url

The url package simply parses url.Values and returns it as structured and organized map.

Why?

Let's look at the payload below (added spaces to make it clearer to read)

value=a & value=b & value=c

Does not require any parsing, url.Values's Get() is more than sufficient to handle it. However when the payload becomes more convoluted like the one below.

slice2[1]=a & slice2[2]=b & map0[key]=c & map1[key][subKey]=d & map2[sub][1]=e & map2[sub][2]=f

Parsing it becomes complexer and most of the time it requires ad-hoc code to handle the specific structure of the data. Hence the package.

Install

go get github.com/thetechpanda/url

Usage

package webStuff

import (
    URL "github.com/thetechpanda/url"
)

func dummyHandler(w http.ResponseWriter, r *http.Request) {

    valueMap, err := URL.ParseValues(r.PostForm())
    if err != nil {
        panic(err)
    }

    // verbose
    v, err := valueMap.GetValue("key","sub")
    if err != nil {
        panic(err)
    }

    s, ok := v.String()
    if !ok {
        panic("not a string")
    }

    // concise
    s := valueMap.GetString("key","sub")
    if s == "" {
        panic("invalid key or empty value")
    }

    // do something
}

Documentation

Please refer to the comments in main.go or go to package documentation

ParseValues()

Parse processes url.Values and returns a Map interface

This function does its best to aggregate and organize the data parsing url.Values For example the input below:

    "input4[key0] = value4[key0]"
    "input4[key1][subKey1] = value4[key1][subKey1]"
    "input4[key1][subKey2] = value4[key1][subKey2]"
    "input4[key2][0] = value4[key2][0]"
    "input4[key2][1] = value4[key2][1]"

Will be interpreted as

    { "input4" :
      { "key0": "value4[key0]",
        "key1": { "subKey1": "value4[key1][subKey1]",
                  "subKey2": "value4[key1][subKey2]" }
        "key2": [ "value4[key2][0]", "value4[key2][1]"]
      }
    }

Testing and Benchmark

You can use make test or make bench to run the benchmarks. Please note that in order to stress the tool main_test.go generates random data each time it runs.

Benchmark results in my lab

make bench
mkdir -p bin/ prof/
go test -c -o bin/url.test
bin/url.test -test.cpu 1 -test.benchmem -test.run=./... -test.bench=./... -test.cpuprofile=prof/cpuprof -test.memprofile=prof/memprof
goos: linux
goarch: amd64
pkg: github.com/thetechpanda/url
cpu: Intel(R) Xeon(R) CPU E5-2678 v3 @ 2.50GHz
BenchmarkParse/url.Values_count/1               305806          7160 ns/op         2664 B/op          41 allocs/op
BenchmarkParse/url.Values_count/10               12826         94128 ns/op        27455 B/op         460 allocs/op
BenchmarkParse/url.Values_count/100               2023        987757 ns/op       281261 B/op        4449 allocs/op
BenchmarkParse/url.Values_count/1000               126      10016646 ns/op      2672150 B/op       40996 allocs/op
BenchmarkParse/url.Values_count/10000               16      97722209 ns/op     25745560 B/op      399359 allocs/op
BenchmarkParse/url.Values_count/100000               1    1228338786 ns/op    258457944 B/op     4036187 allocs/op
PASS

Examples

Please refer to main_test.go

Contributing

Fork, hack and submit the pull request :)

• Found a bug? Issue tracker is the way to go.
• Suggestion on how to optimize the code? Yes please!