Catch a bug then write a test

[drizk1]

Alright this PR got a little bigger than expected ... but here we are

• Total of 62 tests

• adds support for multiple joins using TidierDB queries to be chained together.

  • Inspiration was from this duckplyr blog post by Hannes Mühleisen .. at some point I will download the 1gb and reprex it.
  • adds tests for this too (2 left joins a well as a full join then right join)

• adds mdy, ymd, dmy and ability to easily add time intervals

  • had to deviate a bit from tidierdata/julia logic here because of conflicts with sql ie day(1) is used in different ways for base sql for extractions, so it's just date_col + interval2days for example and it will work - not ideal but most accessible method without lot of parsing needing to be written - examples in the tests but will add docs at some point
  • will possibly go down the rabbit hole of parsing but it would require recognizing the call of - or + and then each unit of interval because the interval would otherwise conflict with the days
  • adds tests for dates, filtering on dates, conditionals with dates

• fixes edge case with joining 2 summarized tables without any CTEs

Only for DuckDB (because other sqls dont support)

• summarize can be used to quickly preview a table when no other args are provided ie

julia> DB.@collect DB.@summarize(DB.t(test_db))
4×12 DataFrame
 Row │ column_name  column_type  min     max     approx_unique  avg      std                 q25      q50      ⋯
     │ String       String       String  String  Int64          String?  String?             String?  String?  ⋯
─────┼──────────────────────────────────────────────────────────────────────────────────────────────────────────
   1 │ id           VARCHAR      AA      AJ                 11  missing  missing             missing  missing  ⋯
   2 │ groups       VARCHAR      aa      bb                  2  missing  missing             missing  missing 
   3 │ value        BIGINT       1       5                   5  3.0      1.4907119849998598  2        3
   4 │ percent      DOUBLE       0.1     1.0                10  0.55     0.3027650354097492  0.3      0.55
                                                                                               3 columns omitted