Enrico Forestieri wrote:
On Wed, Sep 03, 2008 at 10:25:13AM +0200, Abdelrazak Younes wrote:
[EMAIL PROTECTED] wrote:
Author: forenr
Date: Wed Sep 3 10:17:00 2008
New Revision: 26290
URL: http://www.lyx.org/trac/changeset/26290
Log:
We may have buffers with different encodings, that, in turn, may also
be different from the language default ones.
Modified:
lyx-devel/trunk/src/mathed/InsetMathHull.cpp
Modified: lyx-devel/trunk/src/mathed/InsetMathHull.cpp
URL:
http://www.lyx.org/trac/file/lyx-devel/trunk/src/mathed/InsetMathHull.cpp?rev=26290
==============================================================================
--- lyx-devel/trunk/src/mathed/InsetMathHull.cpp (original)
+++ lyx-devel/trunk/src/mathed/InsetMathHull.cpp Wed Sep 3 10:17:00 2008
@@ -450,7 +450,9 @@
// in text mode (such as $\text{$\phi$}$) gets processed twice. The
// first time as a whole, and the second time only the inner math.
// In this last case inset.buffer() would be invalid.
- static Encoding const * encoding =
inset.buffer().language()->encoding();
+ static Encoding const * encoding = 0;
This use of 'static' looks fragile....
Why?
For the reason below.
+ if (inset.isBufferValid())
+ encoding =&(inset.buffer().params().encoding());
WriteStream wi(ls, false, true, false, encoding);
So encoding will keep the previous value if isBufferValid() return
false. Are we sure that isBufferValid() will always return true the
first time latexString() is called?
Yes. See the comment.
OK. But this deserves a FIXME tag IMHO.
Abdel.
