| from __future__ import annotations | |
| import re | |
| def classify_resume_noise(text: str) -> dict[str, object]: | |
| lower = text.lower() | |
| signals: list[str] = [] | |
| if "indeed.com/r/" in lower or "authorized to work in" in lower or "willing to relocate" in lower: | |
| signals.append("job_board_scrape") | |
| if lower.count("work experience") > 1 or lower.count("experience") >= 3 and lower.count("experience") / max(len(lower.split()), 1) > 0.02: | |
| signals.append("repeated_sections") | |
| if re.search(r"\b[a-z]{2,}\s+[a-z]{2,}\s+[a-z]{2,}\b", lower) and any(token in lower for token in [" to present", " octoberober ", " novemberember ", " junee ", " februaryruary "]): | |
| signals.append("ocr_like_dates") | |
| if re.search(r'"[^"]+"\s+internship', lower) or re.search(r'"[^"]+"', lower): | |
| signals.append("quoted_fragments") | |
| if re.search(r"\bhttps?://|www\.", lower): | |
| signals.append("embedded_urls") | |
| if len(re.findall(r"\|", text)) >= 4: | |
| signals.append("dense_delimiters") | |
| if re.search(r"\b[a-z]+/[a-z]+/[a-z]+\b", lower) or re.search(r"\b[a-z]+&[a-z]+\b", lower): | |
| signals.append("compound_tokens") | |
| if re.search(r"\bco\s+septembertember\b|\btx\s+may\b|\bto\s+present\b", lower): | |
| signals.append("span_noise") | |
| noisy = len(signals) >= 2 or "job_board_scrape" in signals or "ocr_like_dates" in signals | |
| bucket = "noisy" if noisy else "clean" | |
| return {"bucket": bucket, "signals": signals} | |