Bug: binary symbols appear in code-slicer output
code-slicer can sometimes output unreadable binary-looking symbols instead of clean source text.
Expected
Output should only contain readable source files.
If a file is binary, unreadable, or not valid text, code-slicer should skip it and show a warning with the file path.
Actual
The output can contain large blocks of corrupted/binary-looking characters, which makes the generated prompt context unusable.
Suggested fix
Add a guard before printing file contents:
- detect obvious binary/non-text files
- skip them
- print a warning like:
Warning: skipped unreadable or binary file: path/to/file
Acceptance criteria
- Binary/unreadable files are not printed
- A warning identifies skipped files
- Normal source files are still included
- Add a regression test for this case
Bug: binary symbols appear in code-slicer output
code-slicercan sometimes output unreadable binary-looking symbols instead of clean source text.Expected
Output should only contain readable source files.
If a file is binary, unreadable, or not valid text,
code-slicershould skip it and show a warning with the file path.Actual
The output can contain large blocks of corrupted/binary-looking characters, which makes the generated prompt context unusable.
Suggested fix
Add a guard before printing file contents:
Acceptance criteria