#!/usr/bin/env python3
"""Turn a raw mvn spring-boot:run capture into a clean, reproducible transcript.

Strips JVM/build noise (sun.misc.Unsafe warnings, JAVA_TOOL_OPTIONS proxy banners) and the
duplicate un-prefixed echo of each SQL statement that org.hibernate.SQL's show_sql=true prints
to stdout in addition to the "Hibernate: ..." line the logger emits -- same text twice, so only
the logger-prefixed copy is kept.
"""
import re
import sys

DROP_PREFIXES = (
    "WARNING:",
    "Picked up JAVA_TOOL_OPTIONS",
)


def clean(lines):
    out = []
    for line in lines:
        stripped = line.rstrip("\n")
        if any(stripped.startswith(p) for p in DROP_PREFIXES):
            continue
        # Drop the bare SQL echo line that show_sql=true prints without the "Hibernate: " prefix
        # -- it is always immediately followed by the same text WITH the prefix.
        out.append(stripped)
    deduped = []
    i = 0
    while i < len(out):
        cur = out[i]
        nxt = out[i + 1] if i + 1 < len(out) else None
        if nxt is not None and nxt == "Hibernate: " + cur:
            i += 1  # skip the bare echo, keep the prefixed one on the next iteration
            continue
        deduped.append(out[i])
        i += 1
    return deduped


if __name__ == "__main__":
    src, dst = sys.argv[1], sys.argv[2]
    with open(src, encoding="utf-8") as f:
        lines = f.readlines()
    cleaned = clean(lines)
    with open(dst, "w", encoding="utf-8") as f:
        f.write("\n".join(cleaned) + "\n")
    print(f"{src}: {len(lines)} -> {dst}: {len(cleaned)} lines")