(backported from v2.5)

no actual change

[makedoc]

See #435

[makedoc]

(cherry picked from commit 781e5677)

[makedoc]

see 8ed9697a

Can be found with the following python snippet

import re
plumedDictionary={}
with open("/opt/local/lib/plumed/vim/syntax/plumed.vim") as fin:
    for line in fin:
        if re.match("^let b:plumedDictionary\[.*$",line):
            line=re.sub("^let b:","",line)
            exec(line,{'__builtins__': None},{'plumedDictionary':plumedDictionary})
for key in plumedDictionary.keys():
    diction={}
    for opt in plumedDictionary[key]:
        if(opt['word'] in diction):
            print(key,opt['word'])
        diction[opt['word']]=1

In addition to numpy arrays, I make it possible to pass plain python arrays.

I think the inclusion of numpy headers was related to non used variables
forgotten from an intermediate implementation.
I just removed them. In this manner it is not necessary that numpy is
already installed when plumed extension is compiled.

Numpy would be anyway required to run our test.

I think this is more robust. If an MD code uses a negative value with a
specific meaning (e.g. -1 to imply some choice) casting to unsigned
will translate -1 to 4294967295, which is positive

This fixes #446

Requires adding a cpp-suppress