Invalid Javadoc can be generated from SBE xml files for > and <

[RichardWarburton]

SBE generates javadoc from description tags in SBE xml files. It converts HTML escaped < and > characters < and > into plaintext ones, which then generates invalid javadoc. If the escaped HTML characters were preserved then correct javadoc would be generated. Examples below:

SBE description tag
description="Price of this fill reason or allocation. Required if NoFills(1362) > 0. Same as LastPx(31)"

Currently generated Javadoc
* @return PRICE9Decoder : Price of this fill reason or allocation. Required if NoFills(1362) > 0. Same as LastPx(31)

Javadoc tool error
/path/to/SomeFile.java:2594: error: bad use of '>' * @return PRICE9Encoder : Price of this fill reason or allocation. Required if NoFills(1362) > 0. Same as LastPx(31)

What I would expect for Javadoc
* @return PRICE9Decoder : Price of this fill reason or allocation. Required if NoFills(1362) > 0. Same as LastPx(31)

[mjpt777]

This can bring up a complex topic. How far do we go for all possible formatting in javadoc? Beyond escaping HTML characters there are the special characters in javadoc such as @ and all code formatting combinations. I've done the minimum for now.

[RichardWarburton]

Thanks Martin. I've just tested this and it works on my example. IMHO just the HTML escape characters is fine. The thing is that &lt and &gt are characters that people do use in SBE protocol documentation - the error was actually generated from an SBE xml file from CME for example. If you do something weird with code formatting - I think it's reasonably to punt that onto the maintainer of the file XML file to fix.