March 2-6

Monday: 1:25-2:30
Tuesday: 12:00-1:00
Friday: 1:25-2:30

     This week was very split up because we had two two hour delays, two snow days, and an early dismissal. So, on Monday, I finished entering all of the teachers and their room numbers into the database so they could be used in the app mapping area. Additionally, we worked on the path-finding algorithm to navigate the hallways of the school. We had to first decide which way we wanted to go about winding the path. We decided that we would start in the hallway that the start room is contained in. Then, from that hallway, we would find the hallway that is closest to the target and is perpendicular to the current hallway, because the layout of our school facilitates this method. It would continue moving through perpendicular hallways until it reaches the target hallway. After the method of path-finding was decided, my colleague began coding that while I started to tackle the beast of data scraping: the athletic calendar. When we first started to look into scraping, I found an issue with the site our school uses to host our athletic calendars. The calendar is not hard-coded into the source code, it is input with a javascript. This makes scraping much more difficult because now we have to find a way to collect the data without having direct access to it. My first idea is to see if I can trace the origin of the data with Google's Developer tools. If I can, I'll just be able to manipulate the site URL based on the date and get the data from there. However, after a bit of searching, I found that these origins are not reliable and will not work for my purposes.
     Later, I found a module called Spynner that can load webpages within Python. I worked on downloading and installing the module into Python for part of the period, but then took some time to help with the mapping rocess. On the path-finding side of the app, my colleague was having some logic errors. Currently, the algorithm will find a path to the target room, but it is not the most efficient and sometimes isn't even sensible. So, we had to make the map more accurate regarding a small hallway that is almost connected to a larger hallway. Before, we had had the smaller hallway connected, but this ended in incorrect paths because not ever path required going to the end of both the smaller and larger hallways. So, we had to change it.
     On Friday, I was still troubleshooting the installation of Spynner. So, after downloading and installing it, I tried to work out how to use Spynner to run the javascript on our athletic calendar's site. However, it turns out that Spynner is not capable of running javascript. Instead, it can manipulate inputs on a webpage. So, I was back to the drawing board. After the frustration of another dead-end, I helped to test the working path-finding program. Then, my coleague decided to start on a drop-down menu to list the possible rooms to start and end in, and I continued my quest of scraping the dreaded athletic calendar.