From b5788134b6bdbeb845906a88bc856ea088e742fa Mon Sep 17 00:00:00 2001
From: Kelly Brazil <kellyjonbrazil@gmail.com>
Date: Tue, 19 Jul 2022 15:23:42 -0700
Subject: [PATCH] add jwt string parser

---
 CHANGELOG         |   2 +
 jc/lib.py         |   1 +
 jc/parsers/jwt.py | 132 ++++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 135 insertions(+)
 create mode 100644 jc/parsers/jwt.py

diff --git a/CHANGELOG b/CHANGELOG
index a0369407..52cd0bee 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -2,6 +2,8 @@ jc changelog
 
 20220705 v1.20.3
 - Add URL string parser
+- Add Email Address string parser
+- Add JWT string parser
 - Add m3u/m3u8 file parser
 - Add pager functionality to help (parser documentation only)
 - Minor parser performance optimizations
diff --git a/jc/lib.py b/jc/lib.py
index e0461dc3..43e98f2a 100644
--- a/jc/lib.py
+++ b/jc/lib.py
@@ -54,6 +54,7 @@ parsers = [
     'iw-scan',
     'jar-manifest',
     'jobs',
+    'jwt',
     'kv',
     'last',
     'ls',
diff --git a/jc/parsers/jwt.py b/jc/parsers/jwt.py
new file mode 100644
index 00000000..27724ff7
--- /dev/null
+++ b/jc/parsers/jwt.py
@@ -0,0 +1,132 @@
+"""jc - JSON Convert JWT string parser
+
+> Note: `jc` will not check the integrity of the JWT payload.
+
+Usage (cli):
+
+    $ echo "eyJhbGciOiJIUzI1N..." | jc --jwt
+
+Usage (module):
+
+    import jc
+    result = jc.parse('jwt', jwt_string)
+
+Schema:
+
+    {
+      "header": {
+        "alg":                    string,
+        "typ":                    string
+      },
+      "payload": {
+        <key name>:               string/integer/float/boolean/null
+      },
+      "signature":                string  # [0]
+    }
+
+    [0] in colon-delimited hex notation
+
+Examples:
+
+    % echo 'eyJhbGciOiJIUzI1N...' | jc --jwt -p
+    {
+      "header": {
+        "alg": "HS256",
+        "typ": "JWT"
+      },
+      "payload": {
+        "sub": "1234567890",
+        "name": "John Doe",
+        "iat": 1516239022
+      },
+      "signature": "49:f9:4a:c7:04:49:48:c7:8a:28:5d:90:4f:87:f0:a4:c7..."
+    }
+"""
+from base64 import urlsafe_b64decode
+import binascii
+import json
+from typing import Dict
+import jc.utils
+
+
+class info():
+    """Provides parser metadata (version, author, etc.)"""
+    version = '1.0'
+    description = 'JWT string parser'
+    author = 'Kelly Brazil'
+    author_email = 'kellyjonbrazil@gmail.com'
+    compatible = ['linux', 'darwin', 'cygwin', 'win32', 'aix', 'freebsd']
+
+
+__version__ = info.version
+
+
+def _process(proc_data: Dict) -> Dict:
+    """
+    Final processing to conform to the schema.
+
+    Parameters:
+
+        proc_data:   (Dictionary) raw structured data to process
+
+    Returns:
+
+        Dictionary. Structured to conform to the schema.
+    """
+    return proc_data
+
+
+def _b2a(byte_string: bytes) -> str:
+    """Convert a byte string to a colon-delimited hex ascii string"""
+    # need try/except since seperator was only introduced in python 3.8.
+    # provides compatibility for python 3.6 and 3.7.
+    try:
+      return binascii.hexlify(byte_string, ':').decode('utf-8')
+    except TypeError:
+      hex_string = binascii.hexlify(byte_string).decode('utf-8')
+      colon_seperated = ':'.join(hex_string[i:i+2] for i in range(0, len(hex_string), 2))
+      return colon_seperated
+
+
+def parse(
+    data: str,
+    raw: bool = False,
+    quiet: bool = False
+) -> Dict:
+    """
+    Main text parsing function
+
+    Parameters:
+
+        data:        (string)  text data to parse
+        raw:         (boolean) unprocessed output if True
+        quiet:       (boolean) suppress warning messages if True
+
+    Returns:
+
+        Dictionary. Raw or processed structured data.
+    """
+    jc.utils.compatibility(__name__, info.compatible, quiet)
+    jc.utils.input_type_check(data)
+
+    raw_output: Dict = {}
+
+    if jc.utils.has_data(data):
+        data = data.strip()
+        header, payload, signature = data.split('.')
+
+        header = urlsafe_b64decode(header + '==').decode('utf-8')
+        payload = urlsafe_b64decode(payload + '==').decode('utf-8')
+        signature_bytes = urlsafe_b64decode(signature + '==')
+
+        header = json.loads(header)
+        payload = json.loads(payload)
+        signature = _b2a(signature_bytes)
+
+        raw_output = {
+            'header': header,
+            'payload': payload,
+            'signature': signature,
+        }
+
+    return raw_output if raw else _process(raw_output)