Method _extract_from_embedded_json

IntelOSINT.py:14724–14768

(html_text)

                            return ""

                        def _extract_from_embedded_json(html_text):
                            script_blobs = re.findall(r"<script[^>]*>(.*?)</script>", html_text, flags=re.DOTALL | re.IGNORECASE)
                            prefixes = (
                                "window.__NUXT__",
                                "__NUXT__",
                                "window.__INITIAL_STATE__",
                                "__NEXT_DATA__",
                                "window.__NEXT_DATA__",
                                "window.__INITIAL_DATA__",
                            )
                            candidates = []
                            for s in script_blobs:
                                if not s:
                                    continue
                                for pfx in prefixes:
                                    pattern = pfx + r"\s*=\s*"
                                    for m in re.finditer(pattern, s, flags=re.IGNORECASE):
                                        start = m.end()
                                        js = s[start:start + 50000]
                                        obj_txt = _extract_json_obj(js, 0)
                                        if obj_txt:
                                            candidates.append(obj_txt)
                                        if pfx == "__NUXT__" or pfx == "window.__NUXT__":
                                            # nel caso in cui il JSON sia racchiuso in json stringa di nuxt
                                            if "window.__NUXT__=" in pfx:
                                                pass
                                        if "=" in pfx:
                                            continue

                            # fallback anche a oggetti "id/name/username" serializzati in script plain
                            for s in script_blobs:
                                if '"username"' in s and '"id"' in s and ('"avatar"' in s or '"header"' in s):
                                    for m in re.finditer(r"\{.*?\}", s, flags=re.DOTALL):
                                        obj_txt = m.group(0)
                                        if len(obj_txt) < 120:
                                            continue
                                        candidates.append(obj_txt)

                            parsed = []
                            for txt in candidates:
                                try:
                                    parsed.append(json.loads(txt))
                                except Exception:
                                    pass
                            return parsed

                        def _walk_and_pick(obj):
                            if not isinstance(obj, (dict, list)):