René's URL Explorer Experiment

{"@context":"https://schema.org","@type":"DiscussionForumPosting","headline":"Invisible bidi control characters survive parsing and pollute output","articleBody":"Text copied from web pages, PDFs, or Word documents frequently carries invisible Unicode bidirectional control characters — most commonly the right-to-left mark U+200F around RTL names. The parser doesn't strip them, so they end up embedded in the parsed attributes:\n\n```python\n\u003e\u003e\u003e from nameparser import HumanName\n\u003e\u003e\u003e hn = HumanName(\"\\u200fمحمد بن سلمان\\u200f\")\n\u003e\u003e\u003e hn.first\n'\\u200fمحمد'\n\u003e\u003e\u003e hn.first == \"محمد\"\nFalse\n```\n\n(The `\\u200f` escapes are written out here because the character is invisible; real-world input contains the raw character.)\n\nBecause the characters are invisible, the corruption is silent: values print normally but fail equality comparisons, lookups, and deduplication downstream. Constants lookups can also miss (a title or prefix with a mark glued to it no longer matches its set entry).\n\n**Suggested fix:** strip bidi control characters during pre-processing, exactly as emoji are already handled (#58, `re_emoji` in `nameparser/config/regexes.py`). The relevant characters: U+200E (LRM), U+200F (RLM), U+061C (ALM), U+202A–U+202E (embedding/override), U+2066–U+2069 (isolates).\n\nObserved on v1.3.0.\n","author":{"url":"https://github.com/derek73","@type":"Person","name":"derek73"},"datePublished":"2026-07-07T20:43:30.000Z","interactionStatistic":{"@type":"InteractionCounter","interactionType":"https://schema.org/CommentAction","userInteractionCount":0},"url":"https://github.com/266/python-nameparser/issues/266"}