The document could not be opened

Hi Jesse,

I just updated to Bike 1.1 and now Bike can’t open the bike files that I have been using. It just says, "The document X could not be opened. "

What can I do to fix this situation?

Anyone else seeing this?

Is it possible to zip and then email me (jesse@hogbaysoftware.com) one of the documents?

Just did. Thanks, Jesse!

Problem resolved, but could show up in other documents.

I don’t think issue had to do with any specific version of Bike. Instead a unicode control character got into the document (in this case it was “\u{0E}”). Apparently Bike’s XML encoder can write that character, but fails when trying to read it.

Any unicode/xml experts around here? I’m wondering what the correct way to handle these chars is. I have found that:

  1. I can still read documents with such chars if I enable HTML_PARSE_RECOVER when parsing document.

  2. I can copy/paste one of these control chars into an existing Bike document and then save … meaning Bike will write out an invalid XML document, since those chars are not allowed. Should I be doing some sort of clean on text before saving as XML?

Bike 1.2 Preview is better at recovering from these sorts of errors, even with some invalid characters document should still open now.