Fixing how the backend uses data types

Highlights

• The new ServiceX background is better about paying attention to the results-format tag - which was not being properly handled by the front-end stack. This fixes that up.
• Small fixes to metadata, package version limits, etc.

What's Changed

• Be consistent about root file type by @gordonwatts in #57
• New Release Prep by @gordonwatts in #58

Full Changelog: 2.1...2.2

Using file types properly

Fixing up how we propagate file types through the various layers. Should have no impact for anyone using this.

Getting AsParquetFiles to work correctly with the uproot backend

The main goal of this release is to make sure that AsParquetFiles has the right semantics and is compatible with the uproot transformer.

As servicex moves towards having more backend file flexibility, this interface may change (and generate some depreciated warnings). Note that best practices (which allows for most flexibility) is to rewrite something like this:

import pandas as pd
from func_adl_servicex import ServiceXSourceUpROOT

dataset_name = "data15_13TeV:data15_13TeV.00282784.physics_Main.deriv.DAOD_PHYSLITE.r9264_p3083_p4165_tid21568807_00"
src = ServiceXSourceUpROOT(dataset_name, "CollectionTree")
data = src.Select("lambda e: e['AnalysisJetsAuxDyn.pt']") \
    .AsParquetFiles('junk.parquet', ['JetPT']) \
    .value()
df = pd.read_parquet(data[0])
print(df)

as this:

import pandas as pd
from func_adl_servicex import ServiceXSourceUpROOT

dataset_name = "data15_13TeV:data15_13TeV.00282784.physics_Main.deriv.DAOD_PHYSLITE.r9264_p3083_p4165_tid21568807_00"
src = ServiceXSourceUpROOT(dataset_name, "CollectionTree")
data = src.Select("lambda e: {'JetPT': e['AnalysisJetsAuxDyn.pt']}") \
    .AsParquetFiles('junk.parquet') \
    .value()
df = pd.read_parquet(data[0])
print(df)

In short - use dictionaries in your query to label the columns that come back.

What's Changed

• Fix markdown syntax in readme by @klieret in #48
• Make sure AsParquetFiles works correctly with uproot by @gordonwatts in #53
• Unclear sentence in readme by @gordonwatts in #54

New Contributors

• @klieret made their first contribution in #48

Full Changelog: 2.0...2.1

Uproot and AsParquetFiles interoperability

What's Changed

• Fix markdown syntax in readme by @klieret in #48
• Make sure AsParquetFiles works correctly with uproot by @gordonwatts in #53

New Contributors

• @klieret made their first contribution in #48

Full Changelog: 2.0...2.1b1

Typed Datasets, Local Executor

This is a new release! Major improvements:

• Datasets can be typed, allowing type checkers like pylance and mypy to follow types through the system. This should aid in the editing experience.
• A local executor for xAOD's has been added.

What's Changed

• Typed Datasets & Local Executor by @gordonwatts in #37
• No Col Specified, No Col to access by @gordonwatts in #46
• Fix up the README by @gordonwatts in #45

Full Changelog: 1.1.3...2.0

Better Single Column Access

What's Changed

• No Col Specified, No Col to access by @gordonwatts in #46
• Fix up the README by @gordonwatts in #45

Full Changelog: 2.0b1...2.0b2

Typed Datasets, Local Executor

What's Changed

• Typed Datasets & Local Executor by @gordonwatts in #37

Full Changelog: 1.1.3...2.0b1

Support Python 3.10

Add Python 3.10 support

Typed Datasets and MetaData

• Typed datasets
• Metadata can be sent down to the backend

Bug Fix: dealing with uproot column names backend

• Uproot transformer can't deal with explicit column names - so this code was removed. User might still mis-use this, and cause a crash...