I am using an algorithm for tree certificates described for example here (p. 24-29).
Let's say I have two trees: A and B, and each tree has it's certificate produced by the algorithm above (C1 and C2).
Is it true, that if C1 contains C2 (exact sequence anywhere), it means A contains B as a subtree (B can be basically concentrated and considered as a leaf node of A)? If not, could you state a counter-example?
--edit--
Algorithm: (please take a look at the linked document for examples):
Label all vertices with string 01
While there are more than 2 vertices in G:
for each non-leaf x do:
If there is only one vertex x left, report x's label as the certificate.
If there are 2 vertices x and y left, concentrate x and y in increasing lexicographic order, and report it as the cerfificate.
Yes, it is true.
Assuming the certificate is correct, there is no possibility a certificate would contain another certificate and it wouldn't be it's subtree.