Re: Mismatched end tag but it seems perfectly regular

I applied the instructions of Greg Peterson and converted all files to utf-=
8 unix. Now, when I open each of them by itself, nXML does declare valid th=
e file. When I drag all their icons on emacs, by opening them =E2=80=9Cin p=
arallel=E2=80=9D, usually 5 on 6 are declared still invalid, each time chan=
ging which file is considered valid and which are considered invalid.
If I disable validation with C-c C-v for a file, and then re-enable it with=
 the same command, then the file is marked as Valid.

So my conclusion is that I have to open the files one at time, since Emacs =
does not have threads, and the validation code of nXML has some problem if =
executed concurrently, at least on my emacs/operating system pair.

Thanks a lot.

Renzo Orsini



> On 14 Mar 2016, at 04:34 , Greg Peterson [email protected] [emacs-nxml=
-mode] <[email protected]> wrote:
>=20
>=20
>=20
> On 2016=E5=B9=B403=E6=9C=8812=E6=97=A5 17:09, Renzo Orsini [email protected]=
t [emacs-nxml-mode] wrote:
>> Here is the folder with 6 files. When I open them with emacs (Emacs.app,=
 version 24.5 (9.0) on Mac Os X 10.11.3 (15D21),=20
>> are randomly marked as invalid, even if they should be all valid. For in=
stance, the last time that I opened them by dragging all of them on the Ema=
cs.app icon, ALL of them were marked invalid. Instead, if I open them indiv=
idually, not all of them are marked invalid. Note that they do not any DTD =
or schema.
>>=20
>=20
> I downloaded your ZIP file ) to my computer running Fedora 23 (GNU/Linux)=
, unzipped it without any options ("unzip xml-txt.zip"),  and opened each w=
ith emacs (GNU Emacs 24.5.1). According to nXML-mode on my computer, all ar=
e shown as "Valid" (well-formed XML) with no schema ("Using vacuous schema"=
).
>=20
> The only "abnormality" that I can see is MS-DOS file format. The mode lin=
e begins, "U (DOS):" and output from "M-x describe-current-coding-system" s=
ays that your files are DOS files (utf-8-dos). If they are DOS files on you=
r system, and if you will process them only on a Mac or Linux, FreeBSD, etc=
. (not on a Microsoft Windows PC), you might try saving them as Unix files:=
 "M-x set-buffer-file-coding-system" or "C-X RET f": "utf-8-unix". It might=
 not help at all, but it cannot hurt to try since MacOS X uses Unix file fo=
rmat. Otherwise your data looks fine to me.
>=20
>=20
>=20



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