Per spider DNS_RESOLVER doesn't work

[synodriver]

Description

The spider with following settings

class ExampleSpider(scrapy.Spider):
    name = "example"
    allowed_domains = ["example.com"]
    start_urls = ["https://example.com"]
    custom_settings = {
         "DNS_RESOLVER": "test1.resolver.CachingAsyncDohResolver"
     }

do not follow the custom_settings

Steps to Reproduce

1. Use the following code (merely an example)

import asyncio
import json
import re
import socket
from typing import List, Optional

import aiodns
import aiohttp
import scrapy.crawler
from scrapy.resolver import CachingThreadedResolver, dnscache
from scrapy.utils.defer import deferred_from_coro
from twisted.internet.interfaces import IResolverSimple
from zope.interface.declarations import implementer


@implementer(IResolverSimple)
class CachingAsyncResolver(CachingThreadedResolver):
    """
    Async caching resolver. Require aiodns
    """

    def __init__(
        self,
        reactor,
        cache_size,
        timeout,
        nameservers: Optional[List[str]] = None,
        **kwargs,
    ):
        super().__init__(reactor, cache_size, timeout)
        self._resolver = aiodns.DNSResolver(nameservers, None, **kwargs)
        self._pattern = re.compile(
            r"((\d|[1-9]\d|1\d\d|2[0-4]\d|25[0-5])\.){3}(1\d\d|2[0-4]\d|25[0-5]|[1-9]\d|\d)"
        )

    @classmethod
    def from_crawler(cls, crawler: scrapy.crawler.Crawler, reactor):
        if crawler.settings.getbool("DNSCACHE_ENABLED"):
            cache_size = crawler.settings.getint("DNSCACHE_SIZE")
        else:
            cache_size = 0
        return cls(
            reactor,
            cache_size,
            crawler.settings.getfloat("DNS_TIMEOUT"),
            crawler.settings.getlist("AIODNS_NAMESERVERS", None),
            **crawler.settings.getdict("AIODNS_KW", {}),
        )

    def getHostByName(self, name, timeout=None):
        print(f"resolving name {name}")
        return deferred_from_coro(self._getHostByName(name, timeout))

    async def _getHostByName(self, name, timeout=None):
        if self._pattern.match(name) is not None:  # just an ip
            return name
        if name in dnscache:
            return dnscache[name]
        try:
            resp = await asyncio.wait_for(
                self._resolver.gethostbyname(name, socket.AF_INET), timeout
            )
            result = resp.addresses[0]
            self._cache_result(result, name)
            return result
        except asyncio.TimeoutError:
            raise
        except aiodns.error.DNSError as exc:
            msg = exc.args[1] if len(exc.args) >= 1 else "DNS lookup failed"
            raise OSError(msg) from exc

And run spider

Expected behavior: The log by that resolver is seen

Actual behavior: The crawler didn't call that resolver

Reproduces how often: Always, both on linux and windows

Versions

Scrapy : 2.11.1
lxml : 4.9.2.0
libxml2 : 2.9.12
cssselect : 1.2.0
parsel : 1.8.1
w3lib : 2.1.1
Twisted : 24.3.0
Python : 3.10.12 | packaged by conda-forge | (main, Jun 23 2023, 22:34:57) [MSC v.1936 64 bit (AMD64)]
pyOpenSSL : 23.3.0 (OpenSSL 3.1.4 24 Oct 2023)
cryptography : 41.0.5
Platform : Windows-10-10.0.19045-SP0

Scrapy : 2.11.1
lxml : 4.9.2.0
libxml2 : 2.9.14
cssselect : 1.2.0
parsel : 1.8.1
w3lib : 2.1.1
Twisted : 24.3.0
Python : 3.10.11 (main, Apr 20 2023, 19:02:41) [GCC 11.2.0]
pyOpenSSL : 23.1.1 (OpenSSL 3.1.0 14 Mar 2023)
cryptography : 40.0.2
Platform : Linux-5.10.0-13-amd64-x86_64-with-glibc2.31

Additional context

It works if you put it into global settings.py

[wRAR]

Duplicate of #5988