Parse TopLevelDomains correctly. #485
Conversation
put out of the loop the check for adding dots to the domain. fixes cSploit#21.
|
this won't still fully work, because the TLD array is sorted in a way, that domains of "one component" are checked first:
so www.ansej.org.dz will match first with .dz and thus set the base domain to org.dz. There is another solution, which is to check every part of the domain against the TLD array: accurate, but slower than the current implementation in most cases (avg of ~30ms vs ~100-200ms), . besides, there are some entries in the array that are simply wrong: "more..." and "...". @tux-mind , I know that evilproxy will be developed, but in the meantime a few questions: should I sort the TLD array? and as the array is outdated, should I update it with the latest list https://publicsuffix.org/list/public_suffix_list.dat ? |
reorder TLD domains, to put simple domains (.us) at the end of the compounded domains of the same type (.com.us)
Parse TopLevelDomains correctly.
and put out of the loop the check for adding dots to the domain.
fixes #21.