X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FLexer.cpp;h=098faa9737282c3fc40458ec67665b26b8b881ef;hb=4ea6d81437ec7b5bea165d78ca37cb163e262042;hp=5136c672073d6dc451360a22f6b088915c1a66f3;hpb=e7645f24391648fb08cf6b6d808fcabec76be744;p=lyx.git diff --git a/src/Lexer.cpp b/src/Lexer.cpp index 5136c67207..098faa9737 100644 --- a/src/Lexer.cpp +++ b/src/Lexer.cpp @@ -4,7 +4,7 @@ * Licence details can be found in the file COPYING. * * \author Alejandro Aguilar Sierra - * \author Lars Gullik Bjønnes + * \author Lars Gullik Bjønnes * \author Jean-Marc Lasgouttes * \author John Levon * @@ -14,12 +14,14 @@ #include #include "Lexer.h" +#include "Format.h" #include "support/convert.h" #include "support/debug.h" #include "support/FileName.h" #include "support/filetools.h" #include "support/gzstream.h" +#include "support/lassert.h" #include "support/lstrings.h" #include "support/lyxalgo.h" #include "support/types.h" @@ -146,7 +148,7 @@ public: } }; -} // end of anon namespace +} // namespace Lexer::Pimpl::Pimpl(LexerKeyword * tab, int num) @@ -173,7 +175,9 @@ void Lexer::Pimpl::printError(string const & message) const { string const tmpmsg = subst(message, "$$Token", getString()); lyxerr << "LyX: " << tmpmsg << " [around line " << lineno - << " of file " << to_utf8(makeDisplayPath(name)) << ']' << endl; + << " of file " << to_utf8(makeDisplayPath(name)) + << " current token: '" << getString() << "'" + << " context: '" << context << "']" << endl; } @@ -235,9 +239,7 @@ void Lexer::Pimpl::popTable() bool Lexer::Pimpl::setFile(FileName const & filename) { // Check the format of the file. - string const format = filename.guessFormatFromContents(); - - if (format == "gzip" || format == "zip" || format == "compress") { + if (theFormats().isZippedFile(filename)) { LYXERR(Debug::LYXLEX, "lyxlex: compressed"); // The check only outputs a debug message, because it triggers // a bug in compaq cxx 6.2, where is_open() returns 'true' for @@ -247,9 +249,10 @@ bool Lexer::Pimpl::setFile(FileName const & filename) "file or stream already set."); gz_.open(filename.toFilesystemEncoding().c_str(), ios::in); is.rdbuf(&gz_); - name = filename.absFilename(); + name = filename.absFileName(); lineno = 0; - return gz_.is_open() && is.good(); + if (!gz_.is_open() || !is.good()) + return false; } else { LYXERR(Debug::LYXLEX, "lyxlex: UNcompressed"); @@ -260,12 +263,25 @@ bool Lexer::Pimpl::setFile(FileName const & filename) LYXERR(Debug::LYXLEX, "Error in Lexer::setFile: " "file or stream already set."); } - fb_.open(filename.toFilesystemEncoding().c_str(), ios::in); + fb_.open(filename.toSafeFilesystemEncoding().c_str(), ios::in); is.rdbuf(&fb_); - name = filename.absFilename(); + name = filename.absFileName(); lineno = 0; - return fb_.is_open() && is.good(); + if (!fb_.is_open() || !is.good()) + return false; + } + + // Skip byte order mark. + if (is.peek() == 0xef) { + is.get(); + if (is.peek() == 0xbb) { + is.get(); + LASSERT(is.get() == 0xbf, /**/); + } else + is.unget(); } + + return true; } @@ -304,18 +320,12 @@ bool Lexer::Pimpl::next(bool esc /* = false */) } - unsigned char c = 0; // getc() returns an int char cc = 0; status = 0; while (is && !status) { is.get(cc); - c = cc; + unsigned char c = cc; - // skip ','s - if (esc && c == ',') - continue; - - if (c == commentChar) { // Read rest of line (fast :-) #if 1 @@ -337,9 +347,8 @@ bool Lexer::Pimpl::next(bool esc /* = false */) if (esc) { - bool escaped = false; do { - escaped = false; + bool escaped = false; is.get(cc); c = cc; if (c == '\r') continue; @@ -380,13 +389,13 @@ bool Lexer::Pimpl::next(bool esc /* = false */) break; } - if (!esc && c == ',') + if (c == ',') continue; /* Skip ','s */ - // using relational operators with chars other - // than == and != is not safe. And if it is done - // the type _have_ to be unsigned. It usually a - // lot better to use the functions from cctype + // using relational operators with chars other + // than == and != is not safe. And if it is done + // the type _have_ to be unsigned. It usually a + // lot better to use the functions from cctype if (c > ' ' && is) { buff.clear(); @@ -404,7 +413,7 @@ bool Lexer::Pimpl::next(bool esc /* = false */) status = LEX_TOKEN; } - if (!esc && c == '\r' && is) { + if (c == '\r' && is) { // The Windows support has lead to the // possibility of "\r\n" at the end of // a line. This will stop LyX choking @@ -461,7 +470,7 @@ bool Lexer::Pimpl::eatLine() is.get(cc); c = cc; //LYXERR(Debug::LYXLEX, "Lexer::EatLine read char: `" << c << '\''); - if (c != '\r') + if (c != '\r' && is) buff.push_back(c); } @@ -502,7 +511,7 @@ bool Lexer::Pimpl::nextToken() char cc = 0; is.get(cc); c = cc; - if (c >= ' ' && is) { + if ((c >= ' ' || c == '\t') && is) { buff.clear(); if (c == '\\') { // first char == '\\' @@ -516,7 +525,7 @@ bool Lexer::Pimpl::nextToken() buff.push_back(c); is.get(cc); c = cc; - } while (c >= ' ' && c != '\\' && is); + } while ((c >= ' ' || c == '\t') && c != '\\' && is); } if (c == '\\') @@ -558,7 +567,7 @@ void Lexer::Pimpl::pushToken(string const & pt) ////////////////////////////////////////////////////////////////////// Lexer::Lexer() - : pimpl_(new Pimpl(0, 0)) + : pimpl_(new Pimpl(0, 0)), lastReadOk_(false) {} @@ -639,6 +648,7 @@ void Lexer::setCommentChar(char c) pimpl_->setCommentChar(c); } + int Lexer::lex() { return pimpl_->lex(); @@ -683,23 +693,23 @@ double Lexer::getFloat() const } -string const Lexer::getString() const +string const Lexer::getString(bool trim) const { lastReadOk_ = pimpl_->status == LEX_DATA || pimpl_->status == LEX_TOKEN; if (lastReadOk_) - return pimpl_->getString(); + return trim ? support::trim(pimpl_->getString(), "\t ") : pimpl_->getString(); return string(); } -docstring const Lexer::getDocString() const +docstring const Lexer::getDocString(bool trim) const { lastReadOk_ = pimpl_->status == LEX_DATA || pimpl_->status == LEX_TOKEN; if (lastReadOk_) - return pimpl_->getDocString(); + return trim ? support::trim(pimpl_->getDocString(), "\t ") : pimpl_->getDocString(); return docstring(); } @@ -708,28 +718,27 @@ docstring const Lexer::getDocString() const // I would prefer to give a tag number instead of an explicit token // here, but it is not possible because Buffer::readDocument uses // explicit tokens (JMarc) -string const Lexer::getLongString(string const & endtoken) +docstring Lexer::getLongString(docstring const & endtoken) { - string str; - string prefix; + docstring str; + docstring prefix; bool firstline = true; while (pimpl_->is) { //< eatLine only reads from is, not from pushTok if (!eatLine()) // blank line in the file being read continue; + docstring tmpstr = getDocString(); + docstring const token = trim(tmpstr, " \t"); - string const token = trim(getString(), " \t"); - - LYXERR(Debug::PARSER, "LongString: `" << getString() << '\''); + LYXERR(Debug::PARSER, "LongString: `" << tmpstr << '\''); // We do a case independent comparison, like searchKeyword does. - if (compare_ascii_no_case(token, endtoken) == 0) + if (compare_no_case(token, endtoken) == 0) break; - string tmpstr = getString(); if (firstline) { - size_t i = tmpstr.find_first_not_of(' '); + size_t i = tmpstr.find_first_not_of(from_ascii(" \t")); if (i != string::npos) prefix = tmpstr.substr(0, i); firstline = false; @@ -738,14 +747,14 @@ string const Lexer::getLongString(string const & endtoken) // further lines in long strings may have the same // whitespace prefix as the first line. Remove it. - if (prefix.length() && prefixIs(tmpstr, prefix)) - tmpstr.erase(0, prefix.length() - 1); + if (!prefix.empty() && prefixIs(tmpstr, prefix)) + tmpstr.erase(0, prefix.length()); - str += ltrim(tmpstr, "\t") + '\n'; + str += tmpstr + '\n'; } if (!pimpl_->is) - printError("Long string not ended by `" + endtoken + '\''); + printError("Long string not ended by `" + to_utf8(endtoken) + '\''); return str; } @@ -753,7 +762,7 @@ string const Lexer::getLongString(string const & endtoken) bool Lexer::getBool() const { - string const s = pimpl_->getString(); + string const s = pimpl_->getString(); if (s == "false" || s == "0") { lastReadOk_ = true; return false; @@ -765,6 +774,7 @@ bool Lexer::getBool() const pimpl_->printError("Bad boolean `$$Token'. " "Use \"false\" or \"true\""); lastReadOk_ = false; + return false; } @@ -902,12 +912,24 @@ string Lexer::quoteString(string const & arg) } +// same for docstring +docstring Lexer::quoteString(docstring const & arg) +{ + docstring res; + res += '"'; + res += subst(subst(arg, from_ascii("\\"), from_ascii("\\\\")), + from_ascii("\""), from_ascii("\\\"")); + res += '"'; + return res; +} + + Lexer & Lexer::operator>>(char const * required) { string token; *this >> token; if (token != required) { - LYXERR0("Missing '" << required << "'-tag in " << pimpl_->context + LYXERR0("Missing '" << required << "'-tag in " << pimpl_->context << ". Got " << token << " instead. Line: " << lineNumber()); pushToken(token); }