github.com/neilotoole/sq

Added

• Related to [#270], the output of sq inspect now includes the source's catalog (in JSON and YAML output formats).

Fixed

• MySQL driver didn't populate all expected values for sq inspect --overview.

Changed

• ☢️ Removed unused --exec and --query flags from sq sql command.

Added

• [#270]: Flag --src.schema permits switching the source's schema (and catalog) for the duration of the command. The flag is supported for the sq, sql and inspect commands.
• New SLQ functions catalog() and schema() return the catalog and schema of the DB connection.
• The SLQ unique function now has a synonym uniq.

Changed

• sq src --text now outputs only the handle of the active source. Previously it also printed the driver type and short location. Instead use sq src --text --verbose to see those details.

Fixed

• [#308]: Fix to allow build on 32-bit systems. Thanks @icp.

Added

• 🐥 [#279]: The SQLite driver now has initial support for several SQLite extensions baked in, including Virtual Table and FTS5. Note that this is an early access release of extensions support. Please open an issue if you discover something bad.

Fixed

• sq version was missing a newline in its output.

This release is heavily focused on improvements to Microsoft Excel support. The underlying Excel library has been changed from tealeg/xlsx to qax-os/excelize, largely because tealeg/xlsx is no longer actively maintained. Thus, both the XLSX output writer and the XLSX driver have been rewritten. There should be some performance improvements, but it's also possible that the rewrite introduced bugs. If you discover anything strange, please open an issue.

Added

• [#99]: The CSV and XLSX drivers can now handle duplicate header column names in the ingest data. For example, given a CSV file:

  actor_id,first_name,actor_id
  1,PENELOPE,1
  2,NICK,2
  

  The columns will be renamed to:

  actor_id,first_name,actor_id_1
  

  The renaming behavior is controlled by a new option ingest.column.rename This new option is effectively the ingest counterpart of the existing output option result.column.rename.

• [#191]: The XLSX driver now detects header rows, like the CSV driver already does. Thus, you now typically don't need to specify the --ingest.header flag for Excel files. However, the option remains available in case sq can't figure it out for a particular file.

• The Excel writer has three new config options for controlling date/time output. Note that these format strings are distinct from format.datetime and friends, because Excel has its own format string mechanism.

  • format.excel.datetime: Controls datetime format, e.g. 2023-08-03 16:07:01.
  • format.excel.date: Controls date-only format, e.g. 2023-08-03.
  • format.excel.time: Controls time-only format, e.g. 4:07 pm.

• The ingest kind detectors (e.g. for CSV or XLSX) now detect more date & time formats as kind.Datetime, kind.Date, and kind.Time.

• If an error occurs when the output format is text, a stack trace is printed to stderr when the command is executed with --verbose (-v).

• There's a new option error.format that controls error output format independent of the main format option . The error.format value must be one of text or json.

Changed

• ☢️ The default Excel date format has changed. Previously the format was 11/9/89, and now it is 1989-11-09. The same applies to datetimes, e.g. 11/9/1989 00:00:00 becomes 1989-11-09 00:00.

  This change is made to reduce ambiguity and confusion. sq uses a library to interact with Excel files, and it seems that the library chooses a particular format by default (11/9/89). There are several paths we could take here:

  1. Interrogate the OS, and use the OS locale date format.
  2. Stick with the library default 11/9/89.
  3. Pick a default other than 11/9/89.

  We pick the third option. The first option (locale-dependent) is excluded because, as a general rule, we want sq to produce the same output regardless of locale/system settings. We exclude the second option because month/day confuses most of the world. Thus, we're left with picking a default, and 1989-11-09 is the format used in RFC3339 and friends.

  Whether this is the correct (standard?) approach is still unclear, and feedback is welcome. However, the user can make use of the new config options (format.excel.datetime etc.) to customize the format as they see fit.

• The XLSX writer now outputs header rows in bold text.

• ☢️ The XLSX writer now outputs blob (bytes) cell data as a base64-encoded string, instead of raw bytes.

Fixed

• Fixed bug where source-specific config wasn't being propagated.

This release features a complete overhaul of the join mechanism.

Added

• [#277]: A table selector can now have an alias. This in and of itself is not particularly useful, but it's a building block for multiple joins.

  $ sq '@sakila | .actor:a | .a.first_name'
  

• New option result.column.rename that exposes a template used to rename result set column names before display. The primary use case is to de-duplicate columns names on a SELECT * FROM tbl1 JOIN tbl2, where tbl1 and tbl2 have clashing column names (docs).

• [#157]: Previously only join (INNER JOIN) was available: now the rest of the join types such as left_outer_join, cross_join, etc. are implemented (docs).

Changed

• ☢️ [#12]: The table join mechanism has been completely overhauled. Now there's support for multiple joins. See docs.

  # Previously, only a single join was possible
  $ sq '.actor, .film_actor | join(.actor_id)'
  
  # Now, an arbitrary number of joins
  $ sq '.actor | join(.film_actor, .actor_id) | join(.film, .film_id)'
  

• ☢️ The alias for --jsonl (JSON Lines) has been changed to -J.

Fixed

• Config options weren't being propagated correctly to all parts of the code.

Fixed

• Bug with sq version output on Windows.

Added

• [#263]: sq version now supports --yaml output.
• [#263]: sq version now outputs host OS details with --verbose, --json and --yaml flags. The motivation behind this is bug submission: we want to know which OS/arch the user is on. E.g. for sq version -j:

{
  "version": "v0.38.1",
  "commit": "eedc11ec46d1f0e78628158cc6fd58850601d701",
  "timestamp": "2023-06-21T11:41:34Z",
  "latest_version": "v0.39.0",
  "host": {
    "platform": "darwin",
    "arch": "arm64",
    "kernel": "Darwin",
    "kernel_version": "22.5.0",
    "variant": "macOS",
    "variant_version": "13.4"
  }
}

• [#263]: The output of sq inspect and sq inspect -v has been refactored significantly, and should now be easier to work with (docs).

Fixed

• [#261]: The JSON writer (--json) could get deadlocked when a record contained a large amount of data, triggering an internal Flush() (which is mutex-guarded) from within the mutex-guarded WriteRecords() method.

This release has significant improvements (and breaking changes) to SLQ (sq's query language).

Changed

• ☢️ [#254]: The formerly-implicit "WHERE" mechanism now requires an explicit where() function. This, alas, is a fairly big breaking change. But it's necessary to remove an ambiguity roadblock. See discussion in the issue.

  # Previously
  $ sq '.actor | .actor_id <= 2'
  
  # Now
  $ sq '.actor | where(.actor_id <= 2)'
  

• [#256]: Column-only queries are now possible. This has the neat side effect that sq can now be used as a calculator.

  $ sq 1+2
  1+2
  3
  

  You may want to use --no-header (-H) when using sq as a calculator.

  $ sq -H 1+2
  3
  $ sq -H '(1+2)*3'
  9
  

Fixed

• Literals can now be selected (docs).

  $ sq '.actor | .first_name, "X":middle_name, .last_name | .[0:2]'
  first_name  middle_name  last_name
  PENELOPE    X            GUINESS
  NICK        X            WAHLBERG
  

• Lots of expressions that previously failed badly, now work.

  $ sq '.actor | .first_name, (1+2):addition | .[0:2]'
  first_name  addition
  PENELOPE    3
  NICK        3
  

• [#258]: Column aliases can now be arbitrary strings, instead of only a valid identifier.

  # Previously only valid identifier allowed
  $ sq '.actor | .first_name:given_name | .[0:2]'
  given_name
  PENELOPE
  NICK
  
  # Now, any arbitrary string can be used
  $ sq '.actor | .first_name:"Given Name" | .[0:2]'
  Given Name
  PENELOPE
  NICK
  

Fixed

• [#252]: Handle *uint64 returned from DB.

Added

• [#244]: Shell completion for sq add LOCATION. See docs.

Changed

• ☢️ Proprietary database functions are now invoked by prefixing the function name with an underscore. For example:

  # mysql "date_format" func
  $ sq '@sakila/mysql | .payment | _date_format(.payment_date, "%m")'
  
  # Postgres "date_trunc" func
  $ sq '@sakila/postgres | .payment | _date_trunc("month", .payment_date)'
  

Fixed

• sq diff: Renamed --count flag to --counts as intended.

The major feature is the long-gestating sq diff.

Added

• [#229]: sq diff compares two sources, or tables.
• sq inspect --dbprops is a new mode that returns only the DB properties. Relatedly, the properties mechanism is now implemented for all four supported DB types (previously, it was only implemented for Postgres and MySQL).
• CSV format now colorizes output.

Changed

• sq inspect -v previously returned DB properties in a field named db_variables. This field has been renamed to db_properties. The renaming reflects the fact that some of those properties aren't really variables in the sense that they can be modified (e.g. DB server version or such).
• The structure of the former db_variables (now db_properties) field has changed. Previously it was an array of {"name": "XX", "value": "YY"} values, but now is a map, where the keys are strings, and the values can be either a scalar (bool, int, string, etc.), or a nested value such as an array or map. This change is made because some databases (e.g. SQLite) feature complex data in some property values.
• CSV format now renders byte sequences as [777 bytes] instead of dumping the raw bytes.
• ☢️ TSV format (--tsv) no longer has a shorthand form -T. Apparently that shorthand wasn't used much, and -T is needed elsewhere.
• ☢️ Likewise, --xml no longer has shorthand -X. And --markdown has lost alias --md.
• In addition to the format flags --text, --json, etc., there is now a --format=FORMAT flag, e.g. --format=json. This will allow sq to continue to expand the number of output formats, without needing to have a dedicated flag for each format.

Fixed

• sq config edit @source was failing to save any edits.

Added

• [#8]: Results can now be output in YAML.

Fixed

• sq config get OPT --text now prints only the value, not KEY VALUE. If you want to see key and value, consider using --yaml, or --text --verbose.

Fixed

• Both --markdown and the alias --md are now supported.

Fixed

• Fixed a minor issue where sq ls -jv and sq ls -yv produced no output if config contained no explicitly set options.

This release significantly overhauls sq's config mechanism ([#199]). For an overview, see the new config docs.

Alas, this release has several minor breaking changes ☢️.

Added

• sq config ls shows config.

• sq config get gets individual config option.

• sq config set sets config values.

• sq config edit edits config.

  • Editor can be specified via $EDITOR or $SQ_EDITOR.

• sq config location prints the location of the config dir.

• --config flag is now honored globally.

• Many more knobs are exposed in config.

• Logging is much more configurable. There are new knobs:

  $ sq config set log true
  $ sq config set log.level INFO
  $ sq config set log.file /var/log/sq.log
  

  There are also equivalent flags (--log, --log.file and --log.level) and envars (SQ_LOG, SQ_LOG_FILE and SQ_LOG_LEVEL).

• Several more commands support YAML output:

  • sq group
  • sq ls
  • sq mv
  • sq rm
  • sq src

Changed

• The structure of sq's config file (sq.yml) has changed. The config file is automatically upgraded when using the new version.

• The default location of the sq log file has changed. The new location is platform-dependent. Use sq config get log.file -v to view the location, or sq config set log.file /path/to/sq.log to set it.

• ☢️ Envar SQ_CONFIG replaces SQ_CONFIGDIR.

• ☢️ Envar SQ_LOG_FILE replaces SQ_LOGFILE.

• ☢️ Format flag --table is renamed to --text. This is changed because while the output is mostly in table format, sometimes it's just plain text. Thus table was not quite accurate.

• ☢️ The flag to explicitly specify a driver when piping input to sq has been renamed from --driver to --ingest.driver. This change aligns the naming of the ingest options and reduces ambiguity.

  # previously
  $ cat mystery.data | sq --driver=csv '.data'
  
  # now
  $ cat mystery.data | sq --ingest.driver=csv '.data'
  

• ☢️ sq add no longer has the generic --opts x=y mechanism. This flag was ambiguous and confusing. Instead, use explicit option flags.

  # previously
  $ sq add ./actor.csv --opts=header=false
  
  # now
  $ sq add ./actor.csv --ingest.header=false
  

• ☢️ The short form of the sq add --handle flag has been changed from -h to -n. While this is not ideal, the -h shorthand is already in use everywhere else as the short form of --header.

  # previously
  $ sq add ./actor.csv -h @actor
  
  # now
  $ sq add ./actor.csv -n @actor
  

• ☢️ The --pretty flag has been removed. Its only previous use was with the json format, where if --pretty=false would output the JSON in compact form. To better align with jq, there is now a --compact / -c flag that behaves identically to jq.

• ☢️ Because of the above --compact / -c flag, the short form of the --csv flag is changing from -c to -C. It's an unfortunate situation, but alignment with jq's behavior is an overarching principle that justifies the change.

The headline feature is source groups. This is the biggest change to the sq CLI in some time, and should make working with lots of sources much easier.

Added

• [#192]: sq now has a mechanism to group sources. A source handle can now be scoped. For example, instead of @sakila_prod, @sakila_staging, etc, you can use @prod/sakila, @staging/sakila. Use sq group prod to set the active group (which sq ls respects). See docs.
• sq group GROUP sets the active group to GROUP.
• sq group returns the active group (default is /, the root group).
• sq ls GROUP lists the sources in GROUP.
• sq ls --group (or sq ls -g) lists all groups.
• sq mv moves/renames sources and groups.

Changed

• sq ls now shows the active item in a distinct color. It no longer adds an asterisk to the active item.
• sq ls now sorts alphabetically when using --table format.
• sq ls now shows the sources in the active group only. But note that the default active group is / (the root group), so the default behavior of sq ls is the same as before.
• sq add hello.csv will now generate the handle @hello instead of @hello_csv. On a second invocation, it will return @hello1 instead of @hello_csv_1. Why this change? Well, with the availability of the source group mechanism, the _ character in the handle somehow looked ugly. And more importantly, _ is a relative pain to type.
• sq ping has changed to support groups. Instead of sq ping --all, you can do sq ping GROUP, e.g. sq ping /.

Added

• [#187]: For csv sources, sq will now try to auto-detect if the CSV file has a header row or not. Previously, this needed to be explicitly specified via an awkward syntax:

  $ sq add ./actor.csv --opts=header=true
  

  This change makes working with CSV files significantly lower friction. A command like the below now almost always works as expected:

  $ cat ./actor.csv | sq .data
  

  Support for Excel/XLSX header detection is in [#191].

Fixed

• sq is now better at detecting the (data) kind of CSV fields. It now more accurately distinguishes between Decimal and Int, and knows how to handle Datetime.

• [#189]: sq now treats CSV empty fields as NULL.

Added

• [#173]: Predefined variables via --arg flag (docs):

  $ sq --arg first TOM '.actor | .first_name == $first'
  

Changes

• Use --md instead of --markdown for outputting Markdown.

Fixed

• [#185]: sq inspect now better handles "too many connections" situations.
• go.mod: Moved to jackc/pgx v5.
• Refactor: switched to slog logging library.

Added

• [#164]: Implemented unique function (docs):

  $ sq '.actor | .first_name | unique'
  

  This is equivalent to:

  SELECT DISTINCT first_name FROM actor
  

• Implemented count_unique function (docs).

  $ sq '.actor | count_unique(.first_name)'
  

Changed

• The count function has been changed (docs)

  • Added no-args version: .actor | count equivalent to SELECT COUNT(*) AS "count" FROM "actor".
  • ☢️ The "star" version (.actor | count(*)) is no longer supported; use the naked version instead.

• Function columns are now named according to the sq token, not the SQL token.

  # previous behavior
  $ sq '.actor | max(.actor_id)'
  max("actor_id")
  200
  
  # now
  $ sq '.actor | max(.actor_id)'
  max(.actor_id)
  200
  

Added

• [#162]: group_by now accepts function arguments.

Changed

• Renamed groupby to group_by to match jq.
• Renamed orderby to order_by to match jq.

Added

• [#160]: Use groupby() to group results. See query guide.

Added

• [#158]: Use orderby() to order results. See query guide.

Added

• [#98]: Whitespace is now allowed in SLQ selector names. You can do @sakila | ."film actor" | ."actor id".

Fixed

• [#155]: sq inspect now populates schema field in JSON for MySQL, SQLite, and SQL Server (Postgres already worked).