Context:
https://groups.google.com/a/chromium.org/d/msg/blink-dev/eu57SCNltls/jdSLfwmQBQAJ
SELECT * FROM (
SELECT page, url, REGEXP_EXTRACT(LOWER(body), r'(<script\s[^>]+>)') AS match
FROM [httparchive:har.2016_06_15_chrome_requests_bodies]
) WHERE REGEXP_MATCH(match, r"\ssrc\s*=") AND
REGEXP_MATCH(match, r"\s(language|type)\s*=") AND
NOT REGEXP_MATCH(match, r'\stype\s*=\s*\\*(["\']|u0022)?\s*(text|application)\\*/(x-)?(java|ecma|j|live)script(1\.\d)?') AND
NOT REGEXP_MATCH(match, r'\slanguage\s*=\s*\\*(["\']|u0022)?(x-)?(java|ecma|j|live)script(1\.\d)?')
973 matches. (Dataset is 496,827 pages; 17,661,140 resources.)