Code fixer corrupts file by appending trailing characters to comment lines

[thomhurst]

Description

The NUnit migration code fixer (TUNU0001) occasionally corrupts source files by appending trailing characters to lines, particularly comment lines. This appears to be a text manipulation bug where characters from elsewhere in the file get appended to unrelated lines.

Reproduction Steps

This issue was discovered during a large-scale migration. After running:

dotnet format analyzers --severity info --diagnostics TUNU0001

A license header comment was corrupted:

Before:

// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and

After:

// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.f
// See the License for the specific language governing permissions and

Note the trailing f character that was appended to line 12.

Expected Behavior

The code fixer should only modify code related to NUnit-to-TUnit migration. Comment lines and unrelated code should remain unchanged.

Environment

• TUnit Version: 1.11.64
• .NET SDK: 10.0.100-preview.1.25103.12
• OS: Windows 11

Additional Context

This appears to be related to other text manipulation bugs reported:

• TUNU0001 code fixer corrupts code when #region directive follows a comment #4487: Code fixer corrupts code when #region directive follows a comment
• TUNU0001 code fixer leaves orphaned #endif when #if directive is removed #4488: Code fixer leaves orphaned #endif when #if directive is removed

These issues suggest the code fixer's text manipulation logic has edge cases that cause character displacement or corruption when processing complex files.

The corruption is subtle and easy to miss during code review, potentially causing issues later.

---

Note: Please add tests when fixing this issue to prevent regression. Consider adding tests that verify the fixer doesn't modify lines it shouldn't touch.

[claude]

This issue is related to other text manipulation bugs in the TUNU0001 code fixer:

• TUNU0001 code fixer corrupts code when #region directive follows a comment #4487: Code fixer corrupts code when #region directive follows a comment
• TUNU0001 code fixer leaves orphaned #endif when #if directive is removed #4488: Code fixer leaves orphaned #endif when #if directive is removed

All three issues suggest underlying problems with the code fixer's text manipulation logic, though each has distinct symptoms and likely different trigger conditions. These should be investigated together to identify if there's a common root cause in how the fixer handles text edits.