#!/usr/bin/env python3
"""Builds the Unicode emoji palette that the shared emoji picker reads.

Input is Unicode's emoji-test.txt (UTS #51), which lists every RGI emoji in
CLDR keyboard order, by group. Output is a small tab-separated file under
commonsUI's compose resources:

    #<license header lines>
    @<group name>
    <emoji>\t<name>[\t<5 skin-tone variants, space separated>]

Only fully-qualified emoji up to --max-version are kept, so the palette does not
offer glyphs the desktop's system emoji font (Segoe UI Emoji, Apple Color Emoji,
Noto Color Emoji) is still too old to draw. Skin-tone variants are folded into
their base emoji: an entry carries the five uniform-tone variants
("<name>: light skin tone", ...) when all five exist, and every other tone
combination is left out. The Component group (bare skin tones, hair) is dropped.

Usage:
    curl -sSO https://unicode.org/Public/emoji/latest/emoji-test.txt
    python3 tools/emoji-data/generate.py emoji-test.txt \
        commonsUI/src/commonMain/composeResources/files/emoji/emoji.tsv
"""

import argparse
import re

TONES = ["light", "medium-light", "medium", "medium-dark", "dark"]
LINE = re.compile(r"^([0-9A-F ]+?)\s*;\s*([a-z-]+)\s*#\s*(\S+)\s+E(\d+\.\d+)\s+(.+)$")

LICENSE = [
    "Generated by tools/emoji-data/generate.py from Unicode emoji-test.txt, version {version}.",
    "Copyright (c) 1991-2026 Unicode, Inc. Distributed under the Unicode License v3:",
    "https://www.unicode.org/license.txt",
]


def version_key(v):
    major, minor = v.split(".")
    return (int(major), int(minor))


def main():
    parser = argparse.ArgumentParser()
    parser.add_argument("source")
    parser.add_argument("output")
    parser.add_argument("--max-version", default="15.0")
    args = parser.parse_args()
    max_version = version_key(args.max_version)

    groups = []  # [(name, [(emoji, name)])]
    by_name = {}  # name -> emoji, for the tone lookup
    source_version = "?"
    group = None

    with open(args.source, encoding="utf-8") as f:
        for raw in f:
            line = raw.rstrip("\n")
            if line.startswith("# Version:"):
                source_version = line.split(":", 1)[1].strip()
                continue
            if line.startswith("# group:"):
                name = line.split(":", 1)[1].strip()
                group = None if name == "Component" else (name, [])
                if group:
                    groups.append(group)
                continue
            m = LINE.match(line)
            if not m or group is None:
                continue
            _, status, emoji, version, name = m.groups()
            if status != "fully-qualified" or version_key(version) > max_version:
                continue
            by_name[name] = emoji
            group[1].append((emoji, name))

    out = ["# " + l.format(version=source_version) for l in LICENSE]
    count = 0
    for group_name, entries in groups:
        out.append("@" + group_name)
        for emoji, name in entries:
            if " skin tone" in name:
                continue  # folded into its base entry below, or dropped
            tones = [by_name.get(f"{name}: {t} skin tone") for t in TONES]
            row = [emoji, name]
            if all(tones):
                row.append(" ".join(tones))
            out.append("\t".join(row))
            count += 1

    with open(args.output, "w", encoding="utf-8") as f:
        f.write("\n".join(out) + "\n")
    print(f"wrote {count} emoji in {len(groups)} groups to {args.output}")


if __name__ == "__main__":
    main()