A step out from Harfbuzz (reduce dependency)

Introduce QCharAttributes and use it instead of HB_CharAttributes everywhere in Qt
(in Harfbuzz, the HB_CharAttributes is only used in the text segmentation algorithm
which has been moved from HB to Qt (well, most of it)).
Rename some members to better reflect their meaning,
remember to keep HB_CharAttributes in sync with QCharAttributes.
Also replace HB_ScriptItem with a (temporary) QUnicodeTools::ScriptItem struct
that will be replaced with a more efficient/friendly solution a bit later.

The soft hyphen and the mandatory break detection has been factored out
of the default text breaking algorithm to a higher level in order to refactor
the QCharAttributes bitfields and to optimize the implementation for the common case.

Change-Id: Ieb365623ae954430f1c8b2dfcd65c82973143eec
Reviewed-by: Lars Knoll <lars.knoll@digia.com>

6 files changed, 25 insertions, 32 deletions

diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-indic.cpp b/src/3rdparty/harfbuzz/src/harfbuzz-indic.cpp
index bbf479e774..0dcddc4f7d 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-indic.cpp
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-indic.cpp
@@ -1880,12 +1880,12 @@ void HB_IndicAttributes(HB_Script script, const HB_UChar16 *text, hb_uint32 from
     while (i < len) {
         bool invalid;
         hb_uint32 boundary = indic_nextSyllableBoundary(script, text, from+i, end, &invalid) - from;
-         attributes[i].charStop = true;
+         attributes[i].graphemeBoundary = true;
 
         if (boundary > len-1) boundary = len;
         i++;
         while (i < boundary) {
-            attributes[i].charStop = false;
+            attributes[i].graphemeBoundary = false;
             ++uc;
             ++i;
         }
diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-khmer.c b/src/3rdparty/harfbuzz/src/harfbuzz-khmer.c
index 958069e866..1f3c7e2bce 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-khmer.c
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-khmer.c
@@ -652,12 +652,12 @@ void HB_KhmerAttributes(HB_Script script, const HB_UChar16 *text, hb_uint32 from
 	HB_Bool invalid;
 	hb_uint32 boundary = khmer_nextSyllableBoundary( text, from+i, end, &invalid ) - from;
 
-	attributes[i].charStop = TRUE;
+    attributes[i].graphemeBoundary = TRUE;
 
 	if ( boundary > len-1 ) boundary = len;
 	i++;
 	while ( i < boundary ) {
-	    attributes[i].charStop = FALSE;
+        attributes[i].graphemeBoundary = FALSE;
 	    ++uc;
 	    ++i;
 	}
diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-myanmar.c b/src/3rdparty/harfbuzz/src/harfbuzz-myanmar.c
index 1327b185b0..b18b83054d 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-myanmar.c
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-myanmar.c
@@ -520,14 +520,14 @@ void HB_MyanmarAttributes(HB_Script script, const HB_UChar16 *text, hb_uint32 fr
 	HB_Bool invalid;
 	hb_uint32 boundary = myanmar_nextSyllableBoundary(text, from+i, end, &invalid) - from;
 
-	attributes[i].charStop = TRUE;
-    attributes[i].lineBreakType = HB_Break;
+    attributes[i].graphemeBoundary = TRUE;
+    attributes[i].lineBreak = TRUE;
 
 	if (boundary > len-1)
             boundary = len;
 	i++;
 	while (i < boundary) {
-	    attributes[i].charStop = FALSE;
+        attributes[i].graphemeBoundary = FALSE;
 	    ++uc;
 	    ++i;
 	}
diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-shaper.h b/src/3rdparty/harfbuzz/src/harfbuzz-shaper.h
index 6dfcdd20a7..2ff52eaf30 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-shaper.h
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-shaper.h
@@ -126,21 +126,14 @@ typedef struct
     hb_uint8 bidiLevel;
 } HB_ScriptItem;
 
-typedef enum {
-    HB_NoBreak,
-    HB_SoftHyphen,
-    HB_Break,
-    HB_ForcedBreak
-} HB_LineBreakType;
-
 
 typedef struct {
-    /*HB_LineBreakType*/ hb_bitfield lineBreakType  :2;
-    /*HB_Bool*/ hb_bitfield whiteSpace              :1;     /* A unicode whitespace character, except NBSP, ZWNBSP */
-    /*HB_Bool*/ hb_bitfield charStop                :1;     /* Valid cursor position (for left/right arrow) */
-    /*HB_Bool*/ hb_bitfield wordBoundary            :1;
-    /*HB_Bool*/ hb_bitfield sentenceBoundary        :1;
-    hb_bitfield unused                  :2;
+    hb_bitfield graphemeBoundary : 1;     /* Valid cursor position (for left/right arrow) */
+    hb_bitfield wordBreak        : 1;
+    hb_bitfield sentenceBoundary : 1;
+    hb_bitfield lineBreak        : 1;
+    hb_bitfield whiteSpace       : 1;     /* A unicode whitespace character */
+    hb_bitfield unused           : 3;
 } HB_CharAttributes;
 
 void HB_GetTailoredCharAttributes(const HB_UChar16 *string, hb_uint32 stringLength,
diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-thai.c b/src/3rdparty/harfbuzz/src/harfbuzz-thai.c
index ecb722d984..70c1d57ff1 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-thai.c
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-thai.c
@@ -395,8 +395,8 @@ static void HB_ThaiAssignAttributes(const HB_UChar16 *string, hb_uint32 len, HB_
     to_tis620(string, len, cstr);
 
     for (i = 0; i < len; ++i) {
-        attributes[i].lineBreakType = HB_NoBreak;
-        attributes[i].wordBoundary = FALSE;
+        attributes[i].lineBreak = FALSE;
+        attributes[i].wordBreak = FALSE;
     }
 
     if (len > 128) {
@@ -410,29 +410,29 @@ static void HB_ThaiAssignAttributes(const HB_UChar16 *string, hb_uint32 len, HB_
     }
 
     if (break_positions) {
-        attributes[0].wordBoundary = TRUE;
+        attributes[0].wordBreak = TRUE;
         numbreaks = th_brk((const unsigned char *)cstr, break_positions, brp_size);
         for (i = 0; i < numbreaks; ++i) {
-            attributes[break_positions[i]].wordBoundary = TRUE;
-            attributes[break_positions[i]].lineBreakType = HB_Break;
+            attributes[break_positions[i]].wordBreak = TRUE;
+            attributes[break_positions[i]].lineBreak = TRUE;
         }
 
         if (break_positions != brp)
             free(break_positions);
     }
 
-    /* manage charStop */
+    /* manage grapheme boundaries */
     i = 0;
     while (i < len) {
         cell_length = th_next_cell((const unsigned char *)cstr + i, len - i, &tis_cell, true);
 
-        attributes[i].charStop = true;
+        attributes[i].graphemeBoundary = true;
         for (j = 1; j < cell_length; j++)
-            attributes[i + j].charStop = false;
+            attributes[i + j].graphemeBoundary = false;
 
-        /* Set charStop for SARA AM */
+        /* Set graphemeBoundary for SARA AM */
         if (cstr[i + cell_length - 1] == (char)0xd3)
-            attributes[i + cell_length - 1].charStop = true;
+            attributes[i + cell_length - 1].graphemeBoundary = true;
 
         i += cell_length;
     }
diff --git a/src/3rdparty/harfbuzz/src/harfbuzz-tibetan.c b/src/3rdparty/harfbuzz/src/harfbuzz-tibetan.c
index e0c263d426..c3570974b3 100644
--- a/src/3rdparty/harfbuzz/src/harfbuzz-tibetan.c
+++ b/src/3rdparty/harfbuzz/src/harfbuzz-tibetan.c
@@ -260,12 +260,12 @@ void HB_TibetanAttributes(HB_Script script, const HB_UChar16 *text, hb_uint32 fr
         HB_Bool invalid;
         hb_uint32 boundary = tibetan_nextSyllableBoundary(text, from+i, end, &invalid) - from;
 
-        attributes[i].charStop = TRUE;
+        attributes[i].graphemeBoundary = TRUE;
 
         if (boundary > len-1) boundary = len;
         i++;
         while (i < boundary) {
-            attributes[i].charStop = FALSE;
+            attributes[i].graphemeBoundary = FALSE;
             ++uc;
             ++i;
         }