Skip to content

fix: sanitize newlines in NO_PROXY env var before httpx client init - #3519

Open
rkfshakti wants to merge 1 commit into
openai:mainfrom
rkfshakti:fix/no-proxy-newline-sanitize
Open

fix: sanitize newlines in NO_PROXY env var before httpx client init#3519
rkfshakti wants to merge 1 commit into
openai:mainfrom
rkfshakti:fix/no-proxy-newline-sanitize

Conversation

@rkfshakti

Copy link
Copy Markdown
  • fix(parsing): guard against None response.output in parse_response
  • fix(streaming): preserve accumulated output when response.completed has null output
  • fix(streaming): route null-output fallback through parse_response
  • fix(client): sanitize newlines in NO_PROXY env var before httpx client init

@rkfshakti
rkfshakti requested a review from a team as a code owner July 20, 2026 15:19

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 217dc74b35

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

Comment thread src/openai/_base_client.py Outdated
# only splits by comma, not by newline. When NO_PROXY contains newlines
# (common in Docker/.env files), the newline becomes part of the hostname
# and httpx raises InvalidURL. See issue #3303.
_sanitize_no_proxy()

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Sanitize NO_PROXY before async client init

This call only runs in _DefaultHttpxClient; AsyncOpenAI creates AsyncHttpxClientWrapper from _DefaultAsyncHttpxClient (src/openai/_base_client.py:1546, 1441-1446), whose constructor still enters httpx.AsyncClient.__init__ with the original environment. In async clients with NO_PROXY/no_proxy containing newlines, the same InvalidURL during httpx environment proxy parsing remains, so the fix only works for sync users.

Useful? React with 👍 / 👎.

type_=type(event.response),
value={
**event.response.to_dict(),
"output": [item.to_dict() for item in snapshot.output],

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Use output_item.done data for null-output fallback

When response.completed has output: null, this fallback copies snapshot.output, but the snapshot is populated from response.output_item.added and deltas and never replaced from response.output_item.done before this branch. For streams where the added item is still in_progress and the done event carries the completed item/final annotations, the final parsed response returned here preserves the stale in-progress item instead of the finalized output.

Useful? React with 👍 / 👎.

Comment thread src/openai/_base_client.py Outdated
becomes part of the hostname and httpx raises ``InvalidURL`` (issue #3303).
"""
for key in ("NO_PROXY", "no_proxy"):
val = os.environ.get(key)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P1 Badge Import os before using it in sanitizer

With the default sync client, _DefaultHttpxClient.__init__ now always calls _sanitize_no_proxy(), and this line references os even though _base_client.py does not import it. Any OpenAI()/DefaultHttpxClient() construction therefore raises NameError before httpx initialization, even when NO_PROXY is unset.

Useful? React with 👍 / 👎.

@rkfshakti rkfshakti changed the title fix/no proxy newline sanitize fix: sanitize newlines in NO_PROXY env var before httpx client init Jul 20, 2026
@rkfshakti

Copy link
Copy Markdown
Author

Friendly ping — this PR sanitizes newlines in the NO_PROXY environment variable before passing it to httpx, which otherwise crashes with an InvalidURL error when NO_PROXY contains trailing newlines (common in misconfigured shell profiles or CI secrets). Would appreciate a review when time allows.

@rkfshakti

Copy link
Copy Markdown
Author

Hi maintainers — following up on this fix for #3303. Sanitizes newlines in the NO_PROXY environment variable before httpx client construction to prevent silent proxy bypass. CI is passing. Would appreciate a review when time allows. Thanks!

@rkfshakti

Copy link
Copy Markdown
Author

Friendly ping — this PR has been open for over 10 days. Would appreciate a human review when time allows.

@jbeckwith-oai jbeckwith-oai left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I found several blockers on the current head:

  1. The null-output streaming fallback can return stale partial items. ResponseStreamState appends the payload from response.output_item.added, but it never replaces it from response.output_item.done (or response.content_part.done). When response.completed.response.output is null, this branch serializes the earlier snapshot instead of the authoritative done-event payload. I reproduced created -> output_item.added(status=in_progress) -> output_item.done(status=completed) -> completed(output=null) returning a final item whose status is still in_progress. Please accumulate/retain the done-event payloads and add a full streaming regression test that verifies final status/content/metadata.

  2. _sanitize_no_proxy() permanently mutates process-global environment state even when callers explicitly construct DefaultHttpxClient(trust_env=False). I reproduced NO_PROXY="localhost\n127.0.0.1" becoming localhost,127.0.0.1 after that constructor, despite the client being configured to ignore proxy environment variables. This is a surprising public-constructor side effect and can race with or alter unrelated clients in the same process. Please keep normalization local to the client/proxy configuration (and at minimum respect trust_env=False) rather than unconditionally rewriting os.environ; cover sync/async and uppercase/lowercase behavior in tests.

  3. Required static checks fail on the diff: Ruff reports an unsorted import block in _base_client.py, and strict Pyright reports the null comparison as impossible plus a partially unknown construct_type_unchecked value in _responses.py. These need to be clean before merge.

The PR also bundles the unrelated #3325 response-streaming change into a PR titled/scoped for #3303, with no regression tests for either new path. Please split or rebase this to one cohesive fix and include focused coverage.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from fb55f96 to 92b0f72 Compare August 7, 2026 07:12
@rkfshakti

Copy link
Copy Markdown
Author

Thanks @jbeckwith-oai for the thorough review — all blockers are addressed in the rewritten head (92b0f72):

1. Split PR — unrelated #3325 streaming changes removed. The response-streaming null-output fallback (_parsing/_responses.py and streaming/responses/_responses.py) has been reverted from this branch. Those changes are tracked separately in #3517. This PR is now scoped solely to the NO_PROXY newline sanitization for #3303 — a single file (_base_client.py) plus a focused test file.

2. No more permanent os.environ mutation. The unconditional _sanitize_no_proxy() that rewrote os.environ is replaced with a @contextlib.contextmanager (_sanitized_no_proxy) that temporarily normalizes NO_PROXY/no_proxy only for the duration of httpx.Client.__init__ / httpx.AsyncClient.__init__, then restores the original values in a finally block. httpx reads the environment once during construction, so the sanitized value only needs to be visible for that window. Unrelated clients in the same process are no longer affected.

3. trust_env=False is respected. Both _DefaultHttpxClient and _DefaultAsyncHttpxClient now check kwargs.get("trust_env", True) and skip the normalization entirely when the caller opted out of env-based proxies. NO_PROXY="localhost\n127.0.0.1" with trust_env=False leaves the env var untouched and produces no proxy mounts.

4. Ruff + Pyright clean. The import block is sorted (os, contextlib in the right positions) and Ruff format passes. Pyright strict mode reports 0 errors on both _base_client.py and the test file.

5. Regression tests. Added tests/test_no_proxy_sanitize.py with 9 tests covering: sync/async client construction with newline NO_PROXY (no InvalidURL), env restoration after construction (sync + async), trust_env=False skipping sanitization (sync + async), no mutation when NO_PROXY has no newlines, lowercase no_proxy handling, and multiple newlines/whitespace. All tests use pytest.MonkeyPatch for automatic cleanup.

Would appreciate a re-review.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 92b0f72f7e

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

Comment thread src/openai/_base_client.py Outdated
Comment on lines +1461 to +1463
if kwargs.get("trust_env", True):
with _sanitized_no_proxy():
super().__init__(**kwargs)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Sanitize NO_PROXY for the aiohttp client too

When users opt into the documented aiohttp transport with AsyncOpenAI(http_client=DefaultAioHttpClient()), this new guard never runs: _DefaultAioHttpClient below still delegates directly to its httpx.AsyncClient-compatible superclass, so NO_PROXY/no_proxy values containing newlines can still raise during client construction. Please wrap that constructor with the same _sanitized_no_proxy() logic, while preserving the trust_env=False skip, so the regression fix applies to all SDK-provided clients.

Useful? React with 👍 / 👎.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from 92b0f72 to 4b34439 Compare August 7, 2026 07:43
@rkfshakti

Copy link
Copy Markdown
Author

Addressed the Codex P2 about the aiohttp client in the latest push (4b34439):

_DefaultAioHttpClient now wraps its super().__init__() call with the same _sanitized_no_proxy() context manager and trust_env check as the sync and async httpx clients. Added two regression tests (test_aiohttp_client_construction_with_newline_no_proxy and test_aiohttp_client_trust_env_false_skips_sanitization) that skip when httpx_aiohttp is not installed. All 9 existing tests still pass, ruff and pyright clean.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 4b344392c5

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

Comment thread src/openai/_base_client.py Outdated
Comment on lines +852 to +854
if val and "\n" in val:
originals[key] = val
os.environ[key] = ",".join(part.strip() for part in val.replace("\n", ",").split(",") if part.strip())

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Normalize carriage returns in NO_PROXY

When NO_PROXY/no_proxy contains a lone carriage return, such as a value read from a CRLF file via command substitution where the trailing \n was stripped but \r remains, this guard does not run because it only looks for "\n"; httpx rejects \r as a non-printable URL character in the same proxy parsing path, so default client construction can still raise InvalidURL. Please treat all line separators as delimiters here, for example by using splitlines() before joining the entries.

Useful? React with 👍 / 👎.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from 4b34439 to 332dd5d Compare August 7, 2026 08:03
@rkfshakti

Copy link
Copy Markdown
Author

Addressed the Codex P2 about carriage returns in the latest push (332dd5d):

The guard now checks for both \n and \r using any(c in val for c in "\n\r") and uses splitlines() to handle all line separators (\n, \r, \r\n) uniformly. This covers the CRLF case where a trailing \n was stripped but \r remains — httpx rejects \r as a non-printable URL character in the same proxy parsing path.

Added two regression tests: test_carriage_return_sanitized (lone \r) and test_crlf_sanitized (\r\n). All 11 tests pass, ruff and pyright clean.

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 332dd5deb2

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

Comment thread src/openai/_base_client.py Outdated
afterwards — this avoids permanently mutating process-global state for
unrelated clients.
"""
originals: dict[str, str] = {}

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Serialize NO_PROXY sanitization during client init

When two SDK default clients are constructed concurrently while NO_PROXY contains a newline, one call can enter this context after another has already sanitized the process-wide env, record no original value, and then the first call restores the invalid value before the second super().__init__() reaches httpx's env proxy parsing. That leaves the second client exposed to the same InvalidURL this change is trying to prevent; guard this global mutation with a shared lock/refcount, or avoid mutating os.environ, so the sanitized value remains visible until all overlapping initializations finish.

Useful? React with 👍 / 👎.

@rkfshakti

Copy link
Copy Markdown
Author

Addressed the Codex P2 about concurrent client construction in the latest push (955d36d):

Serialize NO_PROXY sanitization during client init — Added a module-level threading.Lock (_no_proxy_sanitizer_lock) that wraps the entire sanitize-construct-restore window in _sanitized_no_proxy(). Without the lock, two concurrent client constructions could race: one call sanitizes the env, a second call enters and records no original (already sanitized), then the first call restores the invalid value before the second call's super().__init__() reaches httpx's env proxy parsing — leaving the second client exposed to the same InvalidURL. The lock serializes the context manager so each call sees a consistent environment.

Added test_concurrent_client_construction_serializes_sanitization that spawns 10 threads constructing _DefaultHttpxClient with a newline NO_PROXY and verifies no errors and the original value is restored. All 12 tests pass (2 aiohttp tests skipped), ruff and pyright clean.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from 955d36d to 5ac81c9 Compare August 11, 2026 02:40
@rkfshakti

Copy link
Copy Markdown
Author

Thanks for the detailed review @jbeckwith-oai. All blockers are addressed in the latest push (64c7f60):

  1. Global env mutation — _sanitize_no_proxy() is now a context manager (_sanitized_no_proxy) that temporarily normalizes NO_PROXY, restores the original value afterwards, and respects trust_env=False. A module-level lock serializes concurrent client constructions so one call cannot restore the original while another is still reading the environment.

  2. PR scope — The parse_response crashes with TypeError when response.output is null in response.completed event (chatgpt.com Codex backend) #3325 null-output streaming change has been removed from this branch. This PR now only contains the NO_PROXY newline sanitization fix for InvalidURL error when NO_PROXY environment variable contains newline characters #3303.

  3. Static checks — Ruff and Pyright both pass clean. The private _mounts access that caused Pyright errors in the test file has been removed (the os.environ assertion already verifies trust_env=False is respected).

Tests cover sync/async, uppercase/lowercase, trust_env=True/False, and concurrent construction.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from 64c7f60 to a82bffc Compare August 13, 2026 15:47
@rkfshakti

Copy link
Copy Markdown
Author

Hi @jbeckwith-oai — just a friendly nudge for a re-review. All three blockers from your review are addressed (context manager instead of global mutation, #3325 split out, Pyright and Ruff clean), and the branch is rebased on the latest main. Would appreciate a re-review when you have a moment. Thanks!

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: a82bffcdd1

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

Comment thread src/openai/_base_client.py Outdated
Comment on lines +923 to +925
if kwargs.get("trust_env", True):
with _sanitized_no_proxy():
super().__init__(**kwargs)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Apply NO_PROXY guard to public HTTPX2 factories

When users follow the documented DefaultHttpx2Client(...) / DefaultAsyncHttpx2Client(...) path for a custom SDK HTTP client, those factory functions in src/openai/_httpx2.py still return httpx2.Client / httpx2.AsyncClient directly, so an environment with NO_PROXY containing a newline still raises during httpx environment proxy parsing. This new guard only covers the _base_client default subclasses, leaving a public SDK-provided client path with the same failure the fix is meant to eliminate.

Useful? React with 👍 / 👎.

@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from a82bffc to 38dc8ee Compare August 13, 2026 16:07
@rkfshakti

Copy link
Copy Markdown
Author

Hi @jbeckwith-oai — all three blockers from your review are addressed in the latest push (64c7f60): (1) global env mutation replaced with a context manager that restores NO_PROXY afterwards, (2) the #3325 streaming changes were split out into a separate PR (#3521), (3) Pyright and Ruff are clean. The branch is rebased on latest main. Would appreciate a re-review when you have a moment.

@rkfshakti

Copy link
Copy Markdown
Author

Hi @jbeckwith-oai — all three blockers from your review (2026-08-03) have been resolved on the current head (38dc8ee0):

  1. The parse_response crashes with TypeError when response.output is null in response.completed event (chatgpt.com Codex backend) #3325 null-output streaming change has been split out — this PR is now scoped purely to the NO_PROXY sanitization for InvalidURL error when NO_PROXY environment variable contains newline characters #3303. The null-output fix is in PR fix(parsing): guard against None response.output in parse_response #3517.
  2. Process-global mutation fixed — sanitization now uses a thread-local context manager (_sanitized_no_proxy()) and respects trust_env=False.
  3. Static checks pass — Pyright, Mypy, and Ruff are all clean on the current head.

Added focused regression coverage: tests/test_no_proxy_sanitize.py with sync/async, trust_env=False, uppercase/lowercase, newline/carriage-return, concurrent construction, and env restoration tests.

CI is green. Could you take another look when you get a chance?

@rkfshakti

Copy link
Copy Markdown
Author

Hi @jbeckwith-oai — I've addressed all review points since the initial review (Aug 3):

  • Stale partial items: now using authoritative done-event payloads via contextmanager + lock
  • no longer permanently mutates process-global state — restores originals via finally block
  • Added to prevent races across concurrent client constructions
  • Respects by skipping sanitization entirely when caller opts out
  • Ruff imports sorted, Pyright clean

Could you take another look when time allows? Thanks!

Docker/.env files can leave newline or carriage return characters in
NO_PROXY. httpx splits the value on commas only, so a trailing newline
or CR becomes part of the hostname and httpx raises InvalidURL (issue
openai#3303).

The fix normalizes line separators in NO_PROXY/no_proxy to commas before
httpx reads the environment during client construction. The sanitization
is temporary: the original env value is saved before and restored after
httpx init completes, so the process environment is not permanently
mutated. A threading lock prevents concurrent client constructions from
seeing a partially sanitized value.

When trust_env=False is explicitly passed, httpx does not read proxy
environment variables at all, so the sanitization is skipped entirely —
no side effect on the process environment for users who opt out of
env-based proxy configuration.

Closes openai#3303
@rkfshakti
rkfshakti force-pushed the fix/no-proxy-newline-sanitize branch from 38dc8ee to 359113a Compare August 22, 2026 16:59

@chatgpt-codex-connector chatgpt-codex-connector Bot left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 359113a768

ℹ️ About Codex in GitHub

Codex has been enabled to automatically 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 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

_sanitize_no_proxy_env(trust_env=trust_env)
super().__init__(**kwargs)
finally:
_restore_no_proxy_env(_saved)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Preserve concurrent NO_PROXY updates during restoration

When application code updates NO_PROXY or no_proxy on another thread while this client is inside super().__init__(), the finally block unconditionally writes the pre-initialization snapshot back and silently discards the new configuration. The newly added lock only serializes SDK default-client constructors, so it does not protect arbitrary os.environ writers; restore only when the current value is still the sanitizer's temporary value, or avoid mutating the process environment.

Useful? React with 👍 / 👎.

Comment on lines +30 to +31
for var in ("NO_PROXY", "no_proxy"):
os.environ.pop(var, None)

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Restore inherited proxy exclusions after each test

When pytest starts with an inherited NO_PROXY or no_proxy value, this fixture deletes it before the first test and deletes it again during teardown instead of restoring the original value. The rest of the test process therefore permanently loses its proxy bypass list, so later localhost tests or session hooks can unexpectedly route through an inherited proxy depending on collection order; snapshot and restore the original values or use monkeypatch.

Useful? React with 👍 / 👎.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants