Add some null table handling to grapheme_table_capacity and use that …

…to simplify the iterator code a bit. git-svn-id: https://svn.parrot.org/parrot/branches/gsoc_nfg@47829 d31e2699-5ff4-0310-a27c-f18f2fbe73fe
parrot · Jun 25, 2010 · b5fea9f · b5fea9f
1 parent b2f549d
commit b5fea9f
Show file tree

Hide file tree

Showing 3 changed files with 8 additions and 11 deletions.
diff --git a/src/string/encoding/nfg.c b/src/string/encoding/nfg.c
@@ -548,15 +548,15 @@ nfg_encode_and_advance(PARROT_INTERP, ARGMOD(String_iter *i), UINTVAL c)
         }
         else {
             UErrorCode err = U_ZERO_ERROR;
+            UChar src[3];
             int dst_len = 1;
-            int src_len = 1;
-            UChar src[2];
             UChar dst[2];
             src[0] = s[pos - 1];
             src[1] = s[pos];
+            src[2] = 0;
 
             /* Delegate composition to ICU. */
-            dst_len = unorm_normalize(src, src_len, UNORM_DEFAULT, 0,
+            dst_len = unorm_normalize(src, -1 , UNORM_DEFAULT, 0,
                                       dst, dst_len, &err);
 
             if (U_SUCCESS(err)) {
@@ -566,11 +566,6 @@ nfg_encode_and_advance(PARROT_INTERP, ARGMOD(String_iter *i), UINTVAL c)
             }
 
             /* Composition failed, we need a dynamic codepoint. */
-            if (table == NULL) {
-                table = create_grapheme_table(interp, 1);
-                i->str->extra = table;
-            }
-
             g.len  = 2;
             g.hash = 0xffff;
             g.codepoints = mem_gc_allocate_n_typed(interp, g.len, UChar32);

diff --git a/src/string/grapheme.c b/src/string/grapheme.c
@@ -18,7 +18,9 @@ INTVAL
 grapheme_table_capacity(PARROT_INTERP, grapheme_table *table)
 {
     ASSERT_ARGS(grapheme_table_capacity)
-    return table->size - table->used;
+    if (table)
+        return table->size - table->used;
+    return 0;
 }
 
 grapheme_table *

diff --git a/t/op/string_nfg.t b/t/op/string_nfg.t
@@ -49,11 +49,11 @@ string operations.
     is($I0, $I1, "Bufused is the same.")
 
     # Do the same thing again, without dynamic codepoints.
-    $S2 = utf16:unicode:"O\u0308"
+    $S2 = utf16:unicode:"n\u0303"
     $I0 = find_encoding 'nfg'
     $S2 = trans_encoding $S2, $I0
 
-    $S3 = nfg:unicode:"O\u0308"
+    $S3 = nfg:unicode:"n\u0303"
 
     $I0 = stringinfo $S2, .STRINGINFO_STRLEN
     $I1 = stringinfo $S3, .STRINGINFO_STRLEN