Course Page: https://fenix.tecnico.ulisboa.pt/disciplinas/AVExe/2018-2019/2-semestre/pagina-inicial
Project Instructions: http://grupos.ist.utl.pt/meic-cnv/project/Enunciado_CNV_2018_19.pdf
Project FAQ: https://tinyurl.com/cnv-faq-18-19
The Java code developed by the faculty targets Java version 7. We provide the compiled .class files and the corresponding .java files in this archive in:
pt/
org/
The org/ directory is a helper library which is used by the solver code. You don't need to use or instrument the classes in that directory. If for some reason you compile the original solver .java files, make sure org/ is in the classpath.
For examples in this readme, we assume the .zip archive was extracted to the user's home directory and the directory structure should look like:
$HOME/cnv-project/pt/
$HOME/cnv-project/org/
$HOME/cnv-project/README.md
$HOME/cnv-project/datasets/
$HOME/cnv-project/META-INF
Where $HOME represents the user's home directory. In Amazon EC2 instances, it should look like:
/home/ec2-user/cnv-project/pt/
/home/ec2-user/cnv-project/org/
/home/ec2-user/cnv-project/README.md
/home/ec2-user/cnv-project/datasets/
/home/ec2-user/cnv-project/META-INF
You will launch solver jobs through the faculty web page at http://groups.tecnico.ulisboa.pt/meic-cnv/cnv-project/. In this link there is a simple HTML page where you can choose the map to execute, the search strategy (BFS, DFS, ASTAR) to use, the starting point (xS, yS), the top-left corner (x0, y0) and bottom-right corner (x1, y1) of the bounding-rectangle.
You must also specify the address and port of your Amazon EC2 instance or load balancer so that your browser sends the request to it and receives the solution when it finishes.
When you configure the parameters such as your Amazon EC2 instance (load balancer) address and port, the search algorithm to use, the starting point and the bounding rectangle coordinates and the map on which to execute the solver, you can press the Execute button. Your browser will then send a request to your web server running on Amazon EC2.
We provide an example web server which listens for requests on port 8000 - see below for more information.
- Sample web server (which will be run on your Amazon EC2 instances) which receives requests from the course faculty web page at http://groups.tecnico.ulisboa.pt/meic-cnv/cnv-project/
- Solver with BFS, DFS and ASTAR algorithms.
- Height map generator with different generator strategies.
A simple HTTP server receiving requests. Is is to be run on your Amazon EC2 instances. The server will parse the XMLHttpRequest parameters (x0, y0, x1, y1, xS, yS, s) and pass them to the solver's argument parser. After the solver checks the parameter values and stores them, the solving algorithm strategy (BFS/DFS/ASTAR) is executed.
When the solver finishes, the resulting map (with a drawn path) is saved on the temporary directory of the Amazon EC2 instance that has the web server that received the request. This image is then sent as the response (to the computer where your browser is open) and drawn via JavaScript on the browser.
The following may be performed to run the compiled server class (provided by the faculty). This launches the example server listening on port 8000:
cd $HOME/cnv-project
java pt.ulisboa.tecnico.cnv.server.WebServer
This is the black-box with algorithms to be instrumented in the Amazon EC2 instances and measured by student code using the Java BIT instrumentation tool.
Here is an example to trigger a solver execution over an input map provided locally from the file system:
java pt.ulisboa.tecnico.cnv.solver.SolverMain -d -s BFS -w 512 -h 512 -i 'datasets/RANDOM_HILL_512x512_2019-03-01_10-28-59.dat' -o '$HOME/' -yS 25 -xS 25
This example starts the solver (-d, printing additional debug information) with a BFS strategy (-s) for a map of width 512 (-w) and height 512 (-h) for the map file datasets/RANDOM_HILL_512x512_2019-03-01_10-28-59.dat (-i), to be stored in the user's home directory (-o). The search starts at position x = 25 (-xS) and y = 25 (-yS).
Its argument parser is self-explanatory and a description of its arguments will be printed if one is missing.
usage: utility-name
-d,--debug set debug mode.
-f,--fill should solution path be drawn with dashes?
-g,--gradient <arg> output image's color gradient.
-h,--height <arg> generated image height (default 512
pixels).
-i,--input <arg> path to input gradient image to solve.
-o,--output-directory <arg> output directory for generated images.
-s,--strategy <arg> solver strategy can be one of: BFS, DFS or ASTAR.
-w,--width <arg> generated image width (default 512 pixels).
-x0 <arg> upper-left x coordinate (default 0).
-x1 <arg> lower-right x coordinate (default equal to
image width).
-xS <arg> starting x coordinate (default 0).
-y0 <arg> upper-left y coordinate (default 0).
-y1 <arg> lower-right y coordinate (default equal to
image height).
-yS <arg> starting y coordinate (default 0).
When the solver finishes, it stores an image to disk with an appended timestamp to avoid collisions for the same datasets.
If you want to test locally, you can access existing images at: http://groups.tecnico.ulisboa.pt/meic-cnv/cnv-project/datasets
A generator for height maps with a lot of options. Usage example:
java pt.ulisboa.tecnico.cnv.generator.GeneratorMain -d -s RANDOM_HILL -w 512 -h 512 -o '$HOME'
The shown example outputs debug information (-d) while generating a terrain with a 'RANDOM_HILL' strategy (-s) with width 512 (-w) and height 512 (-h), to be stored in the user's home directory.
usage: utility-name
-d,--debug set debug mode.
-g,--gradient <arg> output image's color gradient.
-h,--height <arg> generated image height (default 512
pixels).
-o,--output-directory <arg> output directory for generated images.
-s,--strategy <arg> generator strategy can be one of: SINUSOIDAL,
RAMP_TEST, PYRAMID, BLUE_NOISE or RANDOM_HILL.
-w,--width <arg> generated image width (default 512 pixels).
To know more about the generator strategies, check these classes:
pt.ulisboa.tecnico.cnv.generator.PyramidStrategy
pt.ulisboa.tecnico.cnv.generator.RampTestStrategy
***pt.ulisboa.tecnico.cnv.generator.RandomHillStrategy***
pt.ulisboa.tecnico.cnv.generator.SinusoidalStrategy