Output Parsing Required #1
Notifications
Total Time Spent: 13 hours 41 minutes
Due Date
sigonasr2
13 hours 41 minutes
No due date set.
Dependencies
No dependencies set.
Reference: sigonasr2/SigPlace#1
Loading…
x
Reference in New Issue
Block a user
No description provided.
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Now that all text has been read by the program, have to parse and interpret all of them into a useable format.
Love Live ParserPop'n Parser
Sound Voltex Parser
DDR Parser
ITG Parser
Output parsing uses midpoints for region detection.
After detecting regions, between JP and EN data, the one with the highest accuracy rating gets used as the data point.
The
toString()
method of these readers will display the relevant data after parsing.Pop'n Sample
Actually decided to remove the initial newline, made no sense.
New sample output:
Because we're mostly recognizing numbers, this method may be promising: https://github.com/eyarng/sdr
Perhaps we write our own implementation or use it directly.
Doesn't work due to python2. However, it gives me an idea. Instead, we just do our own number checking (since numbers always look the same and leave the font detection up to our original detection.)
Makes more sense. Will begin working on refactoring that.
Data format for storing glyphs is:
Missing font glyphs:
Next is to do the actual reading and comparison in the readers (with the ones we have accomplished thus far)
Just for clarity, output from the program:
First testing proves successful. Reading algorithm reports scores like so:
Matches currently allow for 1 pixel of wiggle room in any direction. Scoring system adds 1 point for correct pixels (including transparent ones) while it punishes it -2 for non-matches. This makes sure that incorrect pixels are forced a lot out than just somehow matching some.
Here is what is still missing:
Had to redo Sound Voltex number parsing for large and tiny scores due to not realizing the version change moved some elements. Updated list of missing elements:
All that is left is grade parsing, but that will probably just be an output for displaying on the site as an image.