Merge branch 'branch-24.12' into refactor-dask-cudf

cpp/include/cudf/join.hpp

@@ -573,7 +573,7 @@ class distinct_hash_join {
  * Result: {{1}, {0}}
  * @endcode
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -620,7 +620,7 @@ conditional_inner_join(table_view const& left,
  * Result: {{0, 1, 2}, {None, 0, None}}
  * @endcode
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -666,7 +666,7 @@ conditional_left_join(table_view const& left,
  * Result: {{0, 1, 2, None, None}, {None, 0, None, 1, 2}}
  * @endcode
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -705,7 +705,7 @@ conditional_full_join(table_view const& left,
  * Result: {1}
  * @endcode
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -746,7 +746,7 @@ std::unique_ptr<rmm::device_uvector<size_type>> conditional_left_semi_join(
  * Result: {0, 2}
  * @endcode
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -793,7 +793,7 @@ std::unique_ptr<rmm::device_uvector<size_type>> conditional_left_anti_join(
  * Result: {{1}, {0}}
  * @endcode
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -855,7 +855,7 @@ mixed_inner_join(
  * Result: {{0, 1, 2}, {None, 0, None}}
  * @endcode
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -917,7 +917,7 @@ mixed_left_join(
  * Result: {{0, 1, 2, None, None}, {None, 0, None, 1, 2}}
  * @endcode
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -972,7 +972,7 @@ mixed_full_join(
  * Result: {1}
  * @endcode
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -1022,7 +1022,7 @@ std::unique_ptr<rmm::device_uvector<size_type>> mixed_left_semi_join(
  * Result: {0, 2}
  * @endcode
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -1061,7 +1061,7 @@ std::unique_ptr<rmm::device_uvector<size_type>> mixed_left_anti_join(
  * choose a suitable compare_nulls value AND use appropriate null-safe
  * operators in the expression.
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -1103,7 +1103,7 @@ std::pair<std::size_t, std::unique_ptr<rmm::device_uvector<size_type>>> mixed_in
  * choose a suitable compare_nulls value AND use appropriate null-safe
  * operators in the expression.
  *
- * @throw cudf::logic_error If the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error If the binary predicate outputs a non-boolean result.
  * @throw cudf::logic_error If the number of rows in left_equality and left_conditional do not
  * match.
  * @throw cudf::logic_error If the number of rows in right_equality and right_conditional do not
@@ -1142,7 +1142,7 @@ std::pair<std::size_t, std::unique_ptr<rmm::device_uvector<size_type>>> mixed_le
  * If the provided predicate returns NULL for a pair of rows
  * (left, right), that pair is not included in the output.
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -1167,7 +1167,7 @@ std::size_t conditional_inner_join_size(
  * If the provided predicate returns NULL for a pair of rows
  * (left, right), that pair is not included in the output.
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -1192,7 +1192,7 @@ std::size_t conditional_left_join_size(
  * If the provided predicate returns NULL for a pair of rows
  * (left, right), that pair is not included in the output.
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table
@@ -1217,7 +1217,7 @@ std::size_t conditional_left_semi_join_size(
  * If the provided predicate returns NULL for a pair of rows
  * (left, right), that pair is not included in the output.
  *
- * @throw cudf::logic_error if the binary predicate outputs a non-boolean result.
+ * @throw cudf::data_type_error if the binary predicate outputs a non-boolean result.
  *
  * @param left The left table
  * @param right The right table

cpp/src/join/conditional_join.cu

@@ -28,6 +28,7 @@
 #include <cudf/table/table_device_view.cuh>
 #include <cudf/table/table_view.hpp>
 #include <cudf/types.hpp>
+#include <cudf/utilities/error.hpp>
 #include <cudf/utilities/memory_resource.hpp>
 
 #include <rmm/cuda_stream_view.hpp>
@@ -178,7 +179,8 @@ conditional_join(table_view const& left,
   auto const parser =
     ast::detail::expression_parser{binary_predicate, left, right, has_nulls, stream, mr};
   CUDF_EXPECTS(parser.output_type().id() == type_id::BOOL8,
-               "The expression must produce a boolean output.");
+               "The expression must produce a boolean output.",
+               cudf::data_type_error);
 
   auto left_table  = table_device_view::create(left, stream);
   auto right_table = table_device_view::create(right, stream);
@@ -330,7 +332,8 @@ std::size_t compute_conditional_join_output_size(table_view const& left,
   auto const parser =
     ast::detail::expression_parser{binary_predicate, left, right, has_nulls, stream, mr};
   CUDF_EXPECTS(parser.output_type().id() == type_id::BOOL8,
-               "The expression must produce a boolean output.");
+               "The expression must produce a boolean output.",
+               cudf::data_type_error);
 
   auto left_table  = table_device_view::create(left, stream);
   auto right_table = table_device_view::create(right, stream);

cpp/src/join/mixed_join.cu

@@ -28,6 +28,7 @@
 #include <cudf/table/table_device_view.cuh>
 #include <cudf/table/table_view.hpp>
 #include <cudf/types.hpp>
+#include <cudf/utilities/error.hpp>
 #include <cudf/utilities/memory_resource.hpp>
 #include <cudf/utilities/span.hpp>
 
@@ -115,7 +116,8 @@ mixed_join(
   auto const parser = ast::detail::expression_parser{
     binary_predicate, left_conditional, right_conditional, has_nulls, stream, mr};
   CUDF_EXPECTS(parser.output_type().id() == type_id::BOOL8,
-               "The expression must produce a boolean output.");
+               "The expression must produce a boolean output.",
+               cudf::data_type_error);
 
   // TODO: The non-conditional join impls start with a dictionary matching,
   // figure out what that is and what it's needed for (and if conditional joins
@@ -381,7 +383,8 @@ compute_mixed_join_output_size(table_view const& left_equality,
   auto const parser = ast::detail::expression_parser{
     binary_predicate, left_conditional, right_conditional, has_nulls, stream, mr};
   CUDF_EXPECTS(parser.output_type().id() == type_id::BOOL8,
-               "The expression must produce a boolean output.");
+               "The expression must produce a boolean output.",
+               cudf::data_type_error);
 
   // TODO: The non-conditional join impls start with a dictionary matching,
   // figure out what that is and what it's needed for (and if conditional joins

python/cudf/cudf/_lib/csv.pyx

@@ -28,7 +28,7 @@ from pylibcudf.libcudf.io.csv cimport (
     write_csv as cpp_write_csv,
 )
 from pylibcudf.libcudf.io.data_sink cimport data_sink
-from pylibcudf.libcudf.io.types cimport compression_type, sink_info
+from pylibcudf.libcudf.io.types cimport sink_info
 from pylibcudf.libcudf.table.table_view cimport table_view
 
 from cudf._lib.io.utils cimport make_sink_info
@@ -148,13 +148,13 @@ def read_csv(
         byte_range = (0, 0)
 
     if compression is None:
-        c_compression = compression_type.NONE
+        c_compression = plc.io.types.CompressionType.NONE
     else:
         compression_map = {
-            "infer": compression_type.AUTO,
-            "gzip": compression_type.GZIP,
-            "bz2": compression_type.BZIP2,
-            "zip": compression_type.ZIP,
+            "infer": plc.io.types.CompressionType.AUTO,
+            "gzip": plc.io.types.CompressionType.GZIP,
+            "bz2": plc.io.types.CompressionType.BZIP2,
+            "zip": plc.io.types.CompressionType.ZIP,
         }
         c_compression = compression_map[compression]
 

python/cudf/cudf/_lib/json.pyx

@@ -9,10 +9,6 @@ from cudf.core.buffer import acquire_spill_lock
 
 from libcpp cimport bool
 
-cimport pylibcudf.libcudf.io.types as cudf_io_types
-from pylibcudf.io.types cimport compression_type
-from pylibcudf.libcudf.io.json cimport json_recovery_mode_t
-from pylibcudf.libcudf.io.types cimport compression_type
 from pylibcudf.libcudf.types cimport data_type, type_id
 from pylibcudf.types cimport DataType
 
@@ -24,15 +20,6 @@ from cudf._lib.utils cimport _data_from_columns, data_from_pylibcudf_io
 import pylibcudf as plc
 
 
-cdef json_recovery_mode_t _get_json_recovery_mode(object on_bad_lines):
-    if on_bad_lines.lower() == "error":
-        return json_recovery_mode_t.FAIL
-    elif on_bad_lines.lower() == "recover":
-        return json_recovery_mode_t.RECOVER_WITH_NULL
-    else:
-        raise TypeError(f"Invalid parameter for {on_bad_lines=}")
-
-
 cpdef read_json(object filepaths_or_buffers,
                 object dtype,
                 bool lines,
@@ -41,7 +28,7 @@ cpdef read_json(object filepaths_or_buffers,
                 bool keep_quotes,
                 bool mixed_types_as_string,
                 bool prune_columns,
-                object on_bad_lines):
+                str on_bad_lines):
     """
     Cython function to call into libcudf API, see `read_json`.
 
@@ -64,19 +51,24 @@ cpdef read_json(object filepaths_or_buffers,
             filepaths_or_buffers[idx] = filepaths_or_buffers[idx].encode()
 
     # Setup arguments
-    cdef cudf_io_types.compression_type c_compression
-
     if compression is not None:
         if compression == 'gzip':
-            c_compression = cudf_io_types.compression_type.GZIP
+            c_compression = plc.io.types.CompressionType.GZIP
         elif compression == 'bz2':
-            c_compression = cudf_io_types.compression_type.BZIP2
+            c_compression = plc.io.types.CompressionType.BZIP2
         elif compression == 'zip':
-            c_compression = cudf_io_types.compression_type.ZIP
+            c_compression = plc.io.types.CompressionType.ZIP
         else:
-            c_compression = cudf_io_types.compression_type.AUTO
+            c_compression = plc.io.types.CompressionType.AUTO
+    else:
+        c_compression = plc.io.types.CompressionType.NONE
+
+    if on_bad_lines.lower() == "error":
+        c_on_bad_lines = plc.io.types.JSONRecoveryMode.FAIL
+    elif on_bad_lines.lower() == "recover":
+        c_on_bad_lines = plc.io.types.JSONRecoveryMode.RECOVER_WITH_NULL
     else:
-        c_compression = cudf_io_types.compression_type.NONE
+        raise TypeError(f"Invalid parameter for {on_bad_lines=}")
 
     processed_dtypes = None
 
@@ -108,7 +100,7 @@ cpdef read_json(object filepaths_or_buffers,
             keep_quotes = keep_quotes,
             mixed_types_as_string = mixed_types_as_string,
             prune_columns = prune_columns,
-            recovery_mode = _get_json_recovery_mode(on_bad_lines)
+            recovery_mode = c_on_bad_lines
         )
         df = cudf.DataFrame._from_data(
             *_data_from_columns(
@@ -130,7 +122,7 @@ cpdef read_json(object filepaths_or_buffers,
             keep_quotes = keep_quotes,
             mixed_types_as_string = mixed_types_as_string,
             prune_columns = prune_columns,
-            recovery_mode = _get_json_recovery_mode(on_bad_lines)
+            recovery_mode = c_on_bad_lines
         )
 
         df = cudf.DataFrame._from_data(