This is a program to search through Snapstream closed-caption files. For a sample implementation, see this link, http://johnny.sas.upenn.edu/~pemantle/test.html.

Various parts of the application:

• Searching
• Visualizing
• Hot Lists

All the C++ code is in the src folder. All the JavaScript and HTML is in the html folder. The jobs folder contains some optional utility scripts to automate searches.

To compile and build, do the following

1. Make sure you're in the snapstream-reader directory
2. mkdir build && cd build to make the build somewhere else in case something goes wrong
3. cmake .. to build compilation script
4. make to compile
5. ctest to run unit tests

See jobs/get_snap_branch.sh for a sample deployment on a Apache server, which is run with sh get_snap_branch.sh <branch name>.

On your server, you'll want a directory structure like this

html-root (the root of where you want to serve the application out of, can be named whatever)
| cgi-bin 
| tmp

After building there are several statically linked binaries in build/cgi-bin. Copy those binaries to html-root/cgi-bin on Apache. In the build folder, there are several HTML, CSS, JavaScript, and JSON files. Copy them all to the html-root.

Now there will be a dictionary.csv file in build/cgi-bin. Copy this to your html-root/cgi-bin, too. The Snapstream closed-caption files need to go in html-root/cgi-bin/Data to be read by the binaries.

Finally, ensure that the html-root/tmp folder is writable and html-root/cgi-bin is readable. You can run chmod -R 777 html-root/tmp to do this.

In the html folder, you will find find various front ends. index.html allows you to perform a Simple Search, Multiple Search, and Matrix Search. graph.html visualizes the results of a Matrix Search, and time-series.html visualizes the results of a multiple search.