Mismatched end tag but it seems perfectly regular

"Renzo Orsini [email protected] [emacs-nxml-mode]" <[email protected]> Fri, 11 Mar 2016 08:50:06 +0100
Newsgroups gmane.emacs.nxml.general
Message-ID <[email protected]>
I have an xml file with ~3500 lines, without a DTD, in which a single end t=
ag is marked as =E2=80=9CMismatched end tag" (it is the second </word> in t=
his three-iine fragment:


            <word id=3D"w1065">tibi</word>
            <word id=3D"w1066">saecula</word>
            <word id=3D"w1067">finem</word>


Everything else is correct, according to nXML. I looked at the hexadecimal =
representation and everything seems ok to me.


00000000: 2020 2020 2020 2020 2020 2020 3c77 6f72              <wor
00000010: 6420 6964 3d22 7731 3036 3522 3e74 6962  d id=3D"w1065">tib
00000020: 693c 2f77 6f72 643e 0a20 2020 2020 2020  i</word>.      =
=20
00000030: 2020 2020 203c 776f 7264 2069 643d 2277       <word id=3D"w
00000040: 3130 3636 223e 7361 6563 756c 613c 2f77  1066">saecula</w
00000050: 6f72 643e 0a20 2020 2020 2020 2020 2020  ord>.          =
=20
00000060: 203c 776f 7264 2069 643d 2277 3130 3637   <word id=3D"w1067
00000070: 223e 6669 6e65 6d3c 2f77 6f72 643e 0a    ">finem</word>.


Finally, if a copy the file with C-x C-w the new file has the same error, w=
hile if I copy all its content in a new empty file in txt-mode, then change=
 its mode to nXML, the tag is no more mismatched, and the file is considere=
d valid.


Should I assume that if the file has no schema, the validation algorithm of=
 nXML is not always reliable? Or that there is some subtle problem in my fi=
le that I cannot spot? (the file is UTF-8).


Thanks a lot!



Renzo Orsini
[email protected]












------------------------------------
Posted by: Renzo Orsini <[email protected]>
------------------------------------