Add segmentation attributesbreak-tailoring

Add attributes that let us override word and
sentence boundaries (and, indirectly, line breaks).

Tests included.

10 files changed, 50 insertions, 3 deletions

diff --git a/tests/breaks/five.break b/tests/breaks/five.break
new file mode 100644
index 00000000..c9a192dd
--- /dev/null
+++ b/tests/breaks/five.break
@@ -0,0 +1 @@
+<span segment='word'>ab
cd</span>
diff --git a/tests/breaks/five.expected b/tests/breaks/five.expected
new file mode 100644
index 00000000..17ce3a58
--- /dev/null
+++ b/tests/breaks/five.expected
@@ -0,0 +1,7 @@
+Text:         ⁦a⁩ ⁦b⁩ [0x2028]  ⁦c⁩ ⁦d⁩  [0x0a] 
+Breaks:     lc c c        Lc c c       c
+Whitespace:      w             w       w
+Sentences:  bs   e        bs   e       b
+Words:      bs                 be      b
+Graphemes:  b  b b        b  b b       b
+Hyphens:       i             i          
diff --git a/tests/breaks/fourteen.break b/tests/breaks/fourteen.break
new file mode 100644
index 00000000..e270276d
--- /dev/null
+++ b/tests/breaks/fourteen.break
@@ -0,0 +1,2 @@
+# item-spanning line break attrs
+<span allow_breaks='n'>ab<span segment='word'>c α</span>βγ</span>
diff --git a/tests/breaks/fourteen.expected b/tests/breaks/fourteen.expected
new file mode 100644
index 00000000..c8a22617
--- /dev/null
+++ b/tests/breaks/fourteen.expected
@@ -0,0 +1,7 @@
+Text:         ⁦a⁩ ⁦b⁩   ⁦c⁩ [ ] ⁦α⁩   ⁦β⁩ ⁦γ⁩  [0x0a] 
+Breaks:     c    l         l     c       c
+Whitespace:          x           w       w
+Sentences:  bs                   e       b
+Words:      bs   bse       bse   be      b
+Graphemes:  b  b b   b   b b   b b       b
+Hyphens:       i i         i   i          
diff --git a/tests/breaks/thirteen.break b/tests/breaks/thirteen.break
new file mode 100644
index 00000000..a1e5cacf
--- /dev/null
+++ b/tests/breaks/thirteen.break
@@ -0,0 +1,2 @@
+# some line breaks remain in words
+a <span segment='word'>ab​sp​  de­fg</span> b
diff --git a/tests/breaks/thirteen.expected b/tests/breaks/thirteen.expected
new file mode 100644
index 00000000..70742cff
--- /dev/null
+++ b/tests/breaks/thirteen.expected
@@ -0,0 +1,7 @@
+Text:         ⁦a⁩  [ ]  ⁦a⁩ ⁦b⁩ [0x200b]  ⁦s⁩ ⁦p⁩ [0x200b] [ ] [ ]  ⁦d⁩ ⁦e⁩ [0xad]  ⁦f⁩ ⁦g⁩  [ ]  ⁦b⁩  [0x0a] 
+Breaks:     c  c    lc c c        lc c c        c   c   lc c c      lc c c    lc c       c
+Whitespace:    x                                x   x                    x       w       w
+Sentences:  bs                                                                   e       b
+Words:      bs be   bs                                                   be   bs be      b
+Graphemes:  b  b    b  b b        b  b b        b   b   b  b b      b  b b    b  b       b
+Hyphens:               i          i  i                     i        i  i                  
diff --git a/tests/breaks/twelve.break b/tests/breaks/twelve.break
new file mode 100644
index 00000000..a1659be4
--- /dev/null
+++ b/tests/breaks/twelve.break
@@ -0,0 +1,2 @@
+# test line break attributes
+the file <span segment='word'>/path/</span><span segment='word'>to/</span><span segment='word'>my/</span>home is cursed.
diff --git a/tests/breaks/twelve.expected b/tests/breaks/twelve.expected
new file mode 100644
index 00000000..63e14e64
--- /dev/null
+++ b/tests/breaks/twelve.expected
@@ -0,0 +1,7 @@
+Text:         ⁦t⁩ ⁦h⁩ ⁦e⁩  [ ]  ⁦f⁩ ⁦i⁩ ⁦l⁩ ⁦e⁩  [ ]  ⁦/⁩ ⁦p⁩ ⁦a⁩ ⁦t⁩ ⁦h⁩ ⁦/⁩   ⁦t⁩ ⁦o⁩ ⁦/⁩  ⁦m⁩ ⁦y⁩ ⁦/⁩   ⁦h⁩ ⁦o⁩ ⁦m⁩ ⁦e⁩  [ ]  ⁦i⁩ ⁦s⁩  [ ]  ⁦c⁩ ⁦u⁩ ⁦r⁩ ⁦s⁩ ⁦e⁩ ⁦d⁩  ⁦.⁩ [0x0a] 
+Breaks:     c  c c c    lc c c c c    lc c c c c c lc  c c lc c c lc  c c c c    lc c c    lc c c c c c c  c      c
+Whitespace:        x             x                                          x         x                    w      w
+Sentences:  bs                                                                                             e      b
+Words:      bs     be   bs       be   bs           bse     bs     bse       be   bs   be   bs           be b      b
+Graphemes:  b  b b b    b  b b b b    b  b b b b b b   b b b  b b b   b b b b    b  b b    b  b b b b b b  b      b
+Hyphens:       i i         i i i           i i i       i      i       i i i         i         i i i i i            
diff --git a/tests/test-common.c b/tests/test-common.c
index 011b2eef..0e3719e7 100644
--- a/tests/test-common.c
+++ b/tests/test-common.c
@@ -144,6 +144,8 @@ print_attribute (PangoAttribute *attr, GString *string)
     case PANGO_ATTR_SHOW:
     case PANGO_ATTR_TEXT_TRANSFORM:
     case PANGO_ATTR_ABSOLUTE_LINE_HEIGHT:
+    case PANGO_ATTR_WORD:
+    case PANGO_ATTR_SENTENCE:
       g_string_append_printf (string, "%d", ((PangoAttrInt *)attr)->value);
       break;
     case PANGO_ATTR_FONT_DESC:
diff --git a/tests/testattributes.c b/tests/testattributes.c
index f950a204..b9dcc1b3 100644
--- a/tests/testattributes.c
+++ b/tests/testattributes.c
@@ -75,6 +75,8 @@ test_attributes_basic (void)
   test_copy (pango_attr_text_transform_new (PANGO_TEXT_TRANSFORM_UPPERCASE));
   test_copy (pango_attr_line_height_new (1.5));
   test_copy (pango_attr_line_height_new_absolute (3000));
+  test_copy (pango_attr_word_new ());
+  test_copy (pango_attr_sentence_new ());
 }
 
 static void
@@ -125,7 +127,7 @@ test_binding (PangoAttribute *attr)
     INVALID, LANGUAGE, STRING, INT, INT, INT, INT, SIZE, FONT_DESC, COLOR,
     COLOR, INT, INT, INT, SHAPE, FLOAT, INT, INT, COLOR, COLOR, SIZE,
     INT, INT, FONT_FEATURES, INT, INT, INT, INT, INT, INT, COLOR, FLOAT,
-    INT, INT, INT, INT
+    INT, INT, INT, INT, INT, INT
   };
 
   switch (attr_base[attr->klass->type])
@@ -207,6 +209,8 @@ test_binding_helpers (void)
   test_binding (pango_attr_text_transform_new (PANGO_TEXT_TRANSFORM_UPPERCASE));
   test_binding (pango_attr_line_height_new (1.5));
   test_binding (pango_attr_line_height_new_absolute (3000));
+  test_binding (pango_attr_word_new ());
+  test_binding (pango_attr_sentence_new ());
 }
 
 static void
@@ -1213,8 +1217,8 @@ test_merge2 (void)
   pango_attr_list_unref (list);
 }
 
-/* This only prints rise, size and scale, which are the
- * only relevant attributes in the test that uses this
+/* This only prints rise, size, scale, allow_breaks and line_break,
+ * which are the only relevant attributes in the tests that use this
  * function.
  */
 static void
@@ -1240,6 +1244,12 @@ print_tags_for_attributes (PangoAttrIterator *iter,
     g_string_append_printf (s, "[%d, %d]scale=%f\n",
                             attr->start_index, attr->end_index,
                             ((PangoAttrFloat*)attr)->value);
+
+  attr = pango_attr_iterator_get (iter, PANGO_ATTR_ALLOW_BREAKS);
+  if (attr)
+    g_string_append_printf (s, "[%d, %d]allow_breaks=%d\n",
+                            attr->start_index, attr->end_index,
+                            ((PangoAttrInt*)attr)->value);
 }
 
 static void