More FYI: With the new expat1_2.zip (153591 bytes) DLLs, all tests pass on
Windows except for test_sax. No change in symptoms. The failure modes for
test_sax depend on all of:
+ Whether run in release or debug builds.
+ Whether text_sax.py is run directly or via regrtest.py.
+ Whether I delete all .pyc/.pyo files first, or use precomplied ones.
+ In debug builds, whether the test is started from within the
debugger, or I start it via cmdline and attach to the process after
it crashes (with a memory fault).
Here's a new failure mode:
test test_sax crashed -- XMLParserType: no element found: line 1, column 5
So this smells to high heaven of either a nasty gc problem or referencing
uninitialized memory. Symptoms don't change if I stick
import gc
gc.disable()
at the start of test_sax.py.
Barry, can you try running test_sax under Insure? I've got little chance of
making enough time tonight to figure this out the hard way ...
RetroSearch is an open source project built by @garambo | Open a GitHub Issue
Search and Browse the WWW like it's 1997 | Search results from DuckDuckGo
HTML:
3.2
| Encoding:
UTF-8
| Version:
0.7.4