Fix syllabification.

* configure.ac: Add AC_PROG_YACC
* src/ellinika/phoneme.y: New file.
* src/ellinika/yyrename: New file.
* src/ellinika/syllabificator.c: New file.
* src/ellinika/.gitignore: Update.
* src/ellinika/elchr.c (char_info_st): Move to header.
(el_basic_ctype):
(elchr_info): Remove static qualifier.
Return a pointer to const.
(elchr_letter,elchr_phoneme): New functions.
(elchr_diphthong): Remove.
* src/ellinika/elmorph.c (elstr)<phoneme,phoneme_count>: New members.
(_elstr_syllabize): Rewrite.
(invalidate_maps)" New static function.
(_elstr_alloc): Initialize new fields, take function name
as argument, for diagnostic purposes.
(_elstr_print): Rewrite
(deftab): Update.
(elstr-syllable-prop,elstr-syllable)
(_elstr_set_accent,_elstr_set_accent_on_char): Rewrite.
(elstr-char-phoneme,elstr->phonetic-map): New functions.
* src/ellinika/elmorph.h (CHF_DIPH1,CHF_DIPH2): Remove.
(CHF_DIPHTHONG): New flag.
(PHON_.*): New constants.
(phoneme,syllable): New structures.
(char_info_st)<letter,phoneme>: New members.
(elchr_info,elchr_letter)
(elchr_phoneme,phoneme_map)
(syllable_map): New protos.
(elchr_diphthong): Remove protos.
* src/ellinika/elmorph.scm4: Move public definitions
to elmorph-public.scm; include it here.
* src/ellinika/xlat.scm (ellinika:sounds-like): Rewrite as a
wrapper over elstr->soundslike.

Describe Milesian numbers.

* style.css (img.ellinika-img): New class.
* xml/lingua.conf.in (IMAGE): New tag.
* xml/pl/alfabhta.xml: Describe Milesian numbers.

Various fixes.

* data/dbverb.struct: fix a typo in flection.
Use 'sub' theme for pas/sub/aor.
* data/irregular-verbs.xml: Add more verbs.
* scm/conjugator.scm: Various fixes.
* scm/verbop.scm: Accept empty mood and voice declarations.

1 files changed, 79 insertions, 3 deletions

diff --git a/src/ellinika/elmorph.h b/src/ellinika/elmorph.h
index eacbde5..2399b8a 100644
--- a/src/ellinika/elmorph.h
+++ b/src/ellinika/elmorph.h
@@ -33,10 +33,82 @@
 #define CHF_UPPER         0x02000
 #define CHF_NUMERIC       0x04000
 
-#define CHF_DIPH1         0x10000
-#define CHF_DIPH2         0x20000
+#define CHF_DIPHTHONG     0x08000
 
+/* Phonemes */
+#define PHON_A     1 /* α */ 
+#define PHON_E     2 /* ε αι */ 
+#define PHON_I     3 /* ι η υ ει οι υι */ 
+#define PHON_O     4 /* ο ω */ 
+#define PHON_U     5 /* ου */ 
+
+#define PHON_BH    6 /* β */
+#define PHON_GH    7 /* γ */
+#define PHON_DH    8 /* δ */
+#define PHON_Z     9 /* ζ */
+#define PHON_TH   10 /* θ */ 
+#define PHON_K    11 /* κ */ 
+#define PHON_L    12 /* λ */ 
+#define PHON_M    13 /* μ */ 
+#define PHON_N    14 /* ν */ 
+#define PHON_X    15 /* ξ */ 
+#define PHON_P    16 /* π */ 
+#define PHON_R    17 /* ρ */ 
+#define PHON_S    18 /* σ */ 
+#define PHON_T    19 /* τ */ 
+#define PHON_F    20 /* φ */ 
+#define PHON_H    21 /* χ */
+#define PHON_PS   22 /* ψ */
+
+#define PHON_B    23 /* μπ */
+#define PHON_D    24 /* ντ */
+#define PHON_G    25 /* γγ γκ γχ */
+#define PHON_ZM   26 /* σμ */
+#define PHON_TS   27 /* τσ */
+#define PHON_DZ   28 /* τζ */
+#define PHON_NGZ  29 /* νγζ */
+
+#define PHON_AV   30 /* αυ */
+#define PHON_EV   31 /* ευ */
+
+#define _PHON_MAX 32
+
+struct phoneme {
+	int code;               /* Phoneme code */
+	unsigned start;         /* Start of phoneme */
+	unsigned count;         /* Number of characters in it */
+	int flags;
+};
+
+struct syllable {
+	unsigned char_start;         /* Start of syllable */
+	unsigned char_count;         /* Number of characters in it */
+	unsigned phoneme_start;
+	unsigned phoneme_count;
+	int flags;
+};
+
+struct char_info_st {
+	unsigned ch;           /* Characters */
+	int flags;             /* Flags (see above) */
+	unsigned base;         /* for vowels - a corresponding vowel with
+				  all diacritics removed */
+	unsigned trans;        /* a counter-case equivalent, i.e. a
+				  corresponding uppercase letter if
+				  flags & CHF_LOWER and a corresponding
+				  lowercase letter if flags & CHF_UPPER */
+	unsigned numval;       /* Numeric value */
+	unsigned accented[3];  /* For vowels - corresponding accented variant */
+	unsigned deaccent;     /* For accented vowels with diaeresis -
+				  corresponding non-accented character */
+	int letter;            /* Letter code */
+	int phoneme;           /* Phoneme code */
+};
+
+struct char_info_st const *elchr_info(unsigned ch);
 int elchr_flags(unsigned ch);
+int elchr_letter(unsigned ch);
+int elchr_phoneme(unsigned ch);
 int elchr_isupper(unsigned ch);
 int elchr_islower(unsigned ch);
 int elchr_getaccent(unsigned ch);
@@ -55,8 +127,12 @@ unsigned elchr_tolower(unsigned ch);
 unsigned elchr_base(unsigned ch);
 unsigned elchr_deaccent(unsigned ch);
 unsigned elchr_accent(unsigned ch, int acc);
-int elchr_diphthong(unsigned ch, int state);
 
 
 int elmorph_thema_aoristoy(unsigned *word, size_t len,
 			   unsigned **thema, size_t *tlen);
+
+int phoneme_map(struct phoneme **pph, size_t *plen,
+		unsigned *word, size_t len);
+int syllable_map(struct syllable **psyl, size_t *plen,
+		 struct phoneme *phon, size_t nphon);