Version 5.6.1

Adjusts prefix lengths in containment tries for all individual addresses

Version 5.6.0

This version introduces new types for efficient IP address collections and set operations, along with a few other enhancements

• added IPAddressSeqRangeList, a collection type backed by IP address ranges, offering binary search, an alternative to tries with potentially better CPU cache coherence when searching
• IPAddressSeqRangeList enables the arithmetic set operations add/join, intersect, remove, and complement
• added the corresponding version-specific types IPv4AddressSeqRangeList and IPv6AddressSeqRangeList
• added IPAddressContainmentTrie, an alternative collection option backed by CIDR prefix blocks in a trie, expanding on the existing trie functionality.
• added the corresponding version-specific types IPv4AddressContainmentTrie and IPv6AddressContainmentTrie
• added the IPAddressCollection interface representing collections of individual addresses, implemented by all range list and containment trie types
• added the IPAddressAggregation interface representing all types which can represent a multitude of IP addresses, whether collections, ranges, or subnets
• added new and optimized increment(), decrement() and incrementBoundary() methods to the address types, more efficient than increment(1), increment(-1), incrementBoundary(1)
• extended increment(BigInteger) to the address/subnet types IPAddress and IPv4Address, and not just IPV6Address, for easier polymorphism
• added a complement method to IPAddress returning blocks, and a complement method to IPAddressSeqRange returning ranges, in addition to the complement method included with IPAddressSeqRangeList
• added methods to IPAddressSeqRange to integrate with IPAddressSeqRangeList to create lists efficiently: subtractIntoList, joinIntoList, complementIntoList
• added intoSequentialRangeList methods to IP address and sequential range classes for polymorphic and efficient IPAddressSeqRangeList creation
• added intoContainmentTrie methods to IP address and range classes for polymorphic IPAddressContainmentTrie creation
• added new trie types IPAddressTrie and IPAddressAssociativeTrie, similar to existing types, but version-polymorphic, they can become either IPv4 or IPv6 tries, but not both at the same time
• fixed an issue with merging addresses with prefixes into sequential blocks

Version 5.5.1

• added new merge functions mergeToDualPrefixBlocks and mergeToDualSequentialBlocks for combined IPv4/IPv6 merging
• added public constructors accepting message strings to AddressValueException and IncompatibleAddressException, issue #127
• removed caching functionality from containingFirstIterator, use all-node containingFirstAllNodeIterator instead

Version 5.5.0

• added collection types for dual IPv4/v6 tries and for dual IPv4/v6 associative tries, issue #103
• trie performance improvements for all trie operations
• added shortestPrefixMatch trie methods
• added enumerate methods, the inverse of the increment methods, to find the position of an address in a subnet, or to find the distance between two addresses
• added an increment method accepting a BigInteger argument to IPv6Address and IPv6AddressSection
• added the ability to construct an IPv6Address from two longs
• added replace methods with address section parameter to address classes
• added overlaps methods to check for overlapping subnets, and for checking sequential range overlap with a subnet
• added floor/lower/ceiling/higher methods to the address trie types, these methods previously accessible only from trie sets
• added the extraneous digits IPv4 parsing option allow_inet_aton_extraneous_digits, issue #105.
• includes the fix to invalid radix argument infinite loop, issue #118

Version 5.4.2

• removed unnecessary SDK dependency incompatible with Java 8 that was introduced with version 5.4.1, issue #121
• includes the fix to invalid radix argument infinite loop, issue #118

Version 5.4.1

• fix to invalid radix argument infinite loop, issue #118

Version 5.4.0

• added PrefixBlockAllocator for automatic CIDR prefix block allocation
• added AddedTree and AssociativeAddedTree classes for expanded constructAddedNodesTree methods
• added getBlockSize and getBitsForCount in AddressItem
• added matchUnordered and matchOrdered in Address
• eliminated invalid AddressValueException when joining a range with the IPv4 max value to an IPv6 range, issue #86
• fix to generation of strings from parsed address data, issue #87
• fix to IPv4 address primitive int upper value generation, issue #96

Version 5.3.4

• a fix to avoid ClassCastException when converting IPv4-mapped address to java.net.Inet6Address, issue #73
• fixed the creation of HostName from an unresolved compressed-IPv6 socket address, issue #74
• added IPAddressString.isIPv4Mapped for issue #75
• fixed a class-loading issue with string parsing, issue #79
• removed AddressDivision.getDivisionValueCount() and AddressDivision.getDivisionPrefixCount(int) because of potential overflow with divisions 63 or 64 bits long. If you have a segment, use getValueCount(), or with IPAddressSegment use getPrefixValueCount(int), which both return int. Otherwise, use getPrefixCount(int) or getCount() which return BigInteger
• fixed issues with comparing non-standard division groupings with address comparators
• added verification when host-masking multiple-valued segments
• added verification when splitting IPv6AddressSegment
• fixed check when reversing per-byte multiple-valued division
• fixed trie-printing method option withNonAddedKeys
• fixed an issue with serialization of parsed hosts with qualifiers
• added control over parsing empty zones (a zone specifier followed by no zone)
• added a small fix to host name parsing some unusual host names
• a fix to IPv4AddressSection.toMaxHost
• fixed a rare synchronization issue when generating division strings in AddressDivisionBase

Version 5.3.3

• more flexible and efficient IPv6 zone handling
• addresses issue #48, converting prefixed addresses with zone to Inet6Address
• more efficient merging

Version 5.3.2

• new longestPrefixMatch and longestPrefixMatchNode methods in AddressTrie, AddressTrieMap and AddressTrieSet, for more efficient matching without using the linked list from elementsContaining
• added binary string parsing for IPv4 and IPv6 (see javadoc for IPAddressString for supported formats), and producing binary segmented strings with toSegmentedBinaryString
• added support for parsing strings with uppercase hex 0X and binary 0B
• improved performance of mergeToPrefixBlocks and mergeToSequentialBlocks, also resulting in a different ordering of the result, sorted by lower value. The old ordering remains attainable with the comparator IPAddressSegmentSeries.getPrefixLenComparator
• added asNewTrie to TrieNode to create a new trie which is a copy of the sub-trie with the node as root
• added IPAddressSeqRange.extend(IPAddressRange) to extend ranges more efficiently
• added IPAddress.applyToBounds to use covering and spanning methods with multiple addresses or subnets, not just two
• added IPAddressString.prefixContains(IPAddressString) for quick prefix-based containment checks
• fixed inconsistency between prefixEquals in IPAddressString vs IPAddress with addresses that have no prefix length
• fixed issue in IPAddressString.prefixEquals involving prefix not on segment boundary
• fixed issue with spanWithPrefixBlocks not splitting up subnets with multiple prefix blocks
• fix to string parsing control with allowsReverseRange, also added missing allowBase85
• fix to issue #41, duplicate property key
• fix to issue #44, wrong super method call
• fix to issue #46, wrong boundary check incrementing IPv6 segment at index 2