Refactor data conversion errors and diagnostics, tests, headers and test data directory

[mkgrgis]

In this PR:

• PostgreSQL integer-family out of range checks. SQLite int affinity is 1↔1 with PostgreSQL bigint and also have equal ranges. If we have int or smallint in PostgreSQL some input values which is out of range for this data types are possible.
• sqlite3_column_@(stmt, colid) functions inside of sqlite_convert_to_pg replaced to sqlite3_value_@(val) functions for less parameters and fastest calculations, see sqlite3_value *sqlite3_column_value(sqlite3_stmt *pStmt, int i) and static Mem *columnMem(sqlite3_stmt *pStmt, int i) which will calculated once instead of many sqlite3_column_@(stmt, colid) calls.
• Detach out_of_range test from int4, add int2 TCs https://github.com/pgspider/sqlite_fdw/pull/91/commits/641c539615b36ff203199b0c908524cf069ba358.
• In some cases in bigint PostgreSQL column can come a value with real affinity which is out of range. Add PostgreSQL input core function and TCs for this case.
• Deep refactoring of conversion data error diagnostics to newly implemented ErrorContextCallback function with universal messages about any data conversion problems between PostgreSQL data column and SQLite data value.
• Detach bitstring test from type test https://github.com/pgspider/sqlite_fdw/pull/91/commits/9bc53a470d0a51d40e34e4dfe6d25ad3e45fd3d3.
• Add dtof to float4 PostgreSQL data column branch https://github.com/pgspider/sqlite_fdw/pull/91/commits/f0593dead92aad86ff27e7bea307984a4f3eeb97 (without testing, will tested in other PR about full real affinity support).
• Move SQLite test data to separate /tmp/sqlite_fdw_test directory for less chaos in /tmp https://github.com/pgspider/sqlite_fdw/pull/91/commits/cbba0432a2824c21912dc45f7b760ab4e350d54e . This allow reduce testing time.
• Refactor headers: remove unnecessary headers, ABC other, limit to some versions some of rare used headers. This allow reduce compile time. https://github.com/pgspider/sqlite_fdw/pull/91/commits/e23ba835b19ac894f20bdb56d289f4615db0795d and https://github.com/pgspider/sqlite_fdw/pull/91/commits/932b3e0a0bc69e04cc7cd8b00089d4d51915d25a
• Encapsulate ForeignTable and ForeignServer into ...State structures, because this variables are often calculated near this structures in functions like constructor. https://github.com/pgspider/sqlite_fdw/pull/91/commits/f1a1f89f733dc9f256c1c482018538c9964abdbe

[mkgrgis]

@t-kataym, this simple PR is ready for review.

[mkgrgis]

@t-kataym , could you please write approximate week of review of this PR? Have pgspider team any plans of a new sqlite_fdw release?

[mkgrgis]

I have some comments for your PR, could you please check it?

All your comments have checked, please start 2nd round of review, @nxhai98 !

[mkgrgis]

All comments are checked and fixed, @nxhai98. Thanks for review! What will be later? 3rd round of your review or review by @t-kataym ?

[mkgrgis]

Ping, @nxhai98 ! Are there any results of 2 rounds of your review? What will be later?

[mkgrgis]

@t-kataym , could you please write approximate week of review and status of this PR? Will there 3rd review round?

[mkgrgis]

Hello @t-kataym and @nxhai98 ! Any chances to have feedback or 3rd review round for this PR? What is the status of this PR? Completely PR description after checking all your comments was updated https://github.com/pgspider/sqlite_fdw/pull/91#issue-1994632771

[t-kataym]

@mkgrgis , Thank you for your contribution.
@nxhai98 , Thank you for your confirmation.
I will merge the PR.