This enum won't be noticed by most users, because the country_code_source field of the phone number is only populated with ParseAndKeepRawInput, and in this case, this enum won't be used.

However, it provides a better result for users who incorrectly call parse() and then getCountryCodeSource() and expect to get a sensible answer; now they will get UNSPECIFIED instead of FROM_NUMBER_WITH_PLUS_SIGN which was the default before. In JS this method returns null, but getCountryCodeSourceOrDefault() will now return UNSPECIFIED instead.

The numeric values of the existing enums have not changed (relevant for JS and C++) and in Java the new enum was added to the end so that any stored serialized phone numbers should be able to be restored as they were before.

Suggested in issue #1218

Points people at parseAndKeepRawInput where appropriate.

*   It was confusing and unnecessary. Not every country that has short numbers beginning with a 0 had it, and it is not the only digit that could overlap with a national prefix and hence be interpreted incorrectly.

* Metadata updates for release 8.5.1

* Update release_notes.txt

* Metadata updates for release 8.5.0

* Corrected release notes as US geocding update is mentioned as new addition.

* Make canBeInternationallyDialled_API public in Java and CPP
* update notes in pending_code_changes.txt
* Remove TODO on canBeInternationallyDialled as it is already an public API in JS
* Make JS canBeInternationallyDialled API doc consistent with Java and CPP

Alternate metadata and test metadata for C++ was regenerated not due to changes in this commit, but due to PR #1726 and that they should have been updated then.

1) Changing the BuildMetadataFromXml to only set
same_mobile_and_fixed_line_pattern and
national_prefix_optional_when_formatting if they differ from default
values, and the domestic_carrier_code_formatting_rule,
international_prefix and national_prefix_formatting_rule if there is
one.

2) Adding a test with some golden data for the BuildMetadataJsonFromXml.

3) Adding to the proto a missing explicit default false.

4) Updating a couple of methods in the Phonemetadata.java class to have
names that match those of the auto-generated Phonemetadata used by the
CPP build, now that we use those methods when building the metadata.

5) Fix to ant build file to enable the junit command to still work (instructions on open-source on how to regenerate metadata files refer to this)

Resulting data is smaller for Java and C++ and the resulting BuildMetadataJsonFromXml code is cleaner.

* Handling possible out-of-bounds exception that would be thrown in C++
and Java if a number with a phone-context tag is entered (RFC3966
format) but no actual phone context.

* Metadata updates for release 8.4.1

* Update release_notes.txt

IS_POSSIBLE_LOCAL_ONLY and INVALID_LENGTH are now being returned from IsPossibleNumberWithReason.

The enums were added in v8.3.0

Groups announcement: https://groups.google.com/forum/#!topic/libphonenumber-discuss/sPhYzdzFCmg

* Updating docs for getRegionCodeForNumber to clarify that it does not work for short codes or invalid numbers.

Update doc references to refer to CLDR and the page of CLDR codes where
appropriate - it used to say ISO country codes were used, but they aren't.

Also updated to refer to two *or three* letter language codes since they may be three letters (we use the shortest defined one for a language)

Updating comments about leadingDigits field at territory level as well.

- make them more consistent with each other
- explain why the empty string is sometimes returned when geocoding

Fix for getNationalSignificantNumber to be robust in the face of malicious input (negative number of leading zeros in the phone number proto).