Add README file to the mime package

pkg/mime/README.md

Index: pkg/mime/README.md
diff --git a/pkg/mime/README.md b/pkg/mime/README.md
new file mode 100644
index 0000000000000000000000000000000000000000..d27f3ce00480fac5ce6004825a23b6751b1c47c6
--- /dev/null
+++ b/pkg/mime/README.md
@@ -0,0 +1,49 @@
+#MIME type package
+
+Package for working with MIME type definitions and for processing
+streams of MIME multipart media types.
+
+##Determining the MIME type for a file
+
+The instances of the `MimeTypeResolver` class is used to determine the

[wibling, 2015/02/18 13:44:45]

NIT: I would not stress that it is the instances of, but just say something
like:

The `MimeTypeResolver` class can be used to determine the MIME type of a file.

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+MIME type for file. It supports both using the extention of the file

[wibling, 2015/02/18 13:44:46]

TYPO: extention -> extension.

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+name and possibly looking at magic bytes from the begining of the

[wibling, 2015/02/18 13:44:46]

NIT: I would remove possibly. It is implicit in 'It supports both ...'

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+file.
+
+There is a builtin instance of `MimeTypeResolver` accessible through
+the top level function `lookupMimeType`. This builtin instance have

[wibling, 2015/02/18 13:44:45]

NIT: have -> has

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+the most common file name extensions and magic bytes registered.
+
+    print(lookupMimeType('test.html'));  // Will print text/html
+    print(lookupMimeType('test', [0xFF, 0xD8]));  // Will print image/jpeg
+    print(lookupMimeType('test.html', [0xFF, 0xD8]));  // Will print image/jpeg
+
+You can build you own resolver by instantiating `MimeTypeResolver`

[wibling, 2015/02/18 13:44:46]

NIT: you -> your

NIT: by instantiating ... -> by instantiating the `MimeTypeResolver` class.

[Søren Gjesse, 2015/02/18 14:44:18]

Changed to 'by creating an instance of'

+and adding file name extensions and magic bytes using `addExtension`
+and `addMagicNumber`.
+
+##Processing MIME multipart media types
+
+The class `MimeMultipartTransformer` is used to process a `Stream` of
+bytes encoded using a MIME multipart media types encoding. The
+transformer provides a new `Stream` of `MimeMultipart` objects each of
+which have the headers and the content of each part. The content of a
+part is provided as a strream of bytes.

[wibling, 2015/02/18 13:44:45]

NIT: strream -> stream

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+
+An example of processing an HTTP request just printing the length of
+the content of each part.

[wibling, 2015/02/18 13:44:46]

I would change the above to:

Below is an example showing how to process an HTTP request and print the length
of the content of each part.

[Søren Gjesse, 2015/02/18 14:44:18]

Done.

+
+    // HTTP request with content type multipart/form-data.
+    HttpRequest request = ...;
+    // Determine the boundary form the content type header
+    String boundary = request.headers.contentType.parameters['boundary'];
+
+    // Process the body just calculating the length of each part.
+    request.transform(new MimeMultipartTransformer(boundary))

[wibling, 2015/02/18 13:44:46]

boundary -> _?

[Søren Gjesse, 2015/02/18 14:44:18]

The boundary value is from the variable above.

+        .map((part) => part.fold(0, (p, d) => p + d))
+        .listen((length) => print('Part with length $length'));
+
+Take a look at the `HttpBodyHandler` in the [http_server][1] package for
+handling different content types in a HTTP request.
+
+[1]: https://pub.dartlang.org/packages/http_server