Purpose of this package is to provide a simple wrapper arround jq for different formats. I'm tired of searching a package doing yaml jq, toml jq, ini jq etc. mainly used for scripting.
This script uses:
- @mwilliamson Python bindings on top of @stedolan famous jq lib
- swiss knife for coloration pygments
- binary built with pyinstaller
- easy CLI with click
- for supported types sources, check table
Supported file types
You could install via pipx, pip, brew, deb, rpm or fallback to binary version.
pipx
install and run python applications in isolated environments
pipx install wildq
pip install wildq
brew install ahmet2mir/tap/wildq
brew install ahmet2mir/tap/wq
VERSION=$(curl -s "https://api.github.com/repos/ahmet2mir/wildq/releases/latest" | grep '"tag_name":' | sed -E 's/.*"v([^"]+)".*/\1/')
curl -sL https://github.com/ahmet2mir/wildq/releases/download/v${VERSION}/wildq_${VERSION}-1_amd64.deb -o wildq_${VERSION}-1_amd64.deb
sudo dpkg -i wildq_${VERSION}-1_amd64.deb
wq --help
wildq --help
VERSION=$(curl -s "https://api.github.com/repos/ahmet2mir/wildq/releases/latest" | grep '"tag_name":' | sed -E 's/.*"v([^"]+)".*/\1/')
curl -sL https://github.com/ahmet2mir/wildq/releases/download/v${VERSION}/wildq-${VERSION}-1.x86_64.rpm -o wildq-${VERSION}-1.x86_64.rpm
sudo yum install -y ./wildq-${VERSION}-1.x86_64.rpm
wq --help
wildq --help
@zmedico did a portage on app-misc
emerge -av app-misc/wildq
Compiled using glibc 2.17, it should work on lot of stable/LTS distros.
mkdir -p ~/bin/
VERSION=$(curl -s "https://api.github.com/repos/ahmet2mir/wildq/releases/latest" | grep '"tag_name":' | sed -E 's/.*"v([^"]+)".*/\1/')
curl -sL https://github.com/ahmet2mir/wildq/releases/download/v${VERSION}/wildq-${VERSION}-linux-x86_64.tar.gz -o wildq-${VERSION}-linux-x86_64.tar.gz
tar xvfz wildq-${VERSION}-linux-x86_64.tar.gz -C ~/bin
export PATH="~/bin:$PATH"
wq --help
wildq --help
Wildq use jq.py and it's not yet available on windows platforms. I tried to compile it without windows machine and I failed, and I don't had the time to try to understand how Windows / C binding / Python works. If anybody would contribute, there is an open issue (jq and onigurama are 'compilable' on Windows so I think that someone confortable with that OS could make it);
type | color | ordering | output | source |
---|---|---|---|---|
hcl | json | no | json | pyhcl by @virtuald |
ini | yes | no | yes | ConfigParser |
json | yes | yes | yes | json |
toml | yes | no | yes | toml by @uiri |
xml | yes | no | yes | xmldict by @martinblech |
yaml | yes | yes | yes | pyyaml |
$ wildq --help
Usage: wildq [OPTIONS] JQ_FILTER [FILE]
Options:
-c, --compact-output compact instead of pretty-printed output
-r, --raw output raw strings, not content texts
-C, --color-output colorize content (default), mutally
exclusive with --monochrome-output
-M, --monochrome-output monochrome (don't colorize content), mutally
exclusive with --color-output
--hcl Combine --input hcl --output json, mutally
exclusive with other Combined options
--ini Combine --input ini --output json, mutally
exclusive with other Combined options
--json Combine --input json --output json, mutally
exclusive with other Combined options
--toml Combine --input toml --output json, mutally
exclusive with other Combined options
--xml Combine --input xml --output json, mutally
exclusive with other Combined options
--yaml Combine --input yaml --output json, mutally
exclusive with other Combined options
-i, --input [hcl|ini|json|toml|xml|yaml]
Define the content type of file, mutally
exclusive with Combined option
-o, --output [hcl|ini|json|toml|xml|yaml]
Define the content type of printed output,
mutally exclusive with Combined option
(default input format)
--version Show the version and exit.
--help Show this message and exit.
For backward compatibility in previous version only
--[yaml|json|toml|ini|xml|hcl]
was possible with default to json output. We still keep Monochrome, raw and json output with thoses options. Output was similar tojq -MCr
(no color, no compact and no quote on single value)
But now, by default it's colorized, not raw and if you specify input using -i
or --input
output will be the same format.
There is also a shorter command wq
comming with the package.
Like jq cli
, wildq supports both of stdin and file to the function
See examples to get some example.
Content of examples/json.json
{
"general": {
"user": "admin"
},
"keys": [
{"key": "value1"},
{"key": "value2"},
"alone"
]
}
cat examples/json.json | wildq -i json ".keys[]"
{
"key": "value1"
}
{
"key": "value2"
}
alone
or
wildq -i json ".keys[]" examples/json.json
{
"key": "value1"
}
{
"key": "value2"
}
alone
or
wq -i json ".keys[]" examples/json.json
{
"key": "value1"
}
{
"key": "value2"
}
alone
For TOML
cat examples/toml.toml | wildq -i toml ".keys[]"
{
"key": "value1"
}
{
"key": "value2"
}
alone
For INI (no array)
cat examples/ini.ini | wildq -i ini ".keys"
{
"key1": "value1",
"key2": "value2"
}
For XML
cat examples/xml.xml | wildq -i xml "."
{
"root": {
"general": {
"user": "admin"
},
"keys": {
"element": [
{
"key": "value1"
},
{
"key": "value2"
},
"alone"
]
}
}
}
For YAML
cat examples/yaml.yaml | wildq -i yaml ".keys[]"
{
"key1": "value1"
}
{
"key2": "value2"
}
alone
For HCL
cat examples/hcl.hcl | wildq -i hcl ".keys[]"
{
"key": "value1"
}
{
"key": "value2"
}
Loop on keys in bash without creating a subshell
wildq -i toml "keys[]" examples/toml.toml | while read -r key
do
echo "Getting key ${key}"
done
You could also found some examples on some blogs/websites:
-
a complete guide with advanced examples on how to read/update files with differents processing tools (Original Japanese) (Translate EN)
-
an example of INI manipulation made by AWS ParallelCluster Workshop (workshop) (example 1) (example 2)
- add tests...
- add more control over filters and files
- use click for the CLI
- support different output
- detect automagically filetype
- support all jq types
- ordering
Merge requests are welcome :)
Licensed under the terms of the Apache License, Version 2.0.