Refactor ColumnMethods and its subclasses to remove column argument and require parent argument

python/cudf/cudf/_lib/strings/__init__.py

@@ -0,0 +1,88 @@
+from cudf._lib.nvtext.edit_distance import edit_distance, edit_distance_matrix
+from cudf._lib.nvtext.generate_ngrams import (
+    generate_character_ngrams,
+    generate_ngrams,
+)
+from cudf._lib.nvtext.ngrams_tokenize import ngrams_tokenize
+from cudf._lib.nvtext.normalize import normalize_characters, normalize_spaces
+from cudf._lib.nvtext.replace import filter_tokens, replace_tokens
+from cudf._lib.nvtext.stemmer import (
+    LetterType,
+    is_letter,
+    is_letter_multi,
+    porter_stemmer_measure,
+)
+from cudf._lib.nvtext.subword_tokenize import subword_tokenize_vocab_file
+from cudf._lib.nvtext.tokenize import (
+    _count_tokens_column,
+    _count_tokens_scalar,
+    _tokenize_column,
+    _tokenize_scalar,
+    character_tokenize,
+    detokenize,
+)
+from cudf._lib.strings.attributes import (
+    code_points,
+    count_bytes,
+    count_characters,
+)
+from cudf._lib.strings.capitalize import capitalize, title
+from cudf._lib.strings.case import swapcase, to_lower, to_upper
+from cudf._lib.strings.char_types import (
+    filter_alphanum,
+    is_alnum,
+    is_alpha,
+    is_decimal,
+    is_digit,
+    is_lower,
+    is_numeric,
+    is_space,
+    is_upper,
+)
+from cudf._lib.strings.combine import (
+    concatenate,
+    join,
+    join_lists_with_column,
+    join_lists_with_scalar,
+)
+from cudf._lib.strings.contains import contains_re, count_re, match_re
+from cudf._lib.strings.convert.convert_fixed_point import to_decimal
+from cudf._lib.strings.convert.convert_floats import is_float
+from cudf._lib.strings.convert.convert_integers import is_integer
+from cudf._lib.strings.convert.convert_urls import url_decode, url_encode
+from cudf._lib.strings.extract import extract
+from cudf._lib.strings.find import (
+    contains,
+    contains_multiple,
+    endswith,
+    endswith_multiple,
+    find,
+    rfind,
+    startswith,
+    startswith_multiple,
+)
+from cudf._lib.strings.findall import findall
+from cudf._lib.strings.json import get_json_object
+from cudf._lib.strings.padding import PadSide, center, ljust, pad, rjust, zfill
+from cudf._lib.strings.replace import (
+    insert,
+    replace,
+    replace_multi,
+    slice_replace,
+)
+from cudf._lib.strings.replace_re import (
+    replace_multi_re,
+    replace_re,
+    replace_with_backrefs,
+)
+from cudf._lib.strings.split.partition import partition, rpartition
+from cudf._lib.strings.split.split import (
+    rsplit,
+    rsplit_record,
+    split,
+    split_record,
+)
+from cudf._lib.strings.strip import lstrip, rstrip, strip
+from cudf._lib.strings.substring import get, slice_from, slice_strings
+from cudf._lib.strings.translate import filter_characters, translate
+from cudf._lib.strings.wrap import wrap

python/cudf/cudf/_lib/strings/combine.pyx

@@ -23,15 +23,15 @@ from cudf._lib.cpp.strings.combine cimport (
 
 
 def concatenate(Table source_strings,
-                object py_separator,
-                object py_narep):
+                object sep,
+                object na_rep):
     """
     Returns a Column by concatenating strings column-wise in `source_strings`
-    with the specified `py_separator` between each column and
-    `na`/`None` values are replaced by `py_narep`
+    with the specified `sep` between each column and
+    `na`/`None` values are replaced by `na_rep`
     """
-    cdef DeviceScalar separator = py_separator.device_value
-    cdef DeviceScalar narep = py_narep.device_value
+    cdef DeviceScalar separator = sep.device_value
+    cdef DeviceScalar narep = na_rep.device_value
 
     cdef unique_ptr[column] c_result
     cdef table_view source_view = source_strings.data_view()
@@ -53,16 +53,16 @@ def concatenate(Table source_strings,
 
 
 def join(Column source_strings,
-         object py_separator,
-         object py_narep):
+         object sep,
+         object na_rep):
     """
     Returns a Column by concatenating strings row-wise in `source_strings`
-    with the specified `py_separator` between each column and
-    `na`/`None` values are replaced by `py_narep`
+    with the specified `sep` between each column and
+    `na`/`None` values are replaced by `na_rep`
     """
 
-    cdef DeviceScalar separator = py_separator.device_value
-    cdef DeviceScalar narep = py_narep.device_value
+    cdef DeviceScalar separator = sep.device_value
+    cdef DeviceScalar narep = na_rep.device_value
 
     cdef unique_ptr[column] c_result
     cdef column_view source_view = source_strings.view()

python/cudf/cudf/_lib/transpose.pyx

@@ -36,11 +36,10 @@ def transpose(Table source):
     if is_categorical_dtype(dtype):
         if any(not is_categorical_dtype(c.dtype) for c in source._columns):
             raise ValueError('Columns must all have the same dtype')
-        cats = list(c.cat().categories for c in source._columns)
-        cats = cudf.Series(cudf.concat(cats)).drop_duplicates()._column
+        cats = list(c.categories for c in source._columns)
+        cats = cudf.core.column.concat_columns(cats).unique()
         source = Table(index=source._index, data=[
-            (name, col.cat()._set_categories(
-                col.cat().categories, cats, is_unique=True).codes)
+            (name, col._set_categories(cats, is_unique=True).codes)
             for name, col in source._data.items()
         ])
     elif dtype.kind in 'OU':

python/cudf/cudf/_typing.py

@@ -28,3 +28,4 @@
 BinaryOperand = Union["cudf.Scalar", "cudf.core.column.ColumnBase"]
 
 DataFrameOrSeries = Union["cudf.Series", "cudf.DataFrame"]
+SeriesOrIndex = Union["cudf.Series", "cudf.core.index.BaseIndex"]

python/cudf/cudf/api/types.py

@@ -218,7 +218,7 @@ def _union_categoricals(
         sorted_categories = result_col.categories.sort_by_values(
             ascending=True
         )[0]
-        result_col = result_col.cat().reorder_categories(
+        result_col = result_col.reorder_categories(
             new_categories=sorted_categories
         )
 

python/cudf/cudf/core/column/__init__.py

@@ -1,4 +1,8 @@
 # Copyright (c) 2020-2021, NVIDIA CORPORATION.
+"""
+isort: skip_file
+"""
+
 
 from cudf.core.column.categorical import CategoricalColumn
 from cudf.core.column.column import (
@@ -12,6 +16,7 @@
     column_empty,
     column_empty_like,
     column_empty_like_same_mask,
+    concat_columns,
     deserialize_columns,
     full,
     serialize_columns,
@@ -27,3 +32,4 @@
     Decimal32Column,
     Decimal64Column,
 )
+from cudf.core.column.interval import IntervalColumn  # noqa: F401