Decompiler Regressions in f-strings, Ternaries, and Loops

[sahebbiswas]

Ternary/Compound Condition Mangling (Issues 1 & 2): The "mini-decompiler" (_eval_ternary_branch) used for speculative evaluation of ternary branches and compound conditions is missing support for f-string opcodes (FORMAT_VALUE, FORMAT_SIMPLE, BUILD_STRING). This causes it to silently drop literal string fragments and expressions, resulting in broken assignments like expr = ret_expr if params else ret_expr.
Mangled F-strings (Issue 3): The current _op_build_string implementation incorrectly processes literal string fragments that are already quoted on the stack, leading to nested quote corruption and missing characters.
Integer Dict Keys (Issue 4): A bug in _op_build_const_key_map unconditionally strips quotes from constant keys, causing integer keys to be treated as strings and then re-quoted incorrectly.
Corrupted Loop Unpacking (Issue 5): The peephole logic in _op_for_iter is unaware of combined 3.14 opcodes like STORE_FAST_STORE_FAST. This causes it to fall back to an incomplete _item unpacking strategy that fails to emit the necessary assignment instructions

def ternary_fstring_repro(params, ret_expr):
    # Issue 1: Ternary with f-string
    expr = f"lambda {params}: {ret_expr}" if params else f"lambda: {ret_expr}"
    return expr

def compound_cond_fstring_repro(raw_expr, is_or_jump):
    # Issue 2: Compound condition with f-string
    cond_str = f"{raw_expr} is None" if is_or_jump else f"{raw_expr} is not None"
    return cond_str

def complex_fstring_repro(expr, args):
    # Issue 3: Complex f-string
    return f"({expr})({', '.join(str(a) for a in args)})"

def dict_int_keys_repro():
    # Issue 4: Dict with int keys
    # Note: Modern Python often uses BUILD_CONST_KEY_MAP for this
    _AUG_ASSIGN_MAP = {
        13: "+=",  14: "&=",  15: "//=", 16: "<<=", 17: "@=",
        18: "*=",  19: "%=",  20: "|=",  21: "**=", 22: ">>=",
        23: "-=",  24: "/=",  25: "^=",
    }
    return _AUG_ASSIGN_MAP

def loop_unpacking_repro(instrs):
    # Issue 5: Loop unpacking
    results = []
    for target_off, count in instrs:
        results.append((target_off, count))
    return results

if __name__ == "__main__":
    print(ternary_fstring_repro("x, y", "x + y"))
    print(compound_cond_fstring_repro("val", True))
    print(complex_fstring_repro("func", [1, 2, 3]))
    print(dict_int_keys_repro())
    print(loop_unpacking_repro([(1, 2), (3, 4)]))

[coderabbitai]

🔗 Related PRs

#19 - #18 improve ternary conditional operator decompilation [closed]
#44 - #43 Improve bytecode pattern suppression and control flow handling [merged]
#50 - #49 Enhance bytecode decompiler with slices and tuples [merged]

---

📝 Issue Planner

_{Check the box below or use the @coderabbitai plan command to generate an implementation plan and prompts that you can use with your favorite coding assistant.}

• Create Plan

---

🧪 Issue enrichment is currently in open beta.

You can configure auto-planning by selecting labels in the issue_enrichment configuration.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord!