis

Test	Base Score	PR Score	Change	% Change	Unit	Mode
benchmarks.XmirBench.xmirToEO	144.689	177.312	32.622	22.55%	ms/op	Average Time
benchmarks.XmirBench.xmirToPhi	151.999	190.057	38.058	25.04%	ms/op	Average Time
benchmarks.XmirBench.xmirToSaltyPhi	148.362	126.106	-22.256	-15.00%	ms/op	Average Time

⚠️ Performance loss: benchmarks.XmirBench.xmirToEO is slower by 32.622 ms/op (22.55%)
⚠️ Performance loss: benchmarks.XmirBench.xmirToPhi is slower by 38.058 ms/op (25.04%)
✅ Performance gain: benchmarks.XmirBench.xmirToSaltyPhi is faster by 22.256 ms/op (15.00%)

[volodya-lombrozo]

@maxonfjvipon Is this what we want to achieve in the end?

[maxonfjvipon]

@volodya-lombrozo not really.
For bytes it should be:

<o base="Q.org.eolang.bytes">
  <o>00-...-00</o>
</o>

For numbers and strings:

<o base="Q.org.eolang.string">
  <o base="Q.org.eolang.bytes">
    <o>00-...-00</o>
  </o>
</o>

[maxonfjvipon]

@volodya-lombrozo but maybe yes, you're right about

bytes > app
  2A-

It should be:

<o base="Q.org.eolang.bytes" name="app">
  <o base="Q.org.eolang.bytes">
    <o>2A-</o>
  </o>
</o>

In general we should add extra object between Q.org.eolang.bytes and data

[volodya-lombrozo]

@maxonfjvipon I've moved the test from Java to eo packs. Now they look a bit prettier. Also I've added tests for number and string. What do you think?

[maxonfjvipon]

@volodya-lombrozo LGTM! Thanks

[volodya-lombrozo]

@yegor256 Could you merge that changes, please?

[yegor256]

@volodya-lombrozo once we "resolve" all conversations above, GitHub will merge this one automatically

[volodya-lombrozo]

@yegor256 It's a bit risky to merge after all the conversations are resolved because I can resolve them myself.