StatelessIncrementalEncoder, sys attributes#7195

youknowone

Summary by CodeRabbit

Release Notes

Bug Fixes
- Improved codec compatibility with enhanced fallback support for encoder/decoder operations
- Refined I/O error handling during initialization for better stability
New Features
- Automatic registration of latin-1 encoding aliases for broader compatibility

coderabbitai

📝 Walkthrough

Walkthrough

This pull request introduces stateless codec encoder/decoder wrapper classes as fallback mechanisms in the _io module, refactors the sys module by removing module parameters from pyclass definitions and dynamically computing flag properties, and updates VM initialization logic for codec registration and stdio setup with refined feature gating and error handling.

Changes

Cohort / File(s)	Summary
Codec Wrapper Fallback Implementation `crates/vm/src/stdlib/io.rs`	Introduces `StatelessIncrementalEncoder` and `StatelessIncrementalDecoder` pyclasses that wrap non-incremental codec functions, providing fallback paths when incremental interfaces are unavailable. Each class implements encode/decode methods that delegate to underlying codec functions and state management methods.
Sys Module Refactoring `crates/vm/src/stdlib/sys.rs`	Removes `module = "sys"` parameters from `_BootstrapStderr` and `_SandboxStdio` pyclass definitions. Refactors `FlagsData` by removing `thread_inherit_context` and `context_aware_warnings` fields and converting them to dynamic `pygetset` properties on `PyFlags`. Marks Windows version struct fields with `#[pystruct_sequence(skip)]` to exclude them from Python-facing struct sequence.
VM Initialization & Codec Registration `crates/vm/src/vm/mod.rs`	Adds early registration of latin_1 codec aliases when freeze-stdlib feature is enabled. Refines stdio initialization by adding feature-gating requiring both `host_env` and `stdio` features, adjusting error handler derivation logic to check `stdio_errors` setting and fallback to appropriate defaults.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~25 minutes

Possibly related PRs

#6617: Modifies vm/mod.rs for stdlib encoding registration and stdio initialization/error handling updates.
#7034: Adjusts VM initialization and encoding registration logic in vm/mod.rs for stdlib bootstrap.
#7035: Modifies stdio initialization closure with feature-gating and return type adjustments in vm/mod.rs.

Suggested reviewers

fanninpm
coolreader18
ShaharNaveh

Poem

🐰 A stateless hopper skips through io,
Flags flutter free with property glow,
Latin codecs register their call,
While stdio bootstraps through it all!
Refactor, fallback, feature gates shine,
The VM's foundation now redesigned. ✨

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The PR title directly references two of the three main changes: StatelessIncrementalEncoder (new classes added to io.rs) and sys attributes (modifications to sys.rs attributes and getters). However, it omits the third significant change in vm/mod.rs regarding stdlib encodings registration and stdio initialization.
Docstring Coverage	✅ Passed	Docstring coverage is 83.33% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings (stacked PR)
📝 Generate docstrings (commit on current branch)

🧪 Generate unit tests (beta)

Create PR with unit tests
Post copyable unit tests in a comment

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

github-actions

📦 Library Dependencies

The following Lib/ modules were modified. Here are their dependencies:

[ ] test: cpython/Lib/test/test_compile.py (TODO: 24)
[x] test: cpython/Lib/test/test_compiler_assemble.py
[x] test: cpython/Lib/test/test_compiler_codegen.py
[x] test: cpython/Lib/test/test_peepholer.py (TODO: 32)

dependencies:

dependent tests: (no tests depend on compile)

[x] test: cpython/Lib/test/test_sys.py (TODO: 8)
[ ] test: cpython/Lib/test/test_syslog.py (TODO: 2)
[ ] test: cpython/Lib/test/test_sys_setprofile.py (TODO: 14)
[ ] test: cpython/Lib/test/test_sys_settrace.py (TODO: 85)

dependencies:

dependent tests: (213 tests)

sys: regrtestdata test___all__ test__colorize test__locale test__osx_support test_android test_annotationlib test_argparse test_array test_ast test_asyncio test_audit test_bdb test_bigaddrspace test_bigmem test_bisect test_buffer test_builtin test_bytes test_bz2 test_c_locale_coercion test_calendar test_call test_cmath test_cmd test_cmd_line test_cmd_line_script test_code test_code_module test_codeccallbacks test_codecs test_collections test_compile test_compileall test_complex test_concurrent_futures test_context test_contextlib test_coroutines test_csv test_ctypes test_datetime test_dbm test_dbm_sqlite3 test_descr test_dict test_difflib test_dis test_doctest test_doctest2 test_docxmlrpc test_dtrace test_dynamic test_dynamicclassattribute test_email test_ensurepip test_enum test_enumerate test_eof test_except_star test_exceptions test_faulthandler test_fcntl test_file test_file_eintr test_fileinput test_fileio test_float test_fork1 test_format test_fractions test_frozen test_functools test_future_stmt test_generators test_genericpath test_getopt test_glob test_grammar test_gzip test_hash test_hashlib test_http_cookiejar test_httpservers test_importlib test_inspect test_int test_io test_iter test_itertools test_json test_largefile test_launcher test_list test_locale test_logging test_long test_lzma test_mailbox test_marshal test_math test_memoryio test_memoryview test_metaclass test_mimetypes test_mmap test_msvcrt test_multiprocessing_fork test_multiprocessing_main_handling test_ntpath test_numeric_tower test_operator test_optparse test_ordered_dict test_os test_osx_env test_pathlib test_patma test_peepholer test_pickle test_pkg test_pkgutil test_platform test_plistlib test_posix test_posixpath test_print test_property test_pty test_pwd test_py_compile test_pyclbr test_pydoc test_pyexpat test_quopri test_raise test_range test_re test_regrtest test_repl test_reprlib test_resource test_runpy test_script_helper test_select test_selectors test_shutil test_signal test_site test_slice test_smtplib test_socket test_sqlite3 test_ssl test_stable_abi_ctypes test_stat test_statistics test_str test_strftime test_string_literals test_strtod test_struct test_subprocess test_support test_symtable test_sys test_sys_setprofile test_sys_settrace test_sysconfig test_syslog test_tarfile test_tempfile test_termios test_threadedtempfile test_threading test_threading_local test_threadsignals test_time test_timeit test_tomllib test_tools test_trace test_traceback test_type_comments test_types test_typing test_unicode_file test_unicode_file_functions test_unicodedata test_unittest test_univnewlines test_urllib test_urllib2 test_urllib2net test_urlparse test_utf8_mode test_uuid test_venv test_wait3 test_wait4 test_wave test_weakref test_webbrowser test_winconsoleio test_with test_wsgiref test_xml_etree test_xmlrpc test_zipapp test_zipfile test_zipfile64 test_zipimport test_zlib

Legend:

[+] path exists in CPython
[x] up-to-date, [ ] outdated

coderabbitai

Actionable comments posted: 2

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.

Inline comments:
In `@crates/vm/src/stdlib/io.rs`:
- Around line 2610-2653: The encode wrapper in StatelessIncrementalEncoder
currently ignores the second tuple item from the codec result; update the encode
method to validate that res.try_into_value(vm)? yields a 2-tuple where the
second element is an integer (non-negative if appropriate to your codec
contract) and raise a TypeError via vm.new_type_error when it is missing or not
an int (reusing the existing error message "encoder must return a tuple (object,
integer)"). Locate the encode method and the tuple variable, convert/validate
tuple[1] as an integer (e.g., via try_into_value or checking .isinstance), and
only then return tuple[0].clone(), ensuring malformed codec outputs surface as
errors.
- Around line 2656-2694: In StatelessIncrementalDecoder::decode, validate that
the second tuple element is an integer and raise a TypeError if not; after
converting res to PyTupleRef and checking tuple.len(), try to convert tuple[1]
to an integer (using the VM's integer/index conversion helpers) and return
Err(vm.new_type_error("decoder must return a tuple (object, integer)")) when
conversion fails, otherwise proceed to return tuple[0].clone() as before. Ensure
this check is applied in the decode method to mirror the encoder wrapper's
guard.

StatelessIncrementalEncoder, sys attributes

youknowone added 3 commits February 21, 2026 22:58

Fix WindowsVersionData n_sequence_fields

6a02392

Register latin-1 codec and fix stdio bootstrap

ea8e25c

Move sys.flags thread_inherit_context/context_aware_warnings to getset

d19d523

youknowone force-pushed the sys branch from 05266b6 to d19d523 Compare February 21, 2026 13:58

youknowone marked this pull request as ready for review February 21, 2026 14:46

coderabbitai Bot reviewed Feb 21, 2026

View reviewed changes

youknowone deleted the sys branch February 21, 2026 15:30

coderabbitai Bot mentioned this pull request Feb 22, 2026

Add missing _winapi functions #7185

Merged

youknowone added a commit to youknowone/RustPython that referenced this pull request Mar 22, 2026

Merge pull request RustPython#7195 from youknowone/sys …

c784cc6

StatelessIncrementalEncoder, sys attributes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

StatelessIncrementalEncoder, sys attributes#7195