[FINNA-623] QDC: Add a check for detecting invalid language codes (#60)

NatLibFi · Feb 16, 2024 · 509b478 · 509b478
1 parent d3940ff
commit 509b478
Show file tree

Hide file tree

Showing 3 changed files with 75 additions and 0 deletions.
diff --git a/src/RecordManager/Finna/Record/Qdc.php b/src/RecordManager/Finna/Record/Qdc.php
@@ -144,6 +144,34 @@ protected function getHierarchyFields(array &$data): void
         }
     }
 
+    /**
+     * Get languages
+     *
+     * @return array
+     */
+    protected function getLanguages()
+    {
+        $languages = [];
+        foreach ($this->doc->language as $language) {
+            foreach (explode(' ', trim((string)$language)) as $part) {
+                $check = preg_replace(
+                    '/^http:\/\/lexvo\.org\/id\/iso639-.\/(.*)/',
+                    '$1',
+                    $part
+                );
+                // Check that the language given is in proper form
+                if (mb_strlen($check) > 9 || !ctype_lower($check)) {
+                    $this->storeWarning("unhandled language $check");
+                    continue;
+                }
+                foreach (str_split($check, 3) as $code) {
+                    $languages[] = $code;
+                }
+            }
+        }
+        return $this->metadataUtils->normalizeLanguageStrings($languages);
+    }
+
     /**
      * Get online URLs
      *

diff --git a/tests/RecordManagerTest/Finna/Record/QdcTest.php b/tests/RecordManagerTest/Finna/Record/QdcTest.php
@@ -116,4 +116,38 @@ public function testMediaTypes()
             $fields['media_type_str_mv']
         );
     }
+
+    /**
+     * Test QDC processing warnings handling
+     *
+     * @return void
+     */
+    public function testQdcLanguageWarnings()
+    {
+        $record = $this->createRecord(
+            Qdc::class,
+            'qdc_language_warnings.xml',
+            [],
+            'Finna',
+            [$this->createMock(\RecordManager\Base\Http\ClientManager::class)]
+        );
+        $fields = $record->toSolrArray();
+        $this->compareArray(
+            [
+                'unhandled language Veryodd',
+                'unhandled language verylonglanguagehere',
+                'unhandled language EnGb',
+                'unhandled language caT',
+            ],
+            $record->getProcessingWarnings(),
+            'getProcessingWarnings'
+        );
+        $this->compareArray(
+            [
+                'fi',
+            ],
+            $fields['language'],
+            'LanguageCheckAfterWarnings'
+        );
+    }
 }
diff --git a/tests/fixtures/Finna/record/qdc_language_warnings.xml b/tests/fixtures/Finna/record/qdc_language_warnings.xml
@@ -0,0 +1,13 @@
+<?xml version="1.0"?>
+<qualifieddc schemaLocation="http://purl.org/dc/terms/ http://dublincore.org/schemas/xmls/qdc/2006/01/06/dcterms.xsd http://purl.org/dc/elements/1.1/ http://dublincore.org/schemas/xmls/qdc/2006/01/06/dc.xsd">
+  <title lang="fi">Language test xml</title>
+  <language type="iso">Veryodd</language>
+  <language type="iso">verylonglanguagehere</language>
+  <language type="iso">EnGb</language>
+  <language type="iso">caT</language>
+  <language type="iso">fi</language>
+  <rights>CC BY-NC-ND 4.0</rights>
+  <publisher lang="fi">Sanitation Project, Research Institute for Humanity and Nature</publisher>
+  <permaddress type="doi">http://dx.doi.org/https://doi.org/10.34416/svc.00029</permaddress>
+  <recordID>10138_331330</recordID>
+</qualifieddc>