nagelxz / comicsorter Goto Github PK

Due to a bug/stupid choice of code, I currently do not have the program looking in the Starting Directory for the sort.

When i wrote it in Java the recursion never created the same issue I'm having now, unless its the same thing that's causing the issue over there, where it would look at the same files again and try to move them but can't find the specified (NOTE: I noticed this before I added the actual sorting. Comes from the renaming of the files.) even though it exists right where its looking.

This is my first planned fix.

make it so that at the end of the sorting, once all orphaned files are noted, list the names like ls command (maybe concat the issue numbers next to it) and make it so that you can type add: comic name

also see if there is a way to add the tabbed function on unix systems to autocomplete.

Sometimes [read: ~30%] of files will fail to move if their ending folder exists. The problem is not that they fail, its that its seemingly random and it deletes the original before making sure its copied successfully.

I'm personally not going to update this issue, internally I'm labeling the Java code Depricated. If someone wants to figure it out, I'll gladly pull it into the code base. I'm pretty sure it's something i wrote wrong, probably surrounding the Copy.options[].

I honestly don't remember what I'm logging. I think its only errors. I should try making sure there are no files that deleted accidentally.

Like start the logs with how many files have been found.Then show how many we placed right, how many where orphaned, how many deleted(once I get to that point, another issue I'm adding but not making it for the milestone) and do a comparison of how many files were touched to how many were found.

It should be 2 separate logs, good and bad?

Currently the only file handling I do is for file types I want and don't want. It's easier to do mass downloading than to click through each package for the ones I want than it is for the ones I don't. This is going mostly for the comics that aren't part of the Marvel NOW that I'm interested in, or they got moved into the NOW release cycle and are too far in to start.

Basically want to make a blacklist file for the names that I don't want. Put them in a deleted folder inside the orphaned folder to start with. After knowing it works make it so that it deletes the contents after the end of the cycle run.

Use VS2010's built in profiler to see where the code has the biggest bottleneck. My bet there is the issue with the recursion in the directory to get the files, since that section of the code runs the most.

What does it do to files that don't have a place to go. Currently it leaves them in place. It's already been shown that the current code can move files no problem. Why not make it so that those files will be moved to someplace like "TO BE SORTED INTO RESPECTED FOLDERS PART DEUX" or just the current folder that is there since I don't use it currently. It would be worth utilizing it for something.

Make it so it lists the files that have no home. Just the names to figure out if its an error or if its something new or something I'm not following. Currently I need to look at the error logs to see what failed and track that file down. This is connected to the previous item

Remove files that have extensions that do not fit that of a comic.
not comics(.txt, .doc, .csv, etc...)
comics(.cbr, .cbz, .cb7)

delete .dat files from the beginning

I don't know if I want to make it actively monitor a folder or make it so that every 2 weeks it runs and does it's thing.

Looking at each folder separately inside the "to be sorted" directory and sorting them that way was naive and stupid. It's fine if i have a few [read 5], but as soon as that grows any much larger it increases the runtime dramatically. Changing it so that it gets all the files first, builds the list as it goes, and then sorts is a faster way to go about it.

Change the type of collection used for both the items to be sorted and the list to compare against. It's probably going to be better to use something with a hash based function, like Dictionary. The current way causes the comparisons to be something like O(n^2). With hundreds of files to 40+ directories it'll get painfully slow if I go to run it.