Adopt py-canon and typed runtime assets by soodoku · Pull Request #35 · appeler/naampy

soodoku · 2026-08-17T07:45:22Z

Summary

move the checkpoint out of the wheel to a pinned Hugging Face revision
restore the current py-canon standard: src layout, uv_build, reusable workflows, and flat Sphinx docs without notebook tooling
convert Dataverse CSV.gz transports into validated atomic Parquet runtime caches with an explicit Arrow schema
preserve input indices and output-column contracts, reject invalid state/year filters, and avoid data downloads for empty input
prepare version 0.10.0 with reproducible changelog and documentation

Local verification

42 passed, 27 skipped, 1 live test deselected
Ruff lint and format clean
Pyright: 0 errors
pydoclint: 0 violations
Sphinx HTML and doctest builds clean
preen: all checks passed
full 10,226,249-row v2_1k CSV-to-Parquet conversion and source-count equivalence passed
wheel/sdist build and Twine metadata checks passed

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 51d71da135

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-08-17T07:48:47Z

        # Whitespace-only names strip to "", which is not a name; treat it as missing.
        rdf["__first_name"] = first_name.mask(first_name == "")

+        if rdf.empty:


Validate the dataset before the empty-frame shortcut

When an empty DataFrame is passed with an invalid dataset (for example, dataset="v2-native"), this early return bypasses load_naampy_data() and therefore never performs the new IN_ROLLS_DATA validation; it silently returns a non-native output schema even though the same configuration raises ValueError as soon as a batch contains a row. Validate the dataset key before this shortcut so empty and nonempty batches enforce the same API contract without requiring a download.

Useful? React with 👍 / 👎.

soodoku · 2026-08-17T08:03:27Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: e9e437e109

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-08-17T08:06:33Z

                    "n_third_gender",
                ],
            )
+            if state is not None and state not in set(adf["state"]):


Normalize the Streamlit all-state sentinel

When the deployed Streamlit app's default all option is used, streamlit/streamlit_app.py rewrites it to "" and passes that value as state; this new validation treats every non-None string as a requested state and raises ValueError because the dataset has no empty-string state. As a result, the app's default electoral-roll lookup cannot complete, so normalize blank state values to None or update the wrapper to pass None.

Useful? React with 👍 / 👎.

soodoku · 2026-08-17T08:12:48Z

@codex review

soodoku · 2026-08-17T08:15:49Z

@codex review

chatgpt-codex-connector

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 29f934e850

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

chatgpt-codex-connector · 2026-08-17T08:18:57Z

                    "n_third_gender",
                ],
            )
+            if state is not None and state not in set(adf["state"]):


Align the Streamlit Maharashtra value with the dataset key

In the checked Streamlit flow, selecting Maharashtra still passes state="maharashtra" from streamlit/streamlit_app.py, while the published table uses the maharastra key (also documented by this commit). This new validation therefore raises ValueError for that UI option instead of completing the lookup; update the selector value or normalize the alias before validating.

Useful? React with 👍 / 👎.

soodoku · 2026-08-17T08:23:44Z

@codex review

chatgpt-codex-connector · 2026-08-17T08:27:26Z

Codex Review: Didn't find any major issues. Can't wait for the next one!

Reviewed commit: 4e43044145

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

soodoku · 2026-08-17T08:32:04Z

@codex review exact commit 4e43044

chatgpt-codex-connector · 2026-08-17T08:33:45Z

Codex Review: Didn't find any major issues. More of your lovely PRs please.

Reviewed commit: 4e43044145

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

Open a pull request for review
Mark a draft as ready
Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

soodoku added 2 commits August 16, 2026 00:55

Move gender checkpoint to Hugging Face

d4f63a8

Adopt py-canon and typed runtime assets

51d71da

chatgpt-codex-connector Bot reviewed Aug 17, 2026

View reviewed changes

soodoku added 3 commits August 17, 2026 00:49

Fix release documentation checks

3d75ede

Exclude OS metadata from distributions

8ff7ecc

Validate datasets before empty input shortcut

e9e437e

chatgpt-codex-connector Bot reviewed Aug 17, 2026

View reviewed changes

Fix dataset validator documentation

df2ca5e

Fix Streamlit all-state lookup

29f934e

chatgpt-codex-connector Bot reviewed Aug 17, 2026

View reviewed changes

Map Maharashtra to electoral state key

4e43044

soodoku merged commit c75d251 into master Aug 17, 2026
12 checks passed

soodoku deleted the agent/hf-model-assets branch August 17, 2026 08:32

Uh oh!

Conversation

soodoku commented Aug 17, 2026

Summary

Local verification

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Aug 17, 2026

Choose a reason for hiding this comment

Uh oh!

soodoku commented Aug 17, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Aug 17, 2026

Choose a reason for hiding this comment

Uh oh!

soodoku commented Aug 17, 2026

Uh oh!

soodoku commented Aug 17, 2026

Uh oh!

chatgpt-codex-connector Bot left a comment

Choose a reason for hiding this comment

💡 Codex Review

Uh oh!

chatgpt-codex-connector Bot Aug 17, 2026

Choose a reason for hiding this comment

Uh oh!

soodoku commented Aug 17, 2026

Uh oh!

chatgpt-codex-connector Bot commented Aug 17, 2026

Uh oh!

soodoku commented Aug 17, 2026

Uh oh!

Uh oh!

chatgpt-codex-connector Bot commented Aug 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant