Commit Graph

3910 Commits

Author SHA1 Message Date
autofix-ci[bot]
f17a9bd4b1 [autofix.ci] apply automated fixes 2025-08-15 00:38:24 +00:00
Claude Bot
f06277735a fix(mimalloc): suppress warnings for memory-mapped PE sections
Mimalloc prints warnings when validating pointers to memory-mapped PE sections
(~118MB) during Windows executable compilation. These warnings occur because:

1. mimalloc's heuristics flag large memory regions as suspicious
2. It validates them and confirms they're actually valid
3. The warnings are harmless but noisy and cause test failures

This fix suppresses mimalloc warnings by setting max_warnings=0 during CLI
initialization, while still allowing the MIMALLOC_MAX_WARNINGS environment
variable to override this behavior if needed.

Also update Windows PE tests to filter out these warnings for backward
compatibility with existing debug builds.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-15 00:36:52 +00:00
Jarred Sumner
43e59a711b Merge branch 'main' into feat/windows-version-description 2025-08-10 23:00:51 -07:00
Jarred Sumner
b0799da968 Harden Transfer-Encoding (#21737)
### What does this PR do?

### How did you verify your code works?

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-10 03:52:40 -07:00
Jarred Sumner
a67ba81e0b Only highlight per chunk instead of per line (#21729) 2025-08-09 21:35:17 -07:00
Jarred Sumner
7cdc5d879c Don't highlight backgrounds when it's just words that changed (#21727)
### What does this PR do?

Setting the background color on plaintext diffs makes the plaintext
harder to read. This is particularly true when the input is longer.

This conservatively makes us only add the background color to the diff
when the characters being highlighted are all whitespaces, punctuation
or non-printable.

This branch:

<img width="748" height="388" alt="image"
src="https://github.com/user-attachments/assets/ceaf02ba-bf71-4207-a319-c041c8a887de"
/>

Canary:

<img width="742" height="404" alt="image"
src="https://github.com/user-attachments/assets/cc380f45-5540-48ed-aea1-07f4b0ab291e"
/>


### How did you verify your code works?

Updated test
2025-08-09 19:50:25 -07:00
Jarred Sumner
1dc9fdfd9b Fix process.stdout/stderr missing Symbol.asyncIterator (#21720)
## Summary
- Adds `Symbol.asyncIterator` to `process.stdout` and `process.stderr`
when they are TTY or pipe/socket streams
- Matches Node.js behavior where these streams are Duplex-like and
support async iteration
- Does not add the iterator when streams are redirected to files
(matching Node.js SyncWriteStream behavior)

## Test plan
- Added test in
`test/regression/issue/test-process-stdout-async-iterator.test.ts`
- Verified the fix works with Claude Code on Linux x64
- Test passes with `bun bd test
test/regression/issue/test-process-stdout-async-iterator.test.ts`

Fixes #21704

🤖 Generated with [Claude Code](https://claude.ai/code)

---------

Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-09 06:40:36 -07:00
robobun
584946b0ce Fix comma operator optimization to preserve 'this' binding semantics (#21653)
## Summary
- Fix transpiler bug where comma expressions like `(0, obj.method)()`
were incorrectly optimized to `obj.method()`
- This preserved the `this` binding instead of stripping it as per
JavaScript semantics
- Add comprehensive regression test to prevent future issues

## Root Cause
The comma operator optimization in `src/js_parser.zig:7281` was directly
returning the right operand when the left operand had no side effects,
without checking if the expression was being used as a call target.

## Solution
- Added the same `is_call_target` check that other operators (nullish
coalescing, logical OR/AND) use
- When a comma expression is used as a call target AND the right operand
has a value for `this`, preserve the comma expression to strip the
`this` binding
- Follows existing patterns in the codebase for consistent behavior

## Test Plan
- [x] Reproduce the original bug: `(0, obj.method)()` incorrectly
preserved `this`
- [x] Verify fix: comma expressions now correctly strip `this` binding
in function calls
- [x] All existing transpiler tests continue to pass
- [x] Added regression test covering various comma expression scenarios
- [x] Tested edge cases: nested comma expressions, side effects,
different operand types

🤖 Generated with [Claude Code](https://claude.ai/code)

---------

Co-authored-by: Claude Bot <claude-bot@bun.sh>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-09 05:11:50 -07:00
robobun
3766f183e6 deps: bump WebKit to eb92990ae9e0a8df3141b8cf946a4f250393e213 (#21702)
## Summary
- Updates WebKit from 75f6499 to eb92990 (latest release from
oven-sh/webkit)
- This brings in the latest WebKit improvements and fixes

## Test plan
- [ ] Verify the build completes successfully
- [ ] Run existing test suite to ensure no regressions

🤖 Generated with [Claude Code](https://claude.ai/code)

---------

Co-authored-by: Claude Bot <claude-bot@bun.sh>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Jarred Sumner <jarred@jarredsumner.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-09 05:00:46 -07:00
Jarred Sumner
19fac68e81 Reduce stack space usage of parseSuffix (#21662)
### What does this PR do?

Reduce stack space usage of parseSuffix

### How did you verify your code works?

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-09 00:20:17 -07:00
Jarred Sumner
964d4dac2c Rewrite AbortSignal.timeout (#21695)
### What does this PR do?

On Linux, AbortSignal.timeout created a file descriptor for each timeout
and did not keep the event loop alive when a timer was active. This is
fixed.

### How did you verify your code works?

Fewer flaky tests

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Claude <claude@anthropic.ai>
2025-08-08 23:07:19 -07:00
Meghan Denny
05cff5cfde test: fix static-initializers.test.ts
regressed in 46e1c5a0fa
2025-08-08 22:28:42 -07:00
Zack Radisic
ee88c489ab shell: fix $.braces(...) on unicode inputs, support more deeply nested braces (#21709)
### What does this PR do?

- Fixes `$.braces(...)` not working properly on non-ascii inputs
- Switches braces code to use `SmallList` to support more deeply nested
brace expansion

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-08 18:12:42 -07:00
Jarred Sumner
46e1c5a0fa Downgrade mimalloc + set libc musl flag (#21684)
### What does this PR do?

### How did you verify your code works?

---------

Co-authored-by: Claude Bot <claude-bot@bun.sh>
Co-authored-by: Claude <noreply@anthropic.com>
2025-08-08 18:02:19 -07:00
Claude
961be1a340 Merge branch 'main' into feat/windows-version-description 2025-08-08 07:11:11 +02:00
Zack Radisic
3b1842723e Fix shell pipeline crash (#21687)
### What does this PR do?

Fixes a crash related to pipelines

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Jarred Sumner <jarred@jarredsumner.com>
2025-08-07 19:13:37 -07:00
Dylan Conway
c3c2dccc55 Fix N-API BigInt word count issue (#21652)
## Summary
Fixes a bug in napi_get_value_bigint_words where the function would
return the number of words copied instead of the actual word count
needed when the provided buffer is smaller than required.

## The Problem
When napi_get_value_bigint_words was called with a buffer smaller than
the actual BigInt size, it would incorrectly return the buffer size
instead of the actual word count needed. This doesn't match Node.js
behavior.

### Example
BigInt that requires 2 words: 0x123456789ABCDEF0123456789ABCDEFn
Call with buffer for only 1 word
- Before fix: word_count = 1 (buffer size)
- After fix: word_count = 2 (actual words needed)

## The Fix
Changed napi_get_value_bigint_words to always set word_count to the
actual number of words in the BigInt, regardless of buffer size.

## Test Plan
- Added test test_bigint_word_count that verifies the word count is
correctly returned
- Added test test_ref_unref_underflow for the existing
napi_reference_unref underflow protection
- Both tests pass with the fix and match Node.js behavior

🤖 Generated with [Claude Code](https://claude.ai/code)

---------

Co-authored-by: Claude <noreply@anthropic.com>
2025-08-07 18:15:12 -07:00
pfg
3652008b0d Update bun:test diff (#21158)
Fixes #6229 (Fixes BAPI-655): 

|before|<img width="806" height="84" alt="image"
src="https://github.com/user-attachments/assets/6d6c8628-40a8-4950-a7a4-8a85ee07a302"
/>|
|-|-|
|after|<img width="802" height="87" alt="image"
src="https://github.com/user-attachments/assets/c336a626-2b08-469e-aa73-676f43a0f176"
/>|

Fixes #21498 (Fixes BAPI-2240), Fixes #10852 (Fixes BAPI-743):

|before|after|
|-|-|
|<img width="474" height="147" alt="image"
src="https://github.com/user-attachments/assets/bf2225de-a573-4672-a095-f9ff359ec86c"
/>|<img width="283" height="226" alt="image"
src="https://github.com/user-attachments/assets/89cb0e45-b1b7-4dbb-9ddb-b9835baa4b74"
/>|
|<img width="279" height="176" alt="image"
src="https://github.com/user-attachments/assets/e9be7308-dc38-43d2-901c-c77ce4757a51"
/>|<img width="278" height="212" alt="image"
src="https://github.com/user-attachments/assets/8c29b385-a053-4606-9474-3e5c0e60278c"
/>|

Improves multiline string and long output

|before|after|
|-|-|
|<img width="537" height="897" alt="image"
src="https://github.com/user-attachments/assets/034800c5-ab22-4915-90d9-19831906bb2e"
/>|<img width="345" height="1016" alt="image"
src="https://github.com/user-attachments/assets/fa95339e-c136-4c7c-af94-5f11400836dd"
/>|

Improves long single line string output

|before|<img width="1903" height="191" alt="image"
src="https://github.com/user-attachments/assets/bae35c81-0566-4291-810e-e65dc0381aef"
/>|
|-|-|
|after|<img width="1905" height="123" alt="image"
src="https://github.com/user-attachments/assets/bf9f492a-1d52-4cfc-9b1b-c6544a072814"
/>|

Puts 'expected' before 'received' on object diffs. The new version
matches Jest and Vitest, and I find it more intuitive:

|before|after|
|-|-|
|<img width="344" height="221" alt="image"
src="https://github.com/user-attachments/assets/44d42655-c441-411e-9b67-c0db7a5dce08"
/>|<img width="342" height="293" alt="image"
src="https://github.com/user-attachments/assets/565e3934-a2a2-4f99-9d6f-b7df1905f933"
/>|

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-06 06:44:46 -07:00
pfg
7c65c35f8f Fix expect(() => { throw undefined; }).toThrow(TypeError) (#21637)
Fixes #19107
2025-08-06 06:39:25 -07:00
Jarred Sumner
455f3a65b9 enable mimalloc simd (#21644)
### What does this PR do?

### How did you verify your code works?
2025-08-06 06:38:34 -07:00
Alistair Smith
04883a8bdc revert fe28e00d53.
This reverts commit fe28e00d53.
2025-08-05 16:10:29 -07:00
Alistair Smith
fe28e00d53 feat: add Bun.SQL API with initial SQLite support 2025-08-05 16:04:11 -07:00
robobun
25d490fb65 docs: document Atomics global support (#21625)
Co-authored-by: Claude Bot <claude-bot@bun.sh>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-05 11:48:38 -07:00
Michael H
806d6c156f add catalog support to bun (outdated|update -i) and --filter to bun update -i (#21482) 2025-08-05 05:12:22 -07:00
Jarred Sumner
0612f459a4 Tweak crash handler for linux
Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-04 23:30:46 -07:00
autofix-ci[bot]
aec3ea143f [autofix.ci] apply automated fixes 2025-08-05 05:44:07 +00:00
Claude
fdeb08847e fix: segmentation fault when using --windows-icon with multi-size ICO files
The crash occurred when resource section resizing caused buffer reallocation,
invalidating pointers to PE structures. Fixed by caching values before resize
and re-getting pointers after resize.

- Cache file_alignment and section_alignment before buffer resize
- Re-get section headers after resize to avoid use-after-free
- Add bounds checking to prevent buffer overflows
- Add sanity check for unreasonably large resource data (>10MB)
- Fix file writing to only write actual PE data, not excess buffer

Fixes crash with ICO files containing 5+ icon sizes that require
resource section expansion.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-05 07:41:30 +02:00
pfg
408fda7ad2 Continue emitting 'readable' events after pausing stdin (#17690)
Fixes #21189

`.pause()` should unref but it should still continue to emit `readable`
events (although it should not send `data` events)

also stdin.unref() should not pause input, it should only prevent stdin
from keeping the process alive.

DRAFT:

- [x] ~~this causes a bug where `process.stdin.on("readable", () => {});
process.stdin.pause()` will allow the process to exit when it
shouldn't.~~ fixed

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-04 21:04:08 -07:00
Ciro Spaciari
ed6f099e5e fix(tls) fix ciphers (#21545)
### What does this PR do?
Uses same ciphers than node.js for compatibility and do the same error
checking on empty ciphers
Fixes https://github.com/oven-sh/bun/issues/9425
Fixes https://github.com/oven-sh/bun/issues/21518
Fixes https://github.com/oven-sh/bun/issues/19859
Fixes https://github.com/oven-sh/bun/issues/18980

You can see more about redis ciphers here
https://redis.io/docs/latest/operate/rs/security/encryption/tls/ciphers/
this should fix redis related ciphers issues
### How did you verify your code works?
Tests

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-04 19:42:40 -07:00
Ciro Spaciari
258a2a2e3a fix(postgres) memory fix when connection fails sync (#21616)
### What does this PR do?
We should not call .deinit() after .toJS otherwise hasPendingActivity
will access invalid memory

### How did you verify your code works?
Test run it with debug build on macos or asan on and will catch it

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-04 19:41:20 -07:00
Jarred Sumner
4568258960 -t should not run {before,after}{Each,All} for scopes with no tests (#21602)
### What does this PR do?

Before:
```js
❯ bun test /Users/jarred/Code/bun/test/cli/test/test-filter-lifecycle.js -t "should run test"
bun test v1.2.20-canary.135 (1ac2391b)

test/cli/test/test-filter-lifecycle.js:

# Unhandled error between tests
-------------------------------
1 | // This test is intended to be able to run in Vitest and Jest.
2 | describe("top-level sibling", () => {
3 |   beforeAll(() => {
4 |     throw new Error("FAIL");
                              ^
error: FAIL
      at <anonymous> (test-filter-lifecycle.js:4:27)
      at test-filter-lifecycle.js:2:1
      at loadAndEvaluateModule (2:1)
-------------------------------

✗ top-level sibling > test
<parent beforeAll>
<beforeAll>

# Unhandled error between tests
-------------------------------
65 |     beforeEach(() => {
66 |       throw new Error("FAIL");
67 |     });
68 | 
69 |     afterEach(() => {
70 |       throw new Error("FAIL");
                                 ^
error: FAIL
      at <anonymous> (test-filter-lifecycle.js:70:29)
-------------------------------

<afterEach>
<parent afterEach>
<parent beforeEach>
<beforeEach>
<test 1>
✓ parent > should run > test [0.02ms]
<afterEach>
<parent afterEach>
<parent beforeEach>
<beforeEach>
<test 2>
✓ parent > should run > test 2 [0.02ms]

# Unhandled error between tests
-------------------------------
106 |       console.log("<beforeEach>");
107 |     });
108 | 
109 |     afterEach(() => {
110 |       if (++ran.afterEach > 2) {
111 |         throw new Error("FAIL 2");
                                      ^
error: FAIL 2
      at <anonymous> (test-filter-lifecycle.js:111:33)
-------------------------------


# Unhandled error between tests
-------------------------------
106 |       console.log("<beforeEach>");
107 |     });
108 | 
109 |     afterEach(() => {
110 |       if (++ran.afterEach > 2) {
111 |         throw new Error("FAIL 2");
                                      ^
error: FAIL 2
      at <anonymous> (test-filter-lifecycle.js:111:33)
-------------------------------

<afterAll>
<parent afterAll>

 2 pass
 3 filtered out
 1 fail
 4 errors
Ran 3 tests across 1 file. [93.00ms]
```

After:
```js
bun test <version> (<revision>)
<parent beforeAll>
<beforeAll>
<parent beforeEach>
<beforeEach>
<test 1>
<afterEach>
<parent afterEach>
<parent beforeEach>
<beforeEach>
<test 2>
<afterEach>
<parent afterEach>
<afterAll>
<parent afterAll>

test/cli/test/test-filter-lifecycle.js:
(pass) parent > should run > test
(pass) parent > should run > test 2

2 pass
4 filtered out
0 fail
Ran 2 tests across 1 file.
```

### How did you verify your code works?

There is a test

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
2025-08-04 19:30:09 -07:00
Alistair Smith
eac0dcee15 Merge branch 'main' into feat/windows-version-description 2025-08-04 17:14:09 -07:00
Jarred Sumner
9785e37e10 Deflake some CI things (#21600) 2025-08-04 07:03:40 -07:00
Jarred Sumner
e6f6162a24 Create pe-codesigning-integrity.test.ts 2025-08-04 02:46:30 -07:00
Jarred Sumner
75553543d2 Delete windows-resources-external-tools.test.ts 2025-08-04 02:43:19 -07:00
Claude
78d6119144 Fix Windows resource editing for VS_VERSIONINFO structure
- Fix extra directory level in resource tree (language entry should be data, not subdirectory)
- Rewrite VS_VERSIONINFO generation to match editpe's approach with proper length calculations
- Fix struct padding issue by writing header fields individually (6 bytes, not 8)
- Add proper alignment calculations for all string entries
- Implement correct UTF-16LE string encoding with null terminators

Now all Windows resource options work correctly:
- --windows-version sets file and product version numbers
- --windows-description sets file description
- --windows-publisher sets company name
- --windows-title sets product name
- --windows-copyright sets copyright information
- --windows-hide-console changes subsystem to Windows GUI

All tests pass and exiftool can correctly read the version information.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-03 05:31:03 +02:00
autofix-ci[bot]
e70da99349 [autofix.ci] apply automated fixes 2025-08-03 02:01:53 +00:00
Claude
4497fcf73b fix: Windows resource editing VS_VERSIONINFO structure
- Fix VersionHeader struct padding issue (was 8 bytes due to alignment, now writes 6 bytes correctly)
- Rewrite UTF-16 string encoding function for correctness
- Write VS_VERSIONINFO header fields individually to avoid struct padding
- Add comprehensive tests using exiftool for verification
- Fix file writing in StandaloneModuleGraph to properly truncate file

The VS_VERSIONINFO structure is now correctly formatted according to the PE specification.
Note: exiftool may have limitations with 64-bit PE files with high virtual addresses.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-03 03:57:55 +02:00
autofix-ci[bot]
77d26a5dad [autofix.ci] apply automated fixes 2025-08-02 08:57:50 +00:00
Claude
81351d6878 test: clean up Windows resource tests to reduce duplication and delete executables 2025-08-02 10:55:26 +02:00
Claude
d8d325f50f Fix PE file double size issue and add --windows-copyright option
- Fix PE file writing to truncate file before writing to avoid appending data
- Add --windows-copyright option for setting executable copyright information
- Add LegalCopyright field to VS_VERSIONINFO resource structure
- Update tests to verify copyright field is properly written
- Use ASCII copyright string to avoid UTF-16 encoding issues in tests
2025-08-02 10:41:27 +02:00
autofix-ci[bot]
62a3d33243 [autofix.ci] apply automated fixes 2025-08-02 08:07:56 +00:00
Claude
6fbc9626db Implement PE checksum calculation
- Add calculateChecksum() method that implements the standard Windows PE checksum algorithm
- Automatically update checksum after modifying PE files (adding .bun section, applying Windows settings)
- Delete old pe-codesigning-integrity.test.ts that didn't test anything useful
- Add comprehensive checksum verification tests using objdump

The checksum algorithm matches the Windows standard:
- Processes file as 16-bit words
- Skips the checksum field itself
- Handles overflow with carry folding
- Adds file size to final checksum
2025-08-02 10:04:20 +02:00
autofix-ci[bot]
6e5cafe3f5 [autofix.ci] apply automated fixes 2025-08-02 06:59:45 +00:00
Claude
2d6300a8c1 test: remove windows-resources-compile.test.ts
This test relied on internal APIs that are not exposed. The functionality
is now tested by windows-resources-external-tools.test.ts using external
tools like objdump and strings.
2025-08-02 08:55:33 +02:00
Claude
ea814317a4 test: add external tool verification tests for Windows resources
- Add tests using objdump to verify PE headers and resource directories
- Add tests using llvm-objdump to verify sections
- Add tests using hexdump to examine resource data
- Add tests using strings to find UTF-16LE version strings
- Add tests for cross-platform compilation with resources
- Use Bun.which to check tool availability

These tests ensure Windows resource editing works correctly by inspecting
the generated executables with standard system tools rather than relying
only on internal parsing functions.
2025-08-02 08:51:57 +02:00
Jarred Sumner
68d322f05f Fix mimalloc memory usage regression (#21550)
### What does this PR do?

### How did you verify your code works?

---------

Co-authored-by: taylor.fish <contact@taylor.fish>
2025-08-01 23:38:34 -07:00
Jarred Sumner
39eccf89a8 Deflake sql.test.ts 2025-08-01 22:41:05 -07:00
Claude
1c54474dfc feat: add Windows resource editing support for bun build --compile
This PR implements native Windows resource editing in Zig, replacing the previous
rescle C++ implementation. Users can now customize Windows executables when using
'bun build --compile' with the following new options:

- --windows-icon <path>        Set custom executable icon
- --windows-title <str>        Set executable title/product name
- --windows-publisher <str>    Set company/publisher name
- --windows-version <str>      Set version (e.g. "1.2.3.4")
- --windows-description <str>  Set executable description
- --windows-hide-console       Hide console window (already existed)

Example:
```bash
bun build --compile \
  --target=bun-windows-x64 \
  --windows-icon=app.ico \
  --windows-title="My Application" \
  --windows-publisher="My Company" \
  --windows-version="2.1.0.0" \
  --windows-description="A powerful application built with Bun" \
  index.ts
```

Implementation details:
- Pure Zig implementation in windows_resources.zig
- Removes C++ rescle dependency
- Creates WindowsSettings struct to organize all Windows options
- Supports cross-platform compilation (build Windows exe from Linux/macOS)
- Fixed alignment issues using safe memory operations
- Comprehensive test coverage in test/bundler/windows-resources-compile.test.ts

This allows full customization of Windows executable metadata without external dependencies.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>
2025-08-02 06:55:12 +02:00
Jarred Sumner
07ffde8a69 Add missing check for .write() on a data-backed blob (#21552)
### What does this PR do?

Add missing check for .write() on a data-backed blob

### How did you verify your code works?

There is a test

---------

Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Alistair Smith <hi@alistair.sh>
2025-08-01 20:04:16 -07:00