mbox_tools
Simple tools for summarizing .mbox email archives.
mbox-tools is a small collection of Python scripts to see .mbox files (which contain collections of email messages).
For now, it contains a single command, which transforms .mbox files into .csv digests (where each row represents a single message, and rows include basic information about an email — including its body content and a list of attachment filenames).
Installation
To install, run pip install mbox-tools (preferably in a virtual machine).
Usage
Run the following lines, where input_file_name is the (full or relative) path to your .mbox file and output_file_name is the path (including filename) of the .csv file you wish to generate:
from mbox_tools import generate_csv_digest
generate_csv_digest(input_file_name, output_file_name)If all goes as expected, your CSV should now be populated with as many rows as there were emails in the .mbox file.
The rows will have the following columns, some of which may be empty:
| Column name | May be empty? |
|---|---|
| from | No |
| to | No |
| cc | Yes |
| date | No |
| subject | No |
| body | Yes |
| attachments | Yes |
| priority | Yes |
| importance | Yes |
| sensitivity | Yes |
Prior art / Caveats
mbox-tools would not be possible without Martin Rusev's mailbox/imbox package. Though mailbox appears to have been succeeded by the refactored imbox (which appears to focus more on IMAP/SMTP support, and less on parsing .mbox files), the earlier library does much of the heavy lifting to turn .mbox files into cleanly-formatted Python objects.
Because of this, it may be useful to replicate this parsing within this library in the future — whether through parsing Rusev's existing logic (should we get permission, of course) or creating our own.
Roadmap
In the process of beginning this library, we've already identified several related utilities that would be nice to have. These will be discussed in this repo's issues. Feel free to submit additional suggestions for features in the same way.
Contributing
If you want to add to this effort, please send a pull request or open an issue! We welcome contributions of any type or size.
