-/** The .tex to .lyx converter
- \author André Pönitz (2003)
+/**
+ * \file tex2lyx.C
+ * This file is part of LyX, the document processor.
+ * Licence details can be found in the file COPYING.
+ *
+ * \author André Pönitz
+ *
+ * Full author contact details are available in file CREDITS.
*/
// {[(
#include <config.h>
-#include <algorithm>
+#include "tex2lyx.h"
+#include "context.h"
+
+#include "debug.h"
+#include "lyxtextclass.h"
+
+#include "support/convert.h"
+#include "support/filetools.h"
+#include "support/fs_extras.h"
+#include "support/lstrings.h"
+#include "support/lyxlib.h"
+#include "support/os.h"
+#include "support/package.h"
+#include "support/unicode.h"
+
+#include <boost/function.hpp>
+#include <boost/filesystem/operations.hpp>
+#include <boost/filesystem/path.hpp>
+
#include <cctype>
#include <fstream>
#include <iostream>
-#include <map>
-#include <stack>
#include <string>
+#include <sstream>
#include <vector>
+#include <map>
-#include "Lsstream.h"
-#include "texparser.h"
+namespace lyx {
-using std::count_if;
+using std::endl;
using std::cout;
using std::cerr;
-using std::endl;
-using std::fill;
using std::getline;
-using std::ios;
+
using std::ifstream;
-using std::istream;
+using std::ofstream;
using std::istringstream;
-using std::map;
-using std::ostream;
using std::ostringstream;
-using std::stack;
+using std::stringstream;
using std::string;
using std::vector;
+using std::map;
+using support::changeExtension;
+using support::FileName;
+using support::isStrUnsignedInt;
+using support::ltrim;
+using support::makeAbsPath;
+using support::onlyPath;
+using support::rtrim;
+using support::isFileReadable;
-namespace {
-
-void parse(Parser & p, ostream & os, unsigned flags, const mode_type mode);
-
-char const OPEN = '<';
-char const CLOSE = '>';
-
-// rather brutish way to code table structure in a string:
-//
-// \begin{tabular}{ccc}
-// 1 & 2 & 3\\
-// \multicolumn{2}{c}{4} & 5\\
-// 6 & 7 \\
-// \end{tabular}
-//
-// gets "translated" to:
-//
-// 1 TAB 2 TAB 3 LINE
-// 2 MULT c MULT 4 TAB 5 LINE
-// 5 TAB 7 LINE
-
-char const TAB = '\001';
-char const LINE = '\002';
-char const MULT = '\003';
-
-const char * known_languages[] = { "austrian", "babel", "bahasa", "basque",
-"breton", "british", "bulgarian", "catalan", "croatian", "czech", "danish",
-"dutch", "english", "esperanto", "estonian", "finnish", "francais",
-"frenchb", "galician", "german", "germanb", "greek", "hebcal", "hebfont",
-"hebrew", "hebrew_newcode", "hebrew_oldcode", "hebrew_p", "hyphen",
-"icelandic", "irish", "italian", "latin", "lgrcmr", "lgrcmro", "lgrcmss",
-"lgrcmtt", "lgrenc", "lgrlcmss", "lgrlcmtt", "lheclas", "lhecmr",
-"lhecmss", "lhecmtt", "lhecrml", "lheenc", "lhefr", "lheredis", "lheshold",
-"lheshscr", "lheshstk", "lsorbian", "magyar", "naustrian", "ngermanb",
-"ngerman", "norsk", "polish", "portuges", "rlbabel", "romanian",
-"russianb", "samin", "scottish", "serbian", "slovak", "slovene", "spanish",
-"swedish", "turkish", "ukraineb", "usorbian", "welsh", 0};
-
-char const * known_fontsizes[] = { "10pt", "11pt", "12pt", 0 };
-
-char const * known_headings[] = { "caption", "title", "author", "date",
-"paragraph", "chapter", "section", "subsection", "subsubsection", 0 };
-
-char const * known_math_envs[] = { "equation", "equation*",
-"eqnarray", "eqnarray*", "align", "align*", 0};
-
-char const * known_latex_commands[] = { "ref", "cite", "label", "index",
-"printindex", 0 };
-
-// LaTeX names for quotes
-char const * known_quotes[] = { "glqq", "grqq", 0};
-
-// the same as known_quotes with .lyx names
-char const * known_coded_quotes[] = { "gld", "grd", 0};
-
-
-
-// some ugly stuff
-ostringstream h_preamble;
-string h_textclass = "article";
-string h_options = "";
-string h_language = "english";
-string h_inputencoding = "latin1";
-string h_fontscheme = "default";
-string h_graphics = "default";
-string h_paperfontsize = "default";
-string h_spacing = "single";
-string h_papersize = "default";
-string h_paperpackage = "default";
-string h_use_geometry = "0";
-string h_use_amsmath = "0";
-string h_use_natbib = "0";
-string h_use_numerical_citations = "0";
-string h_paperorientation = "portrait";
-string h_secnumdepth = "3";
-string h_tocdepth = "3";
-string h_paragraph_separation = "indent";
-string h_defskip = "medskip";
-string h_quotes_language = "english";
-string h_quotes_times = "1";
-string h_papercolumns = "1";
-string h_papersides = "1";
-string h_paperpagestyle = "default";
-string h_tracking_changes = "0";
-
-// indicates whether we are in the preamble
-bool in_preamble = true;
-
-// current stack of nested environments
-stack<string> active_environments;
+namespace fs = boost::filesystem;
-string cap(string s)
+IconvProcessor & utf8ToUcs4()
{
- if (s.size())
- s[0] = toupper(s[0]);
- return s;
+ static IconvProcessor iconv(ucs4_codeset, "UTF-8");
+ return iconv;
}
-string const trim(string const & a, char const * p = " \t\n\r")
+// Hacks to allow the thing to link in the lyxlayout stuff
+LyXErr lyxerr(std::cerr.rdbuf());
+
+
+string const trim(string const & a, char const * p)
{
- // lyx::Assert(p);
+ // BOOST_ASSERT(p);
if (a.empty() || !*p)
return a;
}
-void split(string const & s, vector<string> & result, char delim = ',')
+void split(string const & s, vector<string> & result, char delim)
{
- istringstream is(s);
+ //cerr << "split 1: '" << s << "'\n";
+ istringstream is(s);
string t;
while (getline(is, t, delim))
result.push_back(t);
+ //cerr << "split 2\n";
}
-// splits "x=z, y=b" into a map
-map<string, string> split_map(string const & s)
-{
- map<string, string> res;
- vector<string> v;
- split(s, v);
- for (size_t i = 0; i < v.size(); ++i) {
- size_t const pos = v[i].find('=');
- string const index = v[i].substr(0, pos);
- string const value = v[i].substr(pos + 1, string::npos);
- res[trim(index)] = trim(value);
- }
- return res;
-}
-
-
-string join(vector<string> const & input, char delim)
+string join(vector<string> const & input, char const * delim)
{
ostringstream os;
- for (size_t i = 0; i != input.size(); ++i) {
+ for (size_t i = 0; i < input.size(); ++i) {
if (i)
- os << delim;
- os << input[i];
+ os << delim;
+ os << input[i];
}
return os.str();
}
-char const ** is_known(string const & str, char const ** what)
+char const * const * is_known(string const & str, char const * const * what)
{
for ( ; *what; ++what)
if (str == *what)
}
-void handle_opt(vector<string> & opts, char const ** what, string & target)
-{
- if (opts.empty())
- return;
-
- for ( ; *what; ++what) {
- vector<string>::iterator it = find(opts.begin(), opts.end(), *what);
- if (it != opts.end()) {
- //cerr << "### found option '" << *what << "'\n";
- target = *what;
- opts.erase(it);
- return;
- }
- }
-}
-
-
-bool is_math_env(string const & name)
-{
- for (char const ** what = known_math_envs; *what; ++what)
- if (*what == name)
- return true;
- return false;
-}
-
-void begin_inset(ostream & os, string const & name)
-{
- os << "\n\\begin_inset " << name;
-}
+// current stack of nested environments
+vector<string> active_environments;
-void end_inset(ostream & os)
+string active_environment()
{
- os << "\n\\end_inset\n\n";
+ return active_environments.empty() ? string() : active_environments.back();
}
-string curr_env()
-{
- return active_environments.empty() ? string() : active_environments.top();
-}
+CommandMap known_commands;
+CommandMap known_environments;
+CommandMap known_math_environments;
-void handle_ert(ostream & os, string const & s)
+void add_known_command(string const & command, string const & o1,
+ bool o2)
{
- begin_inset(os, "ERT");
- os << "\nstatus Collapsed\n\n\\layout Standard\n\n";
- for (string::const_iterator it = s.begin(), et = s.end(); it != et; ++it) {
- if (*it == '\\')
- os << "\n\\backslash\n";
- else
- os << *it;
+ // We have to handle the following cases:
+ // definition o1 o2 invocation result
+ // \newcommand{\foo}{bar} "" false \foo bar
+ // \newcommand{\foo}[1]{bar #1} "[1]" false \foo{x} bar x
+ // \newcommand{\foo}[1][]{bar #1} "[1]" true \foo bar
+ // \newcommand{\foo}[1][]{bar #1} "[1]" true \foo[x] bar x
+ // \newcommand{\foo}[1][x]{bar #1} "[1]" true \foo[x] bar x
+ unsigned int nargs = 0;
+ vector<ArgumentType> arguments;
+ string const opt1 = rtrim(ltrim(o1, "["), "]");
+ if (isStrUnsignedInt(opt1)) {
+ // The command has arguments
+ nargs = convert<unsigned int>(opt1);
+ if (nargs > 0 && o2) {
+ // The first argument is optional
+ arguments.push_back(optional);
+ --nargs;
+ }
}
- end_inset(os);
+ for (unsigned int i = 0; i < nargs; ++i)
+ arguments.push_back(required);
+ known_commands[command] = arguments;
}
-void handle_tex(ostream & os, string const & s)
-{
- //os << "handle_tex(" << s << ")\n";
- if (in_preamble)
- h_preamble << s;
- else
- handle_ert(os, s);
-}
-
-
-void handle_par(ostream & os)
-{
- if (active_environments.empty())
- return;
- os << "\n\\layout ";
- string s = curr_env();
- if (s == "document" || s == "table") {
- os << "Standard\n\n";
- return;
- }
- if (s == "lyxcode") {
- os << "LyX-Code\n\n";
- return;
- }
- os << cap(s) << "\n\n";
-}
+bool noweb_mode = false;
-void handle_package(string const & name, string const & options)
-{
- //cerr << "handle_package: '" << name << "'\n";
- if (name == "a4wide") {
- h_papersize = "a4paper";
- h_paperpackage = "widemarginsa4";
- } else if (name == "ae")
- h_fontscheme = "ae";
- else if (name == "aecompl")
- h_fontscheme = "ae";
- else if (name == "amsmath")
- h_use_amsmath = "1";
- else if (name == "amssymb")
- h_use_amsmath = "1";
- else if (name == "babel")
- ; // ignore this
- else if (name == "fontenc")
- ; // ignore this
- else if (name == "inputenc")
- h_inputencoding = options;
- else if (name == "makeidx")
- ; // ignore this
- else if (name == "verbatim")
- ; // ignore this
- else if (is_known(name, known_languages)) {
- h_language = name;
- h_quotes_language = name;
- } else {
- if (options.size())
- h_preamble << "\\usepackage[" << options << "]{" << name << "}\n";
- else
- h_preamble << "\\usepackage{" << name << "}\n";
- }
-}
+namespace {
-vector<string> extract_col_align(string const & s)
-{
- vector<string> res;
- for (size_t i = 0; i < s.size(); ++i) {
- switch (s[i]) {
- case 'c': res.push_back("center"); break;
- case 'l': res.push_back("left"); break;
- case 'r': res.push_back("right"); break;
- default : res.push_back("right"); break;
+/*!
+ * Read one command definition from the syntax file
+ */
+void read_command(Parser & p, string command, CommandMap & commands) {
+ if (p.next_token().asInput() == "*") {
+ p.get_token();
+ command += '*';
+ }
+ vector<ArgumentType> arguments;
+ while (p.next_token().cat() == catBegin ||
+ p.next_token().asInput() == "[") {
+ if (p.next_token().cat() == catBegin) {
+ string const arg = p.getArg('{', '}');
+ if (arg == "translate")
+ arguments.push_back(required);
+ else
+ arguments.push_back(verbatim);
+ } else {
+ p.getArg('[', ']');
+ arguments.push_back(optional);
}
}
- return res;
+ commands[command] = arguments;
}
-void handle_tabular(Parser & p, ostream & os, mode_type mode)
+/*!
+ * Read a class of environments from the syntax file
+ */
+void read_environment(Parser & p, string const & begin,
+ CommandMap & environments)
{
- begin_inset(os, "Tabular \n");
- string colopts = p.verbatimItem();
- vector<string> colalign = extract_col_align(colopts);
- ostringstream ss;
- parse(p, ss, FLAG_END, mode);
- vector<string> lines;
- split(ss.str(), lines, LINE);
- const size_t cols = colalign.size();
- const size_t rows = lines.size();
- os << "<lyxtabular version=\"3\" rows=\"" << rows
- << "\" columns=\"" << cols << "\">\n"
- << "<features>\n";
- for (size_t c = 0; c < cols; ++c)
- os << "<column alignment=\"" << colalign[c] << "\""
- << " valignment=\"top\""
- << " width=\"0pt\""
- << ">\n";
- for (size_t r = 0; r < rows; ++r) {
- vector<string> cells;
- split(lines[r], cells, TAB);
- while (cells.size() < cols)
- cells.push_back(string());
- //os << "<row bottomline=\"true\">\n";
- os << "<row>\n";
- for (size_t c = 0; c < cols; ++c) {
- os << "<cell";
- string alignment = "center";
- vector<string> parts;
- split(cells[c], parts, MULT);
- if (parts.size() > 2) {
- os << " multicolumn=\"" << parts[0] << "\"";
- alignment = parts[1];
- }
- os << " alignment=\"" << alignment << "\""
- << " valignment=\"top\""
- << " topline=\"true\""
- << " leftline=\"true\""
- << " usebox=\"none\""
- << ">";
- begin_inset(os, "Text");
- os << "\n\n\\layout Standard\n\n";
- if (parts.size())
- os << parts.back();
- end_inset(os);
- os << "</cell>\n";
+ string environment;
+ while (p.good()) {
+ Token const & t = p.get_token();
+ if (t.cat() == catLetter)
+ environment += t.asInput();
+ else if (!environment.empty()) {
+ p.putback();
+ read_command(p, environment, environments);
+ environment.erase();
+ }
+ if (t.cat() == catEscape && t.asInput() == "\\end") {
+ string const end = p.getArg('{', '}');
+ if (end == begin)
+ return;
}
- os << "</row>\n";
}
- os << "</lyxtabular>\n";
- end_inset(os);
}
-string wrap(string const & cmd, string const & str)
-{
- return OPEN + cmd + ' ' + str + CLOSE;
-}
-
-
-void end_preamble(ostream & os)
-{
- in_preamble = false;
- os << "# tex2lyx 0.0.2 created this file\n"
- << "\\lyxformat 222\n"
- << "\\textclass " << h_textclass << "\n"
- << "\\begin_preamble\n" << h_preamble.str() << "\n\\end_preamble\n"
- << "\\options " << h_options << "\n"
- << "\\language " << h_language << "\n"
- << "\\inputencoding " << h_inputencoding << "\n"
- << "\\fontscheme " << h_fontscheme << "\n"
- << "\\graphics " << h_graphics << "\n"
- << "\\paperfontsize " << h_paperfontsize << "\n"
- << "\\spacing " << h_spacing << "\n"
- << "\\papersize " << h_papersize << "\n"
- << "\\paperpackage " << h_paperpackage << "\n"
- << "\\use_geometry " << h_use_geometry << "\n"
- << "\\use_amsmath " << h_use_amsmath << "\n"
- << "\\use_natbib " << h_use_natbib << "\n"
- << "\\use_numerical_citations " << h_use_numerical_citations << "\n"
- << "\\paperorientation " << h_paperorientation << "\n"
- << "\\secnumdepth " << h_secnumdepth << "\n"
- << "\\tocdepth " << h_tocdepth << "\n"
- << "\\paragraph_separation " << h_paragraph_separation << "\n"
- << "\\defskip " << h_defskip << "\n"
- << "\\quotes_language " << h_quotes_language << "\n"
- << "\\quotes_times " << h_quotes_times << "\n"
- << "\\papercolumns " << h_papercolumns << "\n"
- << "\\papersides " << h_papersides << "\n"
- << "\\paperpagestyle " << h_paperpagestyle << "\n"
- << "\\tracking_changes " << h_tracking_changes << "\n";
-}
-
-
-void parse(Parser & p, ostream & os, unsigned flags, const mode_type mode)
+/*!
+ * Read a list of TeX commands from a reLyX compatible syntax file.
+ * Since this list is used after all commands that have a LyX counterpart
+ * are handled, it does not matter that the "syntax.default" file
+ * has almost all of them listed. For the same reason the reLyX-specific
+ * reLyXre environment is ignored.
+ */
+void read_syntaxfile(FileName const & file_name)
{
+ ifstream is(file_name.toFilesystemEncoding().c_str());
+ if (!is.good()) {
+ cerr << "Could not open syntax file \"" << file_name
+ << "\" for reading." << endl;
+ exit(2);
+ }
+ // We can use our TeX parser, since the syntax of the layout file is
+ // modeled after TeX.
+ // Unknown tokens are just silently ignored, this helps us to skip some
+ // reLyX specific things.
+ Parser p(is);
while (p.good()) {
- Token const & t = p.getToken();
-
-#ifdef FILEDEBUG
- cerr << "t: " << t << " flags: " << flags << "\n";
- //cell->dump();
-#endif
-
- if (flags & FLAG_ITEM) {
- if (t.cat() == catSpace)
- continue;
-
- flags &= ~FLAG_ITEM;
- if (t.cat() == catBegin) {
- // skip the brace and collect everything to the next matching
- // closing brace
- flags |= FLAG_BRACE_LAST;
- continue;
- }
-
- // handle only this single token, leave the loop if done
- flags |= FLAG_LEAVE;
- }
-
-
- if (flags & FLAG_BRACED) {
- if (t.cat() == catSpace)
- continue;
-
- if (t.cat() != catBegin) {
- p.error("opening brace expected");
- return;
- }
-
- // skip the brace and collect everything to the next matching
- // closing brace
- flags = FLAG_BRACE_LAST;
- }
-
-
- if (flags & FLAG_OPTION) {
- if (t.cat() == catOther && t.character() == '[') {
- parse(p, os, FLAG_BRACK_LAST, mode);
+ Token const & t = p.get_token();
+ if (t.cat() == catEscape) {
+ string const command = t.asInput();
+ if (command == "\\begin") {
+ string const name = p.getArg('{', '}');
+ if (name == "environments" || name == "reLyXre")
+ // We understand "reLyXre", but it is
+ // not as powerful as "environments".
+ read_environment(p, name,
+ known_environments);
+ else if (name == "mathenvironments")
+ read_environment(p, name,
+ known_math_environments);
} else {
- // no option found, put back token and we are done
- p.putback();
- }
- return;
- }
-
- //
- // cat codes
- //
- if (t.cat() == catMath) {
- if (mode == TEXT_MODE || mode == MATHTEXT_MODE) {
- // we are inside some text mode thingy, so opening new math is allowed
- if (mode == TEXT_MODE)
- begin_inset(os, "Formula ");
- Token const & n = p.getToken();
- if (n.cat() == catMath) {
- // TeX's $$...$$ syntax for displayed math
- os << "\\[";
- parse(p, os, FLAG_SIMPLE, MATH_MODE);
- os << "\\]";
- p.getToken(); // skip the second '$' token
- } else {
- // simple $...$ stuff
- p.putback();
- os << '$';
- parse(p, os, FLAG_SIMPLE, MATH_MODE);
- os << '$';
- }
- if (mode == TEXT_MODE)
- end_inset(os);
- }
-
- else if (flags & FLAG_SIMPLE) {
- // this is the end of the formula
- return;
- }
-
- else {
- cerr << "mode: " << mode << endl;
- p.error("something strange in the parser\n");
- break;
- }
- }
-
- else if (t.cat() == catLetter)
- os << t.character();
-
- else if (t.cat() == catSpace)
- os << t.character();
-
- else if (t.cat() == catNewline)
- os << ' ';
-
- else if (t.cat() == catSuper)
- os << t.character();
-
- else if (t.cat() == catSub)
- os << t.character();
-
- else if (t.cat() == catParameter) {
- Token const & n = p.getToken();
- os << wrap("macroarg", string(1, n.character()));
- }
-
- else if (t.cat() == catActive) {
- if (t.character() == '~')
- os << (curr_env() == "lyxcode" ? ' ' : '~');
- else
- os << t.asInput();
- }
-
- else if (t.cat() == catBegin) {
- if (mode == TEXT_MODE)
- handle_tex(os, "{");
- else
- os << '{';
- }
-
- else if (t.cat() == catEnd) {
- if (flags & FLAG_BRACE_LAST)
- return;
- if (mode == TEXT_MODE)
- handle_tex(os, "}");
- else
- os << '}';
- }
-
- else if (t.cat() == catAlign) {
- if (mode == TEXT_MODE)
- os << TAB;
- else
- os << t.character();
- }
-
- else if (t.cs() == "tabularnewline") {
- if (mode == TEXT_MODE)
- os << LINE;
- else
- os << t.asInput();
- }
-
- else if (t.cs() == "\\" && mode == MATH_MODE)
- os << t.asInput();
-
- else if (t.cs() == "\\" && mode == TEXT_MODE && curr_env() == "tabular")
- os << LINE;
-
- else if (t.character() == ']' && (flags & FLAG_BRACK_LAST)) {
- //cerr << "finished reading option\n";
- return;
- }
-
- else if (t.cat() == catOther)
- os << string(1, t.character());
-
- else if (t.cat() == catComment) {
- string s;
- while (p.good()) {
- Token const & t = p.getToken();
- if (t.cat() == catNewline)
- break;
- s += t.asString();
- }
- //os << wrap("comment", s);
- p.skipSpaces();
- }
-
- //
- // control sequences
- //
-
- else if (t.cs() == "lyxlock")
- ; // ignored
-
- else if (t.cs() == "makeatletter") {
- p.setCatCode('@', catLetter);
- handle_tex(os, "\\makeatletter\n");
- }
-
- else if (t.cs() == "makeatother") {
- p.setCatCode('@', catOther);
- handle_tex(os, "\\makeatother\n");
- }
-
- else if (t.cs() == "newcommand" || t.cs() == "renewcommand"
- || t.cs() == "providecommand") {
- string const name = p.verbatimItem();
- string const opts = p.getOpt();
- string const body = p.verbatimItem();
- // only non-lyxspecific stuff
- if (name != "\\noun " && name != "\\tabularnewline ") {
- ostringstream ss;
- ss << '\\' << t.cs() << '{' << name << '}'
- << opts << '{' << body << "}\n";
- handle_tex(os, ss.str());
-/*
- ostream & out = in_preamble ? h_preamble : os;
- if (!in_preamble)
- begin_inset(os, "FormulaMacro\n");
- out << "\\" << t.cs() << "{" << name << "}"
- << opts << "{" << body << "}\n";
- if (!in_preamble)
- end_inset(os);
-*/
+ read_command(p, command, known_commands);
}
}
+ }
+}
- else if (t.cs() == "newtheorem") {
- ostringstream ss;
- ss << "\\newtheorem";
- ss << '{' << p.verbatimItem() << '}';
- ss << p.getOpt();
- ss << '{' << p.verbatimItem() << '}';
- ss << p.getOpt();
- ss << '\n';
- handle_tex(os, ss.str());
- }
- else if (t.cs() == "(") {
- begin_inset(os, "Formula");
- os << " \\(";
- parse(p, os, FLAG_SIMPLE2, MATH_MODE);
- os << "\\)";
- end_inset(os);
- }
+string documentclass;
+string syntaxfile;
+bool overwrite_files = false;
- else if (t.cs() == "[") {
- begin_inset(os, "Formula");
- os << " \\[";
- parse(p, os, FLAG_EQUATION, MATH_MODE);
- os << "\\]";
- end_inset(os);
- }
- else if (t.cs() == "protect")
- // ignore \\protect, will hopefully be re-added during output
- ;
-
- else if (t.cs() == "begin") {
- string const name = p.getArg('{', '}');
- active_environments.push(name);
- if (name == "document") {
- end_preamble(os);
- os << "\n\n\\layout Standard\n\n";
- parse(p, os, FLAG_END, mode);
- } else if (name == "abstract") {
- handle_par(os);
- parse(p, os, FLAG_END, mode);
- } else if (is_math_env(name)) {
- begin_inset(os, "Formula ");
- os << "\\begin{" << name << "}";
- parse(p, os, FLAG_END, MATH_MODE);
- os << "\\end{" << name << "}";
- end_inset(os);
- } else if (name == "tabular") {
- if (mode == TEXT_MODE)
- handle_tabular(p, os, mode);
- else {
- os << "\\begin{" << name << "}";
- parse(p, os, FLAG_END, MATHTEXT_MODE);
- os << "\\end{" << name << "}";
- }
- } else if (name == "table" || name == "figure") {
- string opts = p.getOpt();
- begin_inset(os, "Float " + name + "\n");
- if (opts.size())
- os << "placement " << opts << '\n';
- os << "wide false\n"
- << "collapsed false\n"
- << "\n"
- << "\\layout Standard\n";
- parse(p, os, FLAG_END, mode);
- end_inset(os);
- } else if (name == "thebibliography") {
- p.verbatimItem(); // swallow next arg
- parse(p, os, FLAG_END, mode);
- os << "\n\\layout Standard\n\n";
- } else if (mode == MATH_MODE || mode == MATHTEXT_MODE) {
- os << "\\begin{" << name << "}";
- parse(p, os, FLAG_END, mode);
- os << "\\end{" << name << "}";
- } else {
- parse(p, os, FLAG_END, mode);
- }
- }
+/// return the number of arguments consumed
+typedef boost::function<int(string const &, string const &)> cmd_helper;
- else if (t.cs() == "end") {
- if (flags & FLAG_END) {
- // eat environment name
- string const name = p.getArg('{', '}');
- if (name != curr_env())
- p.error("\\end{" + name + "} does not match \\begin{"
- + curr_env() + "}");
- active_environments.pop();
- return;
- }
- p.error("found 'end' unexpectedly");
- }
+int parse_help(string const &, string const &)
+{
+ cerr << "Usage: tex2lyx [ command line switches ] <infile.tex> [<outfile.lyx>]\n"
+ "Command line switches (case sensitive):\n"
+ "\t-help summarize tex2lyx usage\n"
+ "\t-f Force creation of .lyx files even if they exist already\n"
+ "\t-userdir dir try to set user directory to dir\n"
+ "\t-sysdir dir try to set system directory to dir\n"
+ "\t-c textclass declare the textclass\n"
+ "\t-n translate a noweb (aka literate programming) file.\n"
+ "\t-s syntaxfile read additional syntax file" << endl;
+ exit(0);
+}
- else if (t.cs() == "item")
- handle_par(os);
- else if (t.cs() == ")") {
- if (flags & FLAG_SIMPLE2)
- return;
- p.error("found '\\)' unexpectedly");
- }
+int parse_class(string const & arg, string const &)
+{
+ if (arg.empty()) {
+ cerr << "Missing textclass string after -c switch" << endl;
+ exit(1);
+ }
+ documentclass = arg;
+ return 1;
+}
- else if (t.cs() == "]") {
- if (flags & FLAG_EQUATION)
- return;
- p.error("found '\\]' unexpectedly");
- }
- else if (t.cs() == "documentclass") {
- vector<string> opts;
- split(p.getArg('[', ']'), opts, ',');
- handle_opt(opts, known_languages, h_language);
- handle_opt(opts, known_fontsizes, h_paperfontsize);
- h_quotes_language = h_language;
- h_options = join(opts, ',');
- h_textclass = p.getArg('{', '}');
- }
+int parse_syntaxfile(string const & arg, string const &)
+{
+ if (arg.empty()) {
+ cerr << "Missing syntaxfile string after -s switch" << endl;
+ exit(1);
+ }
+ syntaxfile = arg;
+ return 1;
+}
- else if (t.cs() == "usepackage") {
- string const options = p.getArg('[', ']');
- string const name = p.getArg('{', '}');
- if (options.empty() && name.find(',')) {
- vector<string> vecnames;
- split(name, vecnames, ',');
- vector<string>::const_iterator it = vecnames.begin();
- vector<string>::const_iterator end = vecnames.end();
- for (; it != end; ++it)
- handle_package(trim(*it), string());
- } else {
- handle_package(name, options);
- }
- }
- else if (t.cs() == "newenvironment") {
- string const name = p.getArg('{', '}');
- if (name != "lyxcode") {
- ostringstream ss;
- ss << "\\newenvironment{" << name << "}";
- ss << p.getOpt();
- ss << p.getOpt();
- ss << '{' << p.verbatimItem() << '}';
- ss << '{' << p.verbatimItem() << '}';
- ss << '\n';
- handle_tex(os, ss.str());
- }
- }
+// Filled with the command line arguments "foo" of "-sysdir foo" or
+// "-userdir foo".
+string cl_system_support;
+string cl_user_support;
- else if (t.cs() == "newfont") {
- ostringstream ss;
- ss << "\\newfont";
- ss << '{' << p.verbatimItem() << '}';
- ss << '{' << p.verbatimItem() << '}';
- ss << '\n';
- handle_tex(os, ss.str());
- }
- else if (t.cs() == "newcounter") {
- ostringstream ss;
- ss << "\\newcounter";
- ss << '{' << p.verbatimItem() << '}';
- ss << p.getOpt();
- ss << '\n';
- handle_tex(os, ss.str());
- }
+int parse_sysdir(string const & arg, string const &)
+{
+ if (arg.empty()) {
+ cerr << "Missing directory for -sysdir switch" << endl;
+ exit(1);
+ }
+ cl_system_support = arg;
+ return 1;
+}
- else if (t.cs() == "def") {
- string name = p.getToken().cs();
- while (p.nextToken().cat() != catBegin)
- name += p.getToken().asString();
- handle_ert(os, "\\def\\" + name + '{' + p.verbatimItem() + '}');
- }
- else if (t.cs() == "setcounter") {
- string const name = p.getArg('{', '}');
- string const content = p.getArg('{', '}');
- if (name == "secnumdepth")
- h_secnumdepth = content;
- else if (name == "tocdepth")
- h_tocdepth = content;
- else
- h_preamble << "\\setcounter{" << name << "}{" << content << "}\n";
- }
+int parse_userdir(string const & arg, string const &)
+{
+ if (arg.empty()) {
+ cerr << "Missing directory for -userdir switch" << endl;
+ exit(1);
+ }
+ cl_user_support = arg;
+ return 1;
+}
- else if (t.cs() == "setlength") {
- string const name = p.verbatimItem();
- string const content = p.verbatimItem();
- if (in_preamble && name == "parskip")
- h_paragraph_separation = "skip";
- else if (in_preamble && name == "parindent")
- h_paragraph_separation = "skip";
- else
- handle_tex(os, "\\setlength{" + name + "}{" + content + "}\n");
- }
-
- else if (t.cs() == "par")
- handle_par(os);
-
- else if (is_known(t.cs(), known_headings)) {
- string name = t.cs();
- if (p.nextToken().asInput() == "*") {
- p.getToken();
- name += "*";
- }
- os << "\n\n\\layout " << cap(name) << "\n\n";
- parse(p, os, FLAG_ITEM, mode);
- os << "\n\n\\layout Standard\n\n";
- }
- else if (t.cs() == "includegraphics") {
- map<string, string> opts = split_map(p.getArg('[', ']'));
- string name = p.verbatimItem();
- begin_inset(os, "Graphics ");
- os << "\n\tfilename " << name << '\n';
- if (opts.find("width") != opts.end())
- os << "\twidth " << opts["width"] << '\n';
- if (opts.find("height") != opts.end())
- os << "\theight " << opts["height"] << '\n';
- end_inset(os);
- }
+int parse_force(string const &, string const &)
+{
+ overwrite_files = true;
+ return 0;
+}
- else if (t.cs() == "makeindex" || t.cs() == "maketitle")
- ; // swallow this
- else if (t.cs() == "tableofcontents")
- p.verbatimItem(); // swallow this
+int parse_noweb(string const &, string const &)
+{
+ noweb_mode = true;
+ return 0;
+}
- else if (t.cs() == "multicolumn" && mode == TEXT_MODE) {
- // brutish...
- parse(p, os, FLAG_ITEM, mode);
- os << MULT;
- parse(p, os, FLAG_ITEM, mode);
- os << MULT;
- parse(p, os, FLAG_ITEM, mode);
- }
- else if (t.cs() == "textrm") {
- os << '\\' << t.cs() << '{';
- parse(p, os, FLAG_ITEM, MATHTEXT_MODE);
- os << '}';
- }
+void easyParse(int & argc, char * argv[])
+{
+ map<string, cmd_helper> cmdmap;
+
+ cmdmap["-c"] = parse_class;
+ cmdmap["-f"] = parse_force;
+ cmdmap["-s"] = parse_syntaxfile;
+ cmdmap["-help"] = parse_help;
+ cmdmap["--help"] = parse_help;
+ cmdmap["-n"] = parse_noweb;
+ cmdmap["-sysdir"] = parse_sysdir;
+ cmdmap["-userdir"] = parse_userdir;
+
+ for (int i = 1; i < argc; ++i) {
+ std::map<string, cmd_helper>::const_iterator it
+ = cmdmap.find(argv[i]);
+
+ // don't complain if not found - may be parsed later
+ if (it == cmdmap.end())
+ continue;
+
+ string arg(to_utf8(from_local8bit((i + 1 < argc) ? argv[i + 1] : "")));
+ string arg2(to_utf8(from_local8bit((i + 2 < argc) ? argv[i + 2] : "")));
+
+ int const remove = 1 + it->second(arg, arg2);
+
+ // Now, remove used arguments by shifting
+ // the following ones remove places down.
+ argc -= remove;
+ for (int j = i; j < argc; ++j)
+ argv[j] = argv[j + remove];
+ --i;
+ }
+}
- else if ((t.cs() == "emph" || t.cs() == "noun") && mode == TEXT_MODE) {
- os << "\n\\" << t.cs() << " on\n";
- parse(p, os, FLAG_ITEM, mode);
- os << "\n\\" << t.cs() << " default\n";
- }
- else if (is_known(t.cs(), known_latex_commands) && mode == TEXT_MODE) {
- begin_inset(os, "LatexCommand ");
- os << '\\' << t.cs() << '{';
- parse(p, os, FLAG_ITEM, TEXT_MODE);
- os << '}';
- end_inset(os);
- }
+// path of the first parsed file
+string masterFilePath;
+// path of the currently parsed file
+string parentFilePath;
- else if (t.cs() == "bibitem") {
- os << "\n\\layout Bibliography\n\\bibitem ";
- string opt = p.getArg('[',']');
- if (opt.size())
- os << '[' << opt << ']';
- os << '{' << p.getArg('{','}') << '}' << "\n\n";
- }
+} // anonymous namespace
- else if (char const ** where = is_known(t.cs(), known_quotes)) {
- begin_inset(os, "Quotes ");
- os << known_coded_quotes[where - known_quotes];
- end_inset(os);
- }
- else if (t.cs() == "textasciitilde")
- os << '~';
+string getMasterFilePath()
+{
+ return masterFilePath;
+}
- else if (t.cs() == "_" && mode == TEXT_MODE)
- os << '_';
+string getParentFilePath()
+{
+ return parentFilePath;
+}
- else if (t.cs() == "&" && mode == TEXT_MODE)
- os << '&';
- else if (t.cs() == "\"") {
- string const name = p.verbatimItem();
- if (name == "a") os << 'ä';
- else if (name == "o") os << 'ö';
- else if (name == "u") os << 'ü';
- else if (name == "A") os << 'Ä';
- else if (name == "O") os << 'Ö';
- else if (name == "U") os << 'Ü';
- else handle_ert(os, "\"{" + name + "}");
- }
+namespace {
- else if (t.cs() == "ss")
- os << "ß";
+/*!
+ * Reads tex input from \a is and writes lyx output to \a os.
+ * Uses some common settings for the preamble, so this should only
+ * be used more than once for included documents.
+ * Caution: Overwrites the existing preamble settings if the new document
+ * contains a preamble.
+ * You must ensure that \p parentFilePath is properly set before calling
+ * this function!
+ */
+void tex2lyx(std::istream &is, std::ostream &os)
+{
+ Parser p(is);
+ //p.dump();
+
+ stringstream ss;
+ LyXTextClass textclass = parse_preamble(p, ss, documentclass);
+
+ active_environments.push_back("document");
+ Context context(true, textclass);
+ parse_text(p, ss, FLAG_END, true, context);
+ if (Context::empty)
+ // Empty document body. LyX needs at least one paragraph.
+ context.check_layout(ss);
+ context.check_end_layout(ss);
+ ss << "\n\\end_body\n\\end_document\n";
+ active_environments.pop_back();
+ ss.seekg(0);
+ os << ss.str();
+#ifdef TEST_PARSER
+ p.reset();
+ ofstream parsertest("parsertest.tex");
+ while (p.good())
+ parsertest << p.get_token().asInput();
+ // <origfile> and parsertest.tex should now have identical content
+#endif
+}
- else if (t.cs() == "input")
- handle_tex(os, "\\input{" + p.verbatimItem() + "}\n");
- else if (t.cs() == "pagestyle" && in_preamble)
- h_paperpagestyle == p.getArg('{','}');
+/// convert TeX from \p infilename to LyX and write it to \p os
+bool tex2lyx(FileName const & infilename, std::ostream &os)
+{
+ ifstream is(infilename.toFilesystemEncoding().c_str());
+ if (!is.good()) {
+ cerr << "Could not open input file \"" << infilename
+ << "\" for reading." << endl;
+ return false;
+ }
+ string const oldParentFilePath = parentFilePath;
+ parentFilePath = onlyPath(infilename.absFilename());
+ tex2lyx(is, os);
+ parentFilePath = oldParentFilePath;
+ return true;
+}
- else if (t.cs() == "fancyhead") {
- ostringstream ss;
- ss << "\\fancyhead";
- ss << p.getOpt();
- ss << '{' << p.verbatimItem() << "}\n";
- handle_tex(os, ss.str());
- }
+} // anonymous namespace
- else {
- //cerr << "#: " << t << " mode: " << mode << endl;
- if (mode == MATH_MODE || mode == MATHTEXT_MODE) {
- os << t.asInput();
- //cerr << "#: writing: '" << t.asInput() << "'\n";
- }
- else if (in_preamble)
- h_preamble << t.asInput();
- else {
- // heuristic: read up to next non-nested space
- /*
- string s = t.asInput();
- string z = p.verbatimItem();
- while (p.good() && z != " " && z.size()) {
- //cerr << "read: " << z << endl;
- s += z;
- z = p.verbatimItem();
- }
- cerr << "found ERT: " << s << endl;
- handle_ert(os, s + ' ');
- */
- handle_ert(os, t.asInput() + ' ');
- }
- }
- if (flags & FLAG_LEAVE) {
- flags &= ~FLAG_LEAVE;
- break;
+bool tex2lyx(string const &infilename, FileName const &outfilename)
+{
+ if (isFileReadable(outfilename)) {
+ if (overwrite_files) {
+ cerr << "Overwriting existing file "
+ << outfilename << endl;
+ } else {
+ cerr << "Not overwriting existing file "
+ << outfilename << endl;
+ return false;
}
+ } else {
+ cerr << "Creating file " << outfilename << endl;
+ }
+ ofstream os(outfilename.toFilesystemEncoding().c_str());
+ if (!os.good()) {
+ cerr << "Could not open output file \"" << outfilename
+ << "\" for writing." << endl;
+ return false;
}
+#ifdef FILEDEBUG
+ cerr << "Input file: " << infilename << "\n";
+ cerr << "Output file: " << outfilename << "\n";
+#endif
+ return tex2lyx(FileName(infilename), os);
}
-
-
-} // anonymous namespace
+} // namespace lyx
int main(int argc, char * argv[])
{
+ using namespace lyx;
+ fs::path::default_name_check(fs::no_check);
+
+ easyParse(argc, argv);
+
if (argc <= 1) {
- cerr << "Usage: " << argv[0] << " <infile.tex>" << endl;
+ cerr << "Usage: tex2lyx [ command line switches ] <infile.tex> [<outfile.lyx>]\n"
+ "See tex2lyx -help." << endl;
return 2;
}
- ifstream is(argv[1]);
- Parser p(is);
- parse(p, cout, 0, TEXT_MODE);
- cout << "\n\\the_end";
+ lyx::support::os::init(argc, argv);
+ support::init_package(to_utf8(from_local8bit(argv[0])),
+ cl_system_support, cl_user_support,
+ support::top_build_dir_is_two_levels_up);
+
+ // Now every known option is parsed. Look for input and output
+ // file name (the latter is optional).
+ string const infilename = makeAbsPath(to_utf8(from_local8bit(argv[1]))).absFilename();
+ string outfilename;
+ if (argc > 2) {
+ outfilename = to_utf8(from_local8bit(argv[2]));
+ if (outfilename != "-")
+ outfilename = makeAbsPath(to_utf8(from_local8bit(argv[2]))).absFilename();
+ } else
+ outfilename = changeExtension(infilename, ".lyx");
+
+ FileName const system_syntaxfile = lyx::support::libFileSearch("", "syntax.default");
+ if (system_syntaxfile.empty()) {
+ cerr << "Error: Could not find syntax file \"syntax.default\"." << endl;
+ exit(1);
+ }
+ read_syntaxfile(system_syntaxfile);
+ if (!syntaxfile.empty())
+ read_syntaxfile(makeAbsPath(syntaxfile));
+
+ masterFilePath = onlyPath(infilename);
+ parentFilePath = masterFilePath;
- return 0;
-}
+ if (outfilename == "-") {
+ if (tex2lyx(FileName(infilename), cout))
+ return EXIT_SUCCESS;
+ else
+ return EXIT_FAILURE;
+ } else {
+ if (tex2lyx(infilename, FileName(outfilename)))
+ return EXIT_SUCCESS;
+ else
+ return EXIT_FAILURE;
+ }
+}
// }])