DeastinY / CoffeeRecommendationBot

Recommends a coffee to Twitterusers in need !

MIT License

8 stars 14 forks source link

Post a picture of the base coffee along with the main tweet #37

Open shauryachawla opened 8 years ago

shauryachawla commented 8 years ago

I just had this idea. What da' ya' say?

crawling

DeastinY commented 8 years ago

LOVE this. It was one of the very very early ideas I had: Draw a small picture for every aspect a description could have, stack them ontop of each other and generate a complete coffee. The image would be abstract, like this http://imgur.com/C6UgPVO ! Otherwise maybe just google for something along the lines, make sure it's a picture of coffee and add it to the tweet just for the visual appeal :smile:

Anyways, great idea ! Thanks for all of your input @emorres25 !

shauryachawla commented 8 years ago

You are welcome! :)) Now, the question is - we will need to have the images SAVED in the project files, right? Or will it be the same i.e. crawling dynamically and updating the image links or something?

gauravb7090 commented 8 years ago

Hey I wanted to work on this! How are we planning to get the image? I assume the image should be as close as possible to the order generated.

Right now, we are generating a random order on the basis of the "coffeeType.json" file that we generated using crawling.

Do we want to fetch a random image relevant to the order generated during the generation of the order or do we want to prefetch specific images as from the Starbucks website for specific coffee types?

Looking forward to your response.

DeastinY commented 8 years ago

My proposal is, first of all let's decide on:

Put together an image based on single images for the different layers
Search for an image as close as possible on the internet Then talk about the best way to integrate. Both won't be too easy

gauravb7090 commented 8 years ago

I didn't get what you meant by

put together an image based on single images for the different layers.

Can you please elaborate a bit more on that?

I am planning to start by writing a script to fetch closest possible image to the generated coffee.

Can I assign this issue to myself?

shauryachawla commented 8 years ago

Can't really crawl for images of respective available coffees, you will need to add them manually, because the json file will have starbucks coffee + other manually added coffees.

DeastinY commented 8 years ago

I don't really think we can start working on that featue just yet