TimeTable Labelling Files

Format of input graph:
-------------File begins--------------------
Number_of nodes 
node_id: sequences of edges (to_node_id, starting_timestamp, duration, bus_id) -1
.
.
.
------------File ends-----------------------

Each tuple (to_node_id, starting_timestamp, duration, bus_id) indicates that departe from node_id at time starting_timestamp, and arrives at to_node_id at time starting_timestamp+duration, by taking bus bus_id.

Note that the sequence of edges are sorted. In particular, edges are sorted in ascending order of node id, and if node ids are the same, then in ascending order of starting timestamp.


A sample input graph :
------------------File begins-------------------
4
0: 1 2 1 b1 1 3 1 b2 2 3 1 b3 -1
1: 2 4 1 b4 -1
2: 3 5 1 b5 -1
------------------File ends-------------------

In the above example, b1, b2, b3, b4, and b5 are bus ids.



How to compile the code:

g++ TTL.cpp -O3 -o TTL -std=c++11

usage:
./TTL -g graph_file -s coverage/file -dq query_file -i output-index-file

Examples:
#load ordering files. The current stored files stores the ordering  which uses
the heuristic ordering method mentioned in the paper.
Then it generates the index file and do query.

./TTL -g austin.txt -s file heuristic-austin.index.order -dq query-austin.txt


#Ordering on the fly instead of loading ordering files. It will use the
adaptive coverage centrality based heuristic ordering method mentioned in the paper. 

./TTL -g austin.txt -s coverage 

#output the index to specified files
./TTL -g austin.txt -s coverage -i austin.txt.index