Update for boost::regex v4, add test

git-svn-id: svn://svn.lyx.org/lyx/lyx-devel/trunk@13306 a592a061-630c-0410-9148-cb99ea01b6c8

src/frontends/controllers/ChangeLog

@@ -1,3 +1,9 @@
+2006-03-06  Lars Gullik Bjønnes  <larsbj@lyx.org>
+
+	* biblio.C (escape_special_chars): Update for boost::regex v4
+	(RegexMatch): Update for bosot::regex v4
+	(searchKeys): Handle exception if it occurs.
+
 2005-12-12  Jürgen Spitzmüller  <j.spitzmueller@gmx.de>
 
 	* ControlPrefs.[Ch]: new helper functions toPaperSize and

src/frontends/controllers/biblio.C

@@ -370,24 +370,29 @@ namespace {
 // Escape special chars.
 // All characters are literals except: '.|*?+(){}[]^$\'
 // These characters are literals when preceded by a "\", which is done here
+// @todo: This function should be moved to support, and then the test in tests
+//        should be moved there as well.
 string const escape_special_chars(string const & expr)
 {
 	// Search for all chars '.|*?+(){}[^$]\'
 	// Note that '[' and '\' must be escaped.
 	// This is a limitation of boost::regex, but all other chars in BREs
 	// are assumed literal.
-	boost::RegEx reg("[].|*?+(){}^$\\[\\\\]");
+	boost::regex reg("[].|*?+(){}^$\\[\\\\]");
 
-	// $& is a perl-like expression that expands to all of the current match
+	// $& is a perl-like expression that expands to all
+	// of the current match
 	// The '$' must be prefixed with the escape character '\' for
 	// boost to treat it as a literal.
 	// Thus, to prefix a matched expression with '\', we use:
-	return reg.Merge(expr, "\\\\$&");
+	return boost::regex_replace(expr, reg, "\\\\$&");
 }
 
 
 // A functor for use with std::find_if, used to ascertain whether a
 // data entry matches the required regex_
+// @throws: boost::regex_error if the supplied regex pattern is not valid
+// @todo: This function should be moved to support.
 class RegexMatch : public std::unary_function<string, bool>
 {
 public:
@@ -397,9 +402,6 @@ public:
 		: map_(m), regex_(re, icase) {}
 
 	bool operator()(string const & key) const {
-		if (!validRE())
-			return false;
-
 		// the data searched is the key + its associated BibTeX/biblio
 		// fields
 		string data = key;
@@ -409,14 +411,11 @@ public:
 
 		// Attempts to find a match for the current RE
 		// somewhere in data.
-		return regex_.Search(data);
+		return boost::regex_search(data, regex_);
 	}
-
-	bool validRE() const { return regex_.error_code() == 0; }
-
 private:
 	InfoMap const map_;
-	mutable boost::RegEx regex_;
+	mutable boost::regex regex_;
 };
 
 } // namespace anon
@@ -442,27 +441,31 @@ searchKeys(InfoMap const & theMap,
 	if (type == SIMPLE)
 		// We must escape special chars in the search_expr so that
 		// it is treated as a simple string by boost::regex.
-		expr = escape_special_chars(expr);
+	        expr = escape_special_chars(expr);
 
-	// Build the functor that will be passed to find_if.
-	RegexMatch const match(theMap, expr, !caseSensitive);
-	if (!match.validRE())
+	try {
+		// Build the functor that will be passed to find_if.
+		RegexMatch const match(theMap, expr, !caseSensitive);
+
+		// Search the vector of 'keys' from 'start' for one
+		// that matches the predicate 'match'. Searching can
+		// be forward or backward from start.
+		if (dir == FORWARD)
+			return std::find_if(start, keys.end(), match);
+
+		vector<string>::const_reverse_iterator rit(start);
+		vector<string>::const_reverse_iterator rend = keys.rend();
+		rit = std::find_if(rit, rend, match);
+		
+		if (rit == rend)
+			return keys.end();
+		// This is correct and always safe.
+		// (See Meyer's Effective STL, Item 28.)
+		return (++rit).base();
+	}
+	catch (boost::regex_error & regerr) {
 		return keys.end();
-
-	// Search the vector of 'keys' from 'start' for one that matches the
-	// predicate 'match'. Searching can be forward or backward from start.
-	if (dir == FORWARD)
-		return std::find_if(start, keys.end(), match);
-
-	vector<string>::const_reverse_iterator rit(start);
-	vector<string>::const_reverse_iterator rend = keys.rend();
-	rit = std::find_if(rit, rend, match);
-
-	if (rit == rend)
-		return keys.end();
-	// This is correct and always safe.
-	// (See Meyer's Effective STL, Item 28.)
-	return (++rit).base();
+	}
 }
 
 

src/frontends/controllers/tests/Makefile.am

@@ -0,0 +1,23 @@
+include $(top_srcdir)/config/common.am
+
+EXTRA_DIST = pch.h test_biblio
+
+BUILT_SOURCES = $(PCH_FILE)
+
+TESTS = \
+	test_biblio
+
+check_PROGRAMS = \
+	biblio
+
+AM_CPPFLAGS += $(BOOST_INCLUDES)
+
+biblio_LDADD = $(BOOST_REGEX)
+biblio_SOURCES = \
+	biblio.C \
+	boost.C
+
+makeregfiles: ${check_PROGRAMS}
+	for all in ${check_PROGRAMS} ; do \
+		./$$all > ${srcdir}/regfiles/$$all ; \
+	done

src/frontends/controllers/tests/biblio.C

@@ -0,0 +1,95 @@
+#include <config.h>
+
+#include <iostream>
+#include <map>
+
+#include <boost/regex.hpp>
+
+using std::cout;
+using std::endl;
+using std::string;
+
+// Escape special chars.
+// All characters are literals except: '.|*?+(){}[]^$\'
+// These characters are literals when preceded by a "\", which is done here
+// This function is unfortunately copied from ../biblio.C, so we should
+// try to make sure to keep the two in sync.
+string const escape_special_chars(string const & expr)
+{
+	// Search for all chars '.|*?+(){}[^$]\'
+	// Note that '[' and '\' must be escaped.
+	// This is a limitation of boost::regex, but all other chars in BREs
+	// are assumed literal.
+	boost::regex reg("[].|*?+(){}^$\\[\\\\]");
+
+	// $& is a perl-like expression that expands to all of the current match
+	// The '$' must be prefixed with the escape character '\' for
+	// boost to treat it as a literal.
+	// Thus, to prefix a matched expression with '\', we use:
+	return boost::regex_replace(expr, reg, "\\\\$&");
+}
+
+
+typedef std::map<string, string> InfoMap;
+
+// A functor for use with std::find_if, used to ascertain whether a
+// data entry matches the required regex_
+// This class is unfortunately copied from ../biblio.C, so we should
+// try to make sure to keep the two in sync.
+class RegexMatch : public std::unary_function<string, bool>
+{
+public:
+	// re and icase are used to construct an instance of boost::RegEx.
+	// if icase is true, then matching is insensitive to case
+	RegexMatch(InfoMap const & m, string const & re, bool icase)
+		: map_(m), regex_(re, icase) {}
+
+	bool operator()(string const & key) const {
+		// the data searched is the key + its associated BibTeX/biblio
+		// fields
+		string data = key;
+		InfoMap::const_iterator info = map_.find(key);
+		if (info != map_.end())
+			data += ' ' + info->second;
+
+		// Attempts to find a match for the current RE
+		// somewhere in data.
+		return boost::regex_search(data, regex_);
+	}
+private:
+	InfoMap const map_;
+	mutable boost::regex regex_;
+};
+
+
+void test_escape_special_chars()
+{
+	cout << escape_special_chars("abcd") << endl;
+	cout << escape_special_chars("ab&cd") << endl;
+	cout << escape_special_chars(".|*?+(){}[]^$\"") << endl;
+	cout << escape_special_chars("..||**??++(()){{}}[[]]^^$$\\\\") << endl;
+}
+
+
+void test_RegexMatch()
+{
+	InfoMap im;
+	im["hello"] = "hei";
+
+	try {
+		RegexMatch rm(im, "h.*o", false);
+		
+		cout << rm("hello") << endl;
+		cout << rm("hei") << endl;
+	}
+	catch (boost::regex_error & regerr) {
+		cout << regerr.what() << endl;
+	}
+}
+
+
+int main()
+{
+	test_escape_special_chars();
+	test_RegexMatch();
+}

src/frontends/controllers/tests/boost.C

@@ -0,0 +1,33 @@
+/**
+ * \file boost.C
+ * This file is part of LyX, the document processor.
+ * Licence details can be found in the file COPYING.
+ *
+ * \author Lars Gullik Bjønnes
+ *
+ * Full author contact details are available in file CREDITS.
+ */
+
+#include <config.h>
+
+#include <boost/assert.hpp>
+
+#include <cstdlib>
+#include <exception>
+
+namespace boost {
+
+void throw_exception(std::exception const & /*e*/)
+{
+	BOOST_ASSERT(false);
+}
+
+
+void assertion_failed(char const * /*expr*/, char const * /*function*/,
+		      char const * /*file*/, long /*line*/)
+{
+	::abort();
+}
+
+
+}

src/frontends/controllers/tests/pch.h

@@ -0,0 +1,64 @@
+// -*- C++ -*-
+#include <config.h>
+
+#include <boost/assert.hpp>
+#include <boost/crc.hpp>
+// #include <boost/format.hpp> // mult def symbols problem (_1,_2 etc)
+#include <boost/regex.hpp>
+#include <boost/scoped_array.hpp>
+#include <boost/scoped_ptr.hpp>
+#include <boost/shared_ptr.hpp>
+#include <boost/signal.hpp>
+#include <boost/signals/trackable.hpp>
+#include <boost/version.hpp>
+
+#if BOOST_VERSION < 103301
+# include <boost/test/detail/nullstream.hpp>
+#else
+# include <boost/test/utils/nullstream.hpp>
+#endif
+
+#include <boost/tokenizer.hpp>
+#include <boost/utility.hpp>
+
+#include <fcntl.h>
+#ifdef HAVE_SYS_STAT_H
+# include <sys/stat.h>
+#endif
+#ifdef HAVE_SYS_TYPES_H
+# include <sys/types.h>
+#endif
+#include <time.h>
+#ifdef HAVE_SYS_SOCKET_H
+# include <sys/socket.h>
+#endif
+#ifndef _WIN32
+# include <sys/un.h>
+# include <sys/wait.h>
+# ifdef HAVE_UNISTD_H
+#  include <unistd.h>
+# endif
+#endif
+
+#include <algorithm>
+#include <cctype>
+#include <cerrno>
+#include <csignal>
+#include <cstddef>
+#include <cstdio>
+#include <cstdlib>
+#include <ctime>
+#include <deque>
+#include <fstream>
+#include <functional>
+#include <iostream>
+#include <istream>
+#include <iterator>
+#include <list>
+#include <map>
+#include <ostream>
+#include <queue>
+#include <sstream>
+#include <string>
+#include <utility>
+#include <vector>

src/frontends/controllers/tests/regfiles/biblio

@@ -0,0 +1,6 @@
+abcd
+ab&cd
+\.\|\*\?\+\(\)\{\}\[\]\^\$"
+\.\.\|\|\*\*\?\?\+\+\(\(\)\)\{\{\}\}\[\[\]\]\^\^\$\$\\\\
+1
+0

src/frontends/controllers/tests/test_biblio

@@ -0,0 +1,7 @@
+#!/bin/bash
+
+regfile=`cat ${srcdir}/regfiles/biblio`
+output=`./biblio`
+
+test "$regfile" = "$output"
+exit $?