Here is a log of keeping track of automating the steps I use to create a roll20 game from a pathfinder adventure path pdf. (Mainly, so I don't lose focus and have a sounding board for this idea.)

Objective - create a C# application that will take the creature file (manually captured data from the pdf) and map it to a character sheet Observation - first analysis - this sucks, there is an HTML for the character sheet which needs to be analyzed. Then I have the character object (nicely defined, thank you!), then the attributes and abilities object - what the heck are these, and what values do I populate into them? Spent 4 hours searching the internet and reading! The only idea have come up with is to create a javascript that will take an existing creature (npc) and export the attributes and abilities ... maybe there is a way to do it in such a way to mimic C# code - wouldn't that be cool! Then reverse engineer it from there.

Objective - create a C# application that will convert Pathfinder (from Paizo.com) pdf files into Java Script commands to create Handouts for the gm (and possibly the players).

Observation - because the pdf to text is clunky and not all the handouts are organized, the user (currently) will need to manipulate the file before ingestion. File manipulation includes:

• Tagging file as appropriate type (PATH for adventure path, CREATURE for creature stat blocks)
• organizing text by handouts versus being all jumbled together - it is normally a handout that interrupts another handout (i.e. New Items, Author Shout Out, other notes)
• Removing Creature stat blocks (putting them into a creature file)
• Tagging "BoxText:" to each stanza to read to players.

Advantages - coding an Adventure Path is a monster of a task (taking 72 + hours per book) - this is now cut down to almost 3 hours with this application. It does the formatting automatically (probably should make this dynamic - to allow people to modify this).

Objective - create a C# application that will copy the images out of the pdf and save them as jpeg files

Observation - This generates hundreds of duplicate images (they don't have the same id, guid are same for different images) the user (currently) will need to remove "duplicate" images before ingestion (and not all images may be required to run the game).

• User is going to need to sort images by size and remove any that they don't want before uploading to Roll20.

Advantages - since copying images isn't hard to do from the pdf, this is a nice to have, not essential (probably saves about a couple of hours of work)

Honest time - I'm a hack as a developer, on the plus side, I do a lot of testing. I learned html back in the 90s (and kinda stayed current with it). I've done object oriented coding with C#, Java, and PHP (possibly others). I'm weak on Javascript, but being an object oriented language it appears to be simple enough to pick up.