[BUG]cudf.get_dummies fails if symbols ( $,( ) are present in data

[VibhuJawa]

Describe the bug
cudf.get_dummies fails if dollar symbol is present in data

Steps/Code to reproduce bug

import cudf
df = cudf.DataFrame({"a":["$ 1", "$ 2"]})
df['a']=df['a'].astype('category')
cudf.get_dummies(df,columns=['a'])

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/tmp/ipykernel_26534/211579206.py in <module>
      2 df = cudf.DataFrame({"a":["$ 1", "$ 2"]})
      3 df['a']=df['a'].astype('category')
----> 4 cudf.get_dummies(df,columns=['a'])

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/reshape.py in get_dummies(df, prefix, prefix_sep, dummy_na, columns, cats, sparse, drop_first, dtype)
    700                 unique = _get_unique(column=df._data[name], dummy_na=dummy_na)
    701 
--> 702                 col_enc_df = df.one_hot_encoding(
    703                     name,
    704                     prefix=prefix_map.get(name, prefix),

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/dataframe.py in one_hot_encoding(self, column, prefix, cats, prefix_sep, dtype)
   3655             for cat in cats
   3656         ]
-> 3657         newcols = self[column].one_hot_encoding(cats=cats, dtype=dtype)
   3658         outdf = self.copy()
   3659         for name, col in zip(newnames, newcols):

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/series.py in one_hot_encoding(self, cats, dtype)
   3792                 return (self == cat).fillna(False)
   3793 
-> 3794         return [encode(cat).astype(dtype) for cat in cats]
   3795 
   3796     def label_encoding(self, cats, dtype=None, na_sentinel=-1):

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/series.py in <listcomp>(.0)
   3792                 return (self == cat).fillna(False)
   3793 
-> 3794         return [encode(cat).astype(dtype) for cat in cats]
   3795 
   3796     def label_encoding(self, cats, dtype=None, na_sentinel=-1):

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/series.py in encode(cat)
   3790                 return self.__class__(libcudf.unary.is_nan(self._column))
   3791             else:
-> 3792                 return (self == cat).fillna(False)
   3793 
   3794         return [encode(cat).astype(dtype) for cat in cats]

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/frame.py in __eq__(self, other)
   3580     # Binary rich comparison operations.
   3581     def __eq__(self, other):
-> 3582         return self._binaryop(other, "eq")
   3583 
   3584     def __ne__(self, other):

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/series.py in _binaryop(self, other, fn, fill_value, reflect, can_reindex, *args, **kwargs)
   1364 
   1365         # Note that we call the super on lhs, not self.
-> 1366         return super(Series, lhs)._binaryop(other, fn, fill_value, reflect)
   1367 
   1368     def add(self, other, fill_value=None, axis=0):

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/frame.py in _binaryop(self, other, fn, fill_value, reflect, *args, **kwargs)
   3958 
   3959         return self._copy_construct(
-> 3960             data=type(self)._colwise_binop(operands, fn)[result_name],
   3961             name=result_name,
   3962         )

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/frame.py in _colwise_binop(cls, operands, fn)
   3431                 )
   3432             elif not isinstance(right_column, ColumnBase):
-> 3433                 right_column = left_column.normalize_binop_value(right_column)
   3434 
   3435             fn_apply = fn

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/column/categorical.py in normalize_binop_value(self, other)
    931 
    932         ary = cudf.utils.utils.scalar_broadcast_to(
--> 933             self._encode(other), size=len(self), dtype=self.codes.dtype
    934         )
    935         col = column.build_categorical_column(

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/column/categorical.py in _encode(self, value)
   1046 
   1047     def _encode(self, value) -> ScalarLike:
-> 1048         return self.categories.find_first_value(value)
   1049 
   1050     def _decode(self, value: int) -> ScalarLike:

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/column/string.py in find_first_value(self, value, closest)
   5313         self, value: ScalarLike, closest: bool = False
   5314     ) -> int:
-> 5315         return self._find_first_and_last(value)[0]
   5316 
   5317     def find_last_value(self, value: ScalarLike, closest: bool = False) -> int:

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/column/string.py in _find_first_and_last(self, value)
   5306         found_indices = libstrings.contains_re(self, f"^{value}$")
   5307         found_indices = libcudf.unary.cast(found_indices, dtype=np.int32)
-> 5308         first = column.as_column(found_indices).find_first_value(np.int32(1))
   5309         last = column.as_column(found_indices).find_last_value(np.int32(1))
   5310         return first, last

/nvme/0/vjawa/conda/envs/rapids-21.08/lib/python3.8/site-packages/cudf/core/column/numerical.py in find_first_value(self, value, closest)
    421                     raise ValueError("value not found")
    422         elif found == -1:
--> 423             raise ValueError("value not found")
    424         return found
    425 

ValueError: value not found

Expected behavior

I would expect it to work.

Additional context
Looking at the trace it seems like we are not handling regex correctly at line here:

 found_indices = libstrings.contains_re(self, f"^{value}$")

[VibhuJawa]

Also breaks for ( this sybmol.