Merging a dask dataframe with a pandas series fails

```python
In [1]: import dask

In [2]: import dask_expr as dd
...: df = dd.datasets.timeseries()

In [3]: df.head()
Out[3]:
name id x y
timestamp
2000-01-01 00:00:00 Michael 1006 0.927520 -0.442859
2000-01-01 00:00:01 Kevin 1018 -0.411144 -0.037667
2000-01-01 00:00:02 Yvonne 974 -0.648850 -0.515754
2000-01-01 00:00:03 Yvonne 994 0.463103 0.560937
2000-01-01 00:00:04 Yvonne 1002 -0.511311 -0.308211

In [4]: df.merge(df.name.head())
```
```python-traceback
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
in ?()
----> 1 df.merge(df.name.head())

~/workspace/dask-expr/dask_expr/_collection.py in ?(self, right, how, on, left_on, right_on, left_index, right_index, suffixes, indicator, shuffle_method, npartitions, broadcast)
2572 an internal ``shuffle``, because shuffling places all rows that have the same
2573 index in the same partition. To avoid this error, make sure all rows with the
2574 same ``on``-column value can fit on a single partition.
2575 """
-> 2576 return merge(
2577 self,
2578 right,
2579 how,

~/workspace/dask-expr/dask_expr/_collection.py in ?(left, right, how, on, left_on, right_on, left_index, right_index, suffixes, indicator, shuffle_method, npartitions, broadcast)
4779 for o in [on, left_on, right_on]:
4780 if isinstance(o, FrameBase):
4781 raise NotImplementedError()
4782 if not on and not left_on and not right_on and not left_index and not right_index:
-> 4783 on = [c for c in left.columns if c in right.columns]
4784 if not on:
4785 left_index = right_index = True
4786

~/workspace/dask-expr/dask_expr/_collection.py in ?(.0)
-> 4783 def merge(
4784 left,
4785 right,
4786 how="inner",

~/mambaforge/envs/etl-tpch/lib/python3.11/site-packages/pandas/core/generic.py in ?(self, name)
6289 and name not in self._accessors
6290 and self._info_axis._can_hold_identifiers_and_holds_name(name)
6291 ):
6292 return self[name]
-> 6293 return object.__getattribute__(self, name)

AttributeError: 'Series' object has no attribute 'columns'
```
```python
In [5]: df.head().merge(df.name.head())
Out[5]:
name id x y
0 Michael 1006 0.927520 -0.442859
1 Kevin 1018 -0.411144 -0.037667
2 Yvonne 974 -0.648850 -0.515754
3 Yvonne 974 -0.648850 -0.515754
4 Yvonne 974 -0.648850 -0.515754
5 Yvonne 994 0.463103 0.560937
6 Yvonne 994 0.463103 0.560937
7 Yvonne 994 0.463103 0.560937
8 Yvonne 1002 -0.511311 -0.308211
9 Yvonne 1002 -0.511311 -0.308211
10 Yvonne 1002 -0.511311 -0.308211
```

Reproduce df.merge(df.name.head()) from the issue, then inspect the merge entry point in dask_expr/_collection.py around line 4783. Done means merging the Dask dataframe with the pandas Series no longer raises AttributeError and produces behavior consistent with the shown pandas result.

Written by the indexing model from the issue text.