Skip to content

CLI tool to filter JSON and JSON Lines data with Python syntax. (Similar to jq)

License

Notifications You must be signed in to change notification settings

kellyjonbrazil/jello

Repository files navigation

Tests Pypi

Built on jello:

jello

Filter JSON and JSON Lines data with Python syntax

jello is similar to jq in that it processes JSON and JSON Lines data except jello uses standard python dict and list syntax.

JSON or JSON Lines can be piped into jello via STDIN or can be loaded from a JSON file or JSON Lines files (JSON Lines are automatically slurped into a list of dictionaries). Once loaded, the data is available as a python list or dictionary object named '_'. Processed data can be output as JSON, JSON Lines, bash array lines, or a grep-able schema.

For more information on the motivations for this project, see my blog post.

Install

You can install jello via pip, via OS Package Repository, MSI installer for Windows, or by downloading the correct binary for your architecture and running it anywhere on your filesystem.

Pip (macOS, linux, unix, Windows)

For the most up-to-date version and the most cross-platform option, use pip or pip3 to download and install jello directly from PyPi:

Pypi

pip3 install jello

Packages and Binaries

OS Command
Debian/Ubuntu linux apt-get install jello
Fedora linux dnf install jello
Arch linux pacman -S jello
macOS brew install jello

For more OS packages, see https://repology.org/project/jello/versions.

See Releases on Github for MSI packages and binaries.

Usage

cat data.json | jello [OPTIONS] [QUERY | -q <query_file>]

jello [OPTIONS] [QUERY | -q <query_file>] [-f <input_files>]

QUERY is optional and can be most any valid python code. Alternatively, a query file can be specified with -q to load the query from a file. Within the query, _ is the sanitized JSON from STDIN presented as a python dict or list of dicts. If QUERY is omitted then the original JSON input will simply be pretty printed. You can use dot notation or traditional python bracket notation to access key names.

Note: Reserved key names that cannot be accessed using dot notation can be accessed via standard python dictionary notation. (e.g. _.foo["get"] instead of _.foo.get)

A simple query:

cat data.json | jello _.foo

or

jello _.foo -f data.json

or

jello '_["foo"]' -f data.json

Options

  • -c compact print JSON output instead of pretty printing
  • -C force color output even when using pipes (overrides -m and the NO_COLOR env variable)
  • -e empty data (don't process data from STDIN or file)
  • -f load input data from JSON file or JSON Lines files (must be the final option, if used)
  • -i initialize environment with a custom config file
  • -l lines output (suitable for bash array assignment)
  • -m monochrome output
  • -n print selected null values
  • -q load query from a file
  • -r raw output of selected strings (no quotes)
  • -s print the JSON schema in grep-able format
  • -t print type annotations in schema view
  • -h help
  • -v version info

Simple Examples

jello simply pretty prints the JSON if there are no options or query passed:

echo '{"foo":"bar","baz":[1,2,3]}' | jello

{
  "foo": "bar",
  "baz": [
    1,
    2,
    3
  ]
}

If you prefer compact output, use the -c option:

echo '{"foo":"bar","baz":[1,2,3]}' | jello -c

{"foo":"bar","baz":[1,2,3]}

Use the -l option to convert lists/arrays into lines:

echo '{"foo":"bar","baz":[1,2,3]}' | jello -l _.baz

1
2
3

The -l option also allows you to create JSON Lines:

echo '[{"foo":"bar","baz":[1,2,3]},{"fiz":"boo","buz":[4,5,6]}]' | jello -l

{"foo":"bar","baz":[1,2,3]}
{"fiz":"boo","buz":[4,5,6]}

You can print a grep-able schema by using the -s option:

echo '{"foo":"bar","baz":[1,2,3]}' | jello -s

_ = {};
_.foo = "bar";
_.baz = [];
_.baz[0] = 1;
_.baz[1] = 2;
_.baz[2] = 3;

Assigning Results to a Bash Array

Use the -l option to print JSON array output in a manner suitable to be assigned to a bash array. The -r option can be used to remove quotation marks around strings. If you want null values to be printed as null, use the -n option, otherwise they are printed as blank lines.

Bash variable:

variable=($(cat data.json | jello -rl _.foo))

Bash array variable (Bash 4+):

mapfile -t variable < <(cat data.json | jello -rl _.foo)

Bash array variable (older versions of Bash):

variable=()
while read -r value; do
    variable+=("$value")
done < <(cat data.json | jello -rl _.foo)

Setting Custom Colors via Environment Variable

Custom colors can be set via the JELLO_COLORS environment variable. Any colors set in the environment variable will take precedence over any colors set in the initialization file. (see Advanced Usage)

The JELLO_COLORS environment variable takes four comma separated string values in the following format:

JELLO_COLORS=<keyname_color>,<keyword_color>,<number_color>,<string_color>

Where colors are: black, red, green, yellow, blue, magenta, cyan, gray, brightblack, brightred, brightgreen, brightyellow, brightblue, brightmagenta, brightcyan, white, or default

For example, to set to the default colors:

JELLO_COLORS=blue,brightblack,magenta,green

or

JELLO_COLORS=default,default,default,default

Disable Colors via Environment Variable

You can set the NO_COLOR environment variable to any value to disable color output in jello. Note that using the -C option to force color output will override both the NO_COLOR environment variable and the -m option.

Advanced Usage

Here is more Advanced Usage information.

To accelerate filter development and testing, try jellex. jellex is an interactive front-end TUI built on jello that allows you to see your filter results in real-time along with any errors.

Examples:

Printing the Grep-able Schema

$ jc -a | jello -s

_ = {};
_.name = "jc";
_.version = "1.17.2";
_.description = "JSON CLI output utility";
_.author = "Kelly Brazil";
_.author_email = "kellyjonbrazil@gmail.com";
_.website = "https://github.com/kellyjonbrazil/jc";
_.copyright = "© 2019-2021 Kelly Brazil";
_.license = "MIT License";
_.parser_count = 80;
_.parsers = [];
_.parsers[0] = {};
_.parsers[0].name = "acpi";
_.parsers[0].argument = "--acpi";
_.parsers[0].version = "1.2";
_.parsers[0].description = "`acpi` command parser";
_.parsers[0].author = "Kelly Brazil";
_.parsers[0].author_email = "kellyjonbrazil@gmail.com";
_.parsers[0].compatible = [];
_.parsers[0].compatible[0] = "linux";
_.parsers[0].magic_commands = [];
_.parsers[0].magic_commands[0] = "acpi";
_.parsers[1] = {};
_.parsers[1].name = "airport";
_.parsers[1].argument = "--airport";
_.parsers[1].version = "1.3";
...

Printing the Grep-able Schema with type annotations (useful for grepping types)

jc dig example.com | jello -st

_ = [];                                                             //   (array)
_[0] = {};                                                          //  (object)
_[0].id = 23819;                                                    //  (number)
_[0].opcode = "QUERY";                                              //  (string)
_[0].status = "NOERROR";                                            //  (string)
_[0].flags = [];                                                    //   (array)
_[0].flags[0] = "qr";                                               //  (string)
_[0].flags[1] = "rd";                                               //  (string)
_[0].flags[2] = "ra";                                               //  (string)
_[0].query_num = 1;                                                 //  (number)
_[0].answer_num = 1;                                                //  (number)
_[0].authority_num = 0;                                             //  (number)
_[0].additional_num = 1;                                            //  (number)
_[0].opt_pseudosection = {};                                        //  (object)
_[0].opt_pseudosection.edns = {};                                   //  (object)
_[0].opt_pseudosection.edns.version = 0;                            //  (number)
_[0].opt_pseudosection.edns.flags = [];                             //   (array)
_[0].opt_pseudosection.edns.udp = 4096;                             //  (number)
_[0].question = {};                                                 //  (object)
_[0].question.name = "example.com.";                                //  (string)
_[0].question.class = "IN";                                         //  (string)
_[0].question.type = "A";                                           //  (string)
_[0].answer = [];                                                   //   (array)
_[0].answer[0] = {};                                                //  (object)
_[0].answer[0].name = "example.com.";                               //  (string)
_[0].answer[0].class = "IN";                                        //  (string)
_[0].answer[0].type = "A";                                          //  (string)
_[0].answer[0].ttl = 48358;                                         //  (number)
_[0].answer[0].data = "93.184.216.34";                              //  (string)
_[0].query_time = 46;                                               //  (number)
_[0].server = "2600:1700:bab0:d40::1#53(2600:1700:bab0:d40::1)";    //  (string)
_[0].when = "Mon Nov 29 09:41:11 PST 2021";                         //  (string)
_[0].rcvd = 56;                                                     //  (number)
_[0].when_epoch = 1638207671;                                       //  (number)
_[0].when_epoch_utc = null;                                         //    (null)

Printing the Structure of the JSON

jc dig example.com | jello -st | grep '(object)\|(array)'

_ = [];                                                             //   (array)
_[0] = {};                                                          //  (object)
_[0].flags = [];                                                    //   (array)
_[0].opt_pseudosection = {};                                        //  (object)
_[0].opt_pseudosection.edns = {};                                   //  (object)
_[0].opt_pseudosection.edns.flags = [];                             //   (array)
_[0].question = {};                                                 //  (object)
_[0].answer = [];                                                   //   (array)
_[0].answer[0] = {};                                                //  (object)

Lambda Functions and Math

echo '{"t1":-30, "t2":-20, "t3":-10, "t4":0}' | jello '\
keys = _.keys()
vals = _.values()
cel = list(map(lambda x: (float(5)/9)*(x-32), vals))
dict(zip(keys, cel))'

{
  "t1": -34.44444444444444,
  "t2": -28.88888888888889,
  "t3": -23.333333333333336,
  "t4": -17.77777777777778
}
jc -a | jello 'len([entry for entry in _.parsers if "darwin" in entry.compatible])'

45

For Loops

Output as JSON array

jc -a | jello '\
result = []
for entry in _.parsers:
  if "darwin" in entry.compatible:
    result.append(entry.name)
result'

[
  "airport",
  "airport_s",
  "arp",
  "crontab",
  "crontab_u",
  ...
]

Output as bash array

jc -a | jello -rl '\
result = []
for entry in _.parsers:
  if "darwin" in entry.compatible:
    result.append(entry.name)
result'

airport
airport_s
arp
crontab
crontab_u
...

List and Dictionary Comprehension

Output as JSON array

jc -a | jello '[entry.name for entry in _.parsers if "darwin" in entry.compatible]'

[
  "airport",
  "airport_s",
  "arp",
  "crontab",
  "crontab_u",
  ...
]

Output as bash array

jc -a | jello -rl '[entry.name for entry in _.parsers if "darwin" in entry.compatible]'

airport
airport_s
arp
crontab
crontab_u
...

Expressions and Environment Variables

echo '{"login_name": "joeuser"}' | jello 'os.getenv("LOGNAME") == _.login_name'

true

Using 3rd Party Modules

You can import and use your favorite modules to manipulate the data. For example, using glom:

jc -a | jello '\
from glom import *
glom(_, ("parsers", ["name"]))'

[
  "airport",
  "airport_s",
  "arp",
  "blkid",
  "crontab",
  "crontab_u",
  "csv",
  ...
]

Advanced JSON Manipulation

The data from this example comes from https://programminghistorian.org/assets/jq_twitter.json

Under Grouping and Counting, Matthew describes an advanced jq filter against a sample Twitter dataset that includes JSON Lines data. There he describes the following query:

"We can now create a table of users. Let’s create a table with columns for the user id, user name, followers count, and a column of their tweet ids separated by a semicolon."

https://programminghistorian.org/en/lessons/json-and-jq

Here is a simple solution using jello:

cat jq_twitter.json | jello -l '\
user_ids = set()
for tweet in _:
    user_ids.add(tweet.user.id)
result = []
for user in user_ids:
    user_profile = {}
    tweet_ids = []
    for tweet in _:
        if tweet.user.id == user:
            user_profile.update({
                "user_id": user,
                "user_name": tweet.user.screen_name,
                "user_followers": tweet.user.followers_count})
            tweet_ids.append(str(tweet.id))
    user_profile["tweet_ids"] = ";".join(tweet_ids)
    result.append(user_profile)
result'

...
{"user_id": 2696111005, "user_name": "EGEVER142", "user_followers": 1433, "tweet_ids": "619172303654518784"}
{"user_id": 42226593, "user_name": "shirleycolleen", "user_followers": 2114, "tweet_ids": "619172281294655488;619172179960328192"}
{"user_id": 106948003, "user_name": "MrKneeGrow", "user_followers": 172, "tweet_ids": "501064228627705857"}
{"user_id": 18270633, "user_name": "ahhthatswhy", "user_followers": 559, "tweet_ids": "501064204661850113"}
{"user_id": 14331818, "user_name": "edsu", "user_followers": 4220, "tweet_ids": "615973042443956225;618602288781860864"}
{"user_id": 2569107372, "user_name": "SlavinOleg", "user_followers": 35, "tweet_ids": "501064198973960192;501064202794971136;501064214467731457;501064215759568897;501064220121632768"}
{"user_id": 22668719, "user_name": "nodehyena", "user_followers": 294, "tweet_ids": "501064222772445187"}
...