Михаил
Михаил
What is `sqlite_fdw`? I think it's special SQLite client as data input for PostgreSQL. Hence PostgreSQL behaviour with input data is primary. For PostgreSQL data types from `real` family string...
Fixed in https://github.com/pgspider/sqlite_fdw/pull/57. It wasn't very hard to find lines of code for fixing this problem.
> SQLite does not have a concept of 'Infinity' and 'NaN' as a number. @t-kataym your info is outdated. In SQLite there is native ∞ conception as special values with...
@Soni-Harriz , could you give us please some fragments of your testcase with unwanted for You `sqlite_fdw` behaviour? Now I am coding about `NaN` and this test case will be...
@t-kataym , in current tests there is aggregates with `NaN` and expected `0` addition to `sum` for example. If we treat `NaN` from PostgreSQL as `NULL` `sum` will gives `NULL`...
After your message, @t-kataym , i have understand that `±∞` and `NaN` should be different PRs for `float`-like datatypes in our FDW. PR about ∞ will be simpler by conception...
Hello @t-kataym ! After `±∞` support draft in https://github.com/pgspider/sqlite_fdw/pull/97 I can list all `NaN` processing problems in the FDW. Main source of our problems is SQLite [`sqlite3VdbeMemSetDouble`](https://github.com/sqlite/sqlite/blob/6db0b11e078f4b651f0cf00f845f3d77700c1a3a/src/vdbemem.c#L973) where there is...
Special [branch](https://github.com/mkgrgis/sqlite_fdw/tree/NaN) with `NaN` processing tests added. Now `NaN` is unfiltrable/undetectable: ``` --Testcase 329: SELECT * FROM "type_FLOAT_INF+" WHERE f = 'NaN' ORDER BY i; i | f | t...
@t-kataym . I am ready to discussing about this PR and testing environment setup. Look like the problems is near Ubuntu `libspatialite-dev`. Can pgspider team help me with CI testing...
Hello, @son-phamngoc ! Thanks for code and tests reading! > In my understanding, if we want to enable PostGIS to be able to use its data types, we need to...