Use LargeStringArray for casting when writing tables to CSV

Signed-off-by: Tao He <sighingnow@gmail.com>

cpp/src/arrow/csv/writer.cc

@@ -130,8 +130,8 @@ class ColumnPopulator {
     ctx.set_use_threads(false);
     ASSIGN_OR_RAISE(
         std::shared_ptr<Array> casted,
-        compute::Cast(data, /*to_type=*/utf8(), compute::CastOptions(), &ctx));
-    casted_array_ = checked_pointer_cast<StringArray>(casted);
+        compute::Cast(data, /*to_type=*/large_utf8(), compute::CastOptions(), &ctx));
+    casted_array_ = checked_pointer_cast<LargeStringArray>(casted);
     return UpdateRowLengths(row_lengths);
   }
 
@@ -146,7 +146,7 @@ class ColumnPopulator {
 
  protected:
   virtual Status UpdateRowLengths(int64_t* row_lengths) = 0;
-  std::shared_ptr<StringArray> casted_array_;
+  std::shared_ptr<LargeStringArray> casted_array_;
   const std::string end_chars_;
   std::shared_ptr<Buffer> null_string_;
 
@@ -228,7 +228,7 @@ class UnquotedColumnPopulator : public ColumnPopulator {
 
  private:
   // Returns an error status if string array has any structural characters.
-  static Status CheckStringArrayHasNoStructuralChars(const StringArray& array,
+  static Status CheckStringArrayHasNoStructuralChars(const LargeStringArray& array,
                                                      const char delimiter) {
     // scan the underlying string array buffer as a single big string
     const uint8_t* const data = array.raw_data() + array.value_offset(0);
@@ -282,14 +282,14 @@ class QuotedColumnPopulator : public ColumnPopulator {
       : ColumnPopulator(pool, std::move(end_chars), std::move(null_string)) {}
 
   Status UpdateRowLengths(int64_t* row_lengths) override {
-    const StringArray& input = *casted_array_;
+    const LargeStringArray& input = *casted_array_;
 
     row_needs_escaping_.resize(casted_array_->length(), false);
 
     if (NoQuoteInArray(input)) {
       // fast path if no quote
       int row_number = 0;
-      VisitArraySpanInline<StringType>(
+      VisitArraySpanInline<LargeStringType>(
           *input.data(),
           [&](std::string_view s) {
             row_lengths[row_number] += static_cast<int64_t>(s.length()) + kQuoteCount;
@@ -301,7 +301,7 @@ class QuotedColumnPopulator : public ColumnPopulator {
           });
     } else {
       int row_number = 0;
-      VisitArraySpanInline<StringType>(
+      VisitArraySpanInline<LargeStringType>(
           *input.data(),
           [&](std::string_view s) {
             // Each quote in the value string needs to be escaped.
@@ -321,7 +321,7 @@ class QuotedColumnPopulator : public ColumnPopulator {
 
   Status PopulateRows(char* output, int64_t* offsets) const override {
     auto needs_escaping = row_needs_escaping_.begin();
-    VisitArraySpanInline<StringType>(
+    VisitArraySpanInline<LargeStringType>(
         *(casted_array_->data()),
         [&](std::string_view s) {
           // still needs string content length to be added
@@ -355,7 +355,7 @@ class QuotedColumnPopulator : public ColumnPopulator {
 
  private:
   // Returns true if there's no quote in the string array
-  static bool NoQuoteInArray(const StringArray& array) {
+  static bool NoQuoteInArray(const LargeStringArray& array) {
     const uint8_t* data = array.raw_data() + array.value_offset(0);
     const int64_t buffer_size = array.total_values_length();
     return std::memchr(data, '"', buffer_size) == nullptr;