-
-/** 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 <sstream>
-#include <stack>
#include <string>
+#include <sstream>
#include <vector>
+#include <map>
+
+
+namespace lyx {
-using std::atoi;
+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::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 {
+namespace fs = boost::filesystem;
-char const OPEN = '<';
-char const CLOSE = '>';
-
-const char * known_languages[] = { "austrian", "babel", "bahasa",
-"basque", "breton", "bulgarian", "catalan", "croatian", "czech", "danish",
-"dutch", "english", "esperanto", "estonian", "finnish", "francais",
-"frenchb", "galician", "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};
-
-const char * known_fontsizes[] = { "10pt", "11pt", "12pt", 0 };
-
-
-// some ugly stuff
-string h_preamble;
-string h_textclass = "FIXME";
-string h_options = "FIXME";
-string h_language = "FIXME";
-string h_inputencoding = "FIXME";
-string h_fontscheme = "FIXME";
-string h_graphics = "default";
-string h_paperfontsize = "FIXME";
-string h_spacing = "single";
-string h_papersize = "FIXME";
-string h_paperpackage = "FIXME";
-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 = "2";
-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;
+IconvProcessor & utf8ToUcs4()
+{
+ static IconvProcessor iconv(ucs4_codeset, "UTF-8");
+ return iconv;
+}
+
+
+// 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)
+{
+ // BOOST_ASSERT(p);
+
+ if (a.empty() || !*p)
+ return a;
+
+ string::size_type r = a.find_last_not_of(p);
+ string::size_type l = a.find_first_not_of(p);
+
+ // Is this the minimal test? (lgb)
+ if (r == string::npos && l == string::npos)
+ return string();
+
+ return a.substr(l, r - l + 1);
+}
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";
}
-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();
}
-void handle_opt(vector<string> & opts, char const ** what, string & target)
+char const * const * is_known(string const & str, char const * const * what)
{
- 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;
- }
- }
+ for ( ; *what; ++what)
+ if (str == *what)
+ return what;
+ return 0;
}
-void handle_ert(ostream & os, string const & s)
-{
- os << "\n\\begin_inset ERT\nstatus Collapsed\n\n\\layout Standard\n\n";
- os << s;
- os << "\n\\end_inset\n";
-}
-
-string wrap(string const & cmd, string const & str)
-{
- return OPEN + cmd + ' ' + str + CLOSE;
-}
+// current stack of nested environments
+vector<string> active_environments;
-string wrap(string const & cmd, string const & str, string const & str2)
+string active_environment()
{
- return OPEN + cmd + ' ' + str + ' ' + str2 + CLOSE;
+ return active_environments.empty() ? string() : active_environments.back();
}
-enum mode_type {UNDECIDED_MODE, TEXT_MODE, MATH_MODE};
-
-mode_type asMode(mode_type oldmode, string const & str)
-{
- if (str == "mathmode")
- return MATH_MODE;
- if (str == "textmode" || str == "forcetext")
- return TEXT_MODE;
- return oldmode;
-}
+CommandMap known_commands;
+CommandMap known_environments;
+CommandMap known_math_environments;
-// These are TeX's catcodes
-enum CatCode {
- catEscape, // 0 backslash
- catBegin, // 1 {
- catEnd, // 2 }
- cat, // 3 $
- catAlign, // 4 &
- catNewline, // 5 ^^M
- catParameter, // 6 #
- catSuper, // 7 ^
- catSub, // 8 _
- catIgnore, // 9
- catSpace, // 10 space
- catLetter, // 11 a-zA-Z
- catOther, // 12 none of the above
- catActive, // 13 ~
- catComment, // 14 %
- catInvalid // 15 <delete>
-};
-
-CatCode theCatcode[256];
-
-
-inline CatCode catcode(unsigned char c)
+void add_known_command(string const & command, string const & o1,
+ bool o2)
{
- return theCatcode[c];
+ // 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;
+ }
+ }
+ for (unsigned int i = 0; i < nargs; ++i)
+ arguments.push_back(required);
+ known_commands[command] = arguments;
}
-enum {
- FLAG_BRACE_LAST = 1 << 1, // last closing brace ends the parsing
- FLAG_RIGHT = 1 << 2, // next \\right ends the parsing process
- FLAG_END = 1 << 3, // next \\end ends the parsing process
- FLAG_BRACK_LAST = 1 << 4, // next closing bracket ends the parsing
- FLAG_TEXTMODE = 1 << 5, // we are in a box
- FLAG_ITEM = 1 << 6, // read a (possibly braced token)
- FLAG_LEAVE = 1 << 7, // leave the loop at the end
- FLAG_SIMPLE = 1 << 8, // next $ leaves the loop
- FLAG_EQUATION = 1 << 9, // next \] leaves the loop
- FLAG_SIMPLE2 = 1 << 10, // next \) leaves the loop
- FLAG_OPTION = 1 << 11, // read [...] style option
- FLAG_BRACED = 1 << 12 // read {...} style argument
-};
+bool noweb_mode = false;
-void catInit()
-{
- fill(theCatcode, theCatcode + 256, catOther);
- fill(theCatcode + 'a', theCatcode + 'z' + 1, catLetter);
- fill(theCatcode + 'A', theCatcode + 'Z' + 1, catLetter);
-
- theCatcode['\\'] = catEscape;
- theCatcode['{'] = catBegin;
- theCatcode['}'] = catEnd;
- theCatcode['$'] = cat;
- theCatcode['&'] = catAlign;
- theCatcode['\n'] = catNewline;
- theCatcode['#'] = catParameter;
- theCatcode['^'] = catSuper;
- theCatcode['_'] = catSub;
- theCatcode['\7f'] = catIgnore;
- theCatcode[' '] = catSpace;
- theCatcode['\t'] = catSpace;
- theCatcode['\r'] = catNewline;
- theCatcode['~'] = catActive;
- theCatcode['%'] = catComment;
-}
-
+namespace {
-//
-// Helper class for parsing
-//
-
-class Token {
-public:
- ///
- Token() : cs_(), char_(0), cat_(catIgnore) {}
- ///
- Token(char c, CatCode cat) : cs_(), char_(c), cat_(cat) {}
- ///
- Token(string const & cs) : cs_(cs), char_(0), cat_(catIgnore) {}
-
- ///
- string const & cs() const { return cs_; }
- ///
- CatCode cat() const { return cat_; }
- ///
- char character() const { return char_; }
- ///
- string asString() const { return cs_.size() ? cs_ : string(1, char_); }
-
-private:
- ///
- string cs_;
- ///
- char char_;
- ///
- CatCode cat_;
-};
-
-ostream & operator<<(ostream & os, Token const & t)
-{
- if (t.cs().size())
- os << '\\' << t.cs();
- else
- os << '[' << t.character() << ',' << t.cat() << ']';
- return os;
+/*!
+ * 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);
+ }
+ }
+ commands[command] = arguments;
}
-class Parser {
-
-public:
- ///
- Parser(istream & is);
-
- ///
- string parse();
- ///
- string parse(unsigned flags, mode_type mode);
- ///
- int lineno() const { return lineno_; }
- ///
- void putback();
- /// dump contents to screen
- void dump() const;
-
-private:
- ///
- string getArg(char left, char right);
- ///
- char getChar();
- ///
- void error(string const & msg);
- ///
- void tokenize(istream & is);
- ///
- void tokenize(string const & s);
- ///
- void skipSpaceTokens(istream & is, char c);
- ///
- void push_back(Token const & t);
- ///
- void pop_back();
- ///
- Token const & prevToken() const;
- ///
- Token const & nextToken() const;
- ///
- Token const & getToken();
- /// skips spaces if any
- void skipSpaces();
- ///
- void lex(string const & s);
- ///
- bool good() const;
- ///
- string parse_verbatim_item();
- ///
- string parse_verbatim_option();
-
- ///
- int lineno_;
- ///
- vector<Token> tokens_;
- ///
- unsigned pos_;
-};
-
-
-Parser::Parser(istream & is)
- : lineno_(0), pos_(0)
+/*!
+ * Read a class of environments from the syntax file
+ */
+void read_environment(Parser & p, string const & begin,
+ CommandMap & environments)
{
- tokenize(is);
+ 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;
+ }
+ }
}
-void Parser::push_back(Token const & t)
+/*!
+ * 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)
{
- tokens_.push_back(t);
+ 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.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 {
+ read_command(p, command, known_commands);
+ }
+ }
+ }
}
-void Parser::pop_back()
-{
- tokens_.pop_back();
-}
+string documentclass;
+string syntaxfile;
+bool overwrite_files = false;
-Token const & Parser::prevToken() const
-{
- static const Token dummy;
- return pos_ > 0 ? tokens_[pos_ - 1] : dummy;
-}
+/// return the number of arguments consumed
+typedef boost::function<int(string const &, string const &)> cmd_helper;
-Token const & Parser::nextToken() const
+int parse_help(string const &, string const &)
{
- static const Token dummy;
- return good() ? tokens_[pos_] : dummy;
+ 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);
}
-Token const & Parser::getToken()
+int parse_class(string const & arg, string const &)
{
- static const Token dummy;
- //cerr << "looking at token " << tokens_[pos_] << " pos: " << pos_ << '\n';
- return good() ? tokens_[pos_++] : dummy;
+ if (arg.empty()) {
+ cerr << "Missing textclass string after -c switch" << endl;
+ exit(1);
+ }
+ documentclass = arg;
+ return 1;
}
-void Parser::skipSpaces()
+int parse_syntaxfile(string const & arg, string const &)
{
- while (nextToken().cat() == catSpace || nextToken().cat() == catNewline)
- getToken();
+ if (arg.empty()) {
+ cerr << "Missing syntaxfile string after -s switch" << endl;
+ exit(1);
+ }
+ syntaxfile = arg;
+ return 1;
}
-void Parser::putback()
-{
- --pos_;
-}
+// Filled with the command line arguments "foo" of "-sysdir foo" or
+// "-userdir foo".
+string cl_system_support;
+string cl_user_support;
-bool Parser::good() const
+int parse_sysdir(string const & arg, string const &)
{
- return pos_ < tokens_.size();
+ if (arg.empty()) {
+ cerr << "Missing directory for -sysdir switch" << endl;
+ exit(1);
+ }
+ cl_system_support = arg;
+ return 1;
}
-char Parser::getChar()
+int parse_userdir(string const & arg, string const &)
{
- if (!good())
- error("The input stream is not well...");
- return tokens_[pos_++].character();
+ if (arg.empty()) {
+ cerr << "Missing directory for -userdir switch" << endl;
+ exit(1);
+ }
+ cl_user_support = arg;
+ return 1;
}
-string Parser::getArg(char left, char right)
+int parse_force(string const &, string const &)
{
- skipSpaces();
-
- string result;
- char c = getChar();
-
- if (c != left)
- putback();
- else
- while ((c = getChar()) != right && good())
- result += c;
-
- return result;
+ overwrite_files = true;
+ return 0;
}
-void Parser::skipSpaceTokens(istream & is, char c)
+int parse_noweb(string const &, string const &)
{
- // skip trailing spaces
- while (catcode(c) == catSpace || catcode(c) == catNewline)
- if (!is.get(c))
- break;
- //cerr << "putting back: " << c << "\n";
- is.putback(c);
+ noweb_mode = true;
+ return 0;
}
-void Parser::tokenize(istream & is)
+void easyParse(int & argc, char * argv[])
{
- // eat everything up to the next \end_inset or end of stream
- // and store it in s for further tokenization
- string s;
- char c;
- while (is.get(c)) {
- s += c;
- if (s.size() >= 10 && s.substr(s.size() - 10) == "\\end_inset") {
- s = s.substr(0, s.size() - 10);
- break;
- }
+ 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;
}
- // Remove the space after \end_inset
- if (is.get(c) && c != ' ')
- is.unget();
-
- // tokenize buffer
- tokenize(s);
}
-void Parser::tokenize(string const & buffer)
-{
- static bool init_done = false;
-
- if (!init_done) {
- catInit();
- init_done = true;
- }
-
- istringstream is(buffer.c_str(), ios::in | ios::binary);
-
- char c;
- while (is.get(c)) {
- //cerr << "reading c: " << c << "\n";
-
- switch (catcode(c)) {
- case catNewline: {
- ++lineno_;
- is.get(c);
- if (catcode(c) == catNewline)
- push_back(Token("par"));
- else {
- push_back(Token('\n', catNewline));
- is.putback(c);
- }
- break;
- }
-
-/*
- case catComment: {
- while (is.get(c) && catcode(c) != catNewline)
- ;
- ++lineno_;
- break;
- }
-*/
-
- case catEscape: {
- is.get(c);
- if (!is) {
- error("unexpected end of input");
- } else {
- string s(1, c);
- if (catcode(c) == catLetter) {
- // collect letters
- while (is.get(c) && catcode(c) == catLetter)
- s += c;
- skipSpaceTokens(is, c);
- }
- push_back(Token(s));
- }
- break;
- }
-
- case catSuper:
- case catSub: {
- push_back(Token(c, catcode(c)));
- is.get(c);
- skipSpaceTokens(is, c);
- break;
- }
+// path of the first parsed file
+string masterFilePath;
+// path of the currently parsed file
+string parentFilePath;
- case catIgnore: {
- cerr << "ignoring a char: " << int(c) << "\n";
- break;
- }
-
- default:
- push_back(Token(c, catcode(c)));
- }
- }
-
-#ifdef FILEDEBUG
- dump();
-#endif
-}
+} // anonymous namespace
-void Parser::dump() const
+string getMasterFilePath()
{
- cerr << "\nTokens: ";
- for (unsigned i = 0; i < tokens_.size(); ++i) {
- if (i == pos_)
- cerr << " <#> ";
- cerr << tokens_[i];
- }
- cerr << " pos: " << pos_ << "\n";
+ return masterFilePath;
}
-
-void Parser::error(string const & msg)
+string getParentFilePath()
{
- cerr << "Line ~" << lineno_ << ": parse error: " << msg << endl;
- dump();
- //exit(1);
+ return parentFilePath;
}
-string Parser::parse()
+namespace {
+
+/*!
+ * 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)
{
- skipSpaces();
- return parse(0, UNDECIDED_MODE);
+ 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
}
-string Parser::parse_verbatim_option()
+/// convert TeX from \p infilename to LyX and write it to \p os
+bool tex2lyx(FileName const & infilename, std::ostream &os)
{
- string res;
- if (nextToken().character() == '[') {
- Token t = getToken();
- for (Token t = getToken(); t.character() != ']' && good(); t = getToken()) {
- if (t.cat() == catBegin) {
- putback();
- res += '{' + parse_verbatim_item() + '}';
- } else
- res += t.asString();
- }
+ ifstream is(infilename.toFilesystemEncoding().c_str());
+ if (!is.good()) {
+ cerr << "Could not open input file \"" << infilename
+ << "\" for reading." << endl;
+ return false;
}
- return res;
+ string const oldParentFilePath = parentFilePath;
+ parentFilePath = onlyPath(infilename.absFilename());
+ tex2lyx(is, os);
+ parentFilePath = oldParentFilePath;
+ return true;
}
+} // anonymous namespace
+
-string Parser::parse_verbatim_item()
+bool tex2lyx(string const &infilename, FileName const &outfilename)
{
- string res;
- if (nextToken().cat() == catBegin) {
- Token t = getToken();
- for (Token t = getToken(); t.cat() != catEnd && good(); t = getToken()) {
- if (t.cat() == catBegin) {
- putback();
- res += '{' + parse_verbatim_item() + '}';
- }
- else
- res += t.asString();
+ 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;
}
- return res;
-}
-
-
-string Parser::parse(unsigned flags, mode_type mode)
-{
- //int limits = 0;
-
- ostringstream result;
- while (good()) {
- Token const & t = getToken();
-
#ifdef FILEDEBUG
- cerr << "t: " << t << " flags: " << flags << "\n";
- cell->dump();
- cerr << "\n";
+ cerr << "Input file: " << infilename << "\n";
+ cerr << "Output file: " << outfilename << "\n";
#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) {
- error("opening brace expected");
- return result.str();
- }
-
- // 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() == '[') {
- result << parse(FLAG_BRACK_LAST, mode);
- } else {
- // no option found, put back token and we are done
- putback();
- }
- return result.str();
- }
-
- //
- // cat codes
- //
- if (t.cat() == cat) {
- if (mode != MATH_MODE) {
- // we are inside some text mode thingy, so opening new math is allowed
- Token const & n = getToken();
- if (n.cat() == cat) {
- // TeX's $$...$$ syntax for displayed math
- result << wrap("equation", parse(FLAG_SIMPLE, MATH_MODE));
- getToken(); // skip the second '$' token
- } else {
- // simple $...$ stuff
- putback();
- result << wrap("simple", parse(FLAG_SIMPLE, MATH_MODE));
- }
- }
-
- else if (flags & FLAG_SIMPLE) {
- // this is the end of the formula
- return result.str();
- }
-
- else {
- error("something strange in the parser\n");
- break;
- }
- }
-
- else if (t.cat() == catLetter)
- result << t.character();
-
- else if (t.cat() == catSpace && mode != MATH_MODE) {
- //if (result.empty() || result[result.size() - 1] != ' ')
- result << t.character();
- }
-
- else if (t.cat() == catNewline && mode != MATH_MODE)
- result << t.character();
-
- else if (t.cat() == catParameter) {
- Token const & n = getToken();
- result << wrap("macroarg", string(1, n.character()));
- }
-
- else if (t.cat() == catActive)
- result << wrap("active", string(1, t.character()));
-
- else if (t.cat() == catBegin)
- result << wrap("braced", parse(FLAG_BRACE_LAST, mode));
-
- else if (t.cat() == catEnd) {
- if (flags & FLAG_BRACE_LAST)
- return result.str();
- error("found '}' unexpectedly");
- //lyx::Assert(0);
- //add(cell, '}', LM_TC_TEX);
- }
-
-/*
- else if (t.cat() == catAlign) {
- ++cellcol;
- //cerr << " column now " << cellcol << " max: " << grid.ncols() << "\n";
- if (cellcol == grid.ncols()) {
- //cerr << "adding column " << cellcol << "\n";
- grid.addCol(cellcol - 1);
- }
- cell = &grid.cell(grid.index(cellrow, cellcol));
- }
-*/
-
- else if (t.character() == ']' && (flags & FLAG_BRACK_LAST)) {
- //cerr << "finished reading option\n";
- return result.str();
- }
-
- else if (t.cat() == catOther)
- result << string(1, t.character());
-
- else if (t.cat() == catComment) {
- string s;
- while (good()) {
- Token const & t = getToken();
- if (t.cat() == catNewline)
- break;
- s += t.asString();
- }
- //result << wrap("comment", s);
- skipSpaces();
- }
-
- //
- // control sequences
- //
-
- else if (t.cs() == "lyxlock") {
- // ignored;
- }
-
- else if (t.cs() == "newcommand" || t.cs() == "providecommand") {
- string const name = parse_verbatim_item();
- string const opts = getArg('[', ']');
- string const body = parse_verbatim_item();
- // only non-lyxspecific stuff
- if (name != "noun" && name != "tabularnewline") {
- h_preamble += "\\" + t.cs() + "{" + name + "}";
- if (opts.size())
- h_preamble += "[" + opts + "]";
- h_preamble += "{" + body + "}\n";
- }
- }
-
- else if (t.cs() == "(")
- result << wrap("simple", parse(FLAG_SIMPLE2, MATH_MODE));
-
- else if (t.cs() == "[")
- result << wrap("equation", parse(FLAG_EQUATION, MATH_MODE));
-
- else if (t.cs() == "protect")
- // ignore \\protect, will hopefully be re-added during output
- ;
-
- else if (t.cs() == "end") {
- if (flags & FLAG_END) {
- // eat environment name
- string const name = getArg('{', '}');
- if (name != active_environments.top())
- error("\\end{" + name + "} does not match \\begin{"
- + active_environments.top() + "}");
- active_environments.pop();
- return result.str();
- }
- error("found 'end' unexpectedly");
- }
-
- else if (t.cs() == ")") {
- if (flags & FLAG_SIMPLE2)
- return result.str();
- error("found '\\)' unexpectedly");
- }
-
- else if (t.cs() == "]") {
- if (flags & FLAG_EQUATION)
- return result.str();
- error("found '\\]' unexpectedly");
- }
-
-/*
- else if (t.cs() == "\\") {
- grid.vcrskip(LyXLength(getArg('[', ']')), cellrow);
- ++cellrow;
- cellcol = 0;
- if (cellrow == grid.nrows())
- grid.addRow(cellrow - 1);
- if (grid.asHullstring())
- grid.asHullstring()->numbered(cellrow, numbered);
- cell = &grid.cell(grid.index(cellrow, cellcol));
- }
-*/
- else if (t.cs() == "documentclass") {
- vector<string> opts;
- split(getArg('[', ']'), opts, ',');
- handle_opt(opts, known_languages, h_language);
- handle_opt(opts, known_fontsizes, h_paperfontsize);
- h_options = join(opts, ',');
- h_textclass = getArg('{', '}');
- }
-
- else if (t.cs() == "usepackage") {
- string const options = getArg('[', ']');
- string const name = getArg('{', '}');
- if (name == "a4wide") {
- h_papersize = "a4";
- 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 (options.size())
- h_preamble += "\\usepackage[" + options + "]{" + name + "}\n";
- else
- h_preamble += "\\usepackage{" + name + "}\n";
- }
- }
-
- else if (t.cs() == "newenvironment") {
- string const name = getArg('{', '}');
- skipSpaces();
- string const begin = parse_verbatim_item();
- skipSpaces();
- string const end = parse_verbatim_item();
- // ignore out mess
- if (name != "lyxcode")
- result << wrap("newenvironment", begin + end);
- }
-
- else if (t.cs() == "def") {
- string const name = getToken().cs();
- string res;
- while (nextToken().cat() != catBegin)
- res += getToken().asString();
- handle_ert(result, "\\def" + res + '{' + parse_verbatim_item() + '}');
- }
-
- else if (t.cs() == "setcounter") {
- string const name = getArg('{', '}');
- string const content = getArg('{', '}');
- if (name == "secnumdepth")
- h_secnumdepth = content;
- else if (name == "tocdepth")
- h_tocdepth = content;
- else
- h_preamble += "\\setcounter{" + name + "}{" + content + "}\n";
- }
-
- else if (t.cs() == "setlength") {
- string const name = getToken().cs();
- string const content = getArg('{', '}');
- if (name == "parskip")
- h_paragraph_separation = "skip";
- else if (name == "parindent")
- h_paragraph_separation = "skip";
- else
- h_preamble += "\\setcounter{" + name + "}{" + content + "}\n";
- }
-
- else if (t.cs() == "par") {
- if (!active_environments.empty())
- result << "\n\\layout " << active_environments.top() << "\n\n";
- }
-
- else if (t.cs() == "title")
- result << "\\layout Title\n\n" + parse_verbatim_item();
-
- else if (t.cs() == "author")
- result << "\\layout Author\n\n" + parse_verbatim_item();
-
- else if (t.cs() == "abstract")
- result << "\\layout Abstract\n\n" + parse_verbatim_item();
-
- else if (t.cs() == "begin") {
- string const name = getArg('{', '}');
- active_environments.push(name);
- result << parse(FLAG_END, mode);
- }
-
- if (flags & FLAG_LEAVE) {
- flags &= ~FLAG_LEAVE;
- break;
- }
- }
-
- return result.str();
+ 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;
}
- string t;
- ifstream is(argv[1]);
- Parser p(is);
- //p.dump();
- string s = p.parse();
- cout << "# tex2lyx 0.0.2 created this file\n"
- << "\\lyxformat 222\n"
- << "\\textclass " << h_textclass << "\n"
- << "\\begin_preamble\n" << h_preamble << "\\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"
- << s << "\n"
- << "\\the_end";
-
- return 0;
-}
+ 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;
+
+ 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;
+ }
+}
+
+// }])