pasteglint

Find the character that makes copied text behave differently from what you see. Zero runtime dependencies. Python 3.10+. MIT licensed.

Install

python -m pip install pasteglint

For a local source checkout, run python -m pip install . in this directory.

Command line

pasteglint check snippet.py
pasteglint check snippet.py --json
pasteglint diff expected.txt actual.txt
pasteglint check -

An unexpected space might produce:

4:8 U+00A0 NO-BREAK SPACE [space]; possible replacement: ' '

Input files and standard input are UTF-8. A byte-order mark is kept and reported. Newline differences are preserved. A dash means standard input; only one input can use it at a time. All commands also work as python -m pasteglint.

Python

from pasteglint import inspect_text, compare_text

(finding,) = inspect_text("x\u00a0= 1")
assert finding.codepoint == "U+00A0"
assert (finding.line, finding.column) == (1, 2)

changes = compare_text("API_KEY", "\uff21PI_KEY")
assert changes[0].left == "A"
assert changes[0].right == "\uff21"

inspect_text(text) returns immutable Finding objects with index, line, column, codepoint, name, kind and an optional replacement. It reports selected unusual spaces, smart quotes, dashes, fullwidth ASCII, Unicode format characters, unusual controls, and Unicode line/paragraph separators.

compare_text(left, right) returns differing spans and their exact text. Spans are zero-based and end-exclusive. Lines and columns are one-based Unicode code-point positions, not byte offsets or rendered character widths.

Findings are observations: joiners, direction markers and curly punctuation can be legitimate. Nothing is changed automatically. This is a debugging utility, not a complete Unicode security scanner or programming-language parser. Exact diff can be slow on very large or repetitive files; use it for text snippets.

Exit codes

• 0: no configured findings, or identical texts.
• 1: findings or differences.
• 2: invalid input, unreadable files, or invalid UTF-8.

Development

python -m pip install -e .
python -m unittest discover -s tests -v