Memory preserving parse tree based compression with entropy coding

Patent metadata

Field	Value
Publication number	US-10303759-B2
Application number	US-201514958493-A
Country	US
Kind code	B2
Filing date	Dec 3, 2015
Priority date	Dec 3, 2015
Publication date	May 28, 2019
Grant date	May 28, 2019

Abstract

A method, computer program product, and system includes a processor obtaining data including values and generating a value conversion dictionary by applying a parse tree based compression algorithm to the data, where the value conversion dictionary includes dictionary entries that represent the values. The processor obtains a distribution of the values and estimates a likelihood for each based on the distribution. The processor generates a code word to represent each value, a size of each code word is inversely proportional to the likelihood of the word. The processor assigns a rank to each code word, the rank for each represents the likelihood of the value represented by the code word; and based on the rank associated with each code word, the processor reorders each dictionary entry in the value conversion dictionary to associate each dictionary entry with an equivalent rank, the reordered value conversion dictionary comprises an architected dictionary.

First claim

What is claimed is:

1. A computer-implemented method, comprising:
obtaining, by a processor, data comprised of values, wherein the values are of variable size, and generating a value conversion dictionary by applying a parse tree based compression algorithm to the data, wherein the value conversion dictionary is comprised of dictionary entries that represent the values;
obtaining, by the processor, a distribution of the values and estimating a likelihood for each value based on the distribution;
generating, by the processor, a code word to represent each value, wherein a size of each code word is inversely proportional to the likelihood of the value represented by the code word;
assigning, by the processor, a rank to each code word, wherein the rank for each code word represents the likelihood of the value represented by the code word;
based on the rank associated with each code word, reordering, by the processor, each dictionary entry in the value conversion dictionary to associate each dictionary entry with an equivalent rank, wherein the reordered value conversion dictionary comprises an architected dictionary; and
storing, by the processor, the architected dictionary as at least one tree structure in a memory, wherein the processor utilizes the architected dictionary to compress data subsequently obtained by the processor, comprising walking from the data, wherein the data comprises dictionary entries, to code words, without the processor performing a memory lookup, and to decompress code words subsequently obtained by the processor, comprising walking from the subsequently obtained code words to the data comprising dictionary entries, without the processor performing a memory lookup, wherein the dictionary entries are of a fixed size, and wherein utilizing the architected dictionary to compress the data subsequently obtained by the processor and to decompress the code words subsequently obtained by the processor comprises locating ranks relevant to the dictionary entries.

2. The computer-implemented method of claim 1 , further comprising:
obtaining, by the processor, the additional data; and
compressing, by the processor, the additional data utilizing the architected dictionary.

3. The computer-implemented method of claim 1 , further comprising:
obtaining, by the processor, a given code word; and
decompressing, by the processor, the given code word utilizing the architected dictionary.

4. The computer-implemented method of claim 3 , wherein the decompressing comprises walking, by the processor, from the ranks to the architected dictionary.

5. The computer-implemented method of claim 1 , wherein the architected dictionary comprises references for each dictionary entry describing parent and child relationships associated with the dictionary entry.

6. The computer-implemented method of claim 5 , the reordering comprising:
associating, by the processor, each dictionary entry with a rank;
sorting, by the processor, each dictionary entry according to the rank assigned;
updating, by the processor, the references for each dictionary entry; and
discarding, by the processor, locations for each dictionary entry in the value conversion dictionary prior to the updating.

7. The computer-implemented method of claim 6 , the sorting further comprising, retaining, in a memory, the locations for each dictionary entry in the value conversion dictionary.

8. The computer-implemented method of claim 1 , wherein the parse tree based compression algorithm is a Ziv-Lempel compression algorithm.

9. The computer-implemented method of claim 1 , wherein the generating and the assigning comprise generating Canonical Huffman Code.

10. The computer-implemented method of claim 1 , wherein the values are of variable size.

11. A computer program product comprising:
a computer readable storage medium readable by one or more processor and storing instructions for execution by the one or more processor for performing a method comprising:
obtaining, by a processor, data comprised of values, wherein the values are of variable size, and generating a value conversion dictionary by applying a parse tree based compression algorithm to the data, wherein the value conversion dictionary is comprised of dictionary entries that represent the values;
obtaining, by the processor, a distribution of the values and estimating a likelihood for each value based on the distribution;
generating, by the processor, a code word to represent each value, wherein a size of each code word is inversely proportional to the likelihood of the value represented by the code word;
assigning, by the processor, a rank to each code word, wherein the rank for each code word represents the likelihood of the value represented by the code word;
based on the rank associated with each code word, reordering, by the processor, each dictionary entry in the value conversion dictionary to associate each dictionary entry with an equivalent rank, wherein the reordered value conversion dictionary comprises an architected dictionary; and
storing, by the processor, the architected dictionary as at least one tree structure in a memory, wherein the processor utilizes the architected dictionary to compress data subsequently obtained by the processor, comprising walking from the data, wherein the data comprises dictionary entries, to code words, without the processor performing a memory lookup, and to decompress code words subsequently obtained by the processor, comprising walking from the subsequently obtained code words to the data comprising dictionary entries, without the processor performing a memory lookup, wherein the dictionary entries are of a fixed size, and wherein utilizing the architected dictionary to compress the data subsequently obtained by the processor and to decompress the code words subsequently obtained by the processor comprises locating ranks relevant to the dictionary entries.

12. The computer program product of claim 11 , further comprising:
obtaining, by the processor, the additional data; and
compressing, by the processor, the additional data utilizing the architected dictionary, wherein the compressing comprises walking, by the processor, from dictionary entries to ranks.

13. The computer program product of claim 11 , further comprising:
obtaining, by the processor, a given code word; and
decompressing, by the processor, the given code word utilizing the architected dictionary, wherein the decompressing comprises walking, by the processor, from ranks to dictionary entries.

14. The computer program product of claim 11 , wherein the value conversion dictionary comprises references for each dictionary entry describing parent and child relationships associated with the dictionary entry.

15. The computer program product of claim 14 , the reordering comprising:
associating, by the processor, each dictionary entry with a rank;
sorting, by the processor, each dictionary entry according to the rank assigned;
updating, by the processor, the references for each dictionary entry; and
discarding, by the processor, locations for each dictionary entry in the value conversion dictionary prior to the updating.

16. A system comprising:
a memory;
one or more processor in communication with the memory; and
program instructions executable by the one or more processor via the memory to perform a method, the method comprising:
obtaining, by a processor, data comprised of values, wherein the values are of variable size, and generating a value conversion dictionary by applying a parse tree based compression algorithm to the data, wherein th

Assignees

• IBM

Inventors

• Bradbury Jonathan D
• Helms Markus
• Jacobi Christian
• Puranik Aditya N
• Zoellin Christian

Classifications

• G06F3/0638

  Organizing or formatting or addressing of data · CPC title

• G06F16/2365

  Ensuring data consistency and integrity · CPC title

• G06F16/9027

  Trees · CPC title

• H03M7/3088

  employing the use of a dictionary, e.g. LZ78 · CPC title

• G06F3/0608

  Saving storage space on storage systems · CPC title