NFD

One related question: sometimes strategy needs to find a longest prefix match that satisfy certain condition, such as the existence of a certain type of StrategyInfo.

Which design is better?

• add a predicate parameter to findLongestPrefixMatch functions
• add findLongestPrefixMatch(const measurements::Entry&) overload; if the strategy thinks the initial match doesn't satisfy the requirement, it could call this overload to find a longest prefix match of the parent of passed Entry

The first design is cleaner, but it's hard to pass information from the predicate to the caller.

For example, if the predicate calls getStrategyInfo<T>, the caller (who calls findLongestPrefixMatch) will have to call getStrategyInfo<T> again.

http://gerrit.named-data.net/1600 accepts predicate in Measurements::findLongestPrefixMatch.

The next commit will expose findLongestPrefixMatch operation to MeasurementsAccessor.

http://gerrit.named-data.net/1606 adds findLongestPrefixMatch and findExactMatch to MeasurementsAccessor.

However, I realize that I forgot to add Measurements::findLongestPrefixMatch(pit::Entry), which shall come before this Change.

http://gerrit.named-data.net/1607 adds Measurements::findLongestPrefixMatch(pit::Entry).