//* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */ /* This Source Code Form is subject to the terms of the Mozilla Public * License, v. 2.0. If a copy of the MPL was not distributed with this * file, You can obtain one at http://mozilla.org/MPL/2.0/. */ // This service reads a file of rules describing TLD-like domain names. For a // complete description of the expected file format and parsing rules, see // http://wiki.mozilla.org/Gecko:Effective_TLD_Service #include "mozilla/Util.h" #include "nsEffectiveTLDService.h" #include "nsIIDNService.h" #include "nsNetUtil.h" #include "prnetdb.h" #include "mozilla/FunctionTimer.h" using namespace mozilla; NS_IMPL_ISUPPORTS1(nsEffectiveTLDService, nsIEffectiveTLDService) // ---------------------------------------------------------------------- static const ETLDEntry gEntries[] = #include "etld_data.inc" ; // ---------------------------------------------------------------------- nsresult nsEffectiveTLDService::Init() { NS_TIME_FUNCTION; // We'll probably have to rehash at least once, since nsTHashtable doesn't // use a perfect hash, but at least we'll save a few rehashes along the way. // Next optimization here is to precompute the hash using something like // gperf, but one step at a time. :-) mHash.Init(ArrayLength(gEntries) - 1); nsresult rv; mIDNService = do_GetService(NS_IDNSERVICE_CONTRACTID, &rv); if (NS_FAILED(rv)) return rv; // Initialize eTLD hash from static array for (PRUint32 i = 0; i < ArrayLength(gEntries) - 1; i++) { #ifdef DEBUG nsDependentCString name(gEntries[i].domain); nsCAutoString normalizedName(gEntries[i].domain); NS_ASSERTION(NS_SUCCEEDED(NormalizeHostname(normalizedName)), "normalization failure!"); NS_ASSERTION(name.Equals(normalizedName), "domain not normalized!"); #endif nsDomainEntry *entry = mHash.PutEntry(gEntries[i].domain); NS_ENSURE_TRUE(entry, NS_ERROR_OUT_OF_MEMORY); entry->SetData(&gEntries[i]); } return NS_OK; } // External function for dealing with URI's correctly. // Pulls out the host portion from an nsIURI, and calls through to // GetPublicSuffixFromHost(). NS_IMETHODIMP nsEffectiveTLDService::GetPublicSuffix(nsIURI *aURI, nsACString &aPublicSuffix) { NS_ENSURE_ARG_POINTER(aURI); nsCOMPtr innerURI = NS_GetInnermostURI(aURI); NS_ENSURE_ARG_POINTER(innerURI); nsCAutoString host; nsresult rv = innerURI->GetAsciiHost(host); if (NS_FAILED(rv)) return rv; return GetBaseDomainInternal(host, 0, aPublicSuffix); } // External function for dealing with URI's correctly. // Pulls out the host portion from an nsIURI, and calls through to // GetBaseDomainFromHost(). NS_IMETHODIMP nsEffectiveTLDService::GetBaseDomain(nsIURI *aURI, PRUint32 aAdditionalParts, nsACString &aBaseDomain) { NS_ENSURE_ARG_POINTER(aURI); nsCOMPtr innerURI = NS_GetInnermostURI(aURI); NS_ENSURE_ARG_POINTER(innerURI); nsCAutoString host; nsresult rv = innerURI->GetAsciiHost(host); if (NS_FAILED(rv)) return rv; return GetBaseDomainInternal(host, aAdditionalParts + 1, aBaseDomain); } // External function for dealing with a host string directly: finds the public // suffix (e.g. co.uk) for the given hostname. See GetBaseDomainInternal(). NS_IMETHODIMP nsEffectiveTLDService::GetPublicSuffixFromHost(const nsACString &aHostname, nsACString &aPublicSuffix) { // Create a mutable copy of the hostname and normalize it to ACE. // This will fail if the hostname includes invalid characters. nsCAutoString normHostname(aHostname); nsresult rv = NormalizeHostname(normHostname); if (NS_FAILED(rv)) return rv; return GetBaseDomainInternal(normHostname, 0, aPublicSuffix); } // External function for dealing with a host string directly: finds the base // domain (e.g. www.co.uk) for the given hostname and number of subdomain parts // requested. See GetBaseDomainInternal(). NS_IMETHODIMP nsEffectiveTLDService::GetBaseDomainFromHost(const nsACString &aHostname, PRUint32 aAdditionalParts, nsACString &aBaseDomain) { // Create a mutable copy of the hostname and normalize it to ACE. // This will fail if the hostname includes invalid characters. nsCAutoString normHostname(aHostname); nsresult rv = NormalizeHostname(normHostname); if (NS_FAILED(rv)) return rv; return GetBaseDomainInternal(normHostname, aAdditionalParts + 1, aBaseDomain); } // Finds the base domain for a host, with requested number of additional parts. // This will fail, generating an error, if the host is an IPv4/IPv6 address, // if more subdomain parts are requested than are available, or if the hostname // includes characters that are not valid in a URL. Normalization is performed // on the host string and the result will be in UTF8. nsresult nsEffectiveTLDService::GetBaseDomainInternal(nsCString &aHostname, PRUint32 aAdditionalParts, nsACString &aBaseDomain) { if (aHostname.IsEmpty()) return NS_ERROR_INSUFFICIENT_DOMAIN_LEVELS; // chomp any trailing dot, and keep track of it for later bool trailingDot = aHostname.Last() == '.'; if (trailingDot) aHostname.Truncate(aHostname.Length() - 1); // check the edge cases of the host being '.' or having a second trailing '.', // since subsequent checks won't catch it. if (aHostname.IsEmpty() || aHostname.Last() == '.') return NS_ERROR_INVALID_ARG; // Check if we're dealing with an IPv4/IPv6 hostname, and return PRNetAddr addr; PRStatus result = PR_StringToNetAddr(aHostname.get(), &addr); if (result == PR_SUCCESS) return NS_ERROR_HOST_IS_IP_ADDRESS; // Walk up the domain tree, most specific to least specific, // looking for matches at each level. Note that a given level may // have multiple attributes (e.g. IsWild() and IsNormal()). const char *prevDomain = nsnull; const char *currDomain = aHostname.get(); const char *nextDot = strchr(currDomain, '.'); const char *end = currDomain + aHostname.Length(); const char *eTLD = currDomain; while (1) { // sanity check the string we're about to look up: it should not begin with // a '.'; this would mean the hostname began with a '.' or had an // embedded '..' sequence. if (*currDomain == '.') return NS_ERROR_INVALID_ARG; // perform the hash lookup. nsDomainEntry *entry = mHash.GetEntry(currDomain); if (entry) { if (entry->IsWild() && prevDomain) { // wildcard rules imply an eTLD one level inferior to the match. eTLD = prevDomain; break; } else if (entry->IsNormal() || !nextDot) { // specific match, or we've hit the top domain level eTLD = currDomain; break; } else if (entry->IsException()) { // exception rules imply an eTLD one level superior to the match. eTLD = nextDot + 1; break; } } if (!nextDot) { // we've hit the top domain level; use it by default. eTLD = currDomain; break; } prevDomain = currDomain; currDomain = nextDot + 1; nextDot = strchr(currDomain, '.'); } // count off the number of requested domains. const char *begin = aHostname.get(); const char *iter = eTLD; while (1) { if (iter == begin) break; if (*(--iter) == '.' && aAdditionalParts-- == 0) { ++iter; ++aAdditionalParts; break; } } if (aAdditionalParts != 0) return NS_ERROR_INSUFFICIENT_DOMAIN_LEVELS; aBaseDomain = Substring(iter, end); // add on the trailing dot, if applicable if (trailingDot) aBaseDomain.Append('.'); return NS_OK; } // Normalizes the given hostname, component by component. ASCII/ACE // components are lower-cased, and UTF-8 components are normalized per // RFC 3454 and converted to ACE. nsresult nsEffectiveTLDService::NormalizeHostname(nsCString &aHostname) { if (!IsASCII(aHostname)) { nsresult rv = mIDNService->ConvertUTF8toACE(aHostname, aHostname); if (NS_FAILED(rv)) return rv; } ToLowerCase(aHostname); return NS_OK; }