If an arena in the MemPoolForRooSets is not referenced anymore, it
should delete the memory it has allocated.

When doing I/O of RVec objects, TClassEdit::STLArgs was accessing
an element one after the end of a static array. asan rightly complains.

This commit fixes https://github.com/root-project/root/issues/7903,
which contains more details.

Deleting a chunk of repeated comment in the description of TH1::GetMean()

Before this patch, given a TTree with a branch with name _different_
from its leaf, e.g. like this:

```
*Br    0 :NUD_total_ADC : nud_total_adc/D
```

RDataFrame only added "branchname.leafname" to the list of available
columns (i.e. "NUD_total_ADC.nud_total_adc" in the example).

In comparison, in a similar situation `TTree::Draw` also accepts
just "NUD_total_ADC" as it assumes that the desired leaf is the
first sub-leaf of the specified branch.

With this patch, RDataFrame also considers "NUD_total_ADC" as a valid
column name as long as it has only one sub-leaf.

This fixes ROOT-9558.

* Port libAfterImage on Win64

Add `libAfterImage/CMakeLists.txt` to facilitate the port of libAfterImage on Win64

* Update CMake minimal version (thanks Oksana)

The distributed RDataFrame constructor accepts an optional `npartitions` keyword argument. Previously, if this argument was provided by the user, it set the number of partitions in which the rdf would split the distributed computations.
But then, right before starting the execution, the distributed backend implementation tried to optimize this number. In the case of Spark, an educated guess for the number of partitions would be spark.executor.cores * spark.executor.instances, that is the number of distributed nodes times the number of cores used for each node.
If we let this optimization happen just before the start of the execution, it means we completely disregard the user provided value for `npartitions`. Instead, the backend guessing at a number of partitions should happen only if the user doesn't supply one.
This commit addresses the issue by moving the call to `backend.optimize_npartitions` inside the initialization of the distributed dataframe object, plus adds a couple of tests to check the behaviour in the Spark backend.

Solved missing include of std::string referenced in error handling. Needed to compile using clang and libc++

Is now proportional to the browser size.

Setting the move constructor and move assignment operator to = default
causes linker errors because the copy assignment operator is not
implemented in the TIterator base class.

Having the RooLinkedListIter move constructor and assignment operators
work is important to keep user code compatibility.

Rename RooHistFunc::getObservables to not hide
RooAbsArg::getObservables.

Template taken from root-project/root's issue template

Before this patch, given a TTree with a branch with name different
from its leaf, e.g. like this:

```
*Br    0 :NUD_total_ADC : nud_total_adc/D
```

TTreeReaderValue failed to retrieve the leaf when the named passed
to the constructor was just "NUD_total_ADC" (while it worked fine
with "NUD_total_ADC.nud_total_adc"). In comparison, in a similar
situation `TTree::Draw` "tries harder" and it assumes that the
desired leaf is the first sub-leaf of the specified branch.

With this patch, TTreeReaderValue tries `FindLeaf` after `GetLeaf`
and as a last resort it picks the branch sub-leaf if it exists and
it is unique.

This fixes #6881.

* [core] Next bunch of changes for Win64

- Replace some more `long` types by `Longptr_t`
- Fix pointer formatting (use `%zx` and `size_t` for architecture dependent format)
- Fix several `warning C4267: 'argument': conversion from 'size_t' to 'Ssiz_t', possible loss of data`

* [skip-ci] Fix compilation warning on i386 platforms

Fix warning: format ‘%zx’ expects argument of type ‘size_t’, but argument 2 has type ‘Long_t’ {aka ‘long int’} [-Wformat=]

The custom hash table implementation RooHashTable is only used in
RooLinkedList for a transient data member. By using std::unordered_map
instead, RooHashTable gets obsolete and could be deprecated.

After use of asign operator some fields can be destroyed twice

This PR makes the TFile Layout documentation visible in
the Doxygen ROOT reference Guide.

- The .txt file have been turned into markdown
- Doxyfile have io/doc/TFile folder as input
- References/links between the pages have been setup
- The \tableofcontents Doxygen functionality  has been turned ON
- Some typos have been fixed

Still two links need to be resolved:
  http://root.cern.ch/root/RootDoc.html
  http://root.cern.ch/root/htmldoc/examples/Version302.news.html

Events are double counted if one accidentally defines overlapping ranges
and uses them in a multi-range fit. This happened for example in Jira
issue ROOT-9548 where the whole dataset was double counted, leading to
the parameter uncertainties being underestimated by a factor of sqrt(2).

That situation should be avoided. This commit introduces a check for
overlapping ranges before the multi-range likelihood is created.

The RooFit test statistic classes give wrong results if they are
constructed from multiple comma-separated ranges. This can cause errors,
such as the one reported in Jira issue ROOT-10038.

This commit suggests to throw an exception if multiple comma-separated
ranges are detected.

Fix many warnings, use jquery dialog instead of prompt() or alert()

Let run ping.cxx from any other subdir