DTD convert escape to valid characters about translate

[dwaynebailey]

Also there are quote a few of these:

“Headers & Footers”

And other enties with "

It would be great if these were converted to the correct character for the
translators so that they don’t have to learn what all these names character
variables mean. During the reverse process it would simply be a matter of
converting the character back into the named variable as needed.

[dwaynebailey]

I’ve marked this as a 129 blocker. It isn’t really but the value of deciding
which html escapes should e covnerted to normal Unicode is great.

How do we determine what escapes should be escaped? Need to check occurances in
.proeprties and .dtd files to check.

[dwaynebailey]

For now we're focusing only on DTD escapes see http://www.iro.umontreal.ca/~lapalme/ForestInsteadOfTheTrees/HTML/ch03s01.html

i.e

<

&

"

'

>

I think that is all

[unho]

Partially fixed in commits between 126590d and 9c5ba92 Several of the commits in this range include fixes for other tests, or other code affected by the changes in DTD store introduced to fix this bug.

Done:

• Converting &#37, &#37 and % to % character, and converting back to % always.
• Converting " to " character, and converting back (except when =" is present in string in order to avoid escaping things like HTML attributes, for example href one in HTML anchor tag).

Partially done:

• Converting ' to ' character, and converting back, is not done in all cases. Just made minor adjustments in code but keeping the previous behavior. The previous behavior was kept because it is necessary for not breaking special behavior needed for roundtrip conversion in Mozilla DTD files. See bug http://bugs.locamotion.org/show_bug.cgi?id=2740

Not done:

• Converting & to & character, and converting back, since this has implications on accesskeys roundtrip conversion. See bug http://bugs.locamotion.org/show_bug.cgi?id=2742
• Converting < to < character, and converting back, since this introduces a lot of unwanted changes. See bug http://bugs.locamotion.org/show_bug.cgi?id=2741
• Converting > to > character, and converting back, since this introduces a lot of unwanted changes. See bug http://bugs.locamotion.org/show_bug.cgi?id=2741