T2: NameNormalizer testy + transliteratePolish, token-strip titles [checkpoint]

szpitale-graph/src/main/java/com/developx/szpitale/ingest/normalize/NameNormalizer.java

@@ -20,12 +20,31 @@ public class NameNormalizer {
     public String makeSlug(String fullName) {
         String normalized = Normalizer.normalize(fullName, Normalizer.Form.NFKD);
         normalized = normalized.replaceAll("\\p{M}", "");
-        normalized = normalized.replaceAll("[^a-zA-Z0-9\\s-]", " ");
-        normalized = normalized.toLowerCase().replaceAll("\\s+", "-").replaceAll("-+", "-");
+        normalized = normalized.replaceAll("[^\\p{L}\\s-]", "");
+        normalized = transliteratePolish(normalized.toLowerCase());
+        normalized = normalized.replaceAll("\\s+", "-").replaceAll("-+", "-");
         normalized = normalized.replaceAll("^-|-$", "");
         return normalized;
     }
 
+    private String transliteratePolish(String input) {
+        Map<String, String> map = new HashMap<>();
+        map.put("ł", "l");  map.put("Ł", "l");
+        map.put("ę", "e");  map.put("Ĕ", "e");
+        map.put("ó", "o");  map.put("Ō", "o");
+        map.put("ą", "a");  map.put("Ą", "a");
+        map.put("ś", "s");  map.put("Ś", "s");
+        map.put("ź", "z");  map.put("Ź", "z");
+        map.put("ż", "z");  map.put("Ż", "z");
+        map.put("ć", "c");  map.put("Ć", "c");
+        map.put("ń", "n");  map.put("Ń", "n");
+        String result = input;
+        for (Map.Entry<String, String> entry : map.entrySet()) {
+            result = result.replace(entry.getKey(), entry.getValue());
+        }
+        return result;
+    }
+
     public String extractFullName(String rawName) {
         return extractGivenAndFamilyNames(rawName);
     }
@@ -62,12 +81,25 @@ public class NameNormalizer {
                 .collect(Collectors.toList());
     }
 
+    static final List<String> TITLE_PREFIXES = List.of(
+            "prof. dr hab.", "prof.dr hab.", "prof. dr.", "prof. dr",
+            "dr hab.", "dr.hab.", "dr. hab.", "dr.hab",
+            "mgr inż.", "mgr inż", "mgr.inż.", "mgr.inż",
+            "prof.", "prof", "dr.", "dr",
+            "mgr.", "mgr", "inż.", "inż",
+            "lek.", "lek.med.", "n. med.", "n. o zdr.",
+            "licencjat inż.", "licencjat inż"
+    );
+
     private String extractGivenAndFamilyNames(String rawName) {
-        Pattern pattern = Pattern.compile(
-                "(?i)(prof\\.\\s*|dr\\s*hab\\.\\s*|dr\\.\\s*|mgr\\s*|in\u017c\\.\\s*)",
-                Pattern.MULTILINE | Pattern.DOTALL
-        );
-        Matcher matcher = pattern.matcher(rawName);
-        return matcher.replaceAll("").trim();
+        String name = rawName.trim();
+        String lower = name.toLowerCase();
+        for (String prefix : TITLE_PREFIXES) {
+            if (lower.startsWith(prefix.toLowerCase())) {
+                name = name.substring(prefix.length()).trim();
+                lower = name.toLowerCase();
+            }
+        }
+        return name;
     }
 }

szpitale-graph/src/test/java/com/developx/szpitale/ingest/normalize/NameNormalizerTest.java

@@ -0,0 +1,132 @@
+package com.developx.szpitale.ingest.normalize;
+
+import org.junit.jupiter.api.Test;
+import org.junit.jupiter.params.ParameterizedTest;
+import org.junit.jupiter.params.provider.CsvSource;
+
+import java.util.List;
+
+import static org.junit.jupiter.api.Assertions.*;
+
+class NameNormalizerTest {
+
+    private final NameNormalizer normalizer = new NameNormalizer();
+
+    @Test
+    void normalize_polishDiacritics_strippedToAsciiSlug() {
+        String slug = normalizer.makeSlug("Łukasz Żółć");
+        assertEquals("lukasz-zolc", slug);
+    }
+
+    @Test
+    void normalize_slug_lowercaseSeparator() {
+        String slug = normalizer.makeSlug("Jan Kowalski");
+        assertEquals("jan-kowalski", slug);
+    }
+
+    @Test
+    void normalize_slug_multipleSpacesToOne() {
+        String slug = normalizer.makeSlug("Jan    Kowalski");
+        assertEquals("jan-kowalski", slug);
+    }
+
+    @Test
+    void normalize_slug_removesSpecialChars() {
+        String slug = normalizer.makeSlug("Dr. Jan Kowalski Jr.");
+        assertEquals("dr-jan-kowalski-jr", slug);
+    }
+
+    @ParameterizedTest
+    @CsvSource({
+        "'prof. dr hab. Jan Kochanowicz', 'Jan Kochanowicz'",
+        "'dr Jan Nowak', 'Jan Nowak'",
+        "'mgr inż. Anna Kowalska', 'Anna Kowalska'",
+        "'Jan Wiśniewski', 'Jan Wiśniewski'",
+        "'prof. Wojciech Łuczaj', 'Wojciech Łuczaj'"
+    })
+    void normalize_displayName_keepsOriginal(String rawName, String expectedFullName) {
+        String fullName = normalizer.extractFullName(rawName);
+        assertEquals(expectedFullName, fullName);
+    }
+
+    @ParameterizedTest
+    @CsvSource({
+        "'prof. dr hab. Jan Kochanowicz', 'prof.'",
+        "'prof. dr hab. Jan Kochanowicz', 'dr hab.'",
+        "'dr hab. n. med. Jan Wiśniewski', 'dr hab.'",
+        "'dr hab. n. med. Jan Wiśniewski', 'n. med.'",
+        "'mgr inż. Anna Kowalska', 'mgr inż.'"
+    })
+    void normalize_titlesContainExpected(String rawName, String expectedTitle) {
+        List<String> titles = normalizer.extractTitles(rawName);
+        assertTrue(titles.contains(expectedTitle), "Expected title '" + expectedTitle + "' not found in " + titles);
+    }
+
+    @Test
+    void normalize_titlesEmptyForPlainName() {
+        List<String> titles = normalizer.extractTitles("Jan Nowak");
+        assertTrue(titles.isEmpty());
+    }
+
+    @Test
+    void normalize_emptyTitlesWhenNoTitles() {
+        List<String> titles = normalizer.extractTitles("Jan Kowalski");
+        assertTrue(titles.isEmpty());
+    }
+
+    @Test
+    void makePersonId_createsPersonSlug() {
+        String id = normalizer.makePersonId("Jan Kowalski");
+        assertEquals("person:jan-kowalski", id);
+    }
+
+    @Test
+    void makePerson_idUsesSameSlugAsMakeSlug() {
+        String id = normalizer.makePersonId("Łukasz Żółć");
+        assertEquals("person:lukasz-zolc", id);
+    }
+
+    @Test
+    void makeHospitalId_createsHospitalSlug() {
+        String id = normalizer.makeHospitalId("podlaskie", "Uniwersytecki Szpital Kliniczny");
+        assertEquals("hosp:podlaskie:uniwersytecki-szpital-kliniczny", id);
+    }
+
+    @Test
+    void fuzzySimilarity_identicalStringsReturn1() {
+        double similarity = normalizer.fuzzySimilarity("Jan Kowalski", "Jan Kowalski");
+        assertEquals(1.0, similarity, 0.0);
+    }
+
+    @Test
+    void fuzzySimilarity_typoReturnsHigh() {
+        double similarity = normalizer.fuzzySimilarity("Jan Kowalski", "Jan Kowaski");
+        assertTrue(similarity > 0.9, "Typo 'Kowalski' vs 'Kowaski' should have very high similarity: " + similarity);
+    }
+
+    @Test
+    void fuzzySimilarity_differentNamesReturnLower() {
+        double similarity = normalizer.fuzzySimilarity("Jan Kowalski", "Anna Nowak");
+        assertTrue(similarity < 0.7, "Two completely different names should have moderate similarity: " + similarity);
+    }
+
+    @Test
+    void fuzzySimilarity_handlesPolishCharacters() {
+        double similarity = normalizer.fuzzySimilarity("Jan Kowalski", "Jan Kowalscy");
+        assertTrue(similarity > 0.8, "Names differing only by suffix should have high similarity: " + similarity);
+    }
+
+    @Test
+    void findDuplicateCandidates_returnsCandidatesAboveThreshold() {
+        List<String> names = List.of("Jan Kowalski", "Jan Kowaski", "Anna Nowak", "Jan Nowak");
+        List<String> candidates = normalizer.findDuplicateCandidates(names, "Jan Kowalski", 0.7);
+        assertTrue(candidates.contains("Jan Kowaski"), "Should contain similar name");
+    }
+
+    @Test
+    void findDuplicateCandidates_filtersTargetName() {
+        List<String> names = List.of("Jan Kowalski");
+        List<String> candidates = normalizer.findDuplicateCandidates(names, "Jan Kowalski", 0.7);
+        assertTrue(candidates.isEmpty(), "Should not return the target name itself");
+    }
+}