package main

import (
	"regexp"
	"sort"
)

func cleanBadDomains(domains []string) []string {
	// remove duplicates
	total := len(domains)
	all := make(map[string]bool)
	list := []string{}
	for _, item := range domains {
		if _, value := all[item]; !value {
			all[item] = true
			list = append(list, item)
		}
	}
	domains = list
	cfg.Log.Info("hosts removed from blocklist", "reason", "duplicate", "hosts", total-len(domains))

	// remove hosts that are too long
	total = len(domains)
	list = []string{}
	for _, blocklistItem := range domains {
		if len([]rune(blocklistItem)) > 240 {
			continue
		}
		list = append(list, blocklistItem)
	}
	domains = list
	cfg.Log.Info("hosts removed from blocklist", "reason", "too many characters", "hosts", total-len(domains))

	// remove allow-listed matches
	total = len(domains)

	// filter out bad regex
	goodAllowedItemList := []string{}
	for _, allowedItem := range cfg.ConfigFile.AllowLists {
		_, err := regexp.Compile(allowedItem)
		if err != nil {
			cfg.Log.Error("unable to parse allow list item", "error", err, "regex", allowedItem)
			continue
		}
		goodAllowedItemList = append(goodAllowedItemList, allowedItem)
	}

	list = []string{}
	for _, v := range domains {
		addEntry := true
		for _, allowedItem := range goodAllowedItemList {
			if regexp.MustCompile(allowedItem).MatchString(v) {
				cfg.Log.Debug("hosts removed from blocklist", "reason", "allowed host", "match string", allowedItem, "host", v)
				addEntry = false
			}
		}
		if addEntry {
			list = append(list, v)
		}
	}
	domains = list
	cfg.Log.Info("hosts removed from blocklist", "hosts", total-len(domains))

	cfg.Log.Info("total domains in list", "hosts", len(domains))
	sort.Strings(domains)
	return domains
}