fix: the bug of genlib_reader parser

[yu-lin-chen]

Fix Log: genlib_reader constant gate parsing (2026-07-07)

File: include/mockturtle/io/genlib_reader.hpp

Problem:
The genlib_reader incorrectly parses constant gates such as TIEHI (Z=1), TIELO (ZN=0), CONST0, and CONST1. During tokenization, the literal "0" or "1" in the expression is treated as a pin name, producing num_vars=1. When create_from_formula is then called with formula="1" and pin_names={"1"}, the literal collides with the pin name and is interpreted as a variable reference, yielding a 1-variable truth table [0,1] — identical to a buffer. This causes tie-high/tie-low cells to be matched as buffers during technology mapping.

Fix:
After CONST stripping (which converts CONST0/CONST1 to "0"/"1") and after tokenization, check whether the formula is a bare "0" or "1". If so, clear the pin vectors and pin names — these spurious entries were produced by mistaking the constant literal for a variable name. The truth table construction still goes through the standard create_from_formula path, which now correctly receives num_vars=0 and empty pin_names, producing a proper 0-variable constant truth table.

Change: 6 lines added after the CONST stripping block and before num_vars / create_from_formula. No existing code paths altered.
tcbn28hpcplusbwp40p140tt0p9v25c.lib:genlib write by abc
GATE TIEHBWP40P140 0.25 Z=1; PIN * UNKNOWN 1 999 1.00 0.00 1.00 0.00
GATE TIELBWP40P140 0.25 ZN=0; PIN * UNKNOWN 1 999 1.00 0.00 1.00 0.00

[Copilot]

Pull request overview

This PR fixes genlib_reader parsing for constant-output gates whose Boolean expression is a bare literal ("0"/"1"), preventing them from being misinterpreted as 1-input buffers during truth table construction and downstream technology mapping.

Changes:

• Detects bare constant formulas ("0" / "1") after CONST0/CONST1 stripping.
• Clears tokenization-derived pin_names (and corresponding pp) for these constant formulas so create_from_formula builds a 0-variable constant truth table.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.05%. Comparing base (25beb0e) to head (b35aed0).

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #698      +/-   ##
==========================================
- Coverage   84.06%   84.05%   -0.01%     
==========================================
  Files         190      190              
  Lines       29468    29471       +3     
==========================================
  Hits        24773    24773              
- Misses       4695     4698       +3     

☔ View full report in Codecov by Harness.
📢 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.