v0.3.1. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

ATTENTION! The driver was successfully tested with clickhouse-jdbc-0.4.6-shaded.jar (clickhouse-jdbc-0.3.2-patch8-shaded.jar works well too).

As part of our effort to enhance and improve the connector functionality, we tested the connector with the Tableau Data source Verification Tool (TDVT). The changes included in this version cover test fixes.

Changes:

Fix String escaping generated by PostgreSQL90Dialect

v0.3.0. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

ATTENTION! The driver was successfully tested with clickhouse-jdbc-0.4.6-shaded.jar (clickhouse-jdbc-0.3.2-patch8-shaded.jar works well too).

As part of our effort to enhance and improve the connector functionality, we tested the connector with the Tableau Data source Verification Tool (TDVT). The changes included in this version cover test fixes.

Changes:

• Fix spaces function to work with 0 input
• Add round function with real real arguments
• Fix DATENAME week to use ClickHouse' toWeek function

v0.2.6. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch8-shaded.jar

Changes:

• Full support for ISO 8601 week-based calendar, including:
  • Added ISOYEAR, ISOQUARTER, ISOWEEK and ISOWEEKDAY functions
  • Added support for iso-year, iso-quarter, iso-week, iso-weekday date parts in DATEADD, DATEDIFF, DATENAME, DATEPART and DATETRUNC functions
  • Added functionality for choosing ISO 8601 week-based calendar as a standard calendar for a data source (Data -> (data source) -> Date Properties...)
  • More info: ISO-8601 Week-Based Calendar (help.tableau.com)
• Minor fixes and fixed typos

v0.2.5. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch8-shaded.jar

Changes:

• Added 'isoweek' & 'isoyear' literals to DATEPART function
• PROPER function has been added
• Minor fixes and fixed typos

v0.2.4. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch7-shaded.jar

Changes:

• DATENAME, DATEPART, DATETRUNC functions were fixed

v0.2.3. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch7-shaded.jar

Changes:

• SUM_IF(), AVG_IF(), COUNT_IF(), MIN_IF() & MAX_IF() functions have been added.
• Recommendations for installing .taco on Tableau Server were changed, see README.

v0.2.2. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch2-shaded.jar (clickhouse-jdbc-0.3.2-patch1-shaded.jar works well too)

This version will be initial release for Tableau Exchange

Changes:

• Extract creation with Sample N rows (N random rows) option fixed again. In version 0.2.1, you can see a non-uniform sampling, in this release sampling is performed evenly.
• We have added the Analysis tips section to the README so that you can read the important notes and see the description of additional functions in Live mode (non-standard functions do not have a description in the help panel when creating Calculated Fields)

v0.2.1. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.2-patch2-shaded.jar (clickhouse-jdbc-0.3.2-patch1-shaded.jar works well too)

This version will be initial release for Tableau Exchange

Changes:

• Extract creation with Sample N rows (N random rows) option significantly improved (works fast with 1+ million rows)
• Added ability to create Extacts with Sample N percent of rows (N% of random rows) option
• Date functions in the dialect are optimized
• New functions for Calculated Fields in Live mode!
  • New DATE_BIN('day', 10, [my_datetime_or_date]) function (equivalent of toStartOfInterval() in ClickHouse) rounds down a Date or Date & Time to the given interval, for example:

     == my_datetime_or_date == | == DATE_BIN('day', 10, [my_datetime_or_date]) ==
        28.07.2004 06:54:50    |              21.07.2004 00:00:00
        17.07.2004 14:01:56    |              11.07.2004 00:00:00
        14.07.2004 07:43:00    |              11.07.2004 00:00:00
    

  • Native RAND() function returns UInt32 number, for example 3446222955 (equivalent of rand() in ClickHouse)
  • We fond of unofficial RANDOM() Tableau function, which returns Float between 0 and 1, now you can use it
  • Also RAND_CONSTANT() and RAND_CONSTANT([x]) added to dialect (equivalent of randConstant() in ClickHouse)
  • Feel free to use TO_TYPE_NAME([field]) aggregate function to get field type in ClickHouse, for example Nullable(String) (equivalent of toTypeName() in ClickHouse)
  • HEX([my_string]) and UNHEX([my_string]) also available (equivalents of hex() and unhex() in ClickHouse)
  • Get SHA256 hash easily with HEX(SHA256([my_string])) (equivalent of SHA256() in ClickHouse)
  • Use FORMAT_READABLE_QUANTITY([my_int]) to get for example 3.45 billion (equivalent of formatReadableQuantity() in ClickHouse)
  • FORMAT_READABLE_TIMEDELTA([time_period_in_sec], 'days') returns 24112 days, 29 minutes and 16 seconds. Second argument is optional, FORMAT_READABLE_TIMEDELTA([time_period_in_sec]) returns 120 years, 10 months, 8 days, 17 hours, 10 minutes and 10 seconds (equivalent of formatReadableTimeDelta() in ClickHouse)
  • Now you can get server setting by GET_SETTING('join_use_nulls'), returns 1
  • Forget about boring bar charts! Use BAR([my_int], [min_val_int], [max_val_int], [bar_string_length_int]) function instead (equivalent of bar() in ClickHouse). For example, this calculated field returns nice bars as String:

    BAR([my_int], [min_val_int], [max_val_int], [bar_string_length_int]) + "  " + FORMAT_READABLE_QUANTITY([my_int])
    

    == BAR() ==
    ██████████████████▊  327.06 million
    █████  88.02 million
    ███████████████  259.37 million
    

v0.2.0. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

ATTENTION! JDBC Driver required: clickhouse-jdbc-0.3.2-patch1-shaded.jar

Changes:

• The connector has been significantly redesigned for the JDBC driver version 0.3.2-patch1
• Now you don't need to write SET join_use_nulls=1; in Initial SQL for Top-N IN/OUT Sets to work
• Added limited support for data types UInt64, Int128, UInt256, Int256. Details in the README
• Significant changes to the Advanced tab due to the fact that the driver itself has been rewritten. Details in the README
• There is only one option left for reading metadata — TOP_0. The WHERE FALSE option is blocked due to a specific case with Postgres
• Added COUNTD_UNIQ() aggregation - uniq() approximation is used, which is much faster than uniqExact(), which is used by default in COUNTD()
• Errors of conversion from numbers to dates and back are fixed

v0.1.4. The signed taco file (ready to use in prod)

The connector is packaged and signed. Please follow the README to install and run.

JDBC Driver required: clickhouse-jdbc-0.3.1-patch-shaded.jar

Changes:

• Added FINDNTH() function
• Added the Split Statements in the Initial SQL capability
• Added capability of casting bigdecimal to string
• Added (Top N) Sample Records when making an extract
• Default Socket Timeout changed from 30sec to 300sec
• The values of the set_session_id connection parameter have been changed: 0/1 instead of false/true
• Queries are executed in a separate thread (CAP_JDBC_QUERY_ASYNC)
• The dialect is a little cleaned up