Problem with missing characters in CSE Validator Editor

[jgcordes]

Hello Albert,

Today I encountered a very strange problem. I am hoping that it is a software bug, as opposed to bad memory in my computer!

I have a 91 KB HTML file (937 lines). Here is one line from it:
<tr><td>05-02-1822</td><td>_____</td><td>CALDWELL</td><td>Ann Margaret</td><td>17y</td><td>Cole Harbour Rev Charles Ingles</td></tr>

Validator shows this:

<tr><td>05-02-1822</td><td>_____</td><td>CALDWELL/td><td>Ann Margaret</td><td>17y</td><td>Cole Harbour Rev Charles Ingles</td></tr>
and of course produces an Error message -- Note the lack of an opening angle bracket "<" following the word CALDWELL in the Validator editor.

I do my primary editing in Linux, but have also opened this particular file in two Windows editors (Notepad2 and Vim); no other editor besides Validator's has a problem showing this angle bracket.

Any ideas about this weird problem?

Who has a problem: you or me?!

I would be glad to send you this file if you want to examine it.

John Cordes

[Albert Wiersch]

Hi John,

Yes, please email the file to me at support at htmlvalidator dot com. Also, please let me know what version of CSE HTML Validator you are using.

[jgcordes]

Have just emailed the file to you.

John

[Albert Wiersch]

Have just emailed the file to you.

John

Hi John,

Thanks for the file. I think the problem is that CSE HTML Validator is trying to open it as UTF-8 because that is what is specified in a "meta" tag, but your document contains invalid UTF-8 characters so I think that is causing the corruption.

Try using File->Open with Encoding and open the file with Western European (Windows) encoding or just "System codepage" encoding.

The problem character appears to be the '½' character on line 264 and at least one other line. I recommend using the character entity reference "&frac12;" instead. I think if you change all those '½' characters to the character reference and reload the document that it will then load properly in the future.

I suspect it works in other editors because they are not trying to load it as UTF-8, but CSE HTML Validator does because it looks for the "meta" tag that specifies the encoding and it sees "utf-8".

[jgcordes]

Hi Albert,

Thanks -- sounds good. I will try implementing your suggestions, but I think it is getting too late for me tonight.

John

[jgcordes]

Hi Albert,

I couldn't resist checking -- you were absolutely correct. I have changed those fractions to proper HTML entities.
Thanks!
John

[Albert Wiersch]

Great! I'm glad that fixed it.