Split into files

lib/src/exampleSearch.dart

@@ -0,0 +1,108 @@
+import './baseURI.dart';
+import './objects.dart';
+
+import 'package:html/parser.dart';
+import 'package:html/dom.dart';
+
+final RegExp kanjiRegex = RegExp(r'[\u4e00-\u9faf\u3400-\u4dbf]');
+
+String uriForExampleSearch(String phrase) {
+  return '${SCRAPE_BASE_URI}${Uri.encodeComponent(phrase)}%23sentences';
+}
+
+/* TODO: This is the wrong approach. 
+ * Symbols such as 、「」。 are missing in mid sentence
+ * Maybe also JP fullwidth numbers?
+ */
+
+String getEndSymbolsOfExampleSentence(Element ul) {
+  final endSymbols = RegExp(r'<\/li>([^<>]+)$');
+  return endSymbols.firstMatch(ul.innerHtml).group(1);
+}
+
+ExampleResultData getKanjiAndKana(Element div) {
+  final ul = div.querySelector('ul');
+  final contents = ul.children;
+
+  var kanji = '';
+  var kana = '';
+  for (var i = 0; i < contents.length; i += 1) {
+    final content = contents[i];
+    if (content.localName == 'li') {
+      final li = content;
+      final furigana = li.querySelector('.furigana')?.text;
+      final unlifted = li.querySelector('.unlinked')?.text;
+
+      if (furigana != null) {
+        kanji += unlifted;
+        kana += furigana;
+
+        final kanaEnding = [];
+        for (var j = unlifted.length - 1; j > 0; j -= 1) {
+          final char = unlifted[j];
+          if (!kanjiRegex.hasMatch(char)) {
+            kanaEnding.add(char);
+          } else {
+            break;
+          }
+        }
+
+        kana += kanaEnding.reversed.join('');
+      } else {
+        kanji += unlifted;
+        kana += unlifted;
+      }
+    } else {
+      final text = content.text.trim();
+      if (text != null) {
+        kanji += text;
+        kana += text;
+      }
+    }
+  }
+    final endSymbols = getEndSymbolsOfExampleSentence(ul).trim();
+    kanji+= endSymbols;
+    kana += endSymbols;
+
+  return ExampleResultData(
+    kanji: kanji,
+    kana: kana,
+  );
+}
+
+List<ExampleSentencePiece> getPieces(Element sentenceElement) {
+  final pieceElements = sentenceElement.querySelectorAll('li.clearfix');
+  final List<ExampleSentencePiece> pieces = [];
+  for (var pieceIndex = 0; pieceIndex < pieceElements.length; pieceIndex += 1) {
+    final pieceElement = pieceElements[pieceIndex];
+    pieces.add(ExampleSentencePiece(
+      lifted: pieceElement.querySelector('.furigana')?.text,
+      unlifted: pieceElement.querySelector('.unlinked')?.text,
+    ));
+  }
+
+  return pieces;
+}
+
+ExampleResultData parseExampleDiv(Element div) {
+  final result = getKanjiAndKana(div);
+  result.english = div.querySelector('.english').text;
+  result.pieces = getPieces(div) ?? [];
+
+  return result;
+}
+
+ExampleResults parseExamplePageData(String pageHtml, String phrase) {
+  final document = parse(pageHtml);
+  final divs = document.querySelectorAll('.sentence_content');
+
+  final results = divs.map((div) => parseExampleDiv(div)).toList();
+
+  return ExampleResults(
+    query: phrase,
+    found: results.isNotEmpty,
+    results: results ?? [],
+    uri: uriForExampleSearch(phrase),
+    phrase: phrase,
+  );
+}