Updated examples with better comment on expressions. Updated test program to test for apparent Rogur Wave bug (and it's workaround).

[SVN r11390]

example/snippets/regex_grep_example_1.cpp

@@ -31,7 +31,29 @@
 
 typedef std::map<std::string, int, std::less<std::string> > map_type;
 
-boost::regex expression("^(template[[:space:]]*<[^;:{]+>[[:space:]]*)?(class|struct)[[:space:]]*(\\<\\w+\\>([[:blank:]]*\\([^)]*\\))?[[:space:]]*)*(\\<\\w*\\>)[[:space:]]*(<[^;:{]+>)?[[:space:]]*(\\{|:[^;\\{()]*\\{)");
+const char* re = 
+   // possibly leading whitespace:   
+   "^[[:space:]]*" 
+   // possible template declaration:
+   "(template[[:space:]]*<[^;:{]+>[[:space:]]*)?"
+   // class or struct:
+   "(class|struct)[[:space:]]*" 
+   // leading declspec macros etc:
+   "("
+      "\\<\\w+\\>"
+      "("
+         "[[:blank:]]*\\([^)]*\\)"
+      ")?"
+      "[[:space:]]*"
+   ")*" 
+   // the class name
+   "(\\<\\w*\\>)[[:space:]]*" 
+   // template specialisation parameters
+   "(<[^;:{]+>)?[[:space:]]*"
+   // terminate in { or :
+   "(\\{|:[^;\\{()]*\\{)";
+
+boost::regex expression(re);
 
 class IndexClassesPred
 {

example/snippets/regex_grep_example_2.cpp

@@ -32,7 +32,30 @@
 
 typedef std::map<std::string, int, std::less<std::string> > map_type;
 
-boost::regex expression("^(template[[:space:]]*<[^;:{]+>[[:space:]]*)?(class|struct)[[:space:]]*(\\<\\w+\\>([[:blank:]]*\\([^)]*\\))?[[:space:]]*)*(\\<\\w*\\>)[[:space:]]*(<[^;:{]+>)?[[:space:]]*(\\{|:[^;\\{()]*\\{)");
+const char* re = 
+   // possibly leading whitespace:   
+   "^[[:space:]]*" 
+   // possible template declaration:
+   "(template[[:space:]]*<[^;:{]+>[[:space:]]*)?"
+   // class or struct:
+   "(class|struct)[[:space:]]*" 
+   // leading declspec macros etc:
+   "("
+      "\\<\\w+\\>"
+      "("
+         "[[:blank:]]*\\([^)]*\\)"
+      ")?"
+      "[[:space:]]*"
+   ")*" 
+   // the class name
+   "(\\<\\w*\\>)[[:space:]]*" 
+   // template specialisation parameters
+   "(<[^;:{]+>)?[[:space:]]*"
+   // terminate in { or :
+   "(\\{|:[^;\\{()]*\\{)";
+
+
+boost::regex expression(re);
 map_type class_index;
 std::string::const_iterator base;
 

example/snippets/regex_grep_example_3.cpp

@@ -33,6 +33,29 @@
 
 typedef std::map<std::string, int, std::less<std::string> > map_type;
 
+const char* re = 
+   // possibly leading whitespace:   
+   "^[[:space:]]*" 
+   // possible template declaration:
+   "(template[[:space:]]*<[^;:{]+>[[:space:]]*)?"
+   // class or struct:
+   "(class|struct)[[:space:]]*" 
+   // leading declspec macros etc:
+   "("
+      "\\<\\w+\\>"
+      "("
+         "[[:blank:]]*\\([^)]*\\)"
+      ")?"
+      "[[:space:]]*"
+   ")*" 
+   // the class name
+   "(\\<\\w*\\>)[[:space:]]*" 
+   // template specialisation parameters
+   "(<[^;:{]+>)?[[:space:]]*"
+   // terminate in { or :
+   "(\\{|:[^;\\{()]*\\{)";
+
+
 class class_index
 {
    boost::regex expression;
@@ -45,7 +68,7 @@ public:
    void IndexClasses(const std::string& file);
    class_index()
       : index(),
-        expression("^(template[[:space:]]*<[^;:{]+>[[:space:]]*)?(class|struct)[[:space:]]*(\\<\\w+\\>([[:blank:]]*\\([^)]*\\))?[[:space:]]*)*(\\<\\w*\\>)[[:space:]]*(<[^;:{]+>)?[[:space:]]*(\\{|:[^;\\{()]*\\{)")
+        expression(re)
         {}
 };
 

example/snippets/regex_grep_example_4.cpp

@@ -33,6 +33,29 @@
 
 typedef std::map<std::string, int, std::less<std::string> > map_type;
 
+const char* re = 
+   // possibly leading whitespace:   
+   "^[[:space:]]*" 
+   // possible template declaration:
+   "(template[[:space:]]*<[^;:{]+>[[:space:]]*)?"
+   // class or struct:
+   "(class|struct)[[:space:]]*" 
+   // leading declspec macros etc:
+   "("
+      "\\<\\w+\\>"
+      "("
+         "[[:blank:]]*\\([^)]*\\)"
+      ")?"
+      "[[:space:]]*"
+   ")*" 
+   // the class name
+   "(\\<\\w*\\>)[[:space:]]*" 
+   // template specialisation parameters
+   "(<[^;:{]+>)?[[:space:]]*"
+   // terminate in { or :
+   "(\\{|:[^;\\{()]*\\{)";
+
+
 class class_index
 {
    boost::regex expression;
@@ -47,7 +70,7 @@ public:
    void IndexClasses(const std::string& file);
    class_index()
       : index(),
-        expression("^(template[[:space:]]*<[^;:{]+>[[:space:]]*)?(class|struct)[[:space:]]*(\\<\\w+\\>([[:blank:]]*\\([^)]*\\))?[[:space:]]*)*(\\<\\w*\\>)[[:space:]]*(<[^;:{]+>)?[[:space:]]*(\\{|:[^;\\{()]*\\{)")
+        expression(re)
         {}
 };
 

example/snippets/regex_search_example.cpp

@@ -31,7 +31,30 @@
 
 typedef std::map<std::string, int, std::less<std::string> > map_type;
 
-boost::regex expression("^(template[[:space:]]*<[^;:{]+>[[:space:]]*)?(class|struct)[[:space:]]*(\\<\\w+\\>([[:blank:]]*\\([^)]*\\))?[[:space:]]*)*(\\<\\w*\\>)[[:space:]]*(<[^;:{]+>)?[[:space:]]*(\\{|:[^;\\{()]*\\{)");
+const char* re = 
+   // possibly leading whitespace:   
+   "^[[:space:]]*" 
+   // possible template declaration:
+   "(template[[:space:]]*<[^;:{]+>[[:space:]]*)?"
+   // class or struct:
+   "(class|struct)[[:space:]]*" 
+   // leading declspec macros etc:
+   "("
+      "\\<\\w+\\>"
+      "("
+         "[[:blank:]]*\\([^)]*\\)"
+      ")?"
+      "[[:space:]]*"
+   ")*" 
+   // the class name
+   "(\\<\\w*\\>)[[:space:]]*" 
+   // template specialisation parameters
+   "(<[^;:{]+>)?[[:space:]]*"
+   // terminate in { or :
+   "(\\{|:[^;\\{()]*\\{)";
+
+
+boost::regex expression(re);
 
 void IndexClasses(map_type& m, const std::string& file)
 {

test/regress/parse.cpp

@@ -37,11 +37,16 @@ using namespace boost;
 flag_info flag_data[] = {
                           { BOOST_RE_STR("REG_BASIC"), 9, REG_BASIC, 0 },
                           { BOOST_RE_STR("REG_EXTENDED"), 12, REG_EXTENDED, 0 },
+                          { BOOST_RE_STR("REG_ESCAPE_IN_LISTS"), 19, REG_ESCAPE_IN_LISTS, 0 },
                           { BOOST_RE_STR("REG_ICASE"), 9, REG_ICASE, 0 },
                           { BOOST_RE_STR("REG_NOSUB"), 9, REG_NOSUB, 0 },
                           { BOOST_RE_STR("REG_NEWLINE"), 11, REG_NEWLINE, 0 },
-                          { BOOST_RE_STR("REG_NEWLINE"), 11, REG_NEWLINE, 0 },
+                          { BOOST_RE_STR("REG_NOCOLLATE"), 13, REG_NOCOLLATE, 0 },
                           { BOOST_RE_STR("REG_NOSPEC"), 10, REG_NOSPEC, 0 },
+                          { BOOST_RE_STR("REG_NEWLINE_ALT"), 15, REG_NEWLINE_ALT , 0 },
+                          { BOOST_RE_STR("REG_PERL"), 8, REG_PERL, 0 },
+                          { BOOST_RE_STR("REG_AWK"), 7, REG_AWK, 0 },
+                          { BOOST_RE_STR("REG_EGREP"), 9, REG_EGREP, 0 },
 
                           { BOOST_RE_STR("REG_NOTBOL"), 10, REG_NOTBOL, 1 },
                           { BOOST_RE_STR("REG_NOTEOL"), 10, REG_NOTEOL, 1 },

test/regress/tests.txt

@@ -503,8 +503,10 @@ a(b+|((c)*))+d abcd 0 4 2 3 2 3 2 3
 '([^\\']|\\.)*' '\\n' 0 4 1 3
 
 ; now try and test some unicode specific characters:
-- match_default normal REG_EXTENDED REG_UNICODE_ONLY
+- match_default normal REG_PERL REG_UNICODE_ONLY
 [[:unicode:]]+  a\0300\0400z 1 3
+[\x10-\xff] \39135\12409 -1 -1
+[\01-\05]{5} \36865\36865\36865\36865\36865 -1 -1
 
 ; finally try some case insensitive matches:
 - match_default normal REG_EXTENDED REG_ICASE
@@ -880,3 +882,4 @@ a+(?#b+)b+ xaaabbba 1 7
 
 
 
+