This problem has been there since vanilla civ4, and exists because the xmls get parsed for validity with the local codepage, but there are some latin1-characters which aren't correctly escaped.
This problem only sources on with a MBCS-codepage (i.e. japanese, chinese, korean etc.).
The problem is, that some italian strings (and only the italian translations) end with an non-ascii-character, and this character gets parsed with the following < to one multibyte-character, which destroys the tag, so the xml appears to be malformed.
This only affects the italian strings (at least in vanilla civ4), and can be solved by either correctly escaping these characters (by &#<unicode-value> ;, the correct variant) or just adding a space at the end of the problematic strings (the lazy variant
.
In 1.52, these lines need correct escapes:
9358, 9364:
ù -> &#249;
9430, 9452, 9606:
à -> &#224;
The easiest way is just making a search & replace for these two characters. too much escaping won't hurt, just makes it perhaps a bit harder to read (in the xml, the final output is not affected from this).
<rant>
The best solution would be to encode these xmls in UTF-8, as there wouldn't be any escapes needed any more, and as the new languages which seemingly added with a future patch include japanese and chinese, these language will need every single character to be escaped in the current way, which would make it impossible to change or correct these translations.
</rant>
Have much fun,
Inu-Yasha
Edit: Stupid forum parsed my to use as replace escapes
Edit2: okay, but it won't parse &, changed to unicode value
Edit3: okay, there seems to be a bug in the forum which changes any &#38;s to & in edit/preview, screwing any escaped escapes when opening preview or edit.
))