jorgecarleitao commented on a change in pull request #170: URL: https://github.com/apache/parquet-format/pull/170#discussion_r606752379
########## File path: rle-bitpacked.md ########## @@ -0,0 +1,120 @@ +# RLE-Bitpacked hybrid encoder + +The RLE-Bitpacked hybrid encoder is a parquet-specific encoder that combines +two well known encoding strategies, [RLE](https://en.wikipedia.org/wiki/Run-length_encoding) +and bitpacking. Note that "combine" here means this encoder allows both encodings +within the same stream, and, during encoding, it can switch between them. + +This encoder is only used to encode integer values that may either represent definition levels, +representation levels or ids of dictionary-encoded pages. Note that this encoder +supports integers that can be represented in less than 8 bits. + +This document uses [LSB](https://en.wikipedia.org/wiki/Bit_numbering#Least_significant_bit) +to identify bits. In this representation, a byte is represented +by `[b7 b6 b5 b4 b3 b2 b1 b0]` where `b0` is the first bit. + +This document uses MUST, SHOULD, etc. according to [RFC-8174](https://tools.ietf.org/html/rfc8174). + +## Decoding + +Decoding a stream of bytes (denoted as `[a1, a2, a3, ...]`) assumes a specific `bit_width` +indicating the number of bits necessary to represent the largest encoded integer in the stream. + +The `bit_width` MUST be sufficient to represent the largest encoded integer on the +stream or the result is undefined. + +The first 4 bytes of the stream MUST represent a little-endian unsigned integer (`uint32`) +denoting the length of the rest of the stream. For example, `[4u8, 0, 0, 0]` announces +that the stream has a total of `4 + 4 = 8` bytes. The first 4 bytes are only used for this purpose. + +The remaining bytes are divided in "runs", which MUST be either RLE-encoded or bitpacked-encoded. +Each "run" MUST be composed by a header of a variable number of bytes and by a body in sequence. +I.e. `[h11, h12, h13, b11, b12, ...]` where `h11` is the first byte of the header +of run `1`, `h12` the second byte of the header, `b11` is the first byte of the body of the first run. + +The header MUST be a single [ULEB128](https://en.wikipedia.org/wiki/LEB128#Unsigned_LEB128)-encoded `i32`, Review comment: reviewing the spec, I am no longer sure of this. I think that we put no constraint into what type this should be, so I think we should just remove the type from here. There is a typo regardless, it should be unsigned (as we are using unsigned LEB). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
