Add combine_tables function and sakila dataset

[syroBx]

Created combine_tables function to create single cohort table from 2 or more tables for #44 .
And added sakila database to test this function.

Example use of the function:
After load database, create binding_keys

sakila_source <- set_source(as.tblist(sakila))

sakila_source$primary_keys <- primary_keys(
  data_key('actor', 'actor_id'),
  data_key('film', 'film_id'),
  data_key('category', 'category_id'),
  data_key('language', 'language_id')
)

sakila_source$binding_keys <- bind_keys(
  bind_key(update = data_key('actor', 'actor_id'),
           data_key('film_actor', 'actor_id')
  ),
  bind_key(update = data_key('film_actor','film_id'),
           data_key('film','film_id')
  ),
  bind_key(update = data_key('film_category','film_id'),
           data_key('film','film_id')
  ),
  bind_key(update = data_key('category','category_id'),
           data_key('film_category','category_id')
  ),
  bind_key(update = data_key('language','language_id'),
           data_key('film','language_id')
  ),
  bind_key(update = data_key('inventory','film_id'),
           data_key('film','film_id')
  ),
  bind_key(update = data_key('film_actor','actor_id'),
           data_key('actor', 'actor_id')
  ),
  bind_key(update = data_key('film','film_id'),
           data_key('film_actor','film_id')
  )
)

Next create cohort and run function:

coh <- Cohort$new(
  sakila_source,
  step(
    filter(
      type = "range", id = "actor_id", name = "actor.id",
      variable = "actor_id", dataset = "actor", range = c(5, 6)
    )
  ),
  run_flow = TRUE
)

result <- combine_tables(coh,c("actor","category"))

Result:

# A tibble: 2 × 5
  actor_id first_name last_name    last_update         actor_category   
     <dbl> <chr>      <chr>        <chr>               <list>           
1        5 JOHNNY     LOLLOBRIGIDA 2006-02-15 04:34:33 <tibble [13 × 3]>
2        6 BETTE      NICHOLSON    2006-02-15 04:34:33 <tibble [11 × 3]>

result <- combine_tables(coh, c("category","actor","film"))

Result:

# A tibble: 16 × 5
   category_id name        last_update         category_film     category_actor  
         <dbl> <chr>       <chr>               <list>            <list>          
 1           1 Action      2006-02-15 04:46:27 <tibble [4 × 13]> <tibble [2 × 4]>
 2           2 Animation   2006-02-15 04:46:27 <tibble [5 × 13]> <tibble [2 × 4]>
 3           3 Children    2006-02-15 04:46:27 <tibble [2 × 13]> <tibble [2 × 4]>
 4           4 Classics    2006-02-15 04:46:27 <tibble [1 × 13]> <tibble [1 × 4]>
 5           5 Comedy      2006-02-15 04:46:27 <tibble [1 × 13]> <tibble [1 × 4]>
 6           6 Documentary 2006-02-15 04:46:27 <tibble [5 × 13]> <tibble [2 × 4]>
 7           7 Drama       2006-02-15 04:46:27 <tibble [2 × 13]> <tibble [1 × 4]>
 8           8 Family      2006-02-15 04:46:27 <tibble [1 × 13]> <tibble [1 × 4]>
 9           9 Foreign     2006-02-15 04:46:27 <tibble [2 × 13]> <tibble [1 × 4]>
10          10 Games       2006-02-15 04:46:27 <tibble [1 × 13]> <tibble [1 × 4]>
11          11 Horror      2006-02-15 04:46:27 <tibble [3 × 13]> <tibble [1 × 4]>
12          12 Music       2006-02-15 04:46:27 <tibble [2 × 13]> <tibble [1 × 4]>
13          13 New         2006-02-15 04:46:27 <tibble [3 × 13]> <tibble [2 × 4]>
14          14 Sci-Fi      2006-02-15 04:46:27 <tibble [4 × 13]> <tibble [2 × 4]>
15          15 Sports      2006-02-15 04:46:27 <tibble [7 × 13]> <tibble [2 × 4]>
16          16 Travel      2006-02-15 04:46:27 <tibble [5 × 13]> <tibble [2 × 4]>

[syroBx]

Function to draw connection between binding_keys.
To work only need Cohort with defined binding_keys.