Add auto_import test for IMPORT FOREIGN TABLE and test sequence refactoring

[mkgrgis]

In this PR there are absolutely no C code changes.
Content of PR:

• Add auto_import test which shows behavior of IMPORT FOREIGN TABLE command according PostgreSQL internal metadata state.
• Refactor test sequence to different variables - data type test, tests with different results depends on GIS compilation mode, main sequence.
• Reduce DROP ... CASCADE outputs in type tests for both GIS and not GIS modes.

[lamdn1409]

@mkgrgis Can we start reviewing this PR?

[mkgrgis]

@mkgrgis Can we start reviewing this PR?

Yes. There is and there was normal check status yesterday.

[lamdn1409]

@mkgrgis I think there are 03 different purposes in this PR.

1. Add auto_import test which shows behavior of IMPORT FOREIGN TABLE command according PostgreSQL internal metadata state.
2. Refactor test sequence to different variables - data type test, tests with different results depends on GIS compilation mode, main sequence. Reduce DROP ... CASCADE outputs in type tests for both GIS and not GIS modes.
3. Prepare some SQLITE_FOR_TESTING_DIR environment variable infrastructure.

The third purpose is the fix of SQLite CI issue (#114). If correct, please remove related changes.

ifdef SQLITE_FOR_TESTING_DIR
SHLIB_LINK := -L$(SQLITE_FOR_TESTING_DIR)/lib -lsqlite3
PG_CFLAGS += -I$(SQLITE_FOR_TESTING_DIR)/include -Wl,-rpath,$(SQLITE_FOR_TESTING_DIR)/lib
else
SHLIB_LINK := -lsqlite3

Could you create separated PRs for 1) and 2)?

[mkgrgis]

If correct, please remove related changes.

Done in 4b67cfe , please verify.

Could you create separated PRs for 1) and 2)?

I thinks this is very atomic and bureaucracy-related. 2) touches only 2 files and this is transparent change near REGRESS variable value. Both 1) and 2) are about extending of actual testing. Also "reduce DROP ... CASCADE outputs" is about new group of tests organized after auto_import test addition as next version of previous type_postgis and type_nogis tests. Hence 1) and 2) have internal dependency and common purpose.

[mkgrgis]

@lamdn1409 , please recheck all our conversations. Look like we have no other to discuss now expect for #116 (comment) Isn't it?

[mkgrgis]

@lamdn1409 , I cannot due github technical restrictions answer under #116 (comment) , is now no problems in mentioned place?

[lamdn1409]

ifdef ENABLE_GIS
PG_CFLAGS += -DSQLITE_FDW_GIS_ENABLE
GIS_TEST = postgis
GIS_DEP_TESTS_DIR = with_gis_support
$(info  There is PostGIS support for SQLite FDW)
else
GIS_TEST = nogis
GIS_DEP_TESTS_DIR = without_gis_support
$(info  There is NO PostGIS support for SQLite FDW)
endif

postgis.sql is tested if gis is enabled.
nogis.sql is tested if gis is disabled.
So, I think you can move these tests to appropriate folders.

[mkgrgis]

So, I think you can move these tests to appropriate folders.

Moved in 5ad4e56, please recheck.

[mkgrgis]

@lamdn1409, we have successfully tests now. Does it means time to 2nd review round or start merge review?

[lamdn1409]

@mkgrgis Thank you for resolving my comments. I have no more comments. Could you please request @t-kataym to review?

[mkgrgis]

@t-kataym , this PR is ready for pre-merge review. Could you please observe proposed changes?

[t-kataym]

@t-hiroshige Please check it.
@mkgrgis The maintainer of this repository has been changed from me to @t-hiroshige.
I really appreciate your contributions so far. From the next, please contact to him if necessary.

[t-hiroshige]

OK, I confirmed.
I'll merge this PR soon.

[t-hiroshige]

@mkgrgis Thank you for creating the PR.
@lamdn1409 Thank you for your review.

[mkgrgis]

Thanks, @t-hiroshige !

Thanks, @t-kataym ! I was happy to work with you. You was my 1st teacher in C-language culture, @t-kataym - san, ありがとう !

[mkgrgis]

@t-hiroshige , you have forgotten to squash small transitive review commits before merging. Usually PR with more than 3-4 inner commits are merged only as squashed. Now we have very detailed upstream commit tree https://github.com/pgspider/sqlite_fdw/commits/master/

You can fix this and squash all changes after 8f7f4f4 by follow algorithm:

1. Save/copy current correct content expect for .git to any temporary directory
2. git rebase -i HEAD~1 where set s for all commits after first and p or pick for first commit.
3. Delete all content of cloned repository after content conflict, put saved correct content instead of deleted.
4. git commit with current message

Add auto_import test for IMPORT FOREIGN TABLE and test sequence refactoring #116

5. git push --force

I am sorry, I haven't warned you about this case.

Then I'll quickly rebase all of my PRs, this will not hard.