Driving automation from MyInfo data -- is reading the SQLite file directly okay? [Yes]
Posted: Mon Feb 13, 2023 8:43 pm
As noted elsewhere in this Forum, MyInfo stores its data in a SQLite database, and its articles in an HTML-like format. I find this attractive, due to the potential for writing multi-purpose documentation: it's for human readers and programs alike.
For example, if I'm designing and documenting a database or API schema, clearly both kinds of information are present. I could, of course, duplicate the machine-significant parts outside of my MyInfo file. Then I have extra work: two copies to update, and to keep in sync with each other. I would prefer, instead, to maintain a complete, coherent set of definitions in my MyInfo file.
How would I use the machine-significant data? To generate API code, or sanity-checking code. Easy enough, for a C++ and Python programmer, like myself, with a handy SQLite library.
However, MyInfo is not open-source, and its data format may be considered proprietary. Therefore, trying to read my information, directly, from the SQLite file, may be frowned upon. Is there an official opinion on this?
For example, if I'm designing and documenting a database or API schema, clearly both kinds of information are present. I could, of course, duplicate the machine-significant parts outside of my MyInfo file. Then I have extra work: two copies to update, and to keep in sync with each other. I would prefer, instead, to maintain a complete, coherent set of definitions in my MyInfo file.
How would I use the machine-significant data? To generate API code, or sanity-checking code. Easy enough, for a C++ and Python programmer, like myself, with a handy SQLite library.
However, MyInfo is not open-source, and its data format may be considered proprietary. Therefore, trying to read my information, directly, from the SQLite file, may be frowned upon. Is there an official opinion on this?