fix: make stat() Windows-compatible by guarding os.getegid/geteuid

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188

[Rohan5commit]

Summary

Fixes#177 — stat() crashes on Windows because os.getegid() and os.geteuid() do not exist on that platform.

Root Cause

In src/celpy/__main__.py, the stat() function unconditionally calls:

"group_access": celtypes.BoolType(status.st_gid==os.getegid()),
"user_access": celtypes.BoolType(status.st_uid==os.geteuid()),

Both os.getegid() and os.geteuid() are POSIX-only and raise AttributeError on Windows.

Fix

• Added a module-level _IS_WINDOWS = platform.system() == "Windows" flag.
• Wrapped the group_access and user_access assignments in if not _IS_WINDOWS: so they are only included on POSIX platforms.
• On Windows, these two keys are simply absent from the returned MapType, which is consistent with the issue description's recommendation.

Testing

The existing test tests/test_main.py::test_stat_good exercises this path on POSIX. On Windows it previously raised AttributeError; with this change it returns the map without the two POSIX-only keys.

[linux-foundation-easycla]

• ❌ - login: @Rohan5commit / name: Rohan Santhosh Kumar. The commit (26bec6d) is not authorized under a signed CLA. Please click here to be authorized. For further assistance with EasyCLA, please visit our EasyCLA portal and chat with our support bot.

[slott56]

Thanks for helping. This was a speedy effort, but it needs some small changes.

First, use make test to test. Use make to get a list of potential targets. Format the code with ruff format src.

It saves time to actually look over the repository and note the presence of a Makefile. LLM's do a poor job of actually looking at the repository as a whole and are unlikely to correctly parse a Makefile.

Also. The documentation for https://docs.python.org/3/library/os.html#os.getegid and https://docs.python.org/3/library/os.html#os.geteuid are pretty explicit on the platforms supported. Using "Not Windows" will permit WASI, which is not actually supported. Following the Availability link leads to a longer list of platforms that are likely to contain this.

It saves time to read the https://docs.python.org/3/library/intro.html#notes-on-availability section. LLM's do a poor job of actually ingesting all of the information related to this kind of change.

A slightly better bit of logic for exclusions would be helpful, but not ideal.

Handling an exception -- rather than the platform -- would be a much more universal approach to this. Handling an exception is a pattern sometimes called "Easier to Ask Forgiveness Than Permission" (EAFP), which is more consistent with Python. The if statement is a pattern called "Look Before You Leap" (LBYL), which is not very Pythonic. The Notes on Availability suggest that there is no comprehensive list of platforms on which the function is defined.

It saves time to consider the alternative designs. LLM's do a poor job of following design patterns established by a wider community of actual people. Further, the details are open-ended, something LLM's don't deal with, but actual thinking people can handle very well.

[Quantum0uasar]

Thanks for the contribution! Just noting that there's a related PR #179 that addresses the same issue using the EAFP (try/except AttributeError) pattern, which @slott56 recommended here as well. The EAFP approach has an advantage over the platform.system() check since it also covers non-Windows non-POSIX environments like WASI, where os.getegid/os.geteuid may also raise AttributeError.

You might want to coordinate with the author of #179 or update this PR to use the try/except approach to align with the review feedback and improve the robustness of the fix.

[slott56]

Closed by #188