Ticket #424: cfm-v1.diff

pygments/lexers/_mapping.py

@@ -46 +46 @@
     'CheetahLexer': ('pygments.lexers.templates', 'Cheetah', ('cheetah', 'spitfire'), ('*.tmpl', '*.spt'), ('application/x-cheetah', 'application/x-spitfire')),
     'CheetahXmlLexer': ('pygments.lexers.templates', 'XML+Cheetah', ('xml+cheetah', 'xml+spitfire'), (), ('application/xml+cheetah', 'application/xml+spitfire')),
     'ClojureLexer': ('pygments.lexers.agile', 'Clojure', ('clojure', 'clj'), ('*.clj',), ('text/x-clojure', 'application/x-clojure')),
+    'ColdfusionHtmlLexer': ('pygments.lexers.templates', 'Coldufsion HTML', ('cfm',), ('*.cfm', '*.cfml', '*.cfc'), ('application/x-coldfusion',)),
+    'ColdfusionLexer': ('pygments.lexers.templates', 'cfstatement', ('cfs',), (), ()),
     'CommonLispLexer': ('pygments.lexers.functional', 'Common Lisp', ('common-lisp', 'cl'), ('*.cl', '*.lisp', '*.el'), ('text/x-common-lisp',)),
     'CppLexer': ('pygments.lexers.compiled', 'C++', ('cpp', 'c++'), ('*.cpp', '*.hpp', '*.c++', '*.h++', '*.cc', '*.hh', '*.cxx', '*.hxx'), ('text/x-c++hdr', 'text/x-c++src')),
     'CppObjdumpLexer': ('pygments.lexers.asm', 'cpp-objdump', ('cpp-objdump', 'c++-objdumb', 'cxx-objdump'), ('*.cpp-objdump', '*.c++-objdump', '*.cxx-objdump'), ('text/x-cpp-objdump',)),

pygments/lexers/templates.py

@@ -38 +38 @@
            'MakoHtmlLexer', 'MakoXmlLexer', 'MakoJavascriptLexer',
            'MakoCssLexer', 'JspLexer', 'CheetahLexer', 'CheetahHtmlLexer',
            'CheetahXmlLexer', 'CheetahJavascriptLexer',
-           'EvoqueLexer', 'EvoqueHtmlLexer', 'EvoqueXmlLexer']
+           'EvoqueLexer', 'EvoqueHtmlLexer', 'EvoqueXmlLexer',
+           'ColdfusionLexer', 'ColdfusionHtmlLexer']
 
 
 class ErbLexer(Lexer):
@@ -1294 +1295 @@
     def __init__(self, **options):
         super(EvoqueXmlLexer, self).__init__(XmlLexer, EvoqueLexer,
                                              **options)
+
+class ColdfusionLexer(RegexLexer):
+    """
+    Coldfusion statements
+    """
+    name = 'cfstatement'
+    aliases = ['cfs']
+    filenames = []
+    mimetypes = []
+    flags = re.IGNORECASE | re.MULTILINE
+
+    tokens = {
+        'root': [
+            (r'//.*', Comment),
+            (r'\+\+|--', Operator),
+            (r'[-+*/^&=!]', Operator),
+            (r'<=|>=|<|>', Operator),
+            (r'mod\b', Operator),
+            (r'(eq|lt|gt|lte|gte|not|is|and|or)\b', Operator),
+            (r'\|\||&&', Operator),
+            (r'"', String.Double, 'string'),
+            # There is a special rule for allowing html in single quoted
+            # strings, evidently.
+            (r"'.*?'", String.Single),
+            (r'\d+', Number),
+            (r'(if|else|len|var|case|default|break|switch)\b', Keyword),
+            (r'([A-Za-z_$][A-Za-z0-9_.]*)\s*(\()', bygroups(Name.Function, Punctuation)),
+            (r'[A-Za-z_$][A-Za-z0-9_.]*', Name.Variable),
+            (r'[()\[\]{};:,.\\]', Punctuation),
+            (r'\s+', Text),
+        ],
+        'string': [
+            (r'""', String.Double),
+            (r'#.+?#', String.Interp),
+            (r'[^"#]+', String.Double),
+            (r'#', String.Double),
+            (r'"', String.Double, '#pop'),
+        ],
+    }
+
+class ColdfusionMarkupLexer(RegexLexer):
+    """
+    Coldfusion markup only
+    """
+    name = 'Coldfusion'
+    aliases = ['cf']
+    filenames = []
+    mimetypes = []
+
+    tokens = {
+        'root': [
+            (r'[^<]+', Other),
+            include('tags'),
+            (r'<[^<>]*', Other),
+        ],
+        'tags': [
+            (r'(?s)<!---.*?--->', Comment.Multiline),
+            (r'(?s)<!--.*?-->', Comment),
+            (r'<cfoutput.*?>', Name.Builtin, 'cfoutput'),
+            (r'(?s)(<cfscript.*?>)(.+?)(</cfscript.*?>)',
+             bygroups(Name.Builtin, using(ColdfusionLexer), Name.Builtin)),
+            # negative lookbehind is for strings with embedded >
+            (r'(?s)(</?cf(?:component|include|if|else|elseif|loop|return|'
+             r'dbinfo|dump|abort|location|invoke|throw|file|savecontent|'
+             r'mailpart|mail|header|content|zip|image|lock|argument|try|'
+             r'catch|break|directory|http|set|function|param)\b)(.*?)((?<!\\)>)',
+             bygroups(Name.Builtin, using(ColdfusionLexer), Name.Builtin)),
+        ],
+        'cfoutput': [
+            (r'[^#<]+', Other),
+            (r'(#)(.*?)(#)', bygroups(Punctuation, using(ColdfusionLexer),
+                                      Punctuation)),
+            #(r'<cfoutput.*?>', Name.Builtin, '#push'),
+            (r'</cfoutput.*?>', Name.Builtin, '#pop'),
+            include('tags'),
+            (r'(?s)<[^<>]*', Other),
+            (r'#', Other),
+        ],
+    }
+
+
+class ColdfusionHtmlLexer(DelegatingLexer):
+    """
+    Coldfusion markup in html
+    """
+    name = 'Coldufsion HTML'
+    aliases = ['cfm']
+    filenames = ['*.cfm', '*.cfml', '*.cfc']
+    mimetypes = ['application/x-coldfusion']
+
+    def __init__(self, **options):
+        super(ColdfusionHtmlLexer, self).__init__(HtmlLexer, ColdfusionMarkupLexer,
+                                                  **options)
+

(a) /dev/null vs. (b) b/tests/examplefiles/evil.cfm

@@ +1 @@
+<cfloop list="#expression# #reReplace(expression, "[([\])]", " ")#" delimiters=" +=-*/%##" index="token">
+<th<cfif request[resultkey].numErrors neq 0> class="failed"</cfif>>Errors</th>
+<td class="numeric<cfif request[resultkey].numErrors neq 0> failed</cfif>">#request[resultkey].numErrors#</td>
+<cfset loc.regex.plugin = "\<td class=""vt id col_0""\>(.*?)\<\/td\>[[:space:]]*\<td class=""vt col_1""([^\>]*)\>(.*?)\<\/td\>">
+<cfset loc.r = '<div class="errors-found">prepend firstname error1 append</div>'>
+<cfreturn '<select name="ModelUsers1[#arguments.property#]($month)" id="ModelUsers1-#arguments.property#-month"><option value="1">January</option><option value="2">February</option><option value="3">March</option><option value="4">April</option><option value="5">May</option><option value="6">June</option><option value="7">July</option><option value="8">August</option><option value="9">September</option><option value="10">October</option><option value="11" selected="selected">November</option><option value="12">December</option></select>'>