On Sun, Mar 8, 2015 at 6:50 AM, Dirk Laurie <[email protected]> wrote:
> It says: "Run the luaTeX typesetter on TEXNAME, usually creating > TEXNAME.pdf. ... If called as texlua it acts as lua interpreter." > > There are two gotchas here. > > 1. The clear implication is that if called as `luatex` it treats the input > file as TeX. In fact, if the filename ends in `.lua`, it treats it as Lua. > This is good, it is what the user would expect — but the manpage > does not say so. > > 2. Input cannot come from STDIN, a filename must be given, > but `texlua myfile.lua` does not give the same behaviour as > `lua myfile.lua` would have. The difference is subtle. > > The behaviour is as if the lines of `myfile.lua` were entered one > by one in interactive mode. That is to say, the scope of all local > variables ends at the end of the line they are defined on, unless > they are inside an unclosed block. This causes a perfectly valid > Lua script to fail when run by texlua. > > I cannot think of a good reason why this behaviour might be > desirable. It IHMO would be preferable to make `texlua` behave > exactly like `lua`. But if there is some subtle reason for it, maybe > the manpage could warn the user about it. > > hm, could be...can you give an example ? -- luigi
