1. B-tree:expanding the width(merging nodes in BTS) and decresing the height to lower the time of putting individual layer from the disk to the ram.
search:only select a part of the tree to the ram, and cuz the tree is evolved from BTS, so we can view it as a sorted vector and using binary-search to improve effiency. (if this layer dosen't have any, then pull out the next one.[inner node;outer node(in disk)])
m-order B-tree.
m represent the number of the branches.
number of branch:(m/2,m]
prove the efficiency of such stracture(considering the lowest tree height and the maxiumn--conclude that it is much more efficient and stable with different m value)
2. exercise:
KMP:finding out the repetitive substring in a string.
1) finding the max common string(getNext) and caculate the delta length. Considering if it can be divided by the length.
2)finding the substr:
traverse the whole string and if there is one alpha that doesn't match, then go to the previous letter's next value. Continue to search from the letter to the enf.