Command Usage

csvq [options] [subcommand] ["query"|argument]

• Execution of Statements
• Options
• Subcommands
• Configurations
• Special Characters
• Return Code

Execution of Statements

There are three ways to execute a query or statements.

# Pass to the csvq command as an argument
$ csvq 'SELECT id, name FROM users'

# Load from file
$ cat statements.sql
VAR @id := 0;
SELECT @id := @id + 1 AS id,
       name
  FROM users;

$ csvq -s statements.sql

# Execute in the interactive shell
$ csvq
csvq > SELECT id, name FROM users;
+----+-------+
| id | name  |
+----+-------+
| 1  | Louis |
| 2  | Sean  |
+----+-------+
csvq > UPDATE users SET name = 'Mildred' WHERE id = 2;
1 record updated on "/home/mithrandie/docs/csv/users.csv".
csvq > SELECT id, name FROM users;
+----+----------+
| id | name     |
+----+----------+
| 1  | Louis    |
| 2  | Mildred  |
+----+----------+
csvq > COMMIT;
Commit: file "/home/mithrandie/docs/csv/users.csv" is updated.
csvq >

Options

–repository PATH, -r PATH

Directory Path where files are located. The default is the current directory.

–timezone value, -z value

Default Timezone. The default is Local.

Local, UTC or a timezone name in the IANA TimeZone database(in the form of “Area/Location”. e.g. “America/Los_Angeles”).

The timezone database is required in order to use the timezone names. Most Unix-like systems provide the database. But if your system does not provide it and you have not installed Go Lang, then you must put the database file named zoneinfo.zip to the directory “$ZONEINFO” or “$GOROOT/lib/time/”.

–datetime-format value, -t value

Datetime Format to parse strings. Format string is the same as the function DATETIME_FORMAT.

This option can be specified multiple formats using JSON array of strings.

–ansi-quotes, -k

Use double quotation mark (U+0022 ") as identifier enclosure.

–strict-equal, -g

Compare strictly that two values are equal for DISTINCT, GROUP BY and ORDER BY.

–wait-timeout value, -w value

Limit of the waiting time in seconds to wait for locked files to be released. The default is 10.

–source FILE, -s FILE

Load query or statements from FILE.

–import-format value, -i value

Default format to load files. The default is CSV.

Regardless of this option, files with the following extensions will be read in a specific format.

–delimiter value, -d value

Field delimiter for CSV. The default is a comma(U+002C ,).

A delimiter must be one character. Special Characters can be used with backslash escaping.

–allow-uneven-fields

Allow loading CSV files with uneven field length.

–delimiter-positions value, -m value

Delimiter positions for Fixed-Length Format. The default is “SPACES”.

You can specify “SPACES” or a JSON Array of integers. “SPACES” splits lines automatically by spaces. JSON Array is used to specify delimiter positions explicitly.

Delimiter positions indicate the number of bytes from the start of the line. For example, JSON Array “[5, 10, 15]” splits “1234567890abcde” as “12345, 67890, abcde”

If you want to operate Single-Line Fixed-Length Format, then connect a JSON Array to “S”(U+0053) or “s”(U+0073). For example, “S[2, 3, 6]” imports “01aabc02bdef03cghi” as “(‘01’, ‘a’, ‘abc’), (‘02’, ‘b’, ‘def’), (‘03’, ‘c’, ‘ghi’)”.

–json-query QUERY, -j QUERY

QUERY for JSON.

–encoding value, -e value

File encoding. Following encodings are supported. The default is AUTO.

JSON Format is supported only UTF-8.

Whatever the value of this option is, if the first character in a file is a UTF-8 byte order mark, the file will be loaded as UTF-8 encoding.

–no-header, -n

Import the first line as a record.

First line of a CSV file is dealt with as the header line. In case “–no-header” option passed, fields are automatically named as “c” and following sequential number. e.g. “c1”, “c2”, “c3”, …

–without-null, -a

Parse empty fields as empty strings.

In most cases CSV fields are imported as string values, but no-quoted empty fields are imported as nulls. By using the “–without-null” option, no-quoted empty fields are imported as empty string values.

–out FILE, -o FILE

Export result sets of select queries to FILE.

If the output file is not specified, the result sets are written to standard output.

–strip-ending-line-break, -T

Strip line break from the end of files and query results.

–format value, -f value

Format of query results. The default is TEXT, but CSV is used for output to pipe.

When this option is not specified, the file specified by the –out option will be output in a specific format if it has the following extensions.

–write-encoding value, -E value

Character encoding of query results. The default is UTF8.

–write-delimiter value, -D value

Field delimiter for query results in CSV format. The default is a comma(U+002C ,).

–write-delimiter-positions value, -M value

Delimiter positions for query results in Fixed-Length format. The default is “SPACES”.

If the field value is shorter than the length of the field, the missing part is padded with SPACE(U+0020).
For example, JSON Array “[5, 10, 15]” combines “123, abc, def” into “␣␣123abc␣␣def␣␣”.

–without-header, -N

Export result sets of select queries without the header line.

–line-break value, -l value

Line break in query results. One of following values. The default is LF.

–enclose-all, -Q

Enclose all string values in CSV.

–json-escape, -J

JSON escape type. The default is BACKSLASH.

Escaped characters in JSON

–pretty-print, -P

Make JSON output easier to read in query results.

–scientific-notation -SN

Use Scientific Notation for large exponents in output.

–east-asian-encoding, -W

Count ambiguous characters as fullwidth. If not, then that characters are counted as halfwidth.

–count-diacritical-sign, -S

Count diacritical signs as halfwidth. If not, then that characters are counted as zero-width.

–count-format-code, -A

Count format characters and zero-width spaces as halfwidth. If not, then that characters are counted as zero-width.

–color, -c

Use ANSI color escape sequences.

–quiet, -q

Suppress operation log output.

–limit-recursion

Maximum number of iterations for recursive queries. “-1” means no limit. The default is 1000.

–cpu, -p

Hint for the number of cpu cores to be used. The default is the half of the number of cpu cores.

–stats, -x

Show execution time and memory statistics.

Query Execution Time

execution time of one query. select, insert, update, or delete queries are measured.

TotalTime

total execution time

TotalAlloc

cumulative bytes of allocated heap objects

HeapSys

bytes of heap memory obtained from the OS

Mallocs

cumulative count of heap objects allocated

Frees

cumulative count of heap objects freed

–help, -h

Show help

–version, -v

Print the version

If you want to pass “false” to a boolean command option, you can specify it as “–option-name=false”.
Some command options can also be specified in statements by using Set Flag Statements.

Determination of file format

Loading

First, the file extension to be loaded is checked. If the extension is displayed in the following table, the associated file format will be selected. Otherwise, the passed value by the “–import-format” option will be used to load.

The following options are available for loading.

• –delimiter value, -d value
• –delimiter-positions value, -m value
• –json-query QUERY, -j QUERY
• –encoding value, -e value
• –no-header, -n
• –without-null, -a

You can also use Table Object Expressions to specify the format each file. Table Object Expression effects the first loading in a transaction. After the second loading, the specifications in the table object expression are ignored. You must use the ROLLBACK statement to discard all changes in the transaction if you reload the same file.

Updating

The table attributes that were determined when loading will be used to updating. You can modify table attributes by using SET ATTRIBUTE Statement).

Creating

First, the file extension to be loaded is checked. If the extension is displayed in the following table, the associated file format will be selected. Otherwise CSV is used as the format. You can modify table attributes by using SET ATTRIBUTE Statement).

Exporting query results with the “–out” option

The passed value by the “–format” option will be used to export.

The following options are available for exporting.

• –write-encoding value, -E value
• –write-delimiter value, -D value
• –write-delimiter-positions value, -M value
• –without-header, -N
• –line-break value, -l value
• –enclose-all, -Q
• –json-escape, -J
• –pretty-print, -P
• –east-asian-encoding, -W
• –count-diacritical-sign, -S
• –count-format-code, -A

Subcommands

Fields Subcommand

Show fields in a file.

csvq [options] fields CSV_FILE_PATH

Calc Subcommand

Calculate a value from stdin.

csvq [options] calc "expression"

In the expression, the value from stdin is represented as “c1”.

Example:

$ echo 'foo' | csvq calc "base64_encode(c1)"
Zm9v
$ echo 123 | csvq calc "c1 + 100"
223

Syntax Subcommand

Print syntax.

csvq [options] syntax [search_word ...]

Check Update Subcommand

Check for updates.

csvq [options] check-update [subcommand options]

Subcommand Options

–include-pre-release

Check including pre-release version.

Configurations

Before an execution of csvq, the following processings will be performed.

1. Load Environment Configurations.
2. Execute Pre-Load Statements.
3. Overwrite Flags with Command Options.

Environment Configurations

Configuration Files are loaded in the following order.

1. $XDG_CONFIG_HOME/csvq/csvq_env.json or $HOME/.config/csvq/csvq_env.json
2. $HOME/.csvq_env.json
3. $HOME/.csvq/csvq_env.json
4. CURRENT_DIRECTORY/csvq_env.json

Download a JSON file from https://github.com/mithrandie/csvq/blob/master/csvq_env.json and put it to the above paths. If the files exists in the multiple paths, then all existing files will be loaded and configuration items are overwritten with the values in the file that is loaded later.

Configuration Items in JSON

Interactive Shell

Except for Prompt and Continuous Prompt, items in this category are effective only on the following systems.

• darwin dragonfly freebsd linux netbsd openbsd solaris windows

History File

File name to save command history. If it is a relative path, then the path is interpreted as a relative path from your home directory. On UNIX-like systems, you can use environment variable such as $HOME or ${HOME} to specify the path. If not, ‘~’ can be used to represent home directory.

History Limit

Max length of command history. If history_limit is set to -1, then the command history is disabled.

Prompt, Continuous Prompt

Appearance of the prompt on the interactive shell. You can embed some expressions in the appearance by using a Dollar Sign(U+0024 $) and Curly Brackets(U+007B, U+007D {})

Completion

Whether to use completion.

Kill Whole Line

If true then keyboard shortcut “Ctrl+U” will remove the entire line. If not, the shortcut will remove before the current cursor position on the line.

Vi Mode

Whether to use vi-mode.

Effect Object

Effects

Effects are specified with an array of strings. Following strings are supported.

• Reset Bold Faint Italic Underline SlowBlink RapidBlink ReverseVideo Conceal CrossedOut

Colors

Foreground and background colors are specified with one of the following values.

8 Color (String)

Following strings are supported.

• Black Red Green Yellow Blue Magenta Cyan White BrightBlack BrightRed BrightGreen BrightYellow BrightBlue BrightMagenta BrightCyan BrightWhite DefaultColor

256 Color (Number)

Color number from 0 to 255.

RGB Color (Array of 3 numbers)

The first element is the intensity of red between 0 and 255, the second is green, and the third is blue.

Pre-Load Statements

Files in which statements are written will be loaded and executed in the following order.

1. $XDG_CONFIG_HOME/csvq/csvqrc or $HOME/.config/csvq/csvqrc
2. $HOME/.csvqrc
3. $HOME/.csvq/csvqrc
4. CURRENT_DIRECTORY/csvqrc

Special Characters

In command parameters and statements, following strings represent special characters.

Escaped Characters in JSON output

Return Code