This morning I got an acceptably tagged text file out of MS Word. From that moment on, things got much easier.
I made a perl script to remove end tags, and instead put start tags on all lines between a start and end. It also made sure there were no interlinking tag sets. It also put all the start tags in the same format and easily parsable. I hadn't thought to do that when converting out of MS Word -- I had bigger fish to fry at the time. I hadn't marked Normal paragraphs, so my program had to deduce which lines weren't marked already, and put a b_pstyle_normal::: start tag on them. Armed with proper start tags on every line (which is actually a paragraph), it was pretty easy to pipe that through something that added the \begin_layout Whatever and \end_layout commands. At this point I have NOT removed the start or end tags -- I want some redundancy for checking. I also added a little C program to get rid of the '\015' characters that DOS put in. I made a layout with dummy styles for each style I used (sort -u came in very handy for this). Anyway, my program can make the body of a LyX file, and all the Part/Chapter/Section etc works perfectly, and it seems like all the other paragraph styles are working. It's basically a pipeline of little filters creating a LyX file from the text file, and I can do it over and over to my heart's content. I imagine tomorrow I'll add the code to handle character styles, and start making my layout file create effects that look how they're supposed to. That will help in looking at the produced PDF (it already produces a PDF, so the basic code is correct). Bottom line, I now have a text file with tags representing all my document's original style, and I've created perl, awk, sed and C code to convert it to a LyX document with my styles preserved. Anyway, thanks for all the help. SteveT Steve Litt Recession Relief Package http://www.recession-relief.US
