Better scrubbing of file names to allow non-Latin alphabets.

vinumeris · Jan 14, 2015 · cfa5b54 · cfa5b54
1 parent 057e45c
commit cfa5b54
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 6 deletions.
diff --git a/common/src/main/java/lighthouse/protocol/LHUtils.java b/common/src/main/java/lighthouse/protocol/LHUtils.java
@@ -49,7 +49,7 @@ public static boolean pledgeAppearsInClaim(Project forProject, LHProtos.Pledge p
     }
 
     public static String titleToUrlString(String s) {
-        return s.replaceAll("[^a-zA-Z0-9 ]", "").replaceAll(" ", "-").toLowerCase();
+        return s.replaceAll("[!@#$%^&*()_+\\\\';:?=/.,\\[\\] \\n]", "-").replaceAll("-+", "-").replaceAll("(^-+|-+$)", "").toLowerCase();
     }
 
     public static boolean isUnix() {

diff --git a/common/src/test/java/lighthouse/protocol/ProjectTest.java b/common/src/test/java/lighthouse/protocol/ProjectTest.java
@@ -281,11 +281,6 @@ public void completeContract() throws Exception {
         assertEquals(2, contract.getInputs().size());
     }
 
-    @Test
-    public void urlScrubber() throws Exception {
-        assertEquals("a-really-cool-20-title-with-lots-asdf-of-weird--chars", LHUtils.titleToUrlString("A really $cool %20 Title with ;;lots asdf\n of weird // chars"));
-    }
-
     @Test
     public void authKeys() throws Exception {
         details = Project.makeDetails(
@@ -301,4 +296,12 @@ public void authKeys() throws Exception {
             // Expected.
         }
     }
+
+    @Test
+    public void titleToPath() {
+        assertEquals("bbc-14-01-2015-eu-lawyer-approves-ecb-bond-buying-programme", titleToUrlString("BBC 14/01/2015 EU lawyer approves ECB bond-buying programme"));
+        assertEquals("bob-http-example-com-1378-foo-bar", titleToUrlString("Bob http://example.com:1378/?foo=bar"));
+        assertEquals("a-really-cool-20-title-with-lots-asdf-of-weird-chars", titleToUrlString("A really $cool %20 Title with ;;lots asdf\n of weird // chars"));
+        assertEquals("български-език", titleToUrlString("български език"));
+    }
 }