(20241210) listing bug fix.

2024-12-10 10:08:53 +05:30
parent 845827a6bc
commit a3589c4eb7
2 changed files with 143 additions and 4 deletions
@@ -0,0 +1,133 @@
+"""
+
+    AUTHOR:
+
+        Khushal P Soonderji
+
+    DATE:
+
+        Tuesday, 10th Dec., 2024
+
+    OBJECTIVE:
+
+        To parse raw mail bodies and give a structure that is suitable for storing in No-SQL databases like MongoDB. The
+        raw mail's text is expected to be compliant with standard defined in RFC 5322, RFC 2045, and maybe a few more.
+
+    REFERENCES:
+
+        1. GitHub: https://github.com/SpamScope/mail-parser
+        2. RFC 5322: https://datatracker.ietf.org/doc/html/rfc5322
+        3. RFC 2045: https://datatracker.ietf.org/doc/html/rfc2045
+        4. StackOverflow: https://stackoverflow.com/questions/17874360/python-how-to-parse-the-body-from-a-raw-email-given-that-raw-email-does-not
+
+    DOWNLOADS:
+
+        N/A
+
+"""
+
+
+#   *****************************************************************************************************************
+#  *****                                                                                                          ****
+# ***                                                   IMPORT                                                      ***
+#  *****                                                                                                          ****
+#   *****************************************************************************************************************
+
+
+# To make sibling directories accessible for imports:
+import sys
+sys.path.append(".")
+sys.path.append("..")
+
+# System-level activities:
+import io
+
+# My utils:
+from utils_v2.string import json
+from utils_v2.string import regex
+from utils_v2.date_time import date_time
+
+# To work with mails:
+import email
+
+# To parse the HTML content in the mail:
+from bs4 import BeautifulSoup
+
+# To work with datatypes:
+from typing import Any, Dict
+
+# To work with various encodings:
+import base64
+import quopri
+
+
+#   *****************************************************************************************************************
+#  *****                                                                                                          ****
+# ***                                           MACROS / ONE-TIME INIT                                              ***
+#  *****                                                                                                          ****
+#   *****************************************************************************************************************
+
+
+# --- Nothing Yet
+
+
+#   *****************************************************************************************************************
+#  *****                                                                                                          ****
+# ***                                                VARIABLES                                                      ***
+#  *****                                                                                                          ****
+#   *****************************************************************************************************************
+
+
+# --- Nothing Yet
+
+
+#   *****************************************************************************************************************
+#  *****                                                                                                          ****
+# ***                                                  FUNCTIONS                                                    ***
+#  *****                                                                                                          ****
+#   *****************************************************************************************************************
+
+
+def parse(raw_mail: str | bytes) -> Dict[str, Any]:
+
+    """
+    To parse the raw mail text to a usable JSON that can even be stored on a No-SQL database like MongoDB.
+    DOCUMENTATION:
+        1. GitHub: https://github.com/SpamScope/mail-parser
+        2. RFC 5322: https://datatracker.ietf.org/doc/html/rfc5322
+        3. RFC 2045: https://datatracker.ietf.org/doc/html/rfc2045
+    :param raw_mail: The raw mail body that adheres to RFC 5322 and RFC 2045 (among others).
+    :return: The parsed JSON format (dict) of the mail.
+    """
+
+    # Parse the raw format:
+    if isinstance(raw_mail, str): parsed_mail = email.message_from_string(raw_mail)
+    else: parsed_mail = email.message_from_bytes(raw_mail)
+
+    # Make variables:
+    parts = []
+
+    # Iterate through each part of the mail for multipart mails:
+    if parsed_mail.is_multipart():
+        for part in parsed_mail.walk():
+            print("MULTIPART PART:", part)
+
+    # When the mails are not multipart, just plaintext:
+    else: print("PLAINTEXT PART:", parsed_mail.get_payload())
+
+
+#   *****************************************************************************************************************
+#  *****                                                                                                          ****
+# ***                                               MAIN PROGRAM                                                    ***
+#  *****                                                                                                          ****
+#   *****************************************************************************************************************
+
+
+if __name__ == "__main__":
+
+    from utils_v2.system import files
+
+    mail_string_raw = files.read_file(r"/home/developer/Downloads/raw_mail.txt")
+    parse_results = parse(mail_string_raw)
+
+    print(json.to_string(parse_results, default = str))