unilbrk: Update handling of Hebrew letter + hyphen for Unicode 10.0.0.

* lib/unilbrk/lbrktables.h (LBP_HL_BA): New enum value.
* lib/unilbrk/u8-possible-linebreaks.c (u8_possible_linebreaks): Add
code for handling break-after character after Hebrew letter.
* lib/unilbrk/u16-possible-linebreaks.c (u16_possible_linebreaks):
Likewise.
* lib/unilbrk/u32-possible-linebreaks.c (u32_possible_linebreaks):
Likewise.
* tests/unilbrk/test-u8-possible-linebreaks.c (main): Add a test
regarding break-after character after Hebrew letter.
* tests/unilbrk/test-u16-possible-linebreaks.c (main): Likewise.
* tests/unilbrk/test-u32-possible-linebreaks.c (main): Likewise.

3 files changed, 51 insertions, 0 deletions

diff --git a/tests/unilbrk/test-u16-possible-linebreaks.c b/tests/unilbrk/test-u16-possible-linebreaks.c
index 87d712fcc6..fedaa9d842 100644
--- a/tests/unilbrk/test-u16-possible-linebreaks.c
+++ b/tests/unilbrk/test-u16-possible-linebreaks.c
@@ -179,5 +179,22 @@ main ()
     free (p);
   }
 
+  /* Test special behaviour of hyphen/break-after character after
+     Hebrew letter.  */
+  {
+    static const uint16_t input[10] = /* "ab-אב-αβ-ω" */
+      { 'a', 'b', '-', 0x05D0, 0x05D1, '-', 0x03B1, 0x03B2, '-', 0x03C9 };
+    char *p = (char *) malloc (SIZEOF (input));
+    size_t i;
+
+    u16_possible_linebreaks (input, SIZEOF (input), "UTF-8", p);
+    for (i = 0; i < 10; i++)
+      {
+        ASSERT (p[i] == (i == 3 || i == 9 ? UC_BREAK_POSSIBLE :
+                         UC_BREAK_PROHIBITED));
+      }
+    free (p);
+  }
+
   return 0;
 }
diff --git a/tests/unilbrk/test-u32-possible-linebreaks.c b/tests/unilbrk/test-u32-possible-linebreaks.c
index c140ca3b55..1d8dfb122b 100644
--- a/tests/unilbrk/test-u32-possible-linebreaks.c
+++ b/tests/unilbrk/test-u32-possible-linebreaks.c
@@ -179,5 +179,22 @@ main ()
     free (p);
   }
 
+  /* Test special behaviour of hyphen/break-after character after
+     Hebrew letter.  */
+  {
+    static const uint32_t input[10] = /* "ab-אב-αβ-ω" */
+      { 'a', 'b', '-', 0x05D0, 0x05D1, '-', 0x03B1, 0x03B2, '-', 0x03C9 };
+    char *p = (char *) malloc (SIZEOF (input));
+    size_t i;
+
+    u32_possible_linebreaks (input, SIZEOF (input), "UTF-8", p);
+    for (i = 0; i < 10; i++)
+      {
+        ASSERT (p[i] == (i == 3 || i == 9 ? UC_BREAK_POSSIBLE :
+                         UC_BREAK_PROHIBITED));
+      }
+    free (p);
+  }
+
   return 0;
 }
diff --git a/tests/unilbrk/test-u8-possible-linebreaks.c b/tests/unilbrk/test-u8-possible-linebreaks.c
index 0fdb66d367..cad7b7afa3 100644
--- a/tests/unilbrk/test-u8-possible-linebreaks.c
+++ b/tests/unilbrk/test-u8-possible-linebreaks.c
@@ -164,5 +164,22 @@ main ()
     free (p);
   }
 
+  /* Test special behaviour of hyphen/break-after character after
+     Hebrew letter.  */
+  {
+    static const uint8_t input[15] = /* "ab-אב-αβ-ω" */
+      "ab-\327\220\327\221-\316\261\316\262-\317\211";
+    char *p = (char *) malloc (SIZEOF (input));
+    size_t i;
+
+    u8_possible_linebreaks (input, SIZEOF (input), "UTF-8", p);
+    for (i = 0; i < 15; i++)
+      {
+        ASSERT (p[i] == (i == 3 || i == 13 ? UC_BREAK_POSSIBLE :
+                         UC_BREAK_PROHIBITED));
+      }
+    free (p);
+  }
+
   return 0;
 }