MimeMagicSharp

Detects MIME type based on file content or file extension. Nuget package is available. Sample files are included.

Description

This project was originally based on parsing /usr/share/mime/magic (Kali Linux) file (mime database). Location may differ. More information about this file structure could be found here ("the magic files" section).

This file contains mime type definitions with several rule sets, which are applied iteratively to the given file's content to guess mime type(s).

Original file's format (old) is hard to modify according to your needs. I decided to move on to json file format (new) with similary capabilities and some new features. New format could be easily modified and updated with new rule sets and mime type definitions.

Both database formats are supported. Conversion from the old to the new database format is supported. Detection by extension can be used only with new file format. Template original and json files will be included soon.

Structure

Each MIME definition contains a list of rule sets. Each rule set contains an hierarchical list of rules. Rule set pseudocode:

//  Level 0 rules: Rule 0
//  Level 1 rules: Rule 1, Rule 2, Rule 3
//  Level 2 rules: Rule 4
//  Level 3 rules: Rule 5, Rule 6
//	...

When this ruleset is applied, result will be calculated in the following manner:

//  Result (boolean) = 0 && (1 || 2 || 3) && 4 && (5 || 6)...

Rule example:

"Name": "application/x-deb",
"Description": "Debian package",
"Extensions": ["deb"],
"RuleSet": [
    {
    "Rule": [
        {
        "Level": 0,
        "Offset": 0,
        "Range": 0,
        "Data": "213C617263683E",
        "DataUTF8": "!<arch>"
        },
        {
        "Level": 1,
        "Offset": 8,
        "Range": 0,
        "Data": "64656269616E",
        "DataUTF8": "debian"
        }
    ]
    }, "Rule":[...
    ]...

• Level. Explained earlier.
• Offset. Seek start position
• Range. Seek interval (0 means fixed position)
• Data. Data to seek in hex format
• Description and DataUTF8 are for visual purposes only.

Usage

//	Database File (new or old)
string MagicFile = Path.Combine(Environment.CurrentDirectory, "magic");

using (MimeMagicSharp.MimeMagicSharp ms = new MimeMagicSharp.MimeMagicSharp(MimeMagicSharp.EMagicFileType.Json, MagicFile))
{
   try
   {
	foreach (MimeTypeGuess mimeTypeGuess in ms.AssumeMimeType(EMimeTypeBy.Content,
                    	Path.Combine(Environment.CurrentDirectory, "Newtonsoft.Json.xml")))
	{
		//  Iterate over results
	}
    }
    catch (Exception Ex)
    { 
       //   Handle errors
    }
}

Convertion from original (old) format to json (new) is supported

MimeMagicSharp.MimeMagicSharp.ConvertFromOriginalToJson("magic_original", "magic_json_convert_test", out string ConvertError);

Limitations

Only first 4096 bytes of file are read during mime detection procedure. You can extend this limit in source code easily.

Other

Build in vs 2017

You can use or modify the sources however you want