I'd be interested to know of any deficiencies or edge cases I fail, and maybe see about performance. I did this as a character-by-character processors via FSM.
Of course, I'm sure super-csv is more versatile. But it was complicated enough that I felt my own implementation would be more worth my time.
So perhaps, try to improve the documentation or reduce the complexity. A few good examples can go a long way.
Kasper B. Graversen
@dilijev have you even looked at the home page? there is a ton of documentation and examples. What specifically are you missing?
Your implementation lack all the stuff SuperCsv offers out of the box in terms of consistency, converting values from strings, reading directly into objects, handling wrongly escaped "-elements
I did. Both the README on github and the website were unintuitive. In fact, even knowing that what I'm looking for is present, it has taken me 5 minutes to find the first example, and I don't understand what it is supposed to do.
With some effort I believe I would be able to understand it. But the API is not immediately simple or straightforward. Not sure what you mean with the consistency issue. And in my case I did not need to read directly into objects. CSV is mostly strings and I was crunching data as strings. Thus, this simple implementation served my purposes.
I'm not saying this is a bad project by any means. I'm interested in it and I would like to be able to use it in future projects, but it seems to me that it has so many features the API has started to get in the way of itself.
Kasper B. Graversen
Thanks for the feedback. I'm puzzled as to your experience. From the "readme" you choose documentation. Then choose "where to begin" under getting started. Then you select help for reading or writing. Then your select eg reading and you are presented with a lot of ways to do reading. But we always welcome pull requests if you want to contribute with making things clearer. Thanks.
Alright, I may poke around and see if I can think of a way to make things clearer. I'm working on a project right now which is getting close to release and we want to provide some more complex "real-world" type of examples, one of which might be CSV reading, so we'll give your library a shot.
Well the instructions you linked take you to a table which is a feature comparison. And the only obvious links are to javadocs, which is a bad way to understand the usability of an API overall. So I assumed I was in the wrong place and kept looking.