Unify "Database" parameters in the "build" and "classify" workflow elements

To make work with the database parameters consistent in different NGS reads classification workflow elements the following modifications are required:

1. Rename the "Database URL" parameter of the "Build CLARK Database" element to "Database".
2. Rename parameters of the "Build Kraken Database" element:
   • "New database name" > "Database"
   • "Input database name" > "Input database" (the parameter is shown when "Shrink" mode is selected).
3. Currently database files are store is a tool folder. Add additional subfolders for the database. Thus, the "ngs_classification" folders structure, created by the installer, should be the following:

   / ngs_classification
     /clark
       /viral_database
       /bacteria_viral_database  
     /diamond
       /uniref
     /kraken
       /minikraken
     /refseq
      ...
     /taxonomy
      ...
   

4. If the corresponding data are available (= have been downloaded), specify default values for "classify" elements (leave an empty value, if the data are not available):
   • For Kraken-classify - the "minikraken" folder.
   • For CLARK-classify - the "bacteria_viral_database" folder. If it is not present, "viral_database".
   • For DIAMOND-classify - the "uniref/uniref50.dmnd" file. If it is not present, "uniref/uniref90.dmnd".
5. A custom dialog appear when selecting the database parameter for all three "build" elements and CLARK-classify. Replace this dialog by a common dialog.
6. Optionally, fix the issue, specified in UGENE-5835 comment (files inside a folder are not visible on Linux, on Mac OS X it's OK):

   database setup is unclear - one cannot see the folders content