diff --git a/test/metabase/upload_test.clj b/test/metabase/upload_test.clj
index 52136126bbbcde86dd602b5c4cb174b4bab998b9..674e02a73523e12a921e4b864e6c5575cf5a774c 100644
--- a/test/metabase/upload_test.clj
+++ b/test/metabase/upload_test.clj
@@ -519,19 +519,22 @@
"2\t || a |true |1.1\t |2022-01-01|2022-01-01T00:00:00"
"\" 3\"|| b|false|\"$ 1,000.1\"|2022-02-01|2022-02-01T00:00:00"]})
+(defn- auto-pk-column? []
+ (#'upload/auto-pk-column? driver/*driver* (mt/db)))
+
(defn- columns-with-auto-pk [columns]
(cond-> columns
- (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (auto-pk-column?)
(#'upload/columns-with-auto-pk)))
(defn- header-with-auto-pk [header]
(cond->> header
- (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (auto-pk-column?)
(cons @#'upload/auto-pk-column-name)))
(defn- rows-with-auto-pk [rows]
(cond->> rows
- (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (auto-pk-column?)
(map-indexed (fn [i row] (cons (inc i) row)))))
(defn- column-position [table column-name]
@@ -677,30 +680,31 @@
(deftest create-from-csv-offset-datetime-test
(testing "Upload a CSV file with an offset datetime column"
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
- (with-mysql-local-infile-on-and-off
- (with-redefs [driver/db-default-timezone (constantly "Z")
- upload/current-database (constantly (mt/db))]
- (let [transpose (fn [m] (apply mapv vector m))
- [csv-strs expected] (transpose [["2022-01-01T12:00:00-07" "2022-01-01T19:00:00Z"]
- ["2022-01-01T12:00:00-07:00" "2022-01-01T19:00:00Z"]
- ["2022-01-01T12:00:00-07:30" "2022-01-01T19:30:00Z"]
- ["2022-01-01T12:00:00Z" "2022-01-01T12:00:00Z"]
- ["2022-01-01T12:00:00-00:00" "2022-01-01T12:00:00Z"]
- ["2022-01-01T12:00:00+07" "2022-01-01T05:00:00Z"]
- ["2022-01-01T12:00:00+07:00" "2022-01-01T05:00:00Z"]
- ["2022-01-01T12:00:00+07:30" "2022-01-01T04:30:00Z"]
- ["2022-01-01T12:00:00+0730" "2022-01-01T04:30:00Z"]])]
- (testing "Fields exists after sync"
- (with-upload-table!
- [table (create-from-csv-and-sync-with-defaults!
- :file (csv-file-with (into ["offset_datetime"] csv-strs)))]
- (testing "Check the offset datetime column the correct base_type"
- (is (=? :type/DateTimeWithLocalTZ
- (t2/select-one-fn :base_type Field :%lower.name "offset_datetime" :table_id (:id table)))))
- (let [position (column-position table "offset_datetime")
- values (map #(nth % position) (rows-for-table table))]
- (is (= expected
- values)))))))))))
+ (when (driver/upload-type->database-type driver/*driver* :metabase.upload/offset-datetime)
+ (with-mysql-local-infile-on-and-off
+ (with-redefs [driver/db-default-timezone (constantly "Z")
+ upload/current-database (constantly (mt/db))]
+ (let [transpose (fn [m] (apply mapv vector m))
+ [csv-strs expected] (transpose [["2022-01-01T12:00:00-07" "2022-01-01T19:00:00Z"]
+ ["2022-01-01T12:00:00-07:00" "2022-01-01T19:00:00Z"]
+ ["2022-01-01T12:00:00-07:30" "2022-01-01T19:30:00Z"]
+ ["2022-01-01T12:00:00Z" "2022-01-01T12:00:00Z"]
+ ["2022-01-01T12:00:00-00:00" "2022-01-01T12:00:00Z"]
+ ["2022-01-01T12:00:00+07" "2022-01-01T05:00:00Z"]
+ ["2022-01-01T12:00:00+07:00" "2022-01-01T05:00:00Z"]
+ ["2022-01-01T12:00:00+07:30" "2022-01-01T04:30:00Z"]
+ ["2022-01-01T12:00:00+0730" "2022-01-01T04:30:00Z"]])]
+ (testing "Fields exists after sync"
+ (with-upload-table!
+ [table (create-from-csv-and-sync-with-defaults!
+ :file (csv-file-with (into ["offset_datetime"] csv-strs)))]
+ (testing "Check the offset datetime column the correct base_type"
+ (is (=? :type/DateTimeWithLocalTZ
+ (t2/select-one-fn :base_type Field :%lower.name "offset_datetime" :table_id (:id table)))))
+ (let [position (column-position table "offset_datetime")
+ values (map #(nth % position) (rows-for-table table))]
+ (is (= expected
+ values))))))))))))
(deftest create-from-csv-boolean-test
(testing "Upload a CSV file"
@@ -827,7 +831,7 @@
(testing "Check the data was uploaded into the table correctly"
(is (= (header-with-auto-pk ["Cost $" "Cost %" "Cost #"])
(column-display-names-for-table table)))
- (is (= [@#'upload/auto-pk-column-name "cost__" "cost___2" "cost___3"]
+ (is (= (header-with-auto-pk ["cost__" "cost___2" "cost___3"])
(column-names-for-table table))))))))))
(deftest create-from-csv-bool-and-int-test
@@ -863,13 +867,14 @@
"9000000000,Razor Crest,Din Djarin,Spear"])
:auxiliary-sync-steps :synchronous))]
(testing "Check the data was uploaded into the table correctly"
- (is (= [@#'upload/auto-pk-column-name "id" "ship" "name" "weapon"]
+ (is (= (header-with-auto-pk ["id" "ship" "name" "weapon"])
(column-names-for-table table)))
(is (=? {:name #"(?i)id"
:semantic_type :type/PK
:base_type :type/BigInteger
:database_is_auto_increment false}
- (t2/select-one Field :database_position 1 :table_id (:id table))))))))))
+ (let [pos (if (auto-pk-column?) 1 0)]
+ (t2/select-one Field :database_position pos :table_id (:id table)))))))))))
(deftest create-from-csv-auto-pk-column-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads :upload-with-auto-pk)
@@ -1175,7 +1180,7 @@
"size_mb" 3.910064697265625E-5
"num_columns" 2
"num_rows" 2
- "generated_columns" 1
+ "generated_columns" (if (auto-pk-column?) 1 0)
"upload_seconds" pos?}
:user-id (str (mt/user->id :rasta))}
(last (snowplow-test/pop-event-data-and-user-id!)))))
@@ -1211,7 +1216,7 @@
:model-id pos?
:stats {:num-rows 2
:num-columns 2
- :generated-columns 1
+ :generated-columns (if (auto-pk-column?) 1 0)
:size-mb 3.910064697265625E-5
:upload-seconds pos?}}}
(last-audit-event :upload-create)))))))))
@@ -1328,7 +1333,7 @@
:or {table-name (mt/random-name)
schema-name (sql.tx/session-schema driver/*driver*)
col->upload-type (cond->> (ordered-map/ordered-map :name ::upload-types/varchar-255)
- (#'upload/auto-pk-column? driver/*driver* (mt/db))
+ (auto-pk-column?)
(merge (ordered-map/ordered-map
upload/auto-pk-column-keyword ::upload-types/auto-incrementing-int-pk)))
rows [["Obi-Wan Kenobi"]]}}]
@@ -1347,7 +1352,12 @@
{:primary-key [upload/auto-pk-column-keyword]}
{}))
(driver/insert-into! driver db-id schema+table-name insert-col-names rows)
- (sync-upload-test-table! :database (mt/db) :table-name table-name :schema-name schema-name)))
+ (let [table (sync-upload-test-table! :database (mt/db) :table-name table-name :schema-name schema-name)]
+ ;; ensure we have the same display name for the auto-pk-column that a real upload would have
+ (t2/update! :model/Field
+ {:table_id (:id table), :name upload/auto-pk-column-name}
+ {:display_name upload/auto-pk-column-name})
+ table)))
(defn catch-ex-info* [f]
(try
@@ -1481,6 +1491,8 @@
["_mb_row_id,id,extra 1, extra 2,name"]
nil
["extra 1, extra 2"]
+ ;; TODO note that the order of the fields is reversed
+ ;; It would be better if they were alphabetical, or matched the order in the database / file.
(trim-lines "The CSV file is missing columns that are in the table:
- id
- name
@@ -1500,8 +1512,8 @@
- name
There are new columns in the CSV file that are not in the table:
- - _mb_row_id
- - extra_2"))}]
+ - extra_2
+ - _mb_row_id"))}]
(with-upload-table!
[table (create-upload-table!
{:col->upload-type (ordered-map/ordered-map
@@ -1604,53 +1616,54 @@
(deftest update-mb-row-id-csv-only-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
- (doseq [action (actions-to-test driver/*driver*)]
- (testing (action-testing-str action)
- (testing "If the table doesn't have _mb_row_id but the CSV does, ignore the CSV _mb_row_id but create the column anyway"
- (with-upload-table!
- [table (create-upload-table! {:col->upload-type (ordered-map/ordered-map
- :name vchar-type)
- :rows [["Obi-Wan Kenobi"]]})]
- (let [csv-rows ["_MB-row ID,name" "1000,Luke Skywalker"]
- file (csv-file-with csv-rows)]
- (is (= {:row-count 1}
- (update-csv! action {:file file, :table-id (:id table)})))
- ;; Only create auto-pk columns for drivers that supported uploads before auto-pk columns
- ;; were introduced by metabase#36249. Otherwise we can assume that the table was created
- ;; with an auto-pk column.
- (if (driver/create-auto-pk-with-append-csv? driver/*driver*)
- (do
- (testing "Check a _mb_row_id column was created"
- (is (= ["name" "_mb_row_id"]
- (column-names-for-table table))))
- (testing "Check a _mb_row_id column was sync'd"
- (is (=? {:semantic_type :type/PK
- :base_type :type/BigInteger
- :name "_mb_row_id"
- :display_name "_mb_row_id"}
- (t2/select-one :model/Field :table_id (:id table) :name upload/auto-pk-column-name))))
- (testing "Check the data was uploaded into the table, but the _mb_row_id column values were ignored"
+ (when (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (doseq [action (actions-to-test driver/*driver*)]
+ (testing (action-testing-str action)
+ (testing "If the table doesn't have _mb_row_id but the CSV does, ignore the CSV _mb_row_id but create the column anyway"
+ (with-upload-table!
+ [table (create-upload-table! {:col->upload-type (ordered-map/ordered-map
+ :name vchar-type)
+ :rows [["Obi-Wan Kenobi"]]})]
+ (let [csv-rows ["_MB-row ID,name" "1000,Luke Skywalker"]
+ file (csv-file-with csv-rows)]
+ (is (= {:row-count 1}
+ (update-csv! action {:file file, :table-id (:id table)})))
+ ;; Only create auto-pk columns for drivers that supported uploads before auto-pk columns
+ ;; were introduced by metabase#36249. Otherwise we can assume that the table was created
+ ;; with an auto-pk column.
+ (if (driver/create-auto-pk-with-append-csv? driver/*driver*)
+ (do
+ (testing "Check a _mb_row_id column was created"
+ (is (= ["name" "_mb_row_id"]
+ (column-names-for-table table))))
+ (testing "Check a _mb_row_id column was sync'd"
+ (is (=? {:semantic_type :type/PK
+ :base_type :type/BigInteger
+ :name "_mb_row_id"
+ :display_name "_mb_row_id"}
+ (t2/select-one :model/Field :table_id (:id table) :name upload/auto-pk-column-name))))
+ (testing "Check the data was uploaded into the table, but the _mb_row_id column values were ignored"
+ (case action
+ ::upload/append
+ (is (= [["Obi-Wan Kenobi" 1]
+ ["Luke Skywalker" 2]]
+ (rows-for-table table)))
+ ::upload/replace
+ (is (= [["Luke Skywalker" 1]]
+ (rows-for-table table))))))
+ (do
+ (testing "Check a _mb_row_id column wasn't created"
+ (is (= ["name"]
+ (column-names-for-table table))))
(case action
::upload/append
- (is (= [["Obi-Wan Kenobi" 1]
- ["Luke Skywalker" 2]]
+ (is (= [["Obi-Wan Kenobi"]
+ ["Luke Skywalker"]]
(rows-for-table table)))
::upload/replace
- (is (= [["Luke Skywalker" 1]]
+ (is (= [["Luke Skywalker"]]
(rows-for-table table))))))
- (do
- (testing "Check a _mb_row_id column wasn't created"
- (is (= ["name"]
- (column-names-for-table table))))
- (case action
- ::upload/append
- (is (= [["Obi-Wan Kenobi"]
- ["Luke Skywalker"]]
- (rows-for-table table)))
- ::upload/replace
- (is (= [["Luke Skywalker"]]
- (rows-for-table table))))))
- (io/delete-file file))))))))
+ (io/delete-file file)))))))))
(deftest update-no-mb-row-id-failure-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
@@ -1859,26 +1872,13 @@
(deftest update-mb-row-id-csv-and-table-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
- (doseq [action (actions-to-test driver/*driver*)]
- (testing (action-testing-str action)
- (testing "Append succeeds if the table has _mb_row_id and the CSV does too"
- (with-upload-table! [table (create-upload-table!)]
- (let [csv-rows ["_mb_row_id,name" "1000,Luke Skywalker"]
- file (csv-file-with csv-rows (mt/random-name))]
- (is (= {:row-count 1}
- (update-csv! action {:file file, :table-id (:id table)})))
- (testing "Check the data was uploaded into the table, but the _mb_row_id was ignored"
- (is (= (set (updated-contents action
- [["Obi-Wan Kenobi"]]
- [["Luke Skywalker"]]))
- (set (rows-for-table table)))))
- (io/delete-file file)))
-
- ;; TODO we can deduplicate a lot of code in this test
- (testing "with duplicate normalized _mb_row_id columns in the CSV file"
+ (when (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (doseq [action (actions-to-test driver/*driver*)]
+ (testing (action-testing-str action)
+ (testing "Append succeeds if the table has _mb_row_id and the CSV does too"
(with-upload-table! [table (create-upload-table!)]
- (let [csv-rows ["_mb_row_id,name,-MB-ROW-ID" "1000,Luke Skywalker,1001"]
- file (csv-file-with csv-rows)]
+ (let [csv-rows ["_mb_row_id,name" "1000,Luke Skywalker"]
+ file (csv-file-with csv-rows (mt/random-name))]
(is (= {:row-count 1}
(update-csv! action {:file file, :table-id (:id table)})))
(testing "Check the data was uploaded into the table, but the _mb_row_id was ignored"
@@ -1886,7 +1886,21 @@
[["Obi-Wan Kenobi"]]
[["Luke Skywalker"]]))
(set (rows-for-table table)))))
- (io/delete-file file)))))))))
+ (io/delete-file file)))
+
+ ;; TODO we can deduplicate a lot of code in this test
+ (testing "with duplicate normalized _mb_row_id columns in the CSV file"
+ (with-upload-table! [table (create-upload-table!)]
+ (let [csv-rows ["_mb_row_id,name,-MB-ROW-ID" "1000,Luke Skywalker,1001"]
+ file (csv-file-with csv-rows)]
+ (is (= {:row-count 1}
+ (update-csv! action {:file file, :table-id (:id table)})))
+ (testing "Check the data was uploaded into the table, but the _mb_row_id was ignored"
+ (is (= (set (updated-contents action
+ [["Obi-Wan Kenobi"]]
+ [["Luke Skywalker"]]))
+ (set (rows-for-table table)))))
+ (io/delete-file file))))))))))
(deftest update-duplicate-header-csv-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
@@ -1959,8 +1973,8 @@
file (csv-file-with csv-rows)]
(testing "The new row is inserted with the values correctly reordered"
(is (= {:row-count 1} (update-csv! action {:file file, :table-id (:id table)})))
- (is (= ["name" "α"]
- (rest (column-display-names-for-table table))))
+ (is (= (header-with-auto-pk ["name" "α"])
+ (column-display-names-for-table table)))
(is (= (set (updated-contents action
[["Obi-Wan Kenobi" nil]]
[["Everything" "omega"]]))
@@ -1976,7 +1990,7 @@
;; for drivers that insert rows in chunks, we change the chunk size to 1 so that we can test that the
;; inserted rows are rolled back
(binding [driver/*insert-chunk-rows* 1]
- (doseq [auto-pk-column? (if (driver.u/supports? driver/*driver* :upload-with-auto-pk (mt/db))
+ (doseq [auto-pk-column? (if (auto-pk-column?)
[true false]
[false])]
(testing (str "\nFor a table that has " (if auto-pk-column? "an" " no") " automatically generated PK already")
@@ -2138,6 +2152,15 @@
(update!)))))
(io/delete-file file)))))))))))
+(defn- round-floats
+ "Round all floats to have n digits of precision."
+ [digits-precision rows]
+ (let [factor (Math/pow 10 digits-precision)]
+ (mapv (partial mapv #(if (float? %)
+ (/ (Math/round (* factor %)) factor)
+ %))
+ rows)))
+
(deftest update-promotion-multiple-columns-test
(mt/test-drivers (disj (mt/normal-drivers-with-feature :uploads) :redshift) ; redshift doesn't support promotion
(doseq [action (actions-to-test driver/*driver*)]
@@ -2163,8 +2186,10 @@
(testing "\nAppend should succeed"
(is (= {:row-count 1}
(update!))))
- (is (= [[1 coerced coerced]]
- (rows-for-table table))))))))))))))
+ (is (= (rows-with-auto-pk [[coerced coerced]])
+ ;; Clickhouse uses 32bit floats, so we must account for that loss in precision.
+ ;; In this case 2.1 => 2.0999999046325684
+ (round-floats 6 (rows-for-table table)))))))))))))))
(deftest create-from-csv-int-and-float-test
(testing "Creation should handle a mix of int and float-or-int values in any order"
@@ -2176,8 +2201,8 @@
"1, 1.0"
"1.0, 1"]))]
(testing "Check the data was uploaded into the table correctly"
- (is (= [[1 1.0 1.0]
- [2 1.0 1.0]]
+ (is (= (rows-with-auto-pk [[1.0 1.0]
+ [1.0 1.0]])
(rows-for-table table)))))))))
(deftest create-from-csv-int-and-non-integral-float-test
@@ -2190,9 +2215,11 @@
"1, 1.1"
"1.1, 1"]))]
(testing "Check the data was uploaded into the table correctly"
- (is (= [[1 1.0 1.1]
- [2 1.1 1.0]]
- (rows-for-table table)))))))))
+ (is (= (rows-with-auto-pk [[1.0 1.1]
+ [1.1 1.0]])
+ ;; Clickhouse uses 32 bit floats,
+ ;; so 1.1 is approximated by 1.10000002384185791015625
+ (round-floats 7 (rows-for-table table))))))))))
(deftest update-from-csv-int-and-float-test
(mt/test-drivers (mt/normal-drivers-with-feature :uploads)
@@ -2278,10 +2305,10 @@
(map (partial #'upload/normalize-column-name driver/*driver*)))]
(is (every? (set column-names) header-names))))
(testing "We preserve prefixes where_possible"
- (is (= {"_mb_row_" 1
- "a_really" 1
+ (is (= {"a_really" 1
"b_really" 2}
- (frequencies (map #(subs % 0 8) column-names))))))))))))))
+ (dissoc (frequencies (map #(subs % 0 8) column-names))
+ "_mb_row_")))))))))))))
(deftest append-with-really-long-names-test
(testing "Upload a CSV file with unique column names that get sanitized to the same string"
@@ -2300,10 +2327,12 @@
(is (= {:row-count 1}
(update-csv! ::upload/append {:file file, :table-id (:id table)})))
(testing "Check the data was appended into the table"
- (is (= (map second (rows-with-auto-pk
- [(csv/read-csv original-row)
- (csv/read-csv appended-row)]))
- (map rest (rows-for-table table)))))
+ (is (= (set
+ (rows-with-auto-pk
+ (concat
+ (csv/read-csv original-row)
+ (csv/read-csv appended-row))))
+ (set (rows-for-table table)))))
(io/delete-file file))))))))
(deftest append-with-preserved-display-name-test
@@ -2351,8 +2380,8 @@
:data {:status-code 422}}
(catch-ex-info (update-csv! ::upload/append {:file file, :table-id (:id table)}))))
(testing "Check the data was not uploaded into the table"
- (is (= (map second (rows-with-auto-pk [(csv/read-csv original-row)]))
- (map rest (rows-for-table table)))))
+ (is (= (rows-with-auto-pk (csv/read-csv original-row))
+ (rows-for-table table))))
(io/delete-file file))))))))
(driver/register! ::short-column-test-driver)