delete lang engine

2000-11-30  Havoc Pennington  <hp@pobox.com>

* modules/thai/thai.c: delete lang engine

* modules/tamil/tamil.c: delete lang engine
(tamil_engine_x_new): fix type tag for shape engine

* modules/indic/myanmar.c: delete lang engine
(pango_engine_x_new): fix type tag for shape engine

* modules/indic/gurmukhi.c: delete lang engine
(pango_indic_engine_x_new): fix type tag for shape engine

* modules/indic/gujarati.c: delete lang engine
(pango_indic_engine_x_new): fix type tag for shape engine

* modules/indic/devanagari.c: delete lang engine
(pango_indic_engine_x_new): fix type tag for shape engine

* modules/indic/pango-indic-script.h (SCRIPT_ENGINE_DEFINITION):
delete lang engine

* modules/indic/bengali.c: delete the lang engine
(pango_indic_engine_x_new): fix type tag for shape engine

* modules/hangul/hangul.c: delete the lang engine
(hangul_engine_x_new): fix type tag for shape engine

* modules/basic/basic.c: delete the lang engine
(basic_engine_x_new): fix type tag for shape engine

* modules/basic/basic-win32.c: delete the lang engine
(basic_engine_win32_new): this was a shape engine,
use correct type tag

* modules/basic/basic-ft2.c: delete the lang engine

* modules/arabic/arabic.c: Delete the lang engine
(arabic_engine_x_new): this is a shape
engine, not a lang engine, fix type tag

* pango/pango-layout.c (pango_layout_index_to_line_x): handle
the fact that paragraph delimiters aren't in the layout lines
(pango_layout_index_to_pos): update to handle paragraph
delimiters

* pango/break.c (pango_find_paragraph_boundary): New function
to find paragraph boundaries

* pango/pango-layout.c (get_items_log_attrs): don't separate calls
to pango_break() when directional level changes

* pango/pango-layout.h (struct _PangoLayoutLine): put start index
of the line into the struct

 * pango/pango-layout.c (pango_layout_get_cursor_pos): Fixups to
 reflect the fact that paragraph separators are removed from the
 input text.

 * pango/pango-layout.c (can_break_at): don't
 special-case start of line and whitespace-following-alphabetic
 here, because pango_break() already handles that properly

 * tests/testboundaries.c, tests/Makefile.am, tests/runtests.sh:
 Add directory for test programs, and a script to run them all

 * configure.in: Create Makefile in tests

 * pango/break.c (pango_break): Try for a real implementation of
 the Unicode text boundary algorithms
 (pango_get_log_attrs): Allow length to be -1

 * pango/pango-context.c (pango_itemize): use pango_item_new(),
 assert that items added to the list are sane.

 * pango/pango-layout.c (pango_layout_check_lines): Reimplement
 to honor the paragraph boundaries from pango_break()

 * pango/pango-layout.c (process_item): use pango_item_split() here

 * pango/pango-item.c (pango_item_split): New function to split an
 item into two items

1 files changed, 46 insertions, 5 deletions

diff --git a/pango/pango.h b/pango/pango.h
index e0652fd0..3cbc0f59 100644
--- a/pango/pango.h
+++ b/pango/pango.h
@@ -39,14 +39,40 @@ extern "C" {
 #include <pango/pango-layout.h>
 #include <pango/pango-types.h>
 
-/* Logical attributes of a character
+/* Logical attributes of a character.
  */
 struct _PangoLogAttr
 {
-  guint is_break : 1;  /* Break in front of character */
-  guint is_white : 1;
-  guint is_char_stop : 1;
-  guint is_word_stop : 1;
+  guint is_break : 1;           /* Can break line in front of character */
+
+  guint is_mandatory_break : 1; /* Must break line in front of character */
+  
+  guint is_white : 1;           /* Whitespace character */
+
+  /* cursor can appear in front of character (i.e. this is a grapheme
+   * boundary, or the first character in the text)
+   */
+  guint is_cursor_position : 1;
+  
+  /* Note that in degenerate cases, you could have both start/end set on
+   * some text, most likely for sentences (e.g. no space after a period, so
+   * the next sentence starts right away)
+   */
+  
+  guint is_word_start : 1;      /* first character in a word */
+  guint is_word_end   : 1;      /* is first non-word char after a word */
+
+  /* There are two ways to divide sentences. The first assigns all
+   * intersentence whitespace/control/format chars to some sentence,
+   * so all chars are in some sentence; is_sentence_boundary denotes
+   * the boundaries there. The second way doesn't assign
+   * between-sentence spaces, etc. to any sentence, so
+   * is_sentence_start/is_sentence_end mark the boundaries of those
+   * sentences.
+   */
+  guint is_sentence_boundary : 1;
+  guint is_sentence_start : 1;  /* first character in a sentence */
+  guint is_sentence_end : 1;    /* first non-sentence char after a sentence */
 };
 
 /* Determine information about cluster/word/line breaks in a string
@@ -57,6 +83,11 @@ void pango_break (const gchar   *text,
 		  PangoAnalysis *analysis, 
 		  PangoLogAttr  *attrs);
 
+void pango_find_paragraph_boundary (const gchar *text,
+                                    gint         length,
+                                    gint        *paragraph_delimiter_index,
+                                    gint        *next_paragraph_start);
+
 void pango_get_log_attrs (const char    *text,
                           int            length,
                           int            level,
@@ -72,6 +103,16 @@ void pango_shape (const gchar      *text,
 
 GList *pango_reorder_items (GList *logical_items);
 
+/* This is the default break algorithm, used if no language
+ * engine overrides it. Normally you should use pango_break()
+ * instead; this function is mostly useful for chaining up
+ * from a language engine override.
+ */
+void pango_default_break (const gchar   *text,
+                          gint           length,
+                          PangoAnalysis *analysis,
+                          PangoLogAttr  *attrs);
+
 #ifdef __cplusplus
 }
 #endif /* __cplusplus */