import re
import sys

from urllib.parse import urlparse

# Increase the compiled regexp cache limit from 512 to 4096 in case
# someone uses a lot of ignores.
re._MAXCACHE = 4096


class Ignoracle(object):
	"""
	An Ignoracle tests a URL against a list of patterns and returns whether or
	not that URL should be grabbed.

	An Ignoracle's pattern list starts as the empty list.
	"""

	patterns = []

	def set_patterns(self, strings):
		"""
		Given a list of strings, replaces this Ignoracle's pattern state with
		that list.
		"""
		self.patterns = []

		for string in strings:
			if isinstance(string, bytes):
				string = string.decode('utf-8')

			self.patterns.append(string)

	def ignores(self, url, **kwargs):
		"""
		If an ignore pattern matches the given URL, returns that pattern as a string.
		Otherwise, returns False.
		"""
		pu = re.escape(kwargs.get('primary_url', ''))
		ph = re.escape(kwargs.get('primary_netloc', ''))

		for pattern in self.patterns:
			regexp = pattern
			if '{' in regexp:
				regexp = regexp.replace('{primary_url}', pu).replace('{primary_netloc}', ph)
			try:
				if re.search(regexp, url):
					return pattern
			except re.error as error:
				print('Pattern %s is invalid (error: %s).  Ignored.' % (pattern, str(error)), file=sys.stderr)

		return False


def parameterize_record_info(record_info):
	"""
	Given a wpull record_info dict, generates a dict with primary_url and
	primary_netloc keys.  This is meant to be used in Ignoracle.ignores.

	The primary_url key is:

	1. record_info['top_url'], or
	2. record_info['url'] if record_info['level'] is zero, or
	3. None otherwise.

	If primary_url is a valid URL, the primary_netloc key is the network
	location component of primary_url (i.e. for HTTP,
	[user:password@]host[:port]).  Otherwise, primary_netloc is None.
	"""
	primary_url = None
	primary_netloc = None

	if record_info.get('level') == 0:
		primary_url = record_info.get('url')
	else:
		primary_url = record_info.get('top_url')

	if primary_url:
		parsed = urlparse(primary_url)
		primary_netloc = parsed.netloc

	return dict(
		primary_url=primary_url,
		primary_netloc=primary_netloc
	)