From e565b1c94873bc0dbe7adacfc3ecbfd6b7dca425 Mon Sep 17 00:00:00 2001
From: Ken Stevens <khstevens@gmail.com>
Date: Fri, 10 Jul 2020 15:08:50 -0400
Subject: [PATCH] done (#1976)

---
 .../fhir/context/phonetic/ApacheEncoder.java  | 37 +++++++++++++++++++
 .../context/phonetic/PhoneticEncoderTest.java | 29 +++++++++++++++
 2 files changed, 66 insertions(+)
 create mode 100644 hapi-fhir-base/src/test/java/ca/uhn/fhir/context/phonetic/PhoneticEncoderTest.java

diff --git a/hapi-fhir-base/src/main/java/ca/uhn/fhir/context/phonetic/ApacheEncoder.java b/hapi-fhir-base/src/main/java/ca/uhn/fhir/context/phonetic/ApacheEncoder.java
index 11bd84abf0d..2f2c11838df 100644
--- a/hapi-fhir-base/src/main/java/ca/uhn/fhir/context/phonetic/ApacheEncoder.java
+++ b/hapi-fhir-base/src/main/java/ca/uhn/fhir/context/phonetic/ApacheEncoder.java
@@ -22,9 +22,12 @@ package ca.uhn.fhir.context.phonetic;
 
 import org.apache.commons.codec.EncoderException;
 import org.apache.commons.codec.StringEncoder;
+import org.apache.commons.lang3.StringUtils;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
+import java.util.StringJoiner;
+
 public class ApacheEncoder implements IPhoneticEncoder {
 	private static final Logger ourLog = LoggerFactory.getLogger(ApacheEncoder.class);
 
@@ -44,10 +47,44 @@ public class ApacheEncoder implements IPhoneticEncoder {
 	@Override
 	public String encode(String theString) {
 		try {
+			// If the string contains a space, encode alpha parts separately so, for example, numbers are preserved in address lines.
+			if (theString.contains(" ")) {
+				return encodeStringWithSpaces(theString);
+			}
 			return myStringEncoder.encode(theString);
 		} catch (EncoderException e) {
 			ourLog.error("Failed to encode string " + theString, e);
 			return theString;
 		}
 	}
+
+	private String encodeStringWithSpaces(String theString) throws EncoderException {
+		StringJoiner joiner = new StringJoiner(" ");
+
+		// This sub-stack holds the alpha parts
+		StringJoiner alphaJoiner = new StringJoiner(" ");
+
+		for (String part : theString.split("[\\s\\W]+")) {
+			if (StringUtils.isAlpha(part)) {
+				alphaJoiner.add(part);
+			} else {
+				// Once we hit a non-alpha part, encode all the alpha parts together as a single string
+				// This is to allow encoders like METAPHONE to match Hans Peter to Hanspeter
+				alphaJoiner = encodeAlphaParts(joiner, alphaJoiner);
+				joiner.add(part);
+			}
+		}
+		encodeAlphaParts(joiner, alphaJoiner);
+
+		return joiner.toString();
+	}
+
+	private StringJoiner encodeAlphaParts(StringJoiner theJoiner, StringJoiner theAlphaJoiner) throws EncoderException {
+		// Encode the alpha parts as a single string and then flush the alpha encoder
+		if (theAlphaJoiner.length() > 0) {
+			theJoiner.add(myStringEncoder.encode(theAlphaJoiner.toString()));
+			theAlphaJoiner = new StringJoiner(" ");
+		}
+		return theAlphaJoiner;
+	}
 }
diff --git a/hapi-fhir-base/src/test/java/ca/uhn/fhir/context/phonetic/PhoneticEncoderTest.java b/hapi-fhir-base/src/test/java/ca/uhn/fhir/context/phonetic/PhoneticEncoderTest.java
new file mode 100644
index 00000000000..bca150978cb
--- /dev/null
+++ b/hapi-fhir-base/src/test/java/ca/uhn/fhir/context/phonetic/PhoneticEncoderTest.java
@@ -0,0 +1,29 @@
+package ca.uhn.fhir.context.phonetic;
+
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.params.ParameterizedTest;
+import org.junit.jupiter.params.provider.EnumSource;
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import static org.hamcrest.Matchers.startsWith;
+import static org.hamcrest.MatcherAssert.assertThat;
+import static org.hamcrest.Matchers.endsWith;
+
+class PhoneticEncoderTest {
+	private static final Logger ourLog = LoggerFactory.getLogger(PhoneticEncoderTest.class);
+
+	private static final String NUMBER = "123";
+	private static final String STREET = "Nohili St, Suite";
+	private static final String SUITE = "456";
+	private static final String ADDRESS_LINE = NUMBER + " " + STREET + " " + SUITE;
+
+	@ParameterizedTest
+	@EnumSource(PhoneticEncoderEnum.class)
+	public void testEncodeAddress(PhoneticEncoderEnum thePhoneticEncoderEnum) {
+		String encoded = thePhoneticEncoderEnum.getPhoneticEncoder().encode(ADDRESS_LINE);
+		ourLog.info("{}: {}", thePhoneticEncoderEnum.name(), encoded);
+		assertThat(encoded, startsWith(NUMBER + " "));
+		assertThat(encoded, endsWith(" " + SUITE));
+	}
+}