package groups

import (
	"embed"
	"strings"

	"git.kplus.net.ua/yevhen/resource-scraper/internal/releaseparser"
)

// The token dictionaries live under resources/dicts/*.txt — one token (in any
// case) per line; blank lines and lines starting with "#" are ignored. They
// can be edited without recompiling the parsing logic: go:embed pulls them
// into the binary at build time.

//go:embed resources/dicts/*.txt
var dictFS embed.FS

func loadDict(name string) map[string]bool {
	data, err := dictFS.ReadFile("resources/dicts/" + name + ".txt")
	if err != nil {
		panic("release parser: cannot load dictionary " + name + ": " + err.Error())
	}
	out := make(map[string]bool)
	for line := range strings.SplitSeq(string(data), "\n") {
		line = strings.ToUpper(strings.TrimSpace(line))
		if line == "" || strings.HasPrefix(line, "#") {
			continue
		}
		out[line] = true
	}
	return out
}

// Profile holds the shared dictionaries for the release parser. The parser is
// group-agnostic: the group name is read from each release line, never from
// this profile, so there is a single set of dictionaries covering all groups.
var Profile = releaseparser.GroupProfile{
	FormatTokens:     loadDict("format"),
	SourceTokens:     loadDict("source"),
	BitDepthTokens:   loadDict("bitdepth"),
	ResolutionTokens: loadDict("resolution"),
	LanguageMarkers:  loadDict("language"),
	EditionTokens:    loadDict("edition"),
	CommentTokens:    loadDict("comment"),
}