Update spec for delta length byte array and delta byte array encodings.
Storing the encoded byte size of the first half of the encoding makes it
possible to stream decode the values. Otherwise, the only way to figure out
where the second part starts is by completely decoding the first part.
diff --git a/Encodings.md b/Encodings.md
index c3322c2..03b50f5 100644
--- a/Encodings.md
+++ b/Encodings.md
@@ -198,15 +198,15 @@
The data stream looks like:
-<Delta Encoded Lengths> <Byte Array Data>
+<32-bit little endian int for byte size of Delta Encoded Lengths> <Delta Encoded Lengths> <Byte Array Data>
For example, if the data was "Hello", "World", "Foobar", "ABCDEF":
-The encoded data would be DeltaEncoding(5, 5, 6, 6) "HelloWorldFoobarABCDEF"
+The encoded data would be <size of DeltaEncoding(5, 5, 6, 6)> <DeltaEncoding(5, 5, 6, 6)><"HelloWorldFoobarABCDEF">
### Delta Strings: (DELTA_BYTE_ARRAY = 7)
-Supported Types: BYTE_ARRAY
+Supported Types: BYTE_ARRAY, FIXED_LEN_BYTE_ARRAY
This is also known as incremental encoding or front compression: for each element in a
sequence of strings, store the prefix length of the previous entry plus the suffix.
@@ -215,3 +215,8 @@
This is stored as a sequence of delta-encoded prefix lengths (DELTA_BINARY_PACKED), followed by
the suffixes encoded as delta length byte arrays (DELTA_LENGTH_BYTE_ARRAY).
+
+The data stream looks like:
+
+<32-bit little endian int for byte size of Delta Encoded Lengths> <Delta Encoded Lengths> <Delta Length Byte Array Data>
+
