mirror of
https://github.com/qelectrotech/qelectrotech-source-mirror.git
synced 2026-10-14 17:44:13 +02:00
ispyisail
b8390c853c
Every tool now takes a .qetz as it takes a .qet. The tools that read a project's XML read the .qet the configured QElectroTech saves from the .qetz, made once per version of the file (path, size, modification time); without a QElectroTech, they say to set QET_BINARY. The tools that launch QElectroTech hand it the .qetz, and write one when the output path ends in .qetz. qet_scan counts .qetz files too. The server's description and README say so. Reading through QElectroTech instead of the file's own layout keeps the server right while the zipped format still changes; the first read of a big .qetz takes as long as QElectroTech needs to save it. test_qet_mcp.py: the read tools answer the same for a .qetz and its .qet, qet_diff finds nothing between them, qet_scan counts both, a .qetz is converted once, qet_query reads it, qet_edit writes one (four fail when _root() parses the file itself); without QElectroTech the answer names QET_BINARY. Whole suite: 430 tests pass with QET_ENABLE_SCRIPTING=1. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>