reworked document tests

Author: ahmednfwela

lib/src/index.dart

@@ -79,7 +79,12 @@ abstract class MeiliSearchIndex {
 
   /// {@macro meili.add_docs}
   ///
-  /// * The passed [documents] must be a valid CSV string, where each line corresponds to an object.
+  /// *
+  /// {@template meili.csv}
+  /// The passed documents must be a valid CSV string, where the first line contains objects' keys and types, and each subsequent line corresponds to an object.
+  /// [see relevant documentation](https://docs.meilisearch.com/learn/core_concepts/documents.html#csv)
+  /// {@endtemplate}
+  /// 
   /// *
   /// {@macro meili.index_upsert}
   Future<Task> addDocumentsCsv(
@@ -110,7 +115,8 @@ abstract class MeiliSearchIndex {
 
   /// {@macro meili.add_docs_batches}
   ///
-  /// * The passed [documents] must be a valid CSV string, where each line corresponds to an object.
+  /// *
+  /// {@macro meili.csv}
   /// *
   /// {@macro meili.index_upsert}
   Future<List<Task>> addDocumentsCsvInBatches(
@@ -162,7 +168,8 @@ abstract class MeiliSearchIndex {
 
   /// {@macro meili.update_docs}
   ///
-  /// * The passed [documents] must be a valid CSV string, where each line corresponds to an object.
+  /// * 
+  /// {@macro meili.csv}
   /// *
   /// {@macro meili.index_upsert}
   Future<Task> updateDocumentsCsv(

lib/src/index_impl.dart

@@ -237,11 +237,12 @@ class MeiliSearchIndexImpl implements MeiliSearchIndex {
   }) {
     final ls = LineSplitter();
     final split = ls.convert(documents);
-
+    //header is shared for all slices
+    final header = split.first;
     return Future.wait(
-      split.slices(batchSize).map(
+      split.skip(1).slices(batchSize).map(
             (slice) => addDocumentsCsv(
-              slice.join('\n'),
+              [header, ...slice].join('\n'),
               primaryKey: primaryKey,
             ),
           ),
@@ -343,11 +344,13 @@ class MeiliSearchIndexImpl implements MeiliSearchIndex {
   }) {
     final ls = LineSplitter();
     final split = ls.convert(documents);
+    //header is shared for all slices
+    final header = split.first;
 
     return Future.wait(
-      split.slices(batchSize).map(
+      split.skip(1).slices(batchSize).map(
             (slice) => updateDocumentsCsv(
-              slice.join('\n'),
+              [header, ...slice].join('\n'),
               primaryKey: primaryKey,
             ),
           ),