Hi,
Here is a new version of index skip scan patch, based on v8 patch for
UniqueKeys implementation from [1]. I want to start a new thread to
simplify navigation, hopefully I didn't forget anyone who actively
participated in the discussion.
To simplify reviewing I've split it into several parts:
* First two are taken from [1] just for the reference and to make cfbot happy.
* Extend-UniqueKeys consists changes that needs to be done for
UniqueKeys to be used in skip scan. Essentially this is a reduced
version of previous implementation from Jesper & David, based on the
new UniqueKeys infrastructure, as it does the very same thing.
* Index-Skip-Scan contains not am specific code and the overall
infrastructure, including configuration elements and so on.
* Btree-implementation contains btree specific code to implement amskip,
introduced in the previous patch.
* The last one contains just documentation bits.
Interesting enough with a new UniqueKey implementation skipping is
applied in some tests unexpectedly. For those I've disabled
indexskipscan to avoid confusion.
[1]: https://www.postgresql.org/message-id/flat/CAKU4AWrwZMAL%3DuaFUDMf4WGOVkEL3ONbatqju9nSXTUucpp_pw%40mail.gmail.com