fix: [#528] Handle capital greek letters in PubMed authors to pass au…

…thor field validation (#529)

CHANGELOG.adoc

@@ -72,7 +72,6 @@ NOTE: References to user stories are in the form Iteration/Story-Number.
 
 .Added
 
-////
 
 .Changed
 
@@ -81,10 +80,11 @@ NOTE: References to user stories are in the form Iteration/Story-Number.
 .Removed
 
 .Fixed
+- {url-issues}528[#528] Handle authors from PubMed with greek capital letters that otherwise fail author string validation
 
 .Security
 
-////
+
 [[v1.8.9]]
 == {url-tree}1.8.9[1.8.9] -- 2023-04-29
 

core/core-entity/src/main/java/ch/difty/scipamato/core/entity/Paper.java

@@ -52,7 +52,7 @@ public class Paper extends IdScipamatoEntity<Long> implements CodeBoxAware, News
     private static final String RE_S_WW = "\\s" + RE_WW;
 
     /**
-     * Regex verifying the correctness of an Author string. Comprises of:
+     * Regex verifying the correctness of an Author string. Consists of:
      * <ol>
      * <li>a single author, made up of one or more "name words", each made up of
      * <ul>
@@ -87,7 +87,7 @@ public class Paper extends IdScipamatoEntity<Long> implements CodeBoxAware, News
 
     /**
      * Regex to validate DOIs. Does not capture the full range of possible DOIs, but
-     * nearly all of the likely ones.
+     * nearly all the likely ones.
      *
      * <ol>
      * <li>starting with {@literal 10} followed by a period</li>

core/core-entity/src/test/kotlin/ch/difty/scipamato/core/entity/PaperTest.kt

@@ -109,6 +109,22 @@ internal class PaperTest : Jsr303ValidatedEntityTest<Paper>(Paper::class.java) {
         verifyFailedAuthorValidation(p, invalidValue)
     }
 
+    @Test
+    fun validatingPaper_withGreekCapitalAlphaInAuthor_insteadOfRegularA_fails() {
+        val regularA = "A"
+        regularA.first().code shouldBeEqualTo 65
+
+        // as found in Author string in PM ID 35469927
+        val invalidValue = "Sindosi OΑ."
+        val greekCapitalAlpha = invalidValue.split(" ")[1].drop(1)
+        greekCapitalAlpha.first().code shouldBeEqualTo 913
+
+        val p = newValidEntity()
+        p.authors = invalidValue
+
+        verifyFailedAuthorValidation(p, invalidValue)
+    }
+
     @Test
     fun validatingPaper_withSingleAuthorWithoutFirstname_withoutPeriod_fails() {
         val invalidValue = "Turner"

core/core-logic/src/test/kotlin/ch/difty/scipamato/core/logic/parsing/AuthorParsingTest.kt

@@ -6,7 +6,7 @@ import org.amshove.kluent.shouldBeEqualTo
 import org.amshove.kluent.shouldBeInstanceOf
 import org.amshove.kluent.shouldBeNull
 import org.amshove.kluent.shouldContain
-import org.amshove.kluent.shouldContainAll
+import org.amshove.kluent.shouldContainSame
 import org.junit.jupiter.api.Test
 
 internal class PubmedAuthorParserTest {
@@ -65,7 +65,7 @@ internal class PubmedAuthorParserTest {
         p = PubmedAuthorParser("Ln FN 1st, Ln FN 2nd, Ln FN 3rd, Ln FN 4th, Ln FN 5th, Ln FN 100th, Ln FN.")
         p.firstAuthor shouldBeEqualTo "Ln"
         p.authors.map { it.lastName } shouldContain "Ln"
-        p.authors.map { it.firstName } shouldContainAll listOf(
+        p.authors.map { it.firstName } shouldContainSame listOf(
             "FN 1st", "FN 2nd", "FN 3rd", "FN 4th", "FN 5th", "FN 100th", "FN"
         )
     }
@@ -96,17 +96,17 @@ internal class PubmedAuthorParserTest {
         p = PubmedAuthorParser(
             "Turner MC, Cohen A, Jerret M, Gapstur SM, Driver WR, Krewsky D, Beckermann BS, Samet JM."
         )
-        p.authors.map { it.lastName } shouldContainAll listOf(
+        p.authors.map { it.lastName } shouldContainSame listOf(
             "Turner", "Cohen", "Jerret", "Gapstur", "Driver", "Krewsky", "Beckermann", "Samet"
         )
-        p.authors.map { it.firstName } shouldContainAll listOf("MC", "A", "M", "SM", "WR", "D", "BS", "JM")
+        p.authors.map { it.firstName } shouldContainSame listOf("MC", "A", "M", "SM", "WR", "D", "BS", "JM")
     }
 
     @Test
     fun canDoUmlaute() {
         p = PubmedAuthorParser("Flückiger P, Bäni HU.")
-        p.authors.map { it.lastName } shouldContainAll listOf("Flückiger", "Bäni")
-        p.authors.map { it.firstName } shouldContainAll listOf("P", "HU")
+        p.authors.map { it.lastName } shouldContainSame listOf("Flückiger", "Bäni")
+        p.authors.map { it.firstName } shouldContainSame listOf("P", "HU")
     }
 }
 

core/core-pubmed-api/src/main/java/ch/difty/scipamato/core/pubmed/GreekLetterTranslator.kt

@@ -0,0 +1,39 @@
+package ch.difty.scipamato.core.pubmed
+
+/**
+ * Takes care of removing capital greek letters that visually look similar or even identical
+ * to regular latin capital characters and replacing them with the regular ones.
+ *
+ * Needed for processing Author strings from Pubmed that would otherwise not pass the
+ * validation constraint implemented in the Paper class.
+ *
+ * Identified in PM ID 35469927 where one of the authors contains a greek capital alpha.
+ */
+object GreekLetterTranslator {
+
+    @Suppress("MagicNumber")
+    private val greekLetterReplacement = mapOf(
+        913 to 65,
+        914 to 66,
+        917 to 69,
+        918 to 90,
+        919 to 72,
+        921 to 73,
+        922 to 75,
+        924 to 77,
+        925 to 78,
+        927 to 79,
+        929 to 80,
+        932 to 84,
+        933 to 89,
+        935 to 88,
+    ).map { (key, value) -> Char(key) to Char(value) }
+
+    fun replaceGreekLetters(original: String): String {
+        var tmp = original
+        greekLetterReplacement.forEach { (greek, replacement) ->
+            tmp = tmp.replace(greek, replacement)
+        }
+        return tmp
+    }
+}

core/core-pubmed-api/src/main/java/ch/difty/scipamato/core/pubmed/PubmedArticleResult.kt

@@ -10,7 +10,7 @@ private val reasonRegex =
     """.+Reason: <strong>([^<]+)</strong>+.+""".toRegex(RegexOption.DOT_MATCHES_ALL)
 
 /**
- * Data Class providing the [PubmedArticleFacade] or an error specific message
+ * Providing the [PubmedArticleFacade] or an error specific message
  * providing information about the problem that prevented the retrieval of the article.
  */
 class PubmedArticleResult(

core/core-pubmed-api/src/main/java/ch/difty/scipamato/core/pubmed/PubmedXmlService.java

@@ -108,7 +108,7 @@ public PubmedArticleSet unmarshal(@NotNull final String xmlString) throws IOExce
      * <ul>
      * <li>via API, e.g.
      * {@code https://eutils.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?db=pubmed&id=25395026&retmode=xml}</li>
-     * <li>through the Web UI (e.g. https://www.ncbi.nlm.nih.gov/pubmed/25395026)
+     * <li>through the Web UI (e.g. <a href="https://www.ncbi.nlm.nih.gov/pubmed/25395026">...</a>)
      * when sending to {@code file} in format {@code XML}</li>
      * </ul>
      *

core/core-pubmed-api/src/main/java/ch/difty/scipamato/core/pubmed/ScipamatoPubmedArticle.java

@@ -35,8 +35,8 @@ class ScipamatoPubmedArticle extends AbstractPubmedArticleFacade {
         setPmId(pmid.getvalue());
         final AuthorList authorList = article.getAuthorList();
         if (authorList != null) {
-            setAuthors(getAuthorsFrom(authorList));
-            setFirstAuthor(getFirstAuthorFrom(authorList));
+            setAuthors(GreekLetterTranslator.INSTANCE.replaceGreekLetters(getAuthorsFrom(authorList)));
+            setFirstAuthor(GreekLetterTranslator.INSTANCE.replaceGreekLetters(getFirstAuthorFrom(authorList)));
         }
         final boolean isAheadOfPrint = "aheadofprint".equals(pubmedArticle.getPubmedData() != null ?
             pubmedArticle

core/core-pubmed-api/src/test/kotlin/ch/difty/scipamato/core/pubmed/GreekLetterTranslatorTest.kt

@@ -0,0 +1,52 @@
+package ch.difty.scipamato.core.pubmed
+
+import org.amshove.kluent.shouldBeEqualTo
+import org.amshove.kluent.shouldBeFalse
+import org.amshove.kluent.shouldContain
+import org.amshove.kluent.shouldNotBeEqualTo
+import org.amshove.kluent.shouldNotContain
+import org.amshove.kluent.shouldStartWith
+import org.junit.jupiter.api.Test
+
+class GreekLetterTranslatorTest {
+
+    private val translator = GreekLetterTranslator
+
+    @Test
+    fun canParsePmId35469927_resulting_in_validA() {
+        val authorsStringPmId35469927 = "Markozannes G, Pantavou K, Rizos EC, Sindosi OΑ, Tagkas C, " +
+            "Seyfried M, Saldanha IJ, Hatzianastassiou N, Nikolopoulos GK, Ntzani E."
+
+        // Ensure the fourth author has a capital greek letter alpha in it (visually nearly identical to "A")
+        val fourthAuthor = authorsStringPmId35469927.split(", ")[3]
+        fourthAuthor shouldStartWith "Sindosi"
+        val capitalGreekAlpha = Char(913) // Capital Greek Letter Alpha
+        fourthAuthor shouldContain capitalGreekAlpha
+
+        val result = translator.replaceGreekLetters(authorsStringPmId35469927)
+
+        result shouldNotContain capitalGreekAlpha
+        result shouldContain Char(65) // Regular Capital A
+    }
+
+    @Test
+    fun replacesSeveralGreekLetters_with_visuallySimilarLetters() {
+        val wGreeks = "ΑΒΕΖΗΙΚΜΝΟΡΤΥΧ Α."
+        val woGreek = "ABEZHIKMNOPTYX A."
+
+        assertLettersAreNotSame(wGreeks, woGreek)
+        @Suppress("KotlinConstantConditions")
+        (wGreeks == woGreek).shouldBeFalse()
+
+        translator.replaceGreekLetters(wGreeks) shouldBeEqualTo woGreek
+    }
+
+    @Suppress("SameParameterValue")
+    private fun assertLettersAreNotSame(one: String, two: String) {
+        val toIgnore = setOf(' ', '.')
+        val first = one.filterNot { it in toIgnore }
+        val second = two.filterNot { it in toIgnore }
+        first.length shouldBeEqualTo second.length
+        first.forEachIndexed { index, char -> char shouldNotBeEqualTo second[index] }
+    }
+}

justfile

@@ -5,3 +5,6 @@
 run app='core':
     ./gradlew :{{app}}-web:bootRun
 
+
+debug app='core':
+    ./gradlew -Dorg.gradle.jvmargs="-agentlib:jdwp=transport=dt_socket,server=y,suspend=n,address=*:5005" :{{app}}-web:bootRun

public/public-persistence-jooq/build.gradle.kts

@@ -20,6 +20,7 @@ val props = file("src/integration-test/resources/application.properties").asProp
 
 testing {
     suites {
+        @Suppress("UNUSED_VARIABLE")
         val integrationTest by existing {
             dependencies {
                 implementation(libs.bundles.dbTest)