trim nbsp (#5)

bmlt-enabled · Nov 24, 2023 · a4086d5 · a4086d5
1 parent 08c95bb
commit a4086d5
Show file tree

Hide file tree

Showing 4 changed files with 18 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,3 +1,6 @@
+## 1.1.1 (November 24, 2023)
+* Trim NBSP unicode for Portuguese, Italian and Japanese.
+
 ## 1.1.0 (November 1, 2023)
 * This makes content paragraph parsing for Portuguese slightly less brittle.
 

diff --git a/src/FetchMeditation/ItalianJFT.php b/src/FetchMeditation/ItalianJFT.php
@@ -74,6 +74,15 @@ public function fetch(): JFTEntry
         $quoteParts = explode('--', $result['quote']);
         $result['quote'] = $quoteParts[0] ?? '';
         $result['source'] =  $quoteParts[1] ?? $result['source'];
+        $result = array_map(function ($item) {
+            if (is_array($item)) {
+                return array_map(function ($paragraph) {
+                    return trim($paragraph, "\xC2\xA0");
+                }, $item);
+            } else {
+                return trim($item, "\xC2\xA0");
+            }
+        }, $result);
 
         return new JFTEntry(
             $result['date'],

diff --git a/src/FetchMeditation/JapaneseJFT.php b/src/FetchMeditation/JapaneseJFT.php
@@ -79,7 +79,9 @@ public function fetch(): JFTEntry
                 $paragraphs[] .= trim($nextNode->textContent);
                 $nextNode = $nextNode->nextSibling;
             }
-            $result['content'] = array_values(array_filter($paragraphs));
+            $result['content'] = array_map(function ($paragraph) {
+                return trim(preg_replace('/\s+/u', ' ', $paragraph));
+            }, array_values(array_filter($paragraphs)));
         }
 
         return new JFTEntry(

diff --git a/src/FetchMeditation/PortugueseJFT.php b/src/FetchMeditation/PortugueseJFT.php
@@ -73,7 +73,9 @@ public function fetch(): JFTEntry
             }
         }
         $result['page'] = '';
-        $result['content'] = array_map('trim', $paragraphs);
+        $result['content'] = array_map(function ($paragraph) {
+            return trim(preg_replace('/\s+/u', ' ', $paragraph));
+        }, $paragraphs);
 
         return new JFTEntry(
             $result['date'],