module Twitter.Xml
where
-import Data.Maybe
+import Test.HUnit
import Text.Regex (mkRegex, subRegex)
import Text.XML.HaXml
get_char_data _ = Nothing
+-- |A 'CFilter' returning all top-level <status> elements.
+-- The name is due to the fact that if we retrieve more than
+-- one status, they will be wrapped in a <statuses> tag, and
+-- thus not be top-level.
+single_status :: CFilter
+single_status = (tag "status")
+
-- |A 'CFilter' returning all <status> tags within <statuses>.
all_statuses :: CFilter
all_statuses = (tag "statuses" /> tag "status")
-- character represented by that entity.
xml_entities :: [(String, String)]
xml_entities = [("[lr]dquo", "\""),
+ ("quot", "\""),
("[mn]dash", "-"),
("nbsp", " "),
("#8217", "'"),
("amp", "&"),
("lt", "<"),
- ("gt", ">")]
+ ("gt", ">"),
+ ("#8230", "..."),
+ ("hellip", "...")]
-- |Replace all of the XML entities in target.
replace_entities :: String -> String
replacement = (replacements !! 0)
from = "&" ++ (fst replacement) ++ ";"
to = (snd replacement)
+
+
+
+xml_tests :: [Test]
+xml_tests = [ test_replace_entities ]
+
+
+test_replace_entities :: Test
+test_replace_entities =
+ TestCase $ assertEqual "All entities are replaced correctly." expected_text actual_text
+ where
+ actual_text = (replace_entities ""The moon is gay……" said <insert the current president of the United States of America>. “It’s OK—–he’s not a real doctor.”")
+ expected_text = "\"The moon is gay......\" said <insert the current president of the United States of America>. \"It's OK--he's not a real doctor.\""