Skip to content

natefinch/gorram

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

52 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

gorram Codeship Go Report Card

river

image: © SubSuid

It's like go run for any go function.

Automagically understands how to produce an interface from the command line into a Go function.

Sometimes, magic is just someone spending more time on something than anyone else might reasonably expect. -Teller

Demo

Gorram Demo

Installation

go get -u npf.io/gorram

Note: gorram depends on having a working go environment to function, since it dynamically analyzes go code in the stdlib and in your GOPATH.

Usage

Usage: gorram [OPTION] <pkg> <func | var.method> [args...]

Options:
  -t <string>  format output with a go template
  -h, --help   display this help

Executes a go function or an method on a global variable defined in a package in
the stdlib or a package in your GOPATH.  Package must be the full package import
path, e.g. encoding/json.  Only exported functions, methods, and variables may
be called.

Most builtin types are supported, and streams of input (via io.Reader or []byte
for example) may be read from stdin.  If specified as an argument, the argument
to a stream input is expected to be a filename.

Return values are printed to stdout.  If the function has an output argument,
like io.Reader or *bytes.Buffer, it is automatically passed in and then written
to stdout.

If there's no output stream, the return value is simply written to stdout via
fmt.Println.  If the return value is a struct that has an exported field that is
an io.Reader (such as net/http.Request), then that will be treated as the output
value, unless it's empty, in which case we fall back to printing the output
value.

A template specified with -t may either be a template definition (e.g.
{{.Status}}) or a filename, in which case the contents of the file will be used
as the template.

Gorram creates a script file in $GORRAM_CACHE, or, if not set, in
$HOME/.gorram/importpath/Name.go.  Running with -r will re-generate that script
file, otherwise it is reused.

Examples

Pretty print JSON:

$ echo '{ "foo" : "bar" }' | gorram encoding/json Indent "" $'\t'
{
    "foo" : "bar"
}

Calculate a sha256 sum:

$ gorram crypto/sha256 Sum256 foo.gz
abcdef012345678

How it works

The first time you run Gorram with a specific function name, Gorram analyzes the package function and generates a file for use with go run. Gorram intelligently converts stdin and/or cli arguments into arguments for the function. Output is converted similarly to stdout. The code is cached in a local directory so that later runs don't incur the generation overhead.

Heuristics

By default, Gorram just turns CLI args into function args and prints out the return value of a function using fmt's %v. However, there are some special heuristics that it uses to be smarter about inputs and outputs, based on common go idioms.

For example:

usage:
$ cat foo.zip | gorram crypto/sha1 Sum
or
$ gorram crypto/sha1 Sum foo.zip

function:
// crypto/sha1
func Sum(data []byte) [Size]byte

Gorram understands functions that take a single slice of bytes (or an io.Reader) should read from stdin, or if an argument is specified, the argument is treated as a filename to be read.

Return values that are an array of bytes are understood to be intended to be printed with fmt's %x, so that you get 2c37424d58 instead of [44 55 66 77 88].

usage:
$ gorram encoding/json Indent foo.json "" $'\t'
or
$ cat foo.json | gorram encoding/json Indent "" $'\t'

function:
// encoding/json
func Indent(dst *bytes.Buffer, src []byte, prefix, indent string) error

Gorram understands that functions with a src argument that is an io.Reader or []bytes and a dst argument that is a []byte, *bytes.Buffer, or io.Writer will read from stdin (or use an argument as a file to open), and write what is written to dst to stdout.

Gorram understands that if the function returns a non-nil error, the error should be written to stderr, the program exits with a non-zero exit status, and nothing is written to stdout.

Gorram understands that prefix and indent are arguments that need to be specified in the command line.

usage:
$ gorram math Cos 25

function:
// math
func Cos(x float64) float64

Gorram understands how to convert CLI arguments using the stringconv.Parse* functions, and will print outputs with fmt.Printf("%v\n", val).

usage:
$ echo 12345 | gorram encoding/base64 StdEncoding.EncodeToString
MTIzNDU2Cg==

function:
// base64
func (e *Encoding) EncodeToString(b []byte]) string

Gorram understands that packages have global variables that have methods you can call.

usage: 
$ gorram net/http Get https://google.com
<some long html output>

function:
// net/http
func Get(url string) (resp *Response, err error)

Gorram understands that if a function returns a struct, and one of the fields of the struct is an io.Reader, then it will output the contents of that reader.
(if there are no contents or it's nil, the result value will be printed with %v).

Development

See the project page for what's being worked on now.

Hacking

Gorram requires go 1.7 to run the tests (thank you subtests!) But only requires it builds with earlier versions (at least 1.6, I haven't tried earlier ones, but they should be fine, too).