Merge pull request #768 from kachick/implement-matchdata-string,regex…

…p,offset,captures

Implement MatchData#{string,regexp,offset,captures}

lib-topaz/bootstrap.rb

@@ -19,6 +19,7 @@
 load_bootstrap.call("integer.rb")
 load_bootstrap.call("io.rb")
 load_bootstrap.call("kernel.rb")
+load_bootstrap.call("matchdata.rb")
 load_bootstrap.call("numeric.rb")
 load_bootstrap.call("process.rb")
 load_bootstrap.call("range.rb")

lib-topaz/matchdata.rb

@@ -0,0 +1,5 @@
+class MatchData
+  def offset(n)
+    [self.begin(n), self.end(n)]
+  end
+end

spec/tags/core/matchdata/offset_tags.txt

@@ -1,3 +1,2 @@
-fails:MatchData#offset returns a two element array with the begin and end of the nth match
 fails:MatchData#offset returns the offset for multi byte strings
 fails:MatchData#offset returns the offset for multi byte strings with unicode regexp

topaz/objects/regexpobject.py

@@ -150,9 +150,9 @@ def make_ctx(self, s, offset=0):
         endpos = len(s)
         return rsre_core.StrMatchContext(self.code, s, offset, endpos, self.flags)
 
-    def get_match_result(self, space, ctx, found):
+    def get_match_result(self, space, ctx, target, found):
         if found:
-            w_match = W_MatchDataObject(space, self, ctx)
+            w_match = W_MatchDataObject(space, self, ctx, target)
         else:
             w_match = space.w_nil
         space.globals.set(space, "$~", w_match)
@@ -210,7 +210,7 @@ def method_match_operator(self, space, w_s):
         s = Coerce.str(space, w_s)
         ctx = self.make_ctx(s)
         matched = rsre_core.search_context(ctx)
-        self.get_match_result(space, ctx, matched)
+        self.get_match_result(space, ctx, s, matched)
         if matched:
             return space.newint(ctx.match_start)
         else:
@@ -227,13 +227,13 @@ def method_match(self, space, w_s, w_offset=None):
             offset = 0
         ctx = self.make_ctx(s, offset)
         matched = rsre_core.search_context(ctx)
-        return self.get_match_result(space, ctx, matched)
+        return self.get_match_result(space, ctx, s, matched)
 
     @classdef.method("===", s="str")
     def method_eqeqeq(self, space, s):
         ctx = self.make_ctx(s)
         matched = rsre_core.search_context(ctx)
-        self.get_match_result(space, ctx, matched)
+        self.get_match_result(space, ctx, s, matched)
         return space.newbool(matched)
 
     @classdef.method("casefold?")
@@ -256,10 +256,11 @@ def method_options(self, space):
 class W_MatchDataObject(W_Object):
     classdef = ClassDef("MatchData", W_Object.classdef)
 
-    def __init__(self, space, regexp, ctx):
+    def __init__(self, space, regexp, ctx, target):
         W_Object.__init__(self, space)
         self.regexp = regexp
         self.ctx = ctx
+        self.target = target
         self._flatten_cache = None
 
     def size(self):
@@ -290,6 +291,16 @@ def get_span(self, n):
         assert idx >= 0
         return fmarks[idx], fmarks[idx + 1]
 
+    @classdef.method("regexp")
+    def method_regexp(self, space):
+        return self.regexp
+
+    @classdef.method("string")
+    def method_string(self, space):
+        res = space.newstr_fromstr(self.target)
+        space.send(res, "freeze")
+        return res
+
     @classdef.method("[]", n="int")
     def method_subscript(self, space, n):
         if n == 0:
@@ -303,6 +314,13 @@ def method_subscript(self, space, n):
         else:
             return space.w_nil
 
+    @classdef.method("captures")
+    def method_captures(self, space):
+        res_w = []
+        for i in xrange(1, self.size()):
+            res_w.append(space.send(self, "[]", [space.newint(i)]))
+        return space.newarray(res_w)
+
     @classdef.method("to_a")
     def method_to_a(self, space):
         res_w = []

topaz/objects/stringobject.py

@@ -668,7 +668,7 @@ def method_split(self, space, w_sep=None, limit=0):
             last = 0
             string = space.str_w(self)
             ctx = w_sep.make_ctx(string)
-            w_match = w_sep.get_match_result(space, ctx, found=True)
+            w_match = w_sep.get_match_result(space, ctx, string, found=True)
 
             while limit <= 0 or n + 1 < limit:
                 if not self.search_context(space, ctx):
@@ -929,7 +929,7 @@ def scan_regexp(self, space, w_pattern):
         ctx = w_pattern.make_ctx(string)
 
         while last < len(string) and self.search_context(space, ctx):
-            w_matchdata = w_pattern.get_match_result(space, ctx, found=True)
+            w_matchdata = w_pattern.get_match_result(space, ctx, string, found=True)
             if w_matchdata.size() > 1:
                 matches_w = []
                 for num in xrange(1, w_matchdata.size(), 1):
@@ -1021,7 +1021,7 @@ def gsub_regexp(self, space, w_pattern, replacement, w_hash, block, first_only):
         string = space.str_w(self)
         ctx = w_pattern.make_ctx(string)
 
-        w_matchdata = w_pattern.get_match_result(space, ctx, found=True)
+        w_matchdata = w_pattern.get_match_result(space, ctx, string, found=True)
         replacement_parts = None
         if replacement is not None and "\\" in replacement:
             replacement_parts = [s for s in replacement.split("\\") if s]