Whitespace cleanup.

[dead/halcyon.git] / src / Twitter / Xml.hs

diff --git a/src/Twitter/Xml.hs b/src/Twitter/Xml.hs
index e284a0da5ae6601c2a8fac8ee0a9730989325773..871d2164dee05737e9861d45c2e400dfde76e741 100644 (file)
--- a/src/Twitter/Xml.hs
+++ b/src/Twitter/Xml.hs
@@ -10,9 +10,9 @@ import Text.XML.HaXml
 -- |Returns the 'CharData' contained within the given 'Content', or
 -- 'Nothing' if no acceptable CharData was found. It will parse either
 -- a 'CString' ('String') or 'CRef' (XML entity reference).
-get_char_data :: Content -> (Maybe CharData)
-get_char_data (CString _ cd) = Just cd
-get_char_data (CRef ref) = Just (verbatim ref) -- Entities.
+get_char_data :: Content i -> (Maybe CharData)
+get_char_data (CString _ cd _) = Just cd
+get_char_data (CRef ref _) = Just (verbatim ref) -- Entities.
 get_char_data _ = Nothing
 
 
@@ -20,36 +20,46 @@ get_char_data _ = Nothing
 -- 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 :: CFilter i
 single_status = (tag "status")
 
 -- |A 'CFilter' returning all <status> tags within <statuses>.
-all_statuses :: CFilter
+all_statuses :: CFilter i
 all_statuses = (tag "statuses" /> tag "status")
 
 -- |Finds the text of the <id> element contained within some other
 -- content. Called unique_id here because status_id is used elsewhere.
-unique_id :: CFilter
+unique_id :: CFilter i
 unique_id = keep /> (tag "id") /> txt
 
 -- |Finds the text of the <created_at> element contained within some
 -- other element.
-status_created_at :: CFilter
+status_created_at :: CFilter i
 status_created_at = keep /> (tag "created_at") /> txt
 
 -- |Finds the text of the <text> element contained within some
 -- other element.
-status_text :: CFilter
+status_text :: CFilter i
 status_text = keep /> (tag "text") /> txt
 
 -- |Finds the XML of the <user> element contained within some other
 -- element.
-status_user :: CFilter
+status_user :: CFilter i
 status_user = keep /> (tag "user")
 
+-- | Finds the text of the <retweeted> element contained within some
+--   other element.
+status_retweeted :: CFilter i
+status_retweeted = keep /> (tag "retweeted") /> txt
+
+-- | Finds the text of the <in_reply_to_status_id> element contained
+--   within some other element.
+status_reply_to_status_id :: CFilter i
+status_reply_to_status_id = keep /> (tag "in_reply_to_status_id") /> txt
+
 -- |Finds the text of the <screen_name> element contained within some
 -- other element.
-user_screen_name :: CFilter
+user_screen_name :: CFilter i
 user_screen_name = keep /> (tag "screen_name") /> txt
 
 -- |A wrapper around the 'read' function which returns either Nothing
@@ -69,15 +79,20 @@ entity_from_codepoint codepoint =
       Just num -> [(chr num)]
 
 
--- |A list of tuples whose first entry is a regular expression
--- matching XML entities, and whose second entry is the ASCII
--- character represented by that entity.
+-- | A list of tuples whose first entry is a regular expression
+--   matching XML entities, and whose second entry is the ASCII
+--   character represented by that entity.
+--
+--   For some reason, ampersands are escaped twice in the status
+--   text. Rather than unescape everything twice, we just stick "amp"
+--   in the list again.
 xml_entities :: [(String, String)]
 xml_entities = [("[lr]dquo", "\""),
                 ("quot",     "\""),
                 ("[mn]dash", "-"),
                 ("nbsp",     " "),
                 ("amp",      "&"),
+                ("amp",      "&"),
                 ("lt",       "<"),
                 ("gt",       ">"),
                 ("hellip",   "…")]
@@ -120,7 +135,7 @@ unescape_recursive replacements target =
 
 
 xml_tests :: [Test]
-xml_tests = [ test_replace_entities ]
+xml_tests = [ test_replace_entities, test_double_unescape ]
 
 
 test_replace_entities :: Test
@@ -129,3 +144,11 @@ test_replace_entities =
     where
       actual_text = (replace_entities "&quot;The moon is gay&#8230;&hellip;&quot; said &lt;insert the current president of the United States of America&gt;. &ldquo;It&#8217;s OK&mdash;&ndash;he&#8217;s not a real doctor.&rdquo;")
       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.\""
+
+
+test_double_unescape :: Test
+test_double_unescape =
+    TestCase $ assertEqual "The status text is unescaped twice." expected_text actual_text
+    where
+      actual_text = (replace_entities "As a kid, I'd pull a girl's hair to let her know I liked her, but now that I'm older &amp;amp; wiser I simply hit her with my car.")
+      expected_text = "As a kid, I'd pull a girl's hair to let her know I liked her, but now that I'm older & wiser I simply hit her with my car."