Fix for HumanNameParser incorrectly interprets some name components as postnominals

[CarlosManuelRodr]

Rationale

Hi. While using this library I observed that names similar to the default postnominals such as "Ruvin Phidd" cannot be parsed correctly, throwing the exception

com.tupilabs.human_name_parser.ParseException: Couldn't find a last name in '{ruvin}'.

The cause is that the default postnominals

    public static final List<String> DEFAULT_POSTNOMINALS = Collections.unmodifiableList(
            Arrays.asList(
                    "phd",
                    "ph.d.",
                    "ph.d",
                    "esq",
                    "esquire",
                    "apr",
                    "rph",
                    "pe",
                    "md",
                    "ma",
                    "dmd",
                    "cme",
                    "dds",
                    "cpa",
                    "dvm"));

contains periods (.), which are interpreted as wildcard characters when parsed as regular expressions. As a result, unexpected matches occur during name parsing. The fix included on this PR addresses this problem by formatting the default values on the build() step of HumanNameParserBuilder.

Changes

• Added method formatToRegex to HumanNameParserBuilder and modified build() to format the default values.
• Created unit test testLastNameNotMistakenForPostnominal at BuilderTest that catches this problem.

[kinow]

Thank you @CarlosManuelRodr ! I will prepare a new release today or over the weekend.