Fast and efficent Huffman coding for nibbles and bytes in C++?

[88888 Dihedral]

I am wondering is there any good and fast Huffman coding
for nibles of for each segment of 256 to 1K byte

 

[Asger Joergensen]

Hi 88888

I am wondering is there any good and fast Huffman coding
for nibles of for each segment of 256 to 1K byte

Care to explain what You mean ?

Best regards
Asger-P

 

[88888 Dihedral]

在 2012å¹´2月27日星期一UTC+8下åˆ4æ—¶33分40秒，Asger Joergensen写é“：

Hi 88888



Care to explain what You mean ?

Best regards
Asger-P

16 entries of 16 to 32 bytes of the overhead per block to rebuild Huffmancode tree can be very fast for locality

 

[gwowen]

在 2012å¹´2月27日星期一UTC+8下åˆ4æ—¶33分40秒，Asger Joergensen写é“：





16 entries of 16 to 32 bytes of the   overhead per block to rebuild Huffman code tree  can be very fast for locality

Care to explain what you mean?

 

[88888 Dihedral]

在 2012å¹´2月28日星期二UTC+8上åˆ1æ—¶30分16秒，gwowen写é“：

Care to explain what you mean?

在 2012å¹´2月28日星期二UTC+8上åˆ1æ—¶30分16秒，gwowen写é“：

Care to explain what you mean?

I mean that one should carefully sellect basic building units to
develop long live algorithms that can be ported to other languages.

Thus, a list that can hold unknown type of objects dynamically and
a hash table well implemented are the two units which I choose.

But I still use the concepts about set, tree and graph in the abstract
sense to solve problems. Anyway programs in the implementation levels
are different from the abstrct levels.