recsel

Query and filter GNU recutils database records

TLDR

Select all records from a recfile

$ recsel [file.rec]

Select records matching a condition

$ recsel -e "[field] = '[value]'" [file.rec]

Select records with multiple conditions

$ recsel -e "[field1] = '[value]' && [field2] > [number]" [file.rec]

Select specific fields only

$ recsel -p [field1],[field2] [file.rec]

Select records from a specific type

$ recsel -t [record_type] [file.rec]

Count matching records

$ recsel -c -e "[condition]" [file.rec]

Sort records by a field

$ recsel -S [field] [file.rec]

SYNOPSIS

recsel [options] [-e expression] [-t type] [file...]

recsel is part of GNU Recutils, a set of tools for managing plain-text databases using the rec format. It selects and prints records from recfiles based on expressions and criteria.
Recfiles are human-readable text files where each record is a collection of named fields separated by blank lines. The format is simple, versionable with git, and editable with any text editor.
Selection expressions support comparison operators (=, !=, <, >, <=, >=), logical operators (&&, ||, !), pattern matching (~), and field existence checks. Expressions operate on field values within each record.
Output can be restricted to specific fields with -p, sorted with -S, and limited with -n. The tool handles multiple record types within a single file using -t.

PARAMETERS

-e expression

Select records matching the expression

-t type

Select records of specified type only

-p fields

Print only specified fields (comma-separated)

-P fields

Print specified fields without record separators

-c, --count

Print count of matching records

-C, --collapse

Collapse multiple field values into one

-S field

Sort records by specified field

-G field

Group records by specified field

-n max

Return at most max records

-R random

Return random records

-d, --print-descriptors

Print record descriptors

EXPRESSION SYNTAX

field = "value": Exact string match
field ~ "pattern": Regular expression match
field < number: Numeric comparison
field && field2: Logical AND
#field: Field exists
!#field: Field does not exist

CAVEATS

Field names and string values in expressions must follow recutils quoting rules. Strings with spaces require quotes.
Numeric comparisons require fields to contain valid numbers. Non-numeric values cause comparison failures.
Large recfiles may be slow to process compared to proper databases, but the simplicity and portability often outweigh performance concerns for moderate data sizes.