Get the substrings out at the same time as computing the ranges.

app/Main.hs

@@ -65,4 +65,4 @@ replaceLeavesWithWordBranches source term = replaceIn source 0 term
       Fixed f -> Fixed $ replaceIn substring (start range) <$> f
       Keyed k -> Keyed $ replaceIn substring (start range) <$> k
       _ -> syntax
-    makeLeaf source startIndex lineRange categories range = Info range lineRange categories :< Leaf (substring (offsetRange (negate startIndex) range) source)
+    makeLeaf source startIndex lineRange categories (range, substring) = Info range lineRange categories :< Leaf substring

src/Range.hs

@@ -14,13 +14,13 @@ totalRange list = Range 0 $ length list
 offsetRange :: Int -> Range -> Range
 offsetRange i (Range start end) = Range (i + start) (i + end)
 
-rangesAndWordsFrom :: Int -> String -> [Range]
+rangesAndWordsFrom :: Int -> String -> [(Range, String)]
 rangesAndWordsFrom startIndex string = case break (not . isWord) string of
   ([], []) -> []
   ([], rest) -> rangesAndWordsAfterWhitespace startIndex rest
-  (word, rest) -> (Range startIndex $ startIndex + length word) : rangesAndWordsAfterWhitespace (startIndex + length word) rest
+  (word, rest) -> (Range startIndex $ startIndex + length word, word) : rangesAndWordsAfterWhitespace (startIndex + length word) rest
   where
-    rangesAndWordsAfterWhitespace startIndex string | (whitespace, rest) <- break isWord string = rangesOfWordsFrom (startIndex + length whitespace) rest
+    rangesAndWordsAfterWhitespace startIndex string | (whitespace, rest) <- break isWord string = rangesAndWordsFrom (startIndex + length whitespace) rest
     -- | Is this a word character?
     -- | Word characters are defined as in [Ruby’s `\p{Word}` syntax](http://ruby-doc.org/core-2.1.1/Regexp.html#class-Regexp-label-Character+Properties), i.e.:
     -- | > A member of one of the following Unicode general category _Letter_, _Mark_, _Number_, _Connector_Punctuation_

test/Spec.hs

@@ -218,22 +218,22 @@ main = hspec $ do
       rangesAndWordsFrom 0 "  \t\n  " `shouldBe` []
 
     it "should produce a list containing the range of the string for a single-word string" $
-      rangesAndWordsFrom 0 "word" `shouldBe` [ Range 0 4 ]
+      rangesAndWordsFrom 0 "word" `shouldBe` [ (Range 0 4, "word") ]
 
     it "should produce a list of ranges for whitespace-separated words" $
-      rangesAndWordsFrom 0 "wordOne wordTwo" `shouldBe` [ Range 0 7, Range 8 15 ]
+      rangesAndWordsFrom 0 "wordOne wordTwo" `shouldBe` [ (Range 0 7, "wordOne"), (Range 8 15, "wordTwo") ]
 
     it "should skip multiple whitespace characters" $
-      rangesAndWordsFrom 0 "a  b" `shouldBe` [ Range 0 1, Range 3 4 ]
+      rangesAndWordsFrom 0 "a  b" `shouldBe` [ (Range 0 1, "a"), (Range 3 4, "b") ]
 
     it "should skip whitespace at the start" $
-      rangesAndWordsFrom 0 "  a b" `shouldBe` [ Range 2 3, Range 4 5 ]
+      rangesAndWordsFrom 0 "  a b" `shouldBe` [ (Range 2 3, "a"), (Range 4 5, "b") ]
 
     it "should skip whitespace at the end" $
-      rangesAndWordsFrom 0 "a b  " `shouldBe` [ Range 0 1, Range 2 3 ]
+      rangesAndWordsFrom 0 "a b  " `shouldBe` [ (Range 0 1, "a"), (Range 2 3, "b") ]
 
     it "should produce ranges offset by its start index" $
-      rangesAndWordsFrom 100 "a b" `shouldBe` [ Range 100 101, Range 102 103 ]
+      rangesAndWordsFrom 100 "a b" `shouldBe` [ (Range 100 101, "a"), (Range 102 103, "b") ]
 
     where
       rightRowText text = rightRow [ Text text ]