LUCENE-5517: stricter parsing for hunspell parseFlag

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1576389 13f79535-47bb-0310-9956-ffa450edef68

lucene/analysis/common/src/java/org/apache/lucene/analysis/hunspell/Dictionary.java

@@ -822,7 +822,11 @@ public class Dictionary {
      * @return Parsed flag
      */
     char parseFlag(String rawFlag) {
-      return parseFlags(rawFlag)[0];
+      char flags[] = parseFlags(rawFlag);
+      if (flags.length != 1) {
+        throw new IllegalArgumentException("expected only one flag, got: " + rawFlag);
+      }
+      return flags[0];
     }
 
     /**

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/TestAllDictionaries.java

@@ -47,7 +47,7 @@ public class TestAllDictionaries extends LuceneTestCase {
     "bg_BG.zip",                 "bg_BG.dic",           "bg_BG.aff",
     "ca_ANY.zip",                "catalan.dic",         "catalan.aff",
     "ca_ES.zip",                 "ca_ES.dic",           "ca_ES.aff",
-    "cop_EG.zip",                "cop_EG.dic",          "cop_EG.aff",
+// BUG: broken flag "cop_EG.zip",                "cop_EG.dic",          "cop_EG.aff",
     "cs_CZ.zip",                 "cs_CZ.dic",           "cs_CZ.aff",
     "cy_GB.zip",                 "cy_GB.dic",           "cy_GB.aff",
     "da_DK.zip",                 "da_DK.dic",           "da_DK.aff",

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/TestAllDictionaries2.java

@@ -61,7 +61,7 @@ public class TestAllDictionaries2 extends LuceneTestCase {
     "ceske_slovniky_pro_kontrolu_pravopisu-1.0.4-tb+sm+fx.xpi",                       "dictionaries/cs.dic",                "dictionaries/cs.aff",
     "chichewa_spell_checker-0.3-fx+tb+fn+sm+sb.xpi",                                  "dictionaries/ny_MW.dic",             "dictionaries/ny_MW.aff",
     "corrector_de_galego-13.10.0-fn+sm+tb+fx.xpi",                                    "dictionaries/gl_ES.dic",             "dictionaries/gl_ES.aff",
-    "corrector_orthographic_de_interlingua-6.0-fn+sm+tb+fx.xpi",                      "dictionaries/ia-ia.dic",             "dictionaries/ia-ia.aff",
+//BUG: broken flags "corrector_orthographic_de_interlingua-6.0-fn+sm+tb+fx.xpi",                      "dictionaries/ia-ia.dic",             "dictionaries/ia-ia.aff",
     "corrector_ortografico_aragones-0.2-fx+tb+sm.xpi",                                "dictionaries/an_ES.dic",             "dictionaries/an_ES.aff",
     "croatian_dictionary_-_hrvatski_rjecnik-1.0.1-firefox+thunderbird+seamonkey.xpi", "dictionaries/hr.dic",                "dictionaries/hr.aff",
     "croatian_dictionary_hrvatski_rjecnik-1.0.9-an+fx+fn+tb+sm.xpi",                  "dictionaries/hr-HR.dic",             "dictionaries/hr-HR.aff",
@@ -149,7 +149,7 @@ public class TestAllDictionaries2 extends LuceneTestCase {
 //BUG: broken file (hunspell refuses to load, too)    "thamizha_solthiruthitamil_spellchecker-0.8-fx+tb.xpi",                           "dictionaries/ta_IN.dic",             "dictionaries/ta_IN.aff",
     "tsonga_spell_checker-20110323-tb+sm+fx+fn.xpi",                                  "dictionaries/ts-ZA.dic",             "dictionaries/ts-ZA.aff",
     "tswana_spell_checker-20110323-tb+sm+fx+fn.xpi",                                  "dictionaries/tn-ZA.dic",             "dictionaries/tn-ZA.aff",
-    "turkce_yazm_denetimi-3.5-sm+tb+fx.xpi",                                          "dictionaries/tr.dic",                "dictionaries/tr.aff",
+//BUG: missing FLAG declaration "turkce_yazm_denetimi-3.5-sm+tb+fx.xpi",                                          "dictionaries/tr.dic",                "dictionaries/tr.aff",
     "turkmen_spell_checker_dictionary-0.1.6-tb+fx+sm.xpi",                            "dictionaries/tk_TM.dic",             "dictionaries/tk_TM.aff",
     "ukrainian_dictionary-1.7.0-sm+an+fx+fn+tb.xpi",                                  "dictionaries/uk-UA.dic",             "dictionaries/uk-UA.aff",
     "united_states_english_spellchecker-7.0.1-sm+tb+fx+an.xpi",                       "dictionaries/en-US.dic",             "dictionaries/en-US.aff",

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/TestDictionary.java

@@ -131,6 +131,22 @@ public class TestDictionary extends LuceneTestCase {
     dictStream.close();
   }
   
+  // malformed flags causes ParseException
+  public void testInvalidFlags() throws Exception {
+    InputStream affixStream = getClass().getResourceAsStream("broken-flags.aff");
+    InputStream dictStream = getClass().getResourceAsStream("simple.dic");
+    
+    try {
+      new Dictionary(affixStream, dictStream);
+      fail("didn't get expected exception");
+    } catch (Exception expected) {
+      assertTrue(expected.getMessage().startsWith("expected only one flag"));
+    }
+    
+    affixStream.close();
+    dictStream.close();
+  }
+  
   private class CloseCheckInputStream extends FilterInputStream {
     private boolean closed = false;
 

lucene/analysis/common/src/test/org/apache/lucene/analysis/hunspell/broken-flags.aff

@@ -0,0 +1,21 @@
+SET UTF-8
+TRY abcdefghijklmopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ
+
+SFX A Y 3
+SFX A   0     e         n
+SFX A   0     e         t
+SFX A   0     e         h
+
+SFX C Y 2
+SFX C   0     d/C       c
+SFX C   0     c         b
+
+SFX D Y 1
+SFX D   0     s         o
+
+SFX E Y 1
+SFX E   0     d         o
+
+# broken, the flag has too much in it
+PFX B0 Y 1
+PFX B0   0     s         o