Skip to content
Permalink

Comparing changes

Choose two branches to see what’s changed or to start a new pull request. If you need to, you can also or learn more about diff comparisons.

Open a pull request

Create a new pull request by comparing changes across two branches. If you need to, you can also . Learn more about diff comparisons here.
base repository: postgresql-cfbot/postgresql
Failed to load repositories. Confirm that selected base ref is valid, then try again.
Loading
base: cf/5517~1
Choose a base ref
...
head repository: postgresql-cfbot/postgresql
Failed to load repositories. Confirm that selected head ref is valid, then try again.
Loading
compare: cf/5517
Choose a head ref
  • 8 commits
  • 26 files changed
  • 3 contributors

Commits on Nov 12, 2025

  1. Refactor output format of pg_ndistinct. 

    The existing format of pg_ndistinct uses a single-object JSON structure
where each key is itself a comma-separated list of attnums. While this
is a very compact format, it's confusing to read and is difficult to
manipulate values within the object. This wasn't a concern until
statistics import functions were introduced, enabling users to inject
hypothetical statistics into an object to observe their effect on the
query planner.

The new format is an array of objects, each object must have the keys
"attributes", which must contain an array of attnums, and "ndistinct",
which must be an integer. This is a quirk because the underlying
internal storage is a double, but the value stored was always an
integer.

The change in format is described from the changes to
src/test/regress/expected/stats_ext.out.
    @michaelpq
    michaelpq authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    a839ad8 View commit details
    Browse the repository at this point in the history

  2. Refactor output format of pg_dependencies. 

    The existing format of pg_dependencies uses a single-object JSON structure
where each key is itself a comma-separated list of attnums. While this
is a very compact format, it's confusing to read and is difficult to
manipulate values within the object. This wasn't a concern until
statistics import functions were introduced, enabling users to inject
hypothetical statistics into an object to observe their effect on the
query planner.

The new format is an array of objects, each object must have the keys
"attributes", which must contain an array of attnums, "dependency",
which must be an integer, and "degree", which must be a float.

The change in format is adequately described from the changes to
src/test/regress/expected/stats_ext.out so description here is
redundant.
    @michaelpq
    michaelpq authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    8a4d7c0 View commit details
    Browse the repository at this point in the history

  3. Add working input function for pg_ndistinct. 

    This will consume the format that was established when the output
function for pg_ndistinct was recently changed.

This will be needed for importing extended statistics.
    @michaelpq
    michaelpq authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    29c6b89 View commit details
    Browse the repository at this point in the history

  4. Add working input function for pg_dependencies. 

    This will consume the format that was established when the output
function for pg_dependencies was recently changed.

This will be needed for importing extended statistics.
    @michaelpq
    michaelpq authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    ead36b1 View commit details
    Browse the repository at this point in the history

  5. Expose attribute statistics functions for use in extended_stats. 

    Many of the operations of attribute stats have analogous operations in
extended stats.

* get_attr_stat_type() renamed to statatt_get_type()
* init_empty_stats_tuple() renamed to statatt_init_empty_tuple()
* text_to_stavalues()
* get_elem_stat_type() renamed to statatt_get_elem_type()

Also, add comments explaining the function argument index enums, and the
arrays that are indexed by those enums.
    @coreyhuinker
    coreyhuinker authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    c8ead5e View commit details
    Browse the repository at this point in the history

  6. Add extended statistics support functions. 

    Add pg_restore_extended_stats() and pg_clear_extended_stats(). These
functions closely mirror their relation and attribute counterparts,
but for extended statistics (i.e.  CREATE STATISTICS) objects.
    @michaelpq
    michaelpq authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    77c4e37 View commit details
    Browse the repository at this point in the history

  7. Include Extended Statistics in pg_dump. 

    Incorporate the new pg_restore_extended_stats() function into pg_dump.
This detects the existence of extended statistics statistics (i.e.
pg_statistic_ext_data rows).

This handles many of the changes that have happened to extended
statistic statistics over the various versions, including:

* Format change for pg_ndistinct and pg_dependencies in current
  development version. Earlier versions have the format translated via
  the pg_dump SQL statement.

* Inherited extended statistics were introduced in v15.

* Expressions were introduced to extended statistics in v14.

* MCV extended statistics were introduced in v13.

* pg_statistic_ext_data and pg_stats_ext introduced in v12, prior to
  that ndstinct and depdendencies data (the only kind of stats that
  existed were directly on pg_statistic_ext.

* Extended Statistics were introduced in v10, so there is no support for
  prior versions necessary.
    @coreyhuinker
    coreyhuinker authored and Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    df2ccde View commit details
    Browse the repository at this point in the history

  8. [CF 5517] v12 - Extended Statistics set/restore/clear functions 

    This branch was automatically generated by a robot using patches from an
email thread registered at:

https://commitfest.postgresql.org/patch/5517

The branch will be overwritten each time a new patch version is posted to
the thread, and also periodically to check for bitrot caused by changes
on the master branch.

Patch(es): https://www.postgresql.org/message-id/CADkLM=c8-U4GLMw5VdeDdfp1ae6BW=PCfEQqAky04iZMbckCFw@mail.gmail.com
Author(s): Corey Huinker
    Commitfest Bot committed Nov 12, 2025
    Configuration menu
    Copy the full SHA
    96f9c01 View commit details
    Browse the repository at this point in the history
Loading