Re: [PATCH v2] location-importer.in: skip networks with unknown country codes

[Michael Tremer <michael.tremer@ipfire.org>]

[-- Attachment #1: Type: text/plain, Size: 6314 bytes --]

Hello,

I merged this patch, but it has some unwanted side-effects:

Technically it works as designed as we are successfully dropping any countries that are not part of the imported list. I changed our scripts that these will always be imported first now.

I ran a manual import which dropped CS which is Serbia and Montenegro. This used to be a valid country code, but Serbia and Montenegro is not a single country any more. I decided to add it because we would have dropped too many networks without it. Now we are dropping a few networks with country code YU - Yugoslavia.

Montenegro became independent from Serbia in 2006, Yugoslavia became the State Union of Serbia and Montenegro in 2003. For some reasons (probably because I didn’t do research) I thought these events were closer together and therefore thought that all networks with country code CS simply “forgot” to update this, but there never were any that actually existing during the time of Yugoslavia.

Long story short: Would anybody object to add YU to the database although it doesn’t exist as a country any more? I guess we cannot just “rewrite” it because the situation is way too complicated. However, we wanted to give people an idea where some IP address is located and that is kind of does not work if the country does not exist any more. Returning nothing instead is not a great solution either because we are then simply hiding networks that exist.

Or did I overlook an ever better option?

-Michael

> On 30 Mar 2021, at 16:47, Peter Müller <peter.mueller(a)ipfire.org> wrote:
> 
> There is no sense in parsing and storting networks whose country codes
> cannot be found in the ISO-3166-x country code table. This avoids side
> effects in applications using the location database, and introduces
> another sanity check to compensate bogus RIR data.
> 
> On location02, this affects some networks from APNIC (country code: ZZ)
> as well as a bunch of smaller allocations within the RIPE region still
> tagged to CS or YU (Yugoslavia). To my surprise, no network tagged as SU
> (Soviet Union) was found - while the NIC for .su TLD is still
> operational. :-)
> 
> Applying this patch causes the countries to be processed before
> update_whois() is called. In case no countries are present in the SQL
> table, this check is silently omitted.
> 
> Fixes: #12510
> 
> Signed-off-by: Peter Müller <peter.mueller(a)ipfire.org>
> ---
> src/python/location-importer.in | 38 ++++++++++++++++++++++-----------
> 1 file changed, 26 insertions(+), 12 deletions(-)
> 
> diff --git a/src/python/location-importer.in b/src/python/location-importer.in
> index e2f201b..1e08458 100644
> --- a/src/python/location-importer.in
> +++ b/src/python/location-importer.in
> @@ -388,10 +388,17 @@ class CLI(object):
> 				TRUNCATE TABLE networks;
> 			""")
> 
> +			# Fetch all valid country codes to check parsed networks aganist...
> +			rows = self.db.query("SELECT * FROM countries ORDER BY country_code")
> +			validcountries = []
> +
> +			for row in rows:
> +				validcountries.append(row.country_code)
> +
> 			for source in location.importer.WHOIS_SOURCES:
> 				with downloader.request(source, return_blocks=True) as f:
> 					for block in f:
> -						self._parse_block(block)
> +						self._parse_block(block, validcountries)
> 
> 			# Process all parsed networks from every RIR we happen to have access to,
> 			# insert the largest network chunks into the networks table immediately...
> @@ -467,7 +474,7 @@ class CLI(object):
> 				# Download data
> 				with downloader.request(source) as f:
> 					for line in f:
> -						self._parse_line(line)
> +						self._parse_line(line, validcountries)
> 
> 	def _check_parsed_network(self, network):
> 		"""
> @@ -532,7 +539,7 @@ class CLI(object):
> 		# be suitable for libloc consumption...
> 		return True
> 
> -	def _parse_block(self, block):
> +	def _parse_block(self, block, validcountries = None):
> 		# Get first line to find out what type of block this is
> 		line = block[0]
> 
> @@ -542,7 +549,7 @@ class CLI(object):
> 
> 		# inetnum
> 		if line.startswith("inet6num:") or line.startswith("inetnum:"):
> -			return self._parse_inetnum_block(block)
> +			return self._parse_inetnum_block(block, validcountries)
> 
> 		# organisation
> 		elif line.startswith("organisation:"):
> @@ -573,7 +580,7 @@ class CLI(object):
> 			autnum.get("asn"), autnum.get("org"),
> 		)
> 
> -	def _parse_inetnum_block(self, block):
> +	def _parse_inetnum_block(self, block, validcountries = None):
> 		log.debug("Parsing inetnum block:")
> 
> 		inetnum = {}
> @@ -616,10 +623,10 @@ class CLI(object):
> 		if not inetnum or not "country" in inetnum:
> 			return
> 
> -		# Skip objects with bogus country code 'ZZ'
> -		if inetnum.get("country") == "ZZ":
> -			log.warning("Skipping network with bogus country 'ZZ': %s" % \
> -				(inetnum.get("inet6num") or inetnum.get("inetnum")))
> +		# Skip objects with unknown country codes
> +		if validcountries and inetnum.get("country") not in validcountries:
> +			log.warning("Skipping network with bogus country '%s': %s" % \
> +				(inetnum.get("country"), inetnum.get("inet6num") or inetnum.get("inetnum")))
> 			return
> 
> 		# Iterate through all networks enumerated from above, check them for plausibility and insert
> @@ -652,7 +659,7 @@ class CLI(object):
> 			org.get("organisation"), org.get("org-name"),
> 		)
> 
> -	def _parse_line(self, line):
> +	def _parse_line(self, line, validcountries = None):
> 		# Skip version line
> 		if line.startswith("2"):
> 			return
> @@ -667,8 +674,15 @@ class CLI(object):
> 			log.warning("Could not parse line: %s" % line)
> 			return
> 
> -		# Skip any lines that are for stats only
> -		if country_code == "*":
> +		# Skip any lines that are for stats only or do not have a country
> +		# code at all (avoids log spam below)
> +		if not country_code or country_code == '*':
> +			return
> +
> +		# Skip objects with unknown country codes
> +		if validcountries and country_code not in validcountries:
> +			log.warning("Skipping line with bogus country '%s': %s" % \
> +				(country_code, line))
> 			return
> 
> 		if type in ("ipv6", "ipv4"):
> -- 
> 2.26.2