context("Cube queries with on-the-fly expressions")

with_mock_crunch({
    ds <- cachedLoadDataset("test ds")
    expect_JSON <- function(object, expected) {
        expect_identical(
            unclass(toJSON(object)),
            gsub("\n +", "", gsub(": +", ":", expected))
        )
    }
    test_that("formulaToCubeQuery", {
        expect_JSON(
            formulaToCubeQuery(~ gender + bin(birthyr), data = ds),
            '{"dimensions":[
                {"variable":"https://app.crunch.io/api/datasets/1/variables/gender/"},
                {
                    "function":"bin",
                    "args":[
                        {"variable":"https://app.crunch.io/api/datasets/1/variables/birthyr/"}
                    ],
                    "references":{
                        "name": "bin(birthyr)",
                        "alias": "bin(birthyr)"
                    }
                }
                ],
            "measures":{"count":{"function":"cube_count","args":[]}}}'
        )
        expect_JSON(
            formulaToCubeQuery(~ gender + catarray, data = ds),
            '{"dimensions":[
                {"variable":"https://app.crunch.io/api/datasets/1/variables/gender/"},
                {"function":"dimension","args":[
                    {"variable":"https://app.crunch.io/api/datasets/1/variables/catarray/"},
                    {"value":"subvariables"}
                ]},
                {"variable":"https://app.crunch.io/api/datasets/1/variables/catarray/"}
                ],
            "measures":{"count":{"function":"cube_count","args":[]}}}'
        )
        expect_JSON(
            formulaToCubeQuery(~ gender + as_selected(mymrset), data = ds),
            paste0(
                '{"dimensions":[
                    {"variable":"https://app.crunch.io/api/datasets/1/variables/gender/"},
                    {"function":"dimension","args":[
                        {"function":"as_selected","args":[
                            {"variable":"https://app.crunch.io/api/datasets/1/variables/mymrset/"}
                        ]},
                        {"value":"subvariables"}
                    ]},
                    {"function": "as_selected","args": [
                        {"variable":"https://app.crunch.io/api/datasets/1/variables/mymrset/"}]}
                ],
            "measures":{"count":{"function":"cube_count","args":[]}}}'
            )
        )
        expect_error(
            formulaToCubeQuery(~ gender + as_selected(catarray), data = ds),
            paste(
                "Cannot analyze a variable of type",
                dQuote("categorical_array"), "'as_selected'"
            )
        )
        expect_error(
            formulaToCubeQuery(~ categories(gender), data = ds),
            paste(
                "Cannot analyze a variable of type",
                dQuote("categorical"), "using 'categories'"
            )
        )
        expect_error(
            formulaToCubeQuery(~ subvariables(gender), data = ds),
            paste(
                "Cannot analyze a variable of type",
                dQuote("categorical"), "using 'subvariables'"
            )
        )
        expect_JSON(
            formulaToCubeQuery(~ gender + as_array(mymrset), data = ds),
            '{"dimensions":[
                {"variable":"https://app.crunch.io/api/datasets/1/variables/gender/"},
                {"function":"dimension","args":[
                    {"variable":"https://app.crunch.io/api/datasets/1/variables/mymrset/"},
                    {"value":"subvariables"}
                ]},
                {"variable":"https://app.crunch.io/api/datasets/1/variables/mymrset/"}
                ],
            "measures":{"count":{"function":"cube_count","args":[]}}}'
        )
        expect_JSON(
            formulaToCubeQuery(~ gender + (birthyr > 1980), data = ds),
            '{"dimensions":[
                {"variable":"https://app.crunch.io/api/datasets/1/variables/gender/"},
                {
                    "function":">",
                    "args":[
                        {"variable": "https://app.crunch.io/api/datasets/1/variables/birthyr/"},
                        {"value": 1980}
                    ],
                    "references":{
                        "name": "birthyr > 1980",
                        "alias": "birthyr > 1980"
                    }}
                ],
            "measures":{"count":{"function":"cube_count","args":[]}}}'
        )
    })

    test_that("3VL logical cube dims", {
        ## This is `crtabs(~ q1 != "Cat" & !is.na(q1), data=ds)` from below
        cube <- loadCube("cubes/3vl.json")
        expect_identical(
            as.array(cube),
            array(c(7, 6),
                dim = 2L,
                dimnames = list(`q1 != "Cat" & !is.na(q1)` = c("TRUE", "FALSE"))
            )
        )
    })
})

with_test_authentication({
    ds <- newDatasetFromFixture("apidocs")
    test_that("<, <= on numeric", {
        # ndogs
        # 0 1 2 3 6
        # 2 3 7 3 1
        expect_equivalent(as.array(crtabs(~ ndogs < 2, data = ds))["TRUE"], 5)
        expect_equivalent(as.array(crtabs(~ ndogs <= 2, data = ds))["TRUE"], 12)
        expect_equivalent(as.array(crtabs(~ ndogs > 2, data = ds))["TRUE"], 4)
        expect_equivalent(as.array(crtabs(~ ndogs >= 2, data = ds))["TRUE"], 11)
        expect_equivalent(as.array(crtabs(~ ndogs > 1 & ndogs <= 3, data = ds))["TRUE"], 10)
    })

    test_that("%in% with categorical", {
        # q1
        #  Cat  Dog Bird
        #    6    4    3
        skip(paste0(
            "(400) Bad Request: The truth value of an array with more than one ",
            "element is ambiguous. Use a.any() or a.all()"
        ))
        expect_equivalent(as.array(crtabs(~ q1 %in% c("Cat", "Dog"),
            data = ds
        ))["TRUE"], 10)
        expect_equivalent(as.array(crtabs(~ !(q1 %in% c("Cat", "Dog")),
            data = ds
        ))["TRUE"], 3)
    })

    test_that("==, != with categorical", {
        expect_equivalent(as.array(crtabs(~ q1 == "Cat",
            data = ds
        ))["TRUE"], 6)
        expect_equivalent(as.array(crtabs(~ q1 != "Cat" & !is.na(q1),
            data = ds
        ))["TRUE"], 7)
    })
})