Hi,
Some time ago, there was a discussion here about how to handle legacy (8-bits)
encodings for old documents with a reasonable amount of changes to the macro
packages, esp. in the LaTeX world. Since then, experiments have been done (the
lua-inputenc package), and I keep thinking that the only way to a sound solution
would be to have a process_output_buffer callback in LuaTeX.
(Intended use: process_input_buffer converts the input encoding to its bytes
representation, internally everything is bytes as it was with pdfTeX,
process_output_buffer outputs bytes as themselves and not their UTF8
representation, so that when the auxiliary file is opened again, it can been
read correctly. This could be done in lua-inputenc or later inputenc itself
without affecting other LaTeX macro packages.)
I can't find in the archives whether it has already been discussed on the list
(at least not with this name), but here are some arguments and
counter-arguments:
- admittedly, I can't imagine any other use of this callback than for
compatibility with old documents and "old" macro packages, and I know supporting
them is not your priority;
- but adding such a callback costs essentially nothing: the interface keeps
clean (it is symmetric to process_input_buffer) and a possible implementation
doesn't bloat the code at all;
- it keeps in line with LuaTeX philosophy of providing the tools and letting the
macro package authors build whatever they want with it.
In order to help saving your time, I attach a proposed implementation. I ran
quite a few tests with it, and the results look ok.
Please consider including this new callback in LuaTeX.
Thanks,
Manuel.
diff --git a/manual/luatexref-t.tex b/manual/luatexref-t.tex
index b1923da..e3fba27 100644
--- a/manual/luatexref-t.tex
+++ b/manual/luatexref-t.tex
@@ -3536,6 +3536,26 @@ that.
This callback does not replace any internal code.
+\subsubsection{\luatex{process_output_buffer}}
+
+This callback allows you to change the contents of the line output
+buffer just before \LUATEX\ actually starts writing it to a file as the
+result of a \tex{write} command. It is only called for output to an
+actual file (that is, excluding the log, the terminal, and \tex{write18}
+calls).
+
+\startfunctioncall
+function(<string> buffer)
+ return <string> adjusted_buffer
+end
+\stopfunctioncall
+
+If you return \type{nil}, \LUATEX\ will pretend like your callback
+never happened. You can gain a small amount of processing time from
+that.
+
+This callback does not replace any internal code.
+
\subsubsection{\luatex{token_filter}}
This callback allows you to replace the way \LUATEX\ fetches
diff --git a/source/texk/web2c/luatexdir/dvi/dvigen.c b/source/texk/web2c/luatexdir/dvi/dvigen.c
index bf33cf6..d9e3583 100644
--- a/source/texk/web2c/luatexdir/dvi/dvigen.c
+++ b/source/texk/web2c/luatexdir/dvi/dvigen.c
@@ -2126,6 +2126,9 @@ void write_out(halfword p)
integer d; /* number of characters in incomplete current string */
boolean clobbered; /* system string is ok? */
integer ret; /* return value from |runsystem| */
+ char *s, *ss; /* line to be written, as a C string */
+ int callback_id;
+ int lua_retval;
expand_macros_in_tokenlist(p);
old_setting = selector;
j = write_stream(p);
@@ -2138,7 +2141,17 @@ void write_out(halfword p)
selector = log_only;
tprint_nl("");
}
- token_show(def_ref);
+ s = tokenlist_to_cstring(def_ref, false, NULL);
+ if (selector < no_print) { /* selector is a file */
+ /* fix up the output buffer using callbacks */
+ callback_id = callback_defined(process_output_buffer_callback);
+ if (callback_id > 0) {
+ lua_retval = run_callback(callback_id, "S->S", s, &ss);
+ if ((lua_retval == true) && (ss != NULL))
+ s = ss;
+ }
+ }
+ tprint(s);
print_ln();
flush_list(def_ref);
if (j == 18) {
diff --git a/source/texk/web2c/luatexdir/lua/lcallbacklib.c b/source/texk/web2c/luatexdir/lua/lcallbacklib.c
index d593cc7..70ce409 100644
--- a/source/texk/web2c/luatexdir/lua/lcallbacklib.c
+++ b/source/texk/web2c/luatexdir/lua/lcallbacklib.c
@@ -50,7 +50,7 @@ static const char *const callbacknames[] = {
"find_sfd_file", "read_sfd_file",
"find_pk_file", "read_pk_file",
"show_error_hook",
- "process_input_buffer",
+ "process_input_buffer", "process_output_buffer",
"start_page_number", "stop_page_number",
"start_run", "stop_run",
"define_font",
diff --git a/source/texk/web2c/luatexdir/ptexlib.h b/source/texk/web2c/luatexdir/ptexlib.h
index e856266..0f1b040 100644
--- a/source/texk/web2c/luatexdir/ptexlib.h
+++ b/source/texk/web2c/luatexdir/ptexlib.h
@@ -325,7 +325,7 @@ typedef enum {
find_sfd_file_callback, read_sfd_file_callback,
find_pk_file_callback, read_pk_file_callback,
show_error_hook_callback,
- process_input_buffer_callback,
+ process_input_buffer_callback, process_output_buffer_callback,
start_page_number_callback, stop_page_number_callback,
start_run_callback, stop_run_callback,
define_font_callback,
