Conversation
Contributor
Author
|
cc @TomWright if you are curious |
There was a problem hiding this comment.
Pull request overview
This PR aims to harden the internal libyaml scanner/emitter against malformed or truncated UTF-8 so that invalid input returns an error instead of panicking (Fixes #358).
Changes:
- Add predicate-side bounds handling for UTF-8 lookahead in
internal/libyaml/scanner.goand expand predicate test vectors inscanner.yaml. - Add an incomplete UTF-8 sequence check in the emitter’s double-quoted scalar writer to return a clean error instead of panicking.
- Add new scanner/emitter tests intended to validate “no panic” behavior on truncated UTF-8.
Reviewed changes
Copilot reviewed 5 out of 5 changed files in this pull request and generated 7 comments.
Show a summary per file
| File | Description |
|---|---|
| internal/libyaml/testdata/scanner.yaml | Adds new predicate test cases for truncated/edge UTF-8 sequences and adjusts some existing predicate inputs. |
| internal/libyaml/scanner.go | Introduces byteAt and refactors several predicates to avoid out-of-bounds reads. |
| internal/libyaml/scanner_test.go | Adds targeted tests to ensure malformed UTF-8 at EOF and predicate calls do not panic. |
| internal/libyaml/emitter.go | Adds a length guard for truncated UTF-8 sequences while escaping in double-quoted scalars. |
| internal/libyaml/emitter_test.go | Adds emitter tests intended to ensure malformed UTF-8 scalars don’t panic and return errors. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
ccoVeille
force-pushed
the
fix-unicode-error
branch
from
May 25, 2026 16:51
556a6fd to
f4d8fd6
Compare
ccoVeille
commented
May 25, 2026
ingydotnet
force-pushed
the
fix-unicode-error
branch
from
May 27, 2026 19:48
f4d8fd6 to
d3beb6d
Compare
Add security against panic with incomplete multi-byte UTF-8 input so malformed data returns an error instead of panicking.
ccoVeille
force-pushed
the
fix-unicode-error
branch
from
June 14, 2026 15:52
d3beb6d to
3034a7c
Compare
Comment on lines
629
to
+632
| func isPrintable(b []byte, i int) bool { | ||
| return ((b[i] == 0x0A) || // . == #x0A | ||
| (b[i] >= 0x20 && b[i] <= 0x7E) || // #x20 <= . <= #x7E | ||
| (b[i] == 0xC2 && b[i+1] >= 0xA0) || // #0xA0 <= . <= #xD7FF | ||
| (b[i] > 0xC2 && b[i] < 0xED) || | ||
| (b[i] == 0xED && b[i+1] < 0xA0) || | ||
| (b[i] == 0xEE) || | ||
| (b[i] == 0xEF && // #xE000 <= . <= #xFFFD | ||
| !(b[i+1] == 0xBB && b[i+2] == 0xBF) && // && . != #xFEFF | ||
| !(b[i+1] == 0xBF && (b[i+2] == 0xBE || b[i+2] == 0xBF)))) | ||
| c0 := b[i] | ||
| c1 := byteAt(b, i+1) | ||
| c2 := byteAt(b, i+2) |
Comment on lines
751
to
753
| // "? " | ||
| case '?': | ||
| return isBlankOrZero(b, i+1) |
Contributor
Author
There was a problem hiding this comment.
I'm amused by development by the fact it's always the same.
- You take something that is supposed to work.
- You find one small bug, you fix it.
- And then other issues appear all around 😅🤣
At least Copilot helps here. I'm unsure a human would have noticed BT reviewing the PR, as this code was unchanged.
Comment on lines
+58
to
+74
| // isPrintable should not panic on truncated UTF-8 sequences | ||
| notPanic("isPrintable with 0xF0", func() { | ||
| _ = isPrintable([]byte{0xF0}, 0) | ||
| }) | ||
|
|
||
| // Test other predicates with truncated sequences | ||
| notPanic("isLineBreak with 0xC2", func() { | ||
| _ = isLineBreak([]byte{0xC2}, 0) | ||
| }) | ||
|
|
||
| notPanic("isBOM with truncated", func() { | ||
| _ = isBOM([]byte{0xEF, 0xBB}, 0) | ||
| }) | ||
|
|
||
| notPanic("isEndOfScalarInFlowContentChar", func() { | ||
| _ = isEndOfScalarInFlowContentChar([]byte{':'}, 0) | ||
| }) |
Comment on lines
+58
to
+63
| for name, malformed := range map[string][]byte{ | ||
| "Incomplete 2-byte UTF-8 sequence": {0xC2}, | ||
| "Incomplete 3-byte UTF-8 sequence": {0xEF}, | ||
| "Incomplete 4-byte UTF-8 sequence": {0xF0}, | ||
| "truncated BOM sequence": {0xEF, 0xBB}, | ||
| } { |
Comment on lines
+99
to
+100
| t.Run(name, func(t *testing.T) { | ||
| t.Run("double-quoted scalar style", func(t *testing.T) { |
colinjlacy
reviewed
Jun 14, 2026
Contributor
|
@ccoVeille I'm definitely not an expert on UTF-8 sequences, but the suggestions made by copilot regarding additional 3-byte sequences seem legit. Do you want to handle those too? |
Contributor
Author
|
While #422 was AI slop, the approach of using this is interesting // remaining reports whether b[i:] holds at least n bytes.
func remaining(b []byte, i, n int) bool {
return i >= 0 && n >= 0 && i+n <= len(b)
}For example - (b[i] == 0xED && b[i+1] < 0xA0) ||
+ (b[i] == 0xED && remaining(b, i, 2) && b[i+1] < 0xA0) || |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add security against panic with incomplete multi-byte UTF-8 input
so malformed data returns an error instead of panicking.
Fixes #358