CSharp MARC

There has been some confusion about the process of editing and saving files, which hopefully this tutorial will cover. You can edit and save records with C# MARC just fine, but it's a little weird because of how the MARC format is designed. There isn't a really good way to edit records "in place" so it's not as simple as just changing the data. It is also complicated, as I designed the class structure to only decode a single MARC record when it needs to be, rather than attempting to decode an entire file all at once. This makes reading files quicker, but editing them slightly harder. If you're only using the File_MARC object, no changes are saved in the record, as each time you get an individual record it decodes it from the raw record. The main goal of the File_MARC class is to go from a raw MARC file to an actual Record object easily, but that is all it is good for.

//Open the MARC file
FileMARC marcFiles = new FileMARC();
marcFiles.ImportMARC(@"C:\some_order.mrc");

//Open a writer to save the new records.
StreamWriter writer = new StreamWriter(@"C:\edited_order.mrc", false, ASCIIEncoding.Default);

foreach (Record marc in marcFiles)
{
	//Create a new field to insert
	DataField newField = new DataField("852");
	
	//Create some subfields
	Subfield newSubfield = new Subfield('h', "123.45 ABC");
	newField.Subfields.Add(newSubfield);
	newSubfield = new Subfield('p', "123456789");
	newField.Subfields.Add(newSubfield);
	marc.Add(newField);
	
	//Write the marc record out to the file
	writer.Write(marc.ToRaw());
}

//Write the end of file character. This isn't necessary, but some programs will throw an error if there isn't a clear end of file marker after all the records.
writer.Write('\x1A');

//Always clean up your disposables
writer.Close();
writer.Dispose();

List<Record> records = new List<Record>();
foreach (Record marc in marcFiles)
{
	records.Add(marc);
}

using (StreamWriter writer = new StreamWriter(filename, false, ASCIIEncoding.Default))
{
	foreach (Record record in records)
	{
		writer.Write(record.ToRaw());
	}
	//Write end of file character.
	writer.Write('\x1A');
}

Thanks to some feedback from users, I've implemented an update to CSharp MARC to support large files.  In order to support the current API and not break how the FileMARC class was intended to be used (which is loading an entire file into memory and then doing batch processing on it), I added a FileMARCReader class.

The FileMARCReader class simply takes a filename and allows it to be used as an IEnumerator.  It reads the file in small memory-efficient chunks and spits out records one at a time in order.  This will hopefully help those who need to work with large files.

Here's a quick example of its use:

string filename = "manyrecords.mrc";
FileMARCReader reader = new FileMARCReader(filename);
foreach (Record marc in reader)
{
	//Do whatever you'd normally do on a Record object
}
reader.Dispose();

This can also be found in the source code download from SourceForge as "CSharp_MARC Demo"

//Read raw MARC record from a file. 
//The example .mrc files are in the project's root.
string rawMarc = File.ReadAllText("..\\..\\..\\record.mrc");

//The FileMARC class does the actual decoding of a record and splits a string of multiple records into a list object. 
//Decoding is not done until you actually access a single record from the FileMARC object.
//You can import records straight from a string in memory. The string can have one or many MARC records.
FileMARC marcRecords = new FileMARC(rawMarc);

//Or you can import it straight from a file
marcRecords.ImportMARC("..\\..\\..\\record2.mrc");

//You can get how many records were found by using the Count property
Console.WriteLine("Found " + marcRecords.Count + " records.");

//You can access each individual record in it's native MARC format using the RawSource object
Console.WriteLine("Here is the first record:");
Console.WriteLine(marcRecords.RawSource[0]);

//You can access each record manually using array notation.
Record firstRecord = marcRecords[0];

//Or you can loop through them as an Enumerable object
//Note: I recommend only retrieving each record from the FileMARC object once as each time you do it will be decoded. 
int i = 0;
foreach (Record record in marcRecords)
{
	//The Warnings property contains a list of issues that the decoder found with the record. 
	//The decoder attempts to return a valid MARC record to the best of it's ability.
	Console.WriteLine("Book #" + ++i + " has been decoded with " + record.Warnings.Count +" errors!");

	//Once decoded you can easily access specific data within the record, as well as make changes.
	//Array notation we will get the first requested tag in the record, or null if one does not exist.
	//First we'll get the Author.  Since there should only be one author tag array notation is the easiest to use.
	Field authorField = record["100"];

	//Each tag in the record is a field object. To get the data we have to know if it is a DataField or a ControlField and act accordingly.
	if (authorField.IsDataField())
	{
		DataField authorDataField = (DataField)authorField;

		//The author's name is in subfield a.  Once again since there should only be one we can use array notation.
		Subfield authorName = authorDataField['a'];
		Console.WriteLine("The author of this book is " + authorName.Data);
	}
	else if (authorField.IsControlField())
	{
		//Unreachable code!
		Console.WriteLine("Something went horribly wrong. The author field should never be a Control Field.");
	}

	//Now we will get the subjects for this record. Since a book can have multiple subjects we will use GetFields() which returns a List<Field> object.
	//Note: Not passing in a tag number to GetFields will return all the tags in the record.
	List<Field> subjects = record.GetFields("650");
	Console.WriteLine("Here are the subjects for Book #" + i);

	//Here we will assume each Field is actually a DataField since ISBNs should always be a DataField.
	foreach (DataField subject in subjects)
	{
		string subjectText = string.Empty;

		//We also want to loop through each subfield.
		//Just like with GetFields() you can either pass in a subfield value, or nothing to get all the subfields
		foreach (Subfield subfield in subject.GetSubfields())
			subjectText += subfield.Data + " ";
		Console.WriteLine(subjectText);
	}
}