tldts

• New optimizations #255

  • re-use a unique result object for all methods but parse(...)
  • remove memory allocation + callback in packed hash
  • remove unnecessary calls to indexOf and lastIndexOf in packed hash

  These optimizations result in up to 20% faster invocations of parse(...)
  and 40% faster invocations of getPublicSuffix(...)when using
  tldts-experimental. This should also reduce memory pressure in cases where
  specific methods are used (i.e.: all but parse(...)) since no intermediary
  result object needs to be allocated.

• Add minified esm bundle (alongside umd)
• Update public suffix ruleset

• Update public suffix ruleset #231
• Add new getDomainWithoutSuffix(...) method #231

• update public suffix ruleset #207
• fix source maps #207
• make sure tslib does not appear in build artifacts #207

Thanks to @osdiab for catching and reporting the issue with tslib.

• clean-up dead folder + update rules #199

• ship both cjs and es6 source + only minified bundles #183

• Exclude unnecessary files from package.json #177

• #164 Update rules + add fast path for experimental bundle as well

• #155 Update Public Suffix Lists to f651d07

• #151 Update Public Suffix Lists to 456d938

• #150 Update Public Suffix Lists to 7697e3c

• Performance improvements in all methods of public API (up to x2 faster)
  • extractHostname: will now avoid lower-casing the result in some cases
  • extractHostname: handles single or triple '/' after protocol
  • extractHostname: has fast-path for validation of common protocols (e.g. https)
  • isProbablyIpv4: performs first quick check on length of hostname
  • isProbablyIpv6: performs first quick check on length of hostname
  • isValidHostname: make use of charCodeAt instead of codePointAt
  • lookupInTrie: makes use of Trie with more stable structure (faster)
  • lookupInTrie: lazily allocate memory for result
  • suffixLookup: uses fast-path for most common suffixes (massive speed-up)
  • suffixLookup: does not allocate memory for result anymore
  • setDefaults: fast-path in case no argument was provided
  • getSubdomain: fast-path if subdomain is empty
• Add more options to fine-tune behavior and performance
  • detectIp allows to disable IP check
  • mixedInput allows to specify if we expect a mix of URLs and hostnames as
    input. If only hostnames are expected then extractHostname can be set to
    false to speed-up parsing. If only URLs are expected then mixedInputs
    can be set to false. The mixedInputs is only a hint and will not
    change the behavior of the library.
  • validateHostname can be set to false to disable validation and
    speed-up processing further.
• Check that input is string before parsing
• Fix support for reserved keywords in hostnames
• Add tests and bring back coverage to 100%
• Minified bundle is now also tested with the same suite
• Migrate utils scripts from bin/ folder to TypeScript
• Add small tldts cli which can be used to parse URLs
• Update README with more accurate information

• #123 Update Public Suffix Lists to 033221af7f600bcfce38dcbfafe03b9a2269c4cc

• #101 Update Public Suffix Lists
• #101 Update dev dependencies

• #100 Update Public Suffix Lists

• #82 Update Public Suffix Lists

• Update public suffix list to 5d3dfdf7f074ca2831d7c2413cf50b976a019a39

• #35 Update Public Suffix Lists
  • https://github.com/publicsuffix/list/commit/ce0d1a5fba657e55adea3abde4b7f1e50636ff10

This Release introduces some more optimizations both in size of bundles,
memory usage and speed of parsing. Because it introduces some breaking
changes in the API results (host renamed into hostname and deletion
of the isValid attribute), as well as introducing a new experimental
backend (tldts-experimental bundle), this is a major version bump.

• #16 Optimizations + comparison with other libraries (#16)

  • Optimize Trie into a DAWG (size reduction)
  • Implement comparison with other libraries
  • Implement fast path for valid hostnames as arguments
  • Allow to disable hostname parsing and validation using option
  • Add tests for corner-cases URLs parsing
  • Update README

• #13 Implement experimental probabilistic packed suffix structure (#13)

  • Implement packed hash probabilistic structure for more compact
    representation and faster lookups. See ./bin/builders/hashes.js for more
    details about how it works.
  • Create second bundle (tldts-experimental) making use of this new implementation
  • Simplify hostname validation and remove strict IDNA checks
  • Move lookup implementations into 'lookup' sub-folder
  • Move compiled data into 'lookup/data' sub-folder
  • Refactor suffix list parsing out of builders
  • Handle IDNA hostnames at build-time instead of runtime (by indexing
    some suffixes multiple times: once puny-coded and once in unicode form)

• Fix bundles