"""`pattern_matching` task generator."""
import json
import random
import shutil
from collections import defaultdict
from pathlib import Path

from ..base import (
    Generator, _VERIFY_HEADER, _render_verify, pad_to, _write, _para,
    _AUTHORS, _SONG_TITLES, _STUDENTS, _WORDS,
)


class PatternMatching(Generator):
    KEY = "pattern_matching"
    CATEGORY_NAME = "Pattern Matching"
    DIFFICULTY = "L2"
    TAGS = ["file content", "search"]

    KEYWORD = "URGENT"
    OUT = "matches.txt"

    def build(self, env_dir, llm, rng):
        snippets = llm.gen_snippets("project status logs", 5)
        files = []
        for s in snippets:
            name = s["filename"] + ".txt"
            n = rng.randint(4, 7)
            inject = []
            for _ in range(rng.randint(1, 2)):
                idx = rng.randint(0, n - 1)
                inject.append((idx, f"{self.KEYWORD}: " + " ".join(
                    rng.choice(_WORDS) for _ in range(5)).capitalize() + "."))
            lines = _para(rng, n, inject)
            _write(env_dir / name, "\n".join(lines) + "\n")
            files.append(name)
        return {"files": files, "keyword": self.KEYWORD, "out": self.OUT}

    def description(self, spec):
        return (
            "Please use FileSystem tools to finish the following task:\n\n"
            f"### Task: Find lines containing `{self.KEYWORD}`\n\n"
            f"Scan every `.txt` file in the test directory. For each line that "
            f"contains the exact (case-sensitive) substring `{self.KEYWORD}`, record "
            f"a result line in this format:\n\n"
            "```\n<filename>:<line_number>:<full_line_text>\n```\n\n"
            "- `line_number` is 1-based.\n"
            "- `full_line_text` is the matching line with trailing newline removed.\n"
            f"- Sort all result lines by filename, then by line number.\n"
            f"- Write the results to `{self.OUT}` in the test directory root "
            f"(do not include `{self.OUT}` itself in the scan)."
        )

    def verify_src(self, spec):
        body = '''
C = json.loads(__CONSTS__)


def expected(t):
    rows = []
    for f in sorted(t.iterdir()):
        if not (f.is_file() and f.suffix == ".txt" and f.name != C["OUT"]):
            continue
        for i, line in enumerate(f.read_text(encoding="utf-8").splitlines(), 1):
            if C["KEYWORD"] in line:
                rows.append((f.name, i, line))
    rows.sort(key=lambda r: (r[0], r[1]))
    return "\\n".join(f"{n}:{i}:{l}" for n, i, l in rows)


def main():
    t = get_test_dir()
    out = t / C["OUT"]
    if not out.is_file():
        fail(f"output file not found: {C['OUT']}")
    got = out.read_text(encoding="utf-8").strip()
    exp = expected(t).strip()
    if got != exp:
        print("--- expected ---"); print(exp[:800])
        print("--- got ---"); print(got[:800])
        fail("matches.txt does not match expected results")
    ok("all matching lines correctly recorded")
    print("\\U0001f389 All checks passed!")
    sys.exit(0)


if __name__ == "__main__":
    main()
'''
        return _render_verify(body, {"KEYWORD": spec["keyword"], "OUT": spec["out"]})

    def solve(self, work_dir, spec):
        rows = []
        for f in sorted(work_dir.iterdir()):
            if not (f.is_file() and f.suffix == ".txt" and f.name != spec["out"]):
                continue
            for i, line in enumerate(f.read_text(encoding="utf-8").splitlines(), 1):
                if spec["keyword"] in line:
                    rows.append((f.name, i, line))
        rows.sort(key=lambda r: (r[0], r[1]))
        _write(work_dir / spec["out"], "\n".join(f"{n}:{i}:{l}" for n, i, l in rows) + "\n")