Permalink
Commits on Jan 20, 2019
  1. pgindent run and update changelog for last msvc commit.

    rjuju committed Jan 20, 2019
    Also update typedefs.list.
Commits on Dec 11, 2018
  1. Stamp 2.0.0beta

    rjuju committed Dec 11, 2018
Commits on Dec 9, 2018
Commits on Dec 4, 2018
  1. Fix typo in documentation.

    rjuju committed Dec 4, 2018
Commits on Dec 3, 2018
Commits on Nov 24, 2018
Commits on Nov 23, 2018
  1. Reintroduce childrel expanding order for pg10

    rjuju committed Nov 23, 2018
    The partitions has to be sorted by bound order, otherwise many things
    can break.
    
    This commit also introduce specific regression tests for pg10, for which
    the duplucate sanity check would not exhibit the expected error without
    the expanded order.
    
    Some of these new regression tests still fails, due to minor change in
    the join or partition order, we'll see if we can fix the behavior or change
    the expected results.
  2. Comment long #if blocks

    rjuju committed Nov 23, 2018
Commits on Nov 19, 2018
  1. Fix typo in a pg version check

    rjuju committed Nov 19, 2018
Commits on Nov 17, 2018
  1. Allow hypothetical index on hypothetical partition

    rjuju committed Nov 17, 2018
    Regression tests updated.  Also only authorize hypothetical indexes on
    root partition for pg11+ to get the same behavior as postgres.
  2. Check for real table first in hypopg_add_partition()

    rjuju committed Nov 17, 2018
    For now, hypo_table_name_get_entry() can be quite slow if many hypoTable
    are present, so perform cheaper check before it.
  3. Start working on hypothetical partitioning pg10 compatibility.

    rjuju committed Nov 8, 2018
    This is the easy part: get rid of the features added in pg11 such as
    hash partitioning or default partition, and adapt the required imports.
    
    There are still a lot of errors due to the lack of PartitionScheme and
    many fields in RelOptInfo in pg10.
Commits on Nov 15, 2018
Commits on Nov 8, 2018
  1. Store the hypothetical partitions in a hash table.

    rjuju committed Nov 8, 2018
    Previous naive implementation was storing the hypothetical partitions in
    a simple List, which would definitely not scale if many hypothetical
    partitions are stored.
    
    The partition's children oid are also now maintained in each hypoTable
    entry, for performance reason too.
    
    Some minor fixup too while at it.
Commits on Nov 7, 2018
  1. Merge pull request #31 from yuzupy/def_part

    rjuju committed Nov 7, 2018
    Add a hypothetical default partition feature
Commits on Nov 6, 2018
  1. Merge pull request #30 from yuzupy/change_hypoTable_struct

    rjuju committed Nov 6, 2018
    Change hypoTable struct to have its number of tuples
Commits on Nov 4, 2018
  1. Add a hypopg_drop_table(oid) wrapper.

    rjuju committed Nov 4, 2018
    The underlying C function was already existing, I just forgot to add the
    SQL wrapper.
    
    The C function is modified to add an error if the function is called on
    a non hypothetically partitioned table instead of silently doing
    nothing, and add regression tests.
Commits on Oct 13, 2018
  1. Fix selectivity estimation for hash partitioning

    rjuju committed Oct 13, 2018
    Previous implementation assumed that tables hypothetically partitioned by hash
    had all partitions defined and that all partitions had the same modulus.  We
    can compute the fraction of data a leaf partition will have by just looking at
    its modulus and its parent's modulus if any instead, which is also cheaper to
    do, since we don't have to find all child partitions for all hash partitions.
Commits on Oct 12, 2018
  1. Fix hypopg_analyze() behavior for hash partitions.

    rjuju committed Oct 12, 2018
    We do not need (nor can use) statistics for hash partitions, but those can
    exist anywhere in the partitioning tree.  So just forbidding hypopg_analyze()
    if the first level is partitioned by hash was wrong.