added re_group function

src/calibre/utils/formatter.py

@@ -473,10 +473,13 @@ class TemplateFormatter(string.Formatter):
                             (func.arg_count > 1 and func.arg_count != len(args)+1):
                         raise ValueError('Incorrect number of arguments for function '+ fmt[0:p])
                     if func.arg_count == 1:
-                        val = func.eval_(self, self.kwargs, self.book, self.locals, val).strip()
+                        val = func.eval_(self, self.kwargs, self.book, self.locals, val)
+                        if self.strip_results:
+                            val = val.strip()
                     else:
-                        val = func.eval_(self, self.kwargs, self.book, self.locals,
-                                        val, *args).strip()
+                        val = func.eval_(self, self.kwargs, self.book, self.locals, val, *args)
+                        if self.strip_results:
+                            val = val.strip()
                 else:
                     return _('%s: unknown function')%fname
         if val:
@@ -487,10 +490,12 @@ class TemplateFormatter(string.Formatter):
 
     def evaluate(self, fmt, args, kwargs):
         if fmt.startswith('program:'):
-            ans = self._eval_program(None, fmt[8:], self.column_name)
+            ans = self._eval_program(kwargs.get('$', None), fmt[8:], self.column_name)
         else:
             ans = self.vformat(fmt, args, kwargs)
-        return self.compress_spaces.sub(' ', ans).strip()
+        if self.strip_results:
+            return self.compress_spaces.sub(' ', ans).strip()
+        return ans
 
     ########## a formatter that throws exceptions ############
 
@@ -505,7 +510,9 @@ class TemplateFormatter(string.Formatter):
     ########## a formatter guaranteed not to throw an exception ############
 
     def safe_format(self, fmt, kwargs, error_value, book,
-                    column_name=None, template_cache=None):
+                    column_name=None, template_cache=None,
+                    strip_results=True):
+        self.strip_results = strip_results
         self.column_name = column_name
         self.template_cache = template_cache
         self.kwargs = kwargs
@@ -513,9 +520,9 @@ class TemplateFormatter(string.Formatter):
         self.composite_values = {}
         self.locals = {}
         try:
-            ans = self.evaluate(fmt, [], kwargs).strip()
+            ans = self.evaluate(fmt, [], kwargs)
         except Exception as e:
-            if DEBUG and getattr(e, 'is_locking_error', False):
+            if DEBUG: # and getattr(e, 'is_locking_error', False):
                 traceback.print_exc()
             ans = error_value + ' ' + e.message
         return ans

src/calibre/utils/formatter_functions.py

@@ -553,6 +553,39 @@ class BuiltinRe(BuiltinFormatterFunction):
     def evaluate(self, formatter, kwargs, mi, locals, val, pattern, replacement):
         return re.sub(pattern, replacement, val, flags=re.I)
 
+class BuiltinReGroup(BuiltinFormatterFunction):
+    name = 're_group'
+    arg_count = -1
+    category = 'String manipulation'
+    __doc__ = doc = _('re_group(val, pattern, template_for_group_1, for_group_2, ...) -- '
+            'return a string made by applying the reqular expression pattern '
+            'to the val and replacing each matched instance with the string '
+            'computed by replacing each matched group by the value returned '
+            'by the corresponding template. The original matched value for the '
+            'group is available as $. In template program mode, like for '
+            'the template and the eval functions, you use [[ for { and ]] for }.'
+            ' The following example in template program mode looks for series '
+            'with more than one word and uppercases the first word: '
+            "{series:'re_group($, \"(\S* )(.*)\", \"[[$:uppercase()]]\", \"[[$]]\")'}")
+
+    def evaluate(self, formatter, kwargs, mi, locals, *args):
+        from formatter import EvalFormatter
+
+        if (len(args) < 2):
+            return('re_group: incorrect number of arguments')
+        def repl(mo):
+            res = ''
+            for dex in range(1, mo.lastindex+1):
+                if len(args) > dex + 1:
+                    template = args[dex+1].replace('[[', '{').replace(']]', '}')
+                    res += EvalFormatter().safe_format(template,
+                                       {'$': mo.group(dex)}, 'EVAL', None,
+                                       strip_results=False)
+                else:
+                    res += mo.group(dex)
+            return res
+        return re.sub(args[1], repl, args[0], flags=re.I)
+
 class BuiltinSwapAroundComma(BuiltinFormatterFunction):
     name = 'swap_around_comma'
     arg_count = 1
@@ -1341,8 +1374,8 @@ _formatter_builtins = [
     BuiltinListSort(), BuiltinListUnion(), BuiltinLookup(),
     BuiltinLowercase(), BuiltinMultiply(), BuiltinNot(),
     BuiltinOndevice(), BuiltinOr(), BuiltinPrint(), BuiltinRawField(),
-    BuiltinRe(), BuiltinSelect(), BuiltinSeriesSort(), BuiltinShorten(),
-    BuiltinStrcat(), BuiltinStrcatMax(),
+    BuiltinRe(), BuiltinReGroup(), BuiltinSelect(), BuiltinSeriesSort(),
+    BuiltinShorten(), BuiltinStrcat(), BuiltinStrcatMax(),
     BuiltinStrcmp(), BuiltinStrInList(), BuiltinStrlen(), BuiltinSubitems(),
     BuiltinSublist(),BuiltinSubstr(), BuiltinSubtract(), BuiltinSwapAroundComma(),
     BuiltinSwitch(), BuiltinTemplate(), BuiltinTest(), BuiltinTitlecase(),