Numerical character references are not decoded in text elements

[mmcdole]

Numerical character references are not decoded in text elements

[galdor]

I should have a patch for you tomorrow.

[galdor]

For you information, numerical character references are decoded by encoding/xml starting at line 991 of https://golang.org/src/encoding/xml/xml.go.

[mmcdole]

@galdor thanks for the bug report.

For a number of reasons I was unable to utilize go's built in xml chardata decoding. Mostly because I wanted to be able to parse naked XML/HTML for certain elements. This has me using the ,innerxml attribute which bypasses the built-in code you referenced.

So, I think to do this correctly, I'm going to have to re-implement what they do myself. I'll have to replace the string.Replace code that I use in parseutils.go with a proper parser which decodes both the numeric and non-numeric character entities it finds in the text.

I'm currently in the process of rewriting my unit tests for this project (due to an unintentional licensing violation on my part). When #53 is ready to merge with my completed Atom tests, I'll add some new unit tests for both atom/rss that ensures that character entities and numeric character entities get decoded correctly.

[galdor]

So there's only DecodeEntities to modify, i.e. parse the string to decode both types of entities ? I could give it a try if you wish.

[mmcdole]

@galdor sure, give it a try.

I had considered implementing it like I did the other entity decoding, but I was concerned if there was a lot of the numeric entities that we would have to match, it may be an expensive operation to strings.Replace for so many items.

But, perhaps doing it the slow way is better than nothing for now, and I can take time to implement it on a character-by-charcter basis (like go's xml package does) later when I have time.