michael@0: /*
michael@0:  * Copyright © 2011,2012  Google, Inc.
michael@0:  *
michael@0:  *  This is part of HarfBuzz, a text shaping library.
michael@0:  *
michael@0:  * Permission is hereby granted, without written agreement and without
michael@0:  * license or royalty fees, to use, copy, modify, and distribute this
michael@0:  * software and its documentation for any purpose, provided that the
michael@0:  * above copyright notice and the following two paragraphs appear in
michael@0:  * all copies of this software.
michael@0:  *
michael@0:  * IN NO EVENT SHALL THE COPYRIGHT HOLDER BE LIABLE TO ANY PARTY FOR
michael@0:  * DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES
michael@0:  * ARISING OUT OF THE USE OF THIS SOFTWARE AND ITS DOCUMENTATION, EVEN
michael@0:  * IF THE COPYRIGHT HOLDER HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
michael@0:  * DAMAGE.
michael@0:  *
michael@0:  * THE COPYRIGHT HOLDER SPECIFICALLY DISCLAIMS ANY WARRANTIES, INCLUDING,
michael@0:  * BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
michael@0:  * FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS
michael@0:  * ON AN "AS IS" BASIS, AND THE COPYRIGHT HOLDER HAS NO OBLIGATION TO
michael@0:  * PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.
michael@0:  *
michael@0:  * Google Author(s): Behdad Esfahbod
michael@0:  */
michael@0: 
michael@0: #ifndef HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH
michael@0: #define HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH
michael@0: 
michael@0: #include "hb-private.hh"
michael@0: 
michael@0: %%{
michael@0:   machine indic_syllable_machine;
michael@0:   alphtype unsigned char;
michael@0:   write data;
michael@0: }%%
michael@0: 
michael@0: %%{
michael@0: 
michael@0: # Same order as enum indic_category_t.  Not sure how to avoid duplication.
michael@0: X    = 0;
michael@0: C    = 1;
michael@0: V    = 2;
michael@0: N    = 3;
michael@0: H    = 4;
michael@0: ZWNJ = 5;
michael@0: ZWJ  = 6;
michael@0: M    = 7;
michael@0: SM   = 8;
michael@0: VD   = 9;
michael@0: A    = 10;
michael@0: NBSP = 11;
michael@0: DOTTEDCIRCLE = 12;
michael@0: RS    = 13;
michael@0: Coeng = 14;
michael@0: Repha = 15;
michael@0: Ra    = 16;
michael@0: CM    = 17;
michael@0: Avag  = 18;
michael@0: CM2   = 31;
michael@0: 
michael@0: c = (C | Ra);			# is_consonant
michael@0: n = ((ZWNJ?.RS)? (N.N?)?);	# is_consonant_modifier
michael@0: z = ZWJ|ZWNJ;			# is_joiner
michael@0: h = H | Coeng;			# is_halant_or_coeng
michael@0: reph = (Ra H | Repha);		# possible reph
michael@0: 
michael@0: cn = c.ZWJ?.n?;
michael@0: forced_rakar = ZWJ H ZWJ Ra;
michael@0: avagraha = Avag.N?;
michael@0: matra_group = z{0,3}.M.N?.(H | forced_rakar)?;
michael@0: syllable_tail2 = (SM.SM?.ZWNJ?)? (A.A?)? VD?;
michael@0: syllable_tail =  (Coeng (cn|V))? avagraha? syllable_tail2;
michael@0: place_holder = NBSP | DOTTEDCIRCLE;
michael@0: halant_group = (z?.h.(ZWJ.N?)?);
michael@0: final_halant_group = halant_group | h.ZWNJ;
michael@0: medial_group = CM?.CM2?;
michael@0: halant_or_matra_group = (final_halant_group | (h.ZWJ)? matra_group{0,4});
michael@0: 
michael@0: 
michael@0: consonant_syllable =	Repha? (cn.halant_group){0,4} cn medial_group halant_or_matra_group syllable_tail;
michael@0: vowel_syllable =	reph? V.n? (ZWJ | (halant_group.cn){0,4} medial_group halant_or_matra_group syllable_tail);
michael@0: standalone_cluster =	reph? place_holder.n? (halant_group.cn){0,4} medial_group halant_or_matra_group syllable_tail;
michael@0: avagraha_cluster = 	avagraha syllable_tail2;
michael@0: broken_cluster =	reph? n? (halant_group.cn){0,4} medial_group halant_or_matra_group syllable_tail;
michael@0: other =			any;
michael@0: 
michael@0: main := |*
michael@0: 	consonant_syllable	=> { found_syllable (consonant_syllable); };
michael@0: 	vowel_syllable		=> { found_syllable (vowel_syllable); };
michael@0: 	standalone_cluster	=> { found_syllable (standalone_cluster); };
michael@0: 	avagraha_cluster	=> { found_syllable (avagraha_cluster); };
michael@0: 	broken_cluster		=> { found_syllable (broken_cluster); };
michael@0: 	other			=> { found_syllable (non_indic_cluster); };
michael@0: *|;
michael@0: 
michael@0: 
michael@0: }%%
michael@0: 
michael@0: #define found_syllable(syllable_type) \
michael@0:   HB_STMT_START { \
michael@0:     if (0) fprintf (stderr, "syllable %d..%d %s\n", last, p+1, #syllable_type); \
michael@0:     for (unsigned int i = last; i < p+1; i++) \
michael@0:       info[i].syllable() = (syllable_serial << 4) | syllable_type; \
michael@0:     last = p+1; \
michael@0:     syllable_serial++; \
michael@0:     if (unlikely (syllable_serial == 16)) syllable_serial = 1; \
michael@0:   } HB_STMT_END
michael@0: 
michael@0: static void
michael@0: find_syllables (hb_buffer_t *buffer)
michael@0: {
michael@0:   unsigned int p, pe, eof, ts HB_UNUSED, te HB_UNUSED, act HB_UNUSED;
michael@0:   int cs;
michael@0:   hb_glyph_info_t *info = buffer->info;
michael@0:   %%{
michael@0:     write init;
michael@0:     getkey info[p].indic_category();
michael@0:   }%%
michael@0: 
michael@0:   p = 0;
michael@0:   pe = eof = buffer->len;
michael@0: 
michael@0:   unsigned int last = 0;
michael@0:   unsigned int syllable_serial = 1;
michael@0:   %%{
michael@0:     write exec;
michael@0:   }%%
michael@0: }
michael@0: 
michael@0: #endif /* HB_OT_SHAPE_COMPLEX_INDIC_MACHINE_HH */