perf: Improve rename performace for Lazy API

[siddharth-vi]

Closes #17336

For the python rename API, if we pass a dictionary arguments, we convert to LazyFrame and internally call the rename operation; the issue is that we need to get index number for all the columns we are renaming; this is an O(n^2) operation.
If we pass a function as argument to rename, we convert to LazyFrame and internally call the select operation; here we get the index for each column from the Schema, which is an IndexMap and hence overall the operation becomes O(n).
Have tried to fix this issue in a similar fashion by constructing the schema first and then getting index from it.

After the fix :

import polars as pl
df = pl.DataFrame({f"column_{i}": ['x', 'y', 'z'] for i in range(10000)})
cols_map = {f"column_{i}": f"renamed_{i}" for i in range(10000)}
start = time.perf_counter()
_ = df.rename(cols_map)
end = time.perf_counter()
print(f"Time for rename(): {end - start} | Version: {pl.__version__}")

Time for rename(): 0.041061488911509514 | Version: 1.7.1

df2 = df.rename(lambda col: cols_map.get(col, col))
end = time.perf_counter()
print(f"Time for rename(): {end - start} | Version: {pl.__version__}")

Time for rename(): 0.06575666600838304 | Version: 1.7.1

In the Eager API for Rust, this was fixed in #1028 by implementing a hash map.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 79.87%. Comparing base (dddf0b7) to head (314929a).
Report is 75 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #18890      +/-   ##
==========================================
- Coverage   79.88%   79.87%   -0.02%     
==========================================
  Files        1513     1520       +7     
  Lines      203466   206558    +3092     
  Branches     2892     2906      +14     
==========================================
+ Hits       162546   164985    +2439     
- Misses      40372    41025     +653     
  Partials      548      548              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[coastalwhite]

LGTM. If you could quickly fix the clippy error, I feel like this is quite okay to merge.