What is this?
Miller is a command-line toolkit like awk, sed, cut, join, and sort, but built for name-indexed data such as CSV, TSV, JSON, and JSON Lines: you address fields by name instead of counting column positions.
Miller is a command-line toolkit like awk, sed, cut, join, and sort, but built for name-indexed data such as CSV, TSV, JSON, and JSON Lines: you address fields by name instead of counting column positions.So is it a library I import, or a command I run?A standalone command-line executable, not a library. Where classic Unix tools see numbered array slots, Miller sees each record as an insertion-ordered key-value map, and its processing is format-aware.
A standalone command-line executable, not a library. Where classic Unix tools see numbered array slots, Miller sees each record as an insertion-ordered key-value map, and its processing is format-aware.What's the install command on my machine?Prebuilt packages: brew or port on macOS; yum, apt, or snap on Linux; choco, winget, or scoop on Windows (e.g. winget install Miller.Miller). Building from source with make or go install also works.
Prebuilt packages: brew or port on macOS; yum, apt, or snap on Linux; choco, winget, or scoop on Windows (e.g. winget install Miller.Miller). Building from source with make or go install also works.Do I need any runtimes or libraries installed first?None. Miller is written in portable, modern Go with zero runtime dependencies — a single binary you can scp to a faraway machine and expect to work.
None. Miller is written in portable, modern Go with zero runtime dependencies — a single binary you can scp to a faraway machine and expect to work.Will the copy in my package manager be the latest version?Not necessarily: the README notes long-term-support (LtS) releases are likely on older versions, so check README-versions.md or build from source if you need the newest one.
Not necessarily: the README notes long-term-support (LtS) releases are likely on older versions, so check README-versions.md or build from source if you need the newest one.Why pick Miller over pandas or R for data work?They're complements, not rivals: you can clean and prepare data with Miller on the command line, then load the result into R or pandas for deeper statistics and graphics. Its streaming single-pass algorithms also let you shrink very large files before those tools ever see them.
They're complements, not rivals: you can clean and prepare data with Miller on the command line, then load the result into R or pandas for deeper statistics and graphics. Its streaming single-pass algorithms also let you shrink very large files before those tools ever see them.Can it convert files between formats?Yes — its I/O formats include CSV, TSV, JSON, JSON Lines, positionally-indexed, and tabular pretty-printing, and Miller converts between them. Conversion is format-aware: for example, sorting a CSV keeps the header line first.
Yes — its I/O formats include CSV, TSV, JSON, JSON Lines, positionally-indexed, and tabular pretty-printing, and Miller converts between them. Conversion is format-aware: for example, sorting a CSV keeps the header line first.What is the command actually called?mlr — that's the executable you run (mlr.exe on Windows). Building from source produces ./mlr, and go install puts it in GOPATH/bin.
mlr — that's the executable you run (mlr.exe on Windows). Building from source produces ./mlr, and go install puts it in GOPATH/bin.Can I use it like a quick database over my files?For one-off tasks, yes: the README touts setup-free querying of data sitting in disk files when you're in a hurry, and Miller can also slice, dice, and reformat data on its way into or out of a SQL database.
For one-off tasks, yes: the README touts setup-free querying of data sitting in disk files when you're in a hurry, and Miller can also slice, dice, and reformat data on its way into or out of a SQL database.Which topics is it in?data-cleaning, tabular-data, devops-tools, data-processing, csv, statistics, unix, analytics, data-science, cli-tool, json, bsd
data-cleaning, tabular-data, devops-tools, data-processing, csv, statistics, unix, analytics, data-science, cli-tool, json, bsd