On 7/28/25 18:01, The Wanderer wrote:
Reading a document is often easier to implement than writing it out correctly. Reading is also safer, because you're not overwriting any data; if you try to save the document, you might be doing so to an existing filename, and if the resulting format is invalid by the standards of whatever you want to read it next, then you've just caused data loss.
While this is true, there's also another side to it. On writing, you can restrict yourselves to a safe subset, while on reading you need to handle essentially all obscure possiblities: even if you don't really understand it, you need to be able to parse it. Of course, XML (for xlsx) makes this easier. There's a reason pandoc supports more formats for writing than for reading. Detlef