I'm not sure why you'd want to start with json or yaml. Seems like you'd want to start with a document, not a datastore. A datastore would tend to being opinionated and inflexible. What happens when my resume doesn't match a prescribed format? Or what about other documents beyond resumes? And whats with the star bullets? Why mix style with data? What if I want hearts instead of stars?
I only mention this cause I'm working on something similar for my resume. I start with markdown and output pdf, html, etc.
What happens when the code you wrote to parse markdown is unable to understand your new fangled resume material? Then markdown is just as much a datastore as json. The problem is not in json or markdown. It is in the code: your code must be able to understand all reasonable resume material.
Comments
I'm not sure why you'd want to start with json or yaml. Seems like you'd want to start with a document, not a datastore. A datastore would tend to being opinionated and inflexible. What happens when my resume doesn't match a prescribed format? Or what about other documents beyond resumes? And whats with the star bullets? Why mix style with data? What if I want hearts instead of stars?
I only mention this cause I'm working on something similar for my resume. I start with markdown and output pdf, html, etc.
In general, I would say you want to have schema. However, both yaml and json lack this. Still I find this a nice approach.
What happens when the code you wrote to parse markdown is unable to understand your new fangled resume material? Then markdown is just as much a datastore as json. The problem is not in json or markdown. It is in the code: your code must be able to understand all reasonable resume material.
I don't understand. Why would you write a markdown parser? Theres like a million implementations already. I use https://www.npmjs.org/package/markdown-pdf