Full_Name: Fred Nicolls OS: FreeBSD 3.3-STABLE Version: xemacs-21.1.7 mule: no severity: Aggravating Submission from: (NULL) (196.7.0.164) Quite often it seems to happen (at least for me) that the LaTeX mode font-lock loses sync between math and non-math. The primary cause appears to be when fairly complex inline math expressions are split across line boundaries. The following file excerpt exhibits this problem: --- \documentclass[a4paper,10pt]{article} \usepackage{a4wide} \usepackage{amsmath} \begin{document} Basilevsky~\cite{Basilevsky83}, for example, provides such a matrix, which is closely related to the DFT matrix. A second approach, suggested by Johnson~\cite{Johnson94}, is to require conjugate symmetry in the random draws of ${\bf y}$, so $y_k = (y_{2 M + 2 - k})^*$. In that case the transformed data ${\bf z}={\bf F}^\dagger {\bf y}$ will be real, as required. The procedure outlined here may be extended to the generation of 2-D random samples without any further conceptual complications. In this case the covariance of the required data will be doubly block Toeplitz, which can be embedded into a doubly circulant matrix in much the same way as for the 1-D case. The 2-D DFT can be used to efficiently diagonalise the resulting covariance, and samples generated accordingly. \end{document} --- The problem can be reproduced by performing the following steps: 1. Create a file on disk (sg.tex in my case) containing the above code snippet. 2. Load the file using "xemacs -vanilla sg.tex". 3. Enable font-locking "M-x font-lock-mode" 4. The buffer fontifies correctly. However, if anything in the second paragraph is changed, the line switches to fontification as if in math mode. For example, adding a space somewhere in the centre of the third line of the second paragraph switches the entire line to math fontification. The problem also occurs when using Peter Galbraith's font-latex.el package. This seems to suggest a more general font-locking bug. The problem does not occur with GNU emacs. --=-=-=--