ngrams reindexing: add timer (diagnostic: my functionis too slow, negative effect on list importing and free annotation add)

docs view: prepare feature for remote search in abstracts (will be useful when we do abstract indexing in sql etc)

(adds the new 'source' names that were formerly known as 'journal')

[FIX] server cache clean after detached instance error (should make the exception not critical anymore)

[FEAT] new param 'hyperdata_filter' for the main nodes API (ex: api/nodes?types[]=DOCUMENT&fields[]=hyperdata&hyperdata_filter[]=title)

[FIX] API PUT ngram: new param 'testgroup' can test not only if ngram exists but also if ngram has a mainform

[FIX] new ngram indexing: fix cases where ngram was already partially indexed (ex: if extraction was only correct in some documents) (todo: change bulk_insert_ifnotexists for cases like this)

[GRAPH] Number of nodes need to be saved during the graph generation (and not only with cooc matrix); in order to keep all parameters.

add corpus: reactivate load button for file corpus via minimalist a posteriori validation checks in checkReady function (en attendant mieux :) )