perf: replace O(n) name/varname list scan with O(1) dict lookup

[P403n1x87]

_ConvertBytecodeToConcrete.add() used list.index() to find whether a name or varname was already registered, making each lookup O(n) in the number of names accumulated so far. For functions with many locals or names this becomes a non-trivial cost on every HAS_LOCAL / HAS_NAME instruction.

Replace the generic static add() method with two instance methods, add_varname() and add_name(), each backed by a parallel dict (varnames_map / names_map). Lookups are now O(1). The argnames pre-population loop is also updated to go through add_varname() so the map stays in sync.

Benchmark (round-trip test):

	r/s
Baseline	204.1 +/- 4.2
This change	221.3 +/- 2.9

So a +8.4% improvement overall. We also move some more instruction validation checks from iteration to insertion, which also contributes to the total uplift (~ +3% of the +8.4% total).

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 95.39%. Comparing base (8a249bb) to head (a563e70).

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #199      +/-   ##
==========================================
+ Coverage   95.24%   95.39%   +0.14%     
==========================================
  Files           7        7              
  Lines        2083     2107      +24     
  Branches      449      456       +7     
==========================================
+ Hits         1984     2010      +26     
+ Misses         55       54       -1     
+ Partials       44       43       -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[MatthieuDartiailh]

This makes sense I think even though it is a breaking change for CFG. Could you add a changelog entry so that I do not forget to document when doing the next release ?

[P403n1x87]

it is a breaking change for CFG.

In what sense? The only change in CFG is the time when validation happens, but no API has been changed (_ConvertBytecodeToConcrete is to be considered private given the _ prefix I guess). Is this what you would like to have documented?

[MatthieuDartiailh]

The fact the validation of the content now occurs earlier could break some usages (extend node and then break it into sub nodes). It is not a large breakage but worth documenting.

[MatthieuDartiailh]

Looking at codecov report we are actually missing tests for the error paths of append and insert. Would you mind adding those ?

[MatthieuDartiailh]

I promise those are the last comments. Recently I only manage to review from my phone which makes me miss the broader context quite often sorry for this.

[P403n1x87]

I promise those are the last comments.

No worries! Changes made.