X-Git-Url: https://git.lyx.org/gitweb/?a=blobdiff_plain;f=src%2FLexer.cpp;h=25d337a0793287e3ac6ba8e532dc74ca821b076f;hb=0430132aa049f2a97280bcbcff69f71d42ed7d98;hp=546ae0e9e9e865cd7834ebe667421abc30db618b;hpb=c2bdcc6b639a9c238cd44ad0a0ba96c0383bec40;p=lyx.git

diff --git a/src/Lexer.cpp b/src/Lexer.cpp
index 546ae0e9e9..25d337a079 100644
--- a/src/Lexer.cpp
+++ b/src/Lexer.cpp
@@ -14,12 +14,14 @@
 #include <config.h>
 
 #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"
@@ -237,9 +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 (formats.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
@@ -249,9 +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");
 
@@ -262,12 +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;
 }
 
 
@@ -306,12 +320,11 @@ 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;
 
 		if (c == commentChar) {
 			// Read rest of line (fast :-)
@@ -334,9 +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;
@@ -458,7 +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);
 	}
 
@@ -555,7 +567,7 @@ void Lexer::Pimpl::pushToken(string const & pt)
 //////////////////////////////////////////////////////////////////////
 
 Lexer::Lexer()
-	: pimpl_(new Pimpl(0, 0))
+	: pimpl_(new Pimpl(0, 0)), lastReadOk_(false)
 {}
 
 
@@ -681,23 +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();
 }