Add CLI script for resume text/link extraction

scripts/resume_text_extractor.py

@@ -0,0 +1,137 @@
+#!/usr/bin/env python3
+"""Command-line helper to exercise resume text and link extraction."""
+
+from __future__ import annotations
+
+import argparse
+import json
+from dataclasses import dataclass
+from pathlib import Path
+import sys
+from typing import Any
+
+from five08.document_text import extract_document_text
+from five08.resume_extractor import ResumeProfileExtractor
+
+
+@dataclass
+class FileResult:
+    path: str
+    text_length: int
+    extracted_text: str
+    links: list[tuple[str, float]]
+
+
+def _parse_args() -> argparse.Namespace:
+    parser = argparse.ArgumentParser(
+        description=(
+            "Extract resume text with shared helpers and print the links discovered "
+            "by the current resume URL extraction logic."
+        )
+    )
+    parser.add_argument(
+        "files",
+        nargs="+",
+        metavar="FILE",
+        help="One or more resume files to extract.",
+    )
+    parser.add_argument(
+        "--text-max",
+        type=int,
+        default=4000,
+        help="Max characters of extracted text to print. Set to 0 for full text.",
+    )
+    parser.add_argument(
+        "--json",
+        action="store_true",
+        help="Print machine-readable JSON output instead of pretty text.",
+    )
+    return parser.parse_args()
+
+
+def _shorten_text(text: str, limit: int) -> str:
+    if limit <= 0 or len(text) <= limit:
+        return text
+    omitted = len(text) - limit
+    return f"{text[:limit]}\n... ({omitted} more chars omitted)"
+
+
+def extract_website_link_candidates(text: str) -> list[tuple[str, float]]:
+    """Extract normalized website links and confidence scores from resume text."""
+    return ResumeProfileExtractor._extract_website_link_candidates(text)
+
+
+def _process_file(path: Path) -> FileResult:
+    raw = path.read_bytes()
+    text = extract_document_text(raw, filename=path.name)
+    link_candidates = extract_website_link_candidates(text)
+    return FileResult(
+        path=str(path),
+        text_length=len(text),
+        extracted_text=text,
+        links=link_candidates,
+    )
+
+
+def _print_pretty(results: list[FileResult], text_max: int) -> None:
+    for index, result in enumerate(results, start=1):
+        print(f"\n[{index}] {result.path}")
+        print(f"  extracted_text_len={result.text_length}")
+        print("  extracted_text:")
+        print(_shorten_text(result.extracted_text, text_max))
+
+        if result.links:
+            print("  extracted_links:")
+            for link, confidence in result.links:
+                print(f"    - {link} (confidence={confidence:.3f})")
+        else:
+            print("  extracted_links: []")
+
+
+def _print_json(results: list[FileResult], text_max: int) -> None:
+    payload: list[dict[str, Any]] = []
+    for result in results:
+        payload.append(
+            {
+                "path": result.path,
+                "extracted_text_length": result.text_length,
+                "extracted_text": _shorten_text(result.extracted_text, text_max),
+                "extracted_links": [
+                    {"url": url, "confidence": confidence}
+                    for url, confidence in result.links
+                ],
+            }
+        )
+    print(json.dumps(payload, indent=2))
+
+
+def main() -> None:
+    args = _parse_args()
+    results: list[FileResult] = []
+    exit_code = 0
+
+    for file_arg in args.files:
+        path = Path(file_arg)
+        if not path.is_file():
+            print(f"error: missing file: {path}", file=sys.stderr)
+            exit_code = 1
+            continue
+        try:
+            results.append(_process_file(path))
+        except Exception as exc:  # pragma: no cover - runtime diagnostic path
+            print(f"error: failed to process {path}: {exc}", file=sys.stderr)
+            exit_code = 1
+
+    if not results:
+        raise SystemExit(exit_code or 1)
+
+    if args.json:
+        _print_json(results, args.text_max)
+    else:
+        _print_pretty(results, args.text_max)
+
+    raise SystemExit(exit_code)
+
+
+if __name__ == "__main__":
+    main()