Extract versioned-SQL generation into sql.mk; install generated upgrade scripts (fixes #28)

[jnasbyupgrade]

What this is

This PR is a pure build-system refactor: it extracts the versioned-SQL generation machinery out of the Makefile into a new, documented sql.mk, and fixes #28 so generated upgrade scripts land in DATA reliably on a clean build.

Dependency — READ THIS: The #18 pg_upgrade source reconcile that used to be bundled into this PR has been pulled out into its own PR: #38 (fix-omit-source). This refactor regenerates the current-version file in place from sql/cat_tools.sql.in; on master that source still lacks the #18 omit_column additions, so until #38 merges to master and this branch is rebased onto it, this PR's binary pg_upgrade CI legs will be RED. That is expected and correct — the fix lives in #38. Do not merge this before #38; after #38 merges, rebase this branch onto master and the pg_upgrade legs go green.

What moved into sql.mk (new)

Included after include pgxntool/base.mk, sql.mk now owns:

• B, versioned_in/versioned_out, upgrade_scripts_out
• historical_installs — the committed cat_tools--0.1.* install list, derived (glob the committed install-shaped .sql, subtract update scripts and everything generated from .sql.in) instead of the old hardcoded list
• the DATA handling, including the Clean-build 'make test' fails on all PG>=12: generated upgrade scripts excluded from DATA (parse-time wildcard) #28 fix (DATA += $(upgrade_scripts_out) then DATA := $(sort $(DATA)))
• all / installcheck / EXTRA_CLEAN
• the SED version-marker machinery (LT95/LT93, _apply_version_seds), the $B/%.sql pattern rule, and the EXTENSION_VERSION_FILES rules

The Makefile now just include sql.mk (with a pointer comment) and keeps its non-SQL-gen content (the upgrade-test TEST_BUILD_DIR block, old_version/clean_old_version).

Why #28 happens (documented in sql.mk's header)

GNU Make is two-phase. We generate the .sql we install; the generated files are gitignored and absent on a clean tree when the makefile is parsed. Any parse-time $(wildcard) over them — including base.mk's DATA seed $(wildcard sql/*--*--*.sql) — silently comes up short, and Make's cached directory listing means it never notices them later. So on a fresh build the generated update scripts never make it into DATA, make install skips them, and a later ALTER EXTENSION cat_tools UPDATE fails with "no update path". The fix: name-derive the lists from the .sql.in sources, feed the generated names into DATA explicitly, and $(sort). Rule of thumb: glob what git tracks, name-derive what we generate.

Clean-tree #28 proof

On a truly clean tree (make clean, no generated 0.2.* .sql present):

• Before (upstream/master): none of the generated 0.2.0--0.2.1, 0.2.0--0.2.2, 0.2.1--0.2.2 upgrade scripts appear in DATA.
• After (this branch): all three appear in DATA, and a clean make install copies all of them into the extension directory.

make print-DATA is byte-identical on a clean tree vs. a fully-built tree (parse-stable), and historical_installs resolves to exactly the committed cat_tools--0.1.0/0.1.3/0.1.4/0.1.5.sql.

[coderabbitai]

Important

Review skipped

Auto reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: 143aa8c8-dd6b-485e-a29f-079c5e8a2996

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}