Merged PRs

dolt

• 6847: has_ancestor() function
  Motivated by this query:

  select from_x, to_x, to_commit, tag_name as version from dolt_diff_xy,
  LATERAL (
  select tag_name
  from dolt_tags
  where has_ancestor(to_commit, tag_hash)
  order by date desc
  limit 1
  ) tag

  "Get the latest tag for a commit", which we solve by reading the tags table backwards, and returning the first that is an ancestor of our target commit.
• 6845: add --silent option for push, pull, fetch
  Adds a --silent option for push, pull, fetch to suppress output of progress information.
  Resolves: #6828
• 6834: migrate dolt ls to use sql queries
  This change updates dolt ls to use the appropriate sql engine to generate results. This change also creates a system variable that, when enabled, shows dolt system tables in show tables and information_schema.tables.
  Related: #3922
  Resolves: #2073
• 6833: JSON and stats interface refactor
  Bring Dolt into line with the new stats and json interfaces.

go-mysql-server

• 2091: Non-ambiguous ORDER BY col should not error
• 2088: fix off by one for found_rows when limit > count(*)
  We had a bug where we could increment the limit counter before receiving the EOF error.
  fixes #6829
  companion pr: dolthub/vitess#283
• 2085: small fixes for new test harness using server engine
  This PR adds small fixes to server_engine that is used to test existing engine tests over running server.
• 2081: Refactor JSON interfaces, stats interfaces
  This tries to organize some of the stats and json code.
  Stats has interfaces in the sql package to avoid circular dependencies (catalog depends on stats, stats depends on sql.Row, etc). So stats package can depend on sql for row and types, and most of GMS can use the generic sql interfaces without having to take a dependency on the stats package.
  memory and enginetests are the two main places where we depend on the stats concrete implementations. Rather than putting the statistic objects in memory, I chose to put them into a separate package. The two main reasons are 1) the Dolt side has a conversion to the memory implementation as a presentation layer, and it made more sense in my head for the "presentation layer" to be in the stats package rather than memory; 2) I'm going to add logic that operates on histograms, and it seemed convenient to have that in its own package with a concrete implementation there for testing. If this is wrong I can always reverse course.
  Most of the json changes are to try to make custom json types (statistic, histogram) more closely aligned with the default JSONDocument implementation, remove unused interfaces, deduplicate methods that all have a generic implementation. Other than typing additions in helper functions, the biggest change is probably adding the ToInterface{} method, which converts a generic JSONWrapper interface (newly changed from JSONValue to avoid overloading the term) into a map of strings->interface{}. I mostly did this because I was having to go from Statistic->[]byte->map[string]interface{} to use our json serialization primitives. ToInterface{} might be less performant than inlining search, extract, etc on structs (for example, we don't need to convert Statistic an interface to extract the row_count value), but the generic intermediate makes it pretty easy to plug new json types. Again this may be misguided, happy to reverse course if I'm overcomplicating something.
• 2080: drop sort when secondary index is available

vitess

• 284: Include the With clause in walked subtrees for Select statements
  We weren't walking the With clause for select statements, which caused us to not find any bind vars in use there.
  Related to: #6852
• 283: allow query options to appear in any order any number of times
  Allow statements like this to parse:

  select distinct sql_calc_found_rows distinct * from t;

  Fixes #6829
  Companion PR: dolthub/go-mysql-server#2088

Closed Issues

• 6817: [no conn] no binlog connection to source, attempting to establish one {}
• 1256: dolt diff empty output with table filter on newly created tables
• 6828: $ dolt push generates lots of Uploading... text in cron
• 2073: Need a way to list all dolt system tables in SQL context
• 6829: SQL_CALC_FOUND_ROWS doesn't work?