2.2.3

• Issue #23: Fixed Territory.fromString to make sure the parent territory is valid for
  input like "CHE-GR". This returned "MX-GRO" instead of throwing UnknownTerritoryException.
  Added unit test for this type of case.
• Fixed minor JavaDoc issues.

v2.2.2

• Fixed error in Point which in rare cases would allow longitudes outside proper range.

2.2.1

• Fixed unit test. Reduced size of files for unit tests considerably. Improved unit test speed.
• Fixed Point interface.
• Cleaned up Boundary and DataAccess.
• Solved 1-microdegree gap in a few spots on Earth, noticable now extreme precision is possible.
• Replaced floating point by fixed point math.
• Improved speed.
• Enforce Mencode(decode(M)) == M, except at territory border corners.
• Cleaned up source; moved hard-coded data into mminfo.dat.
• Added micro-meter precision (mapcodes can now have eight precision digits).
• Assure that encode(decode(m)) delivers m.
• Renames to bring source more in line with other implementations.

2.0.2

• Renamed isValidPrecisionFormat to sValidMapcodeFormat.
• Removed public microdegree references from Point class. Everything is degrees now.
• Removed ParentTerritory class.

2.0.1

• Reverted Java JDK level to 1.6 (Java 6) from 1.8 (Java 8), so the library can be used on
  Android platforms operating at Java 6 as well.
• Use multi-threading for long running test to speed them up (uses all CPU cores now).
• Added the ability to use a country name for `Territory.fromString().

2.0.0

* Fixes to the data rectangles (primarily intended for ISO proposal).

* Removed functionality to use numeric territory codes; only alpha codes are accepted.

* Note that this release only allows high-precision mapcodes up to 2 additional suffix characters.
A future release will be scheduled to allow up to 8 suffix characters (nanometer accuracy).

1.50.3

• This release breaks compatiblity with earlier releases, to clean up the interface significantly.
• Removed <<<Mapcode.encodeToShortest(lat, lon))>>> as this will produce a randomly chosen territory.
  You must specify a <<>> now.
• Renamed Territory.codez toTerritory.number`.
• Renamed fromTerritoryCode() to fromNumber()).
• Renamed Territory.isState())to Territory.isSubdivision()).
• Renamed Territory.hasStates() to Territory.hasSubdivision()).
• Renamed Alphabet.code to Alphabet.number.
• Renamed romCode()) to fromNumber()).
• Renamed MapcodeFormat to PrecisionForma.
• Deprecated methods have been removed.

1.50.2

• Cleaned up Unicode handling a bit.
• Speed up of reading initialization data.
• RenametoNameFormatinto toAlphaFormat and NAME_FORMAT to ALPHA_FORMAT.

1.50.1

• Bug fix for state IN-DD (in India).

1.50.0

• Major release. This version is not backwards compatible with mapcode 1.4x: is has dropped support for
  Antartica AT0-8 codes and has a changed (improved) way of dealing with the Greek alphabet.
• Added 22-chararcter post-processing of all-digit mapcodes for the Greek alphabet.
• Retired legacy aliases EAZ and SKM, AU-QL, AU-TS, AU-NI and AU-JB.
• Retired legacy Antarctica claims AT0 through AT8.
• Added convencience methods for <<>> to accept <<>> for all encode functions
  as well (not just <<>>, <<>>).
• Added alphabet support to convert mapcodes (both codes and territories) between <<>>s.
• Exceptions have been corrected and documented in code.
• Allowed nullable values in <<>> encode and decode methods to assume reasonable defaults.
• Microdegrees are no longer support publicly in <<>>. Only degrees.
• Latitudes are limited to -90..90 and longitudes are wrapped to -180..180 (non inclusive).