fix getsup (HH)

[luatex.git] / source / texk / web2c / luatexdir / lang / texlang.w

% texlang.w
%
% Copyright 2006-2012 Taco Hoekwater <taco@@luatex.org>
%
% This file is part of LuaTeX.
%
% LuaTeX is free software; you can redistribute it and/or modify it under
% the terms of the GNU General Public License as published by the Free
% Software Foundation; either version 2 of the License, or (at your
% option) any later version.
%
% LuaTeX is distributed in the hope that it will be useful, but WITHOUT
% ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
% FITNESS FOR A PARTICULAR PURPOSE.  See the GNU Lesser General Public
% License for more details.
%
% You should have received a copy of the GNU General Public License along
% with LuaTeX; if not, see <http://www.gnu.org/licenses/>.

@ @c

#include "ptexlib.h"
#include <string.h>
#include "lua/luatex-api.h"

@ Low-level helpers

@ @c
#define unVERBOSE

#define MAX_TEX_LANGUAGES  16384

static struct tex_language *tex_languages[MAX_TEX_LANGUAGES] = { NULL };

static int next_lang_id = 0;

struct tex_language *new_language(int n)
{
    struct tex_language *lang;
    unsigned l;
    if (n >= 0) {
        l = (unsigned) n;
        if (l != (MAX_TEX_LANGUAGES - 1))
            if (next_lang_id <= n)
                next_lang_id = n + 1;
    } else {
        while (tex_languages[next_lang_id] != NULL)
            next_lang_id++;
        l = (unsigned) next_lang_id++;
    }
    if (l < (MAX_TEX_LANGUAGES - 1) && tex_languages[l] == NULL) {
        lang = xmalloc(sizeof(struct tex_language));
        tex_languages[l] = lang;
        lang->id = (int) l;
        lang->exceptions = 0;
        lang->patterns = NULL;
        lang->pre_hyphen_char = '-';
        lang->post_hyphen_char = 0;
        lang->pre_exhyphen_char = 0;
        lang->post_exhyphen_char = 0;
        lang->hyphenation_min = -1;
        if (saving_hyph_codes_par) {
            hj_codes_from_lc_codes(l); /* for now, we might just use specific value for whatever task */
        }
        return lang;
    } else {
        return NULL;
    }
}

struct tex_language *get_language(int n)
{
    if (n >= 0 && n < MAX_TEX_LANGUAGES) {
        if (tex_languages[n] != NULL) {
            return tex_languages[n];
        } else {
            return new_language(n);
        }
    } else {
        return NULL;
    }
}

@ @c
void set_pre_hyphen_char(int n, int v)
{
    struct tex_language *l = get_language((int) n);
    if (l != NULL)
        l->pre_hyphen_char = (int) v;
}

void set_post_hyphen_char(int n, int v)
{
    struct tex_language *l = get_language((int) n);
    if (l != NULL)
        l->post_hyphen_char = (int) v;
}

void set_pre_exhyphen_char(int n, int v)
{
    struct tex_language *l = get_language((int) n);
    if (l != NULL)
        l->pre_exhyphen_char = (int) v;
}

void set_post_exhyphen_char(int n, int v)
{
    struct tex_language *l = get_language((int) n);
    if (l != NULL)
        l->post_exhyphen_char = (int) v;
}

int get_pre_hyphen_char(int n)
{
    struct tex_language *l = get_language((int) n);
    if (l == NULL)
        return -1;
    return (int) l->pre_hyphen_char;
}

int get_post_hyphen_char(int n)
{
    struct tex_language *l = get_language((int) n);
    if (l == NULL)
        return -1;
    return (int) l->post_hyphen_char;
}

int get_pre_exhyphen_char(int n)
{
    struct tex_language *l = get_language((int) n);
    if (l == NULL)
        return -1;
    return (int) l->pre_exhyphen_char;
}

int get_post_exhyphen_char(int n)
{
    struct tex_language *l = get_language((int) n);
    if (l == NULL)
        return -1;
    return (int) l->post_exhyphen_char;
}

void set_hyphenation_min(int n, int v)
{
    struct tex_language *l = get_language((int) n);
    if (l != NULL)
        l->hyphenation_min = (int) v;
}

int get_hyphenation_min(int n)
{
    struct tex_language *l = get_language((int) n);
    if (l == NULL)
        return -1;
    return (int) l->hyphenation_min;
}

void load_patterns(struct tex_language *lang, const unsigned char *buff)
{
    if (lang == NULL || buff == NULL || strlen((const char *) buff) == 0)
        return;
    if (lang->patterns == NULL) {
        lang->patterns = hnj_hyphen_new();
    }
    hnj_hyphen_load(lang->patterns, buff);
}

void clear_patterns(struct tex_language *lang)
{
    if (lang == NULL)
        return;
    if (lang->patterns != NULL) {
        hnj_hyphen_clear(lang->patterns);
    }
}

void load_tex_patterns(int curlang, halfword head)
{
    char *s = tokenlist_to_cstring(head, 1, NULL);
    load_patterns(get_language(curlang), (unsigned char *) s);
}

@ @c
#define STORE_CHAR(l,x) do { \
    unsigned xx = get_hj_code(l,x); \
    if (!xx || xx <= 32) { \
        xx = x; \
    } \
    uindex = uni2string(uindex, xx); \
} while (0)

@ Cleans one word which is returned in |cleaned|, returns the new offset into
|buffer|

@c
const char *clean_hyphenation(int id, const char *buff, char **cleaned)
{
    int items = 0;
    unsigned char word[MAX_WORD_LEN + 1]; /* work buffer for bytes */
    unsigned uword[MAX_WORD_LEN + 1] = { 0 };  /* work buffer for unicode */
    int u = 0; /* unicode buffer value */
    int i = 0; /* index into buffer */
    char *uindex = (char *)word;
    const char *s = buff;

    while (*s && !isspace((unsigned char)*s)) {
        word[i++] = (unsigned)*s;
        s++;
        if ((s-buff)>MAX_WORD_LEN) {
            /* todo: this is too strict, should count unicode, not bytes */
            *cleaned = NULL;
            tex_error("exception too long", NULL);
            return s;
        }
    }
    /* now convert the input to unicode */
    word[i] = '\0';
    utf2uni_strcpy(uword, (const char *)word);

    /* build the new word string */
    i = 0;
    while (uword[i]>0) {
        u = uword[i++];
        if (u == '-') {        /* skip */
        } else if (u == '=') {
            STORE_CHAR(id,'-');
        } else if (u == '{') {
            u = uword[i++];
            items = 0;
            while (u && u != '}') {
                u = uword[i++];
            }
            if (u == '}') {
                items++;
                u = uword[i++];
            }
            while (u && u != '}') {
                u = uword[i++];
            }
            if (u == '}') {
                items++;
                u = uword[i++];
            }
            if (u == '{') {
                u = uword[i++];
            }
            while (u && u != '}') {
                STORE_CHAR(id,u);
                u = uword[i++];
            }
            if (u == '}') {
                items++;
            }
            if (items != 3) {   /* syntax error */
                *cleaned = NULL;
                tex_error("exception syntax error", NULL);
                return s;
            }
        } else {
            STORE_CHAR(id,u);
        }
    }
    *uindex = '\0';
    *cleaned = xstrdup((char *) word);
    return s;
}

@ @c
void load_hyphenation(struct tex_language *lang, const unsigned char *buff)
{
    const char *s;
    const char *value;
    char *cleaned;
    int id ;
    if (lang == NULL)
        return;
    if (lang->exceptions == 0) {
        lua_newtable(Luas);
        lang->exceptions = luaL_ref(Luas, LUA_REGISTRYINDEX);
    }
    lua_rawgeti(Luas, LUA_REGISTRYINDEX, lang->exceptions);
    s = (const char *) buff;
    id = lang->id;
    while (*s) {
        while (isspace((unsigned char)*s))
            s++;
        if (*s) {
            value = s;
            s = clean_hyphenation(id, s, &cleaned);
            if (cleaned != NULL) {
                if ((s - value) > 0) {
                    lua_pushstring(Luas, cleaned);
                    lua_pushlstring(Luas, value, (size_t) (s - value));
                    lua_rawset(Luas, -3);
                }
                free(cleaned);
            } else {
#ifdef VERBOSE
                formatted_warning("hyphenation","skipping invalid hyphenation exception: %s", value);
#endif
            }
        }
    }
}

void clear_hyphenation(struct tex_language *lang)
{
    if (lang == NULL)
        return;
    if (lang->exceptions != 0) {
        luaL_unref(Luas, LUA_REGISTRYINDEX, lang->exceptions);
        lang->exceptions = 0;
    }
}

void load_tex_hyphenation(int curlang, halfword head)
{
    char *s = tokenlist_to_cstring(head, 1, NULL);
    load_hyphenation(get_language(curlang), (unsigned char *) s);
}

@ TODO: clean this up. The |delete_attribute_ref()| statements are not very nice,
but needed. Also, in the post-break, it would be nicer to get the attribute list
from |vlink(n)|. No rush, as it is currently not used much.

@c
halfword insert_discretionary(halfword t, halfword pre, halfword post,
                              halfword replace, int penalty)
{
    halfword g, n;
    int f;
    n = new_node(disc_node, syllable_disc);
    disc_penalty(n) = penalty;
    try_couple_nodes(n, vlink(t));
    couple_nodes(t, n);
    if (replace != null)
        f = font(replace);
    else
        f = get_cur_font();     /* for compound words following explicit hyphens */
    for (g = pre; g != null; g = vlink(g)) {
        font(g) = f;
        if (node_attr(t) != null) {
            delete_attribute_ref(node_attr(g));
            node_attr(g) = node_attr(t);
            attr_list_ref(node_attr(t)) += 1;
        }
    }
    for (g = post; g != null; g = vlink(g)) {
        font(g) = f;
        if (node_attr(t) != null) {
            delete_attribute_ref(node_attr(g));
            node_attr(g) = node_attr(t);
            attr_list_ref(node_attr(t)) += 1;
        }
    }
    for (g = replace; g != null; g = vlink(g)) {
        if (node_attr(t) != null) {
            delete_attribute_ref(node_attr(g));
            node_attr(g) = node_attr(t);
            attr_list_ref(node_attr(t)) += 1;
        }
    }
    if (node_attr(t) != null) {
        delete_attribute_ref(node_attr(vlink(t)));
        node_attr(vlink(t)) = node_attr(t);
        attr_list_ref(node_attr(t)) += 1;
    }
    t = vlink(t);
    set_disc_field(pre_break(t), pre);
    set_disc_field(post_break(t), post);
    set_disc_field(no_break(t), replace);
    return t;
}

halfword insert_syllable_discretionary(halfword t, lang_variables * lan)
{
    halfword g, n;
    n = new_node(disc_node, syllable_disc);
    disc_penalty(n) = hyphen_penalty_par;
    couple_nodes(n, vlink(t));
    couple_nodes(t, n);
    delete_attribute_ref(node_attr(n));
    if (node_attr(t) != null) {
        node_attr(n) = node_attr(t);
        attr_list_ref(node_attr(t))++;
    } else {
        node_attr(n) = null;
    }
    if (lan->pre_hyphen_char > 0) {
        g = raw_glyph_node();
        set_to_character(g);
        character(g) = lan->pre_hyphen_char;
        font(g) = font(t);
        lang_data(g) = lang_data(t);
        if (node_attr(t) != null) {
            node_attr(g) = node_attr(t);
            attr_list_ref(node_attr(t))++;
        }
        set_disc_field(pre_break(n), g);
    }

    if (lan->post_hyphen_char > 0) {
        t = vlink(n);
        g = raw_glyph_node();
        set_to_character(g);
        character(g) = lan->post_hyphen_char;
        font(g) = font(t);
        lang_data(g) = lang_data(t);
        if (node_attr(t) != null) {
            node_attr(g) = node_attr(t);
            attr_list_ref(node_attr(t)) += 1;
        }
        set_disc_field(post_break(n), g);
    }
    return n;
}

halfword insert_word_discretionary(halfword t, lang_variables * lan)
{
    halfword pre = null, pos = null;
    if (lan->pre_exhyphen_char > 0)
        pre = insert_character(null, lan->pre_exhyphen_char);
    if (lan->post_exhyphen_char > 0)
        pos = insert_character(null, lan->post_exhyphen_char);
    return insert_discretionary(t, pre, pos, null,ex_hyphen_penalty_par);
}

@ @c
halfword compound_word_break(halfword t, int clang)
{
    int disc;
    lang_variables langdata;
    langdata.pre_exhyphen_char = get_pre_exhyphen_char(clang);
    langdata.post_exhyphen_char = get_post_exhyphen_char(clang);
    disc = insert_word_discretionary(t, &langdata);
    return disc;
}

halfword insert_complex_discretionary(halfword t, lang_variables * lan,
                                      halfword pre, halfword pos,
                                      halfword replace)
{
    (void) lan;
    return insert_discretionary(t, pre, pos, replace,hyphen_penalty_par);
}

halfword insert_character(halfword t, int c)
{
    halfword p;
    p = new_node(glyph_node, 0);
    set_to_character(p);
    character(p) = c;
    if (t != null) {
        couple_nodes(t, p);
    }
    return p;
}

@ @c
void set_disc_field(halfword f, halfword t)
{
    if (t != null) {
        /*
            couple_nodes(f, t); // better not expose f as prev pointer
        */
        vlink(f) = t ;
        alink(t) = null ;
        tlink(f) = tail_of_list(t);
    } else {
        vlink(f) = null;
        tlink(f) = null;
    }
}

@ @c
static char *hyphenation_exception(int exceptions, char *w)
{
    char *ret = NULL;
    lua_checkstack(Luas, 2);
    lua_rawgeti(Luas, LUA_REGISTRYINDEX, exceptions);
    if (lua_istable(Luas, -1)) {   /* ?? */
        lua_pushstring(Luas, w);   /* word table */
        lua_rawget(Luas, -2);
        if (lua_type(Luas, -1) == LUA_TSTRING) {
            ret = xstrdup(lua_tostring(Luas, -1));
        }
        lua_pop(Luas, 2);
    } else {
        lua_pop(Luas, 1);
    }
    return ret;
}

@ @c
char *exception_strings(struct tex_language *lang)
{
    const char *value;
    size_t size = 0, current = 0;
    size_t l = 0;
    char *ret = NULL;
    if (lang->exceptions == 0)
        return NULL;
    lua_checkstack(Luas, 2);
    lua_rawgeti(Luas, LUA_REGISTRYINDEX, lang->exceptions);
    if (lua_istable(Luas, -1)) {
        /* iterate and join */
        lua_pushnil(Luas);         /* first key */
        while (lua_next(Luas, -2) != 0) {
            value = lua_tolstring(Luas, -1, &l);
            if (current + 2 + l > size) {
                ret = xrealloc(ret, (unsigned) ((size + size / 5) + current + l + 1024));
                size = (size + size / 5) + current + l + 1024;
            }
            *(ret + current) = ' ';
            strcpy(ret + current + 1, value);
            current += l + 1;
            lua_pop(Luas, 1);
        }
    }
    return ret;
}

@ the sequence from |wordstart| to |r| can contain only normal characters it
could be faster to modify a halfword pointer and return an integer

@c
static halfword find_exception_part(unsigned int *j, unsigned int *uword, int len)
{
    halfword g = null, gg = null;
    register unsigned i = *j;
    i++;                        /* this puts uword[i] on the |{| */
    while (i < (unsigned) len && uword[i + 1] != '}') {
        if (g == null) {
            gg = new_char(0, (int) uword[i + 1]);
            g = gg;
        } else {
            halfword s = new_char(0, (int) uword[i + 1]);
            couple_nodes(g, s);
            g = vlink(g);
        }
        i++;
    }
    *j = ++i;
    return gg;
}

static int count_exception_part(unsigned int *j, unsigned int *uword, int len)
{
    int ret = 0;
    register unsigned i = *j;
    i++;                        /* this puts uword[i] on the |{| */
    while (i < (unsigned) len && uword[i + 1] != '}') {
        ret++;
        i++;
    }
    *j = ++i;
    return ret;
}

@ @c
static const char *PAT_ERROR[] = {
    "Exception discretionaries should contain three pairs of braced items.",
    "No intervening spaces are allowed.",
    NULL
};

/*
    The exceptions are taken as-is: no min values are taken into account. One can
    add normal patterns on-the-fly if needed.
*/

static void do_exception(halfword wordstart, halfword r, char *replacement)
{
    unsigned i;
    halfword t;
    unsigned len;
    int clang;
    lang_variables langdata;
    unsigned uword[MAX_WORD_LEN + 1] = { 0 };
    utf2uni_strcpy(uword, replacement);
    len = u_length(uword);
    i = 0;
    t = wordstart;
    clang = char_lang(wordstart);
    langdata.pre_hyphen_char = get_pre_hyphen_char(clang);
    langdata.post_hyphen_char = get_post_hyphen_char(clang);

    for (i = 0; i < len; i++) {
        if (uword[i + 1] == '-') {      /* a hyphen follows */
            while (vlink(t) != r && (type(t) != glyph_node || !is_simple_character(t)))
                t = vlink(t);
            if (vlink(t) == r)
                break;
            insert_syllable_discretionary(t, &langdata);
            t = vlink(t);       /* skip the new disc */
        } else if (uword[i + 1] == '=') {
            /* do nothing ? */
            t = vlink(t);
        } else if (uword[i + 1] == '{') {
            halfword gg, hh, replace = null;
            int repl;
            gg = find_exception_part(&i, uword, (int) len);
            if (i == len || uword[i + 1] != '{') {
                tex_error("broken pattern 1", PAT_ERROR);
            }
            hh = find_exception_part(&i, uword, (int) len);
            if (i == len || uword[i + 1] != '{') {
                tex_error("broken pattern 2", PAT_ERROR);
            }
            repl = count_exception_part(&i, uword, (int) len);
            if (i == len) {
                tex_error("broken pattern 3", PAT_ERROR);
            }
            /*i++;  *//* jump over the last right brace */
            if (vlink(t) == r)
                break;
            if (repl > 0) {
                halfword q = t;
                replace = vlink(q);
                while (repl > 0 && q != null) {
                    q = vlink(q);
                    if (type(q) == glyph_node) {
                        repl--;
                    }
                }
                try_couple_nodes(t, vlink(q));
                vlink(q) = null;
            }
            t = insert_discretionary(t, gg, hh, replace, hyphen_penalty_par);
            t = vlink(t);       /* skip the new disc */
        } else {
            t = vlink(t);
        }
    }
}

@ This is a documentation section from the pascal web file. It is not true any
more, but I do not have time right now to rewrite it -- Taco

When the line-breaking routine is unable to find a feasible sequence of
breakpoints, it makes a second pass over the paragraph, attempting to hyphenate
the hyphenatable words. The goal of hyphenation is to insert discretionary
material into the paragraph so that there are more potential places to break.

The general rules for hyphenation are somewhat complex and technical, because we
want to be able to hyphenate words that are preceded or followed by punctuation
marks, and because we want the rules to work for languages other than English. We
also must contend with the fact that hyphens might radically alter the ligature
and kerning structure of a word.

A sequence of characters will be considered for hyphenation only if it belongs to
a ``potentially hyphenatable part'' of the current paragraph. This is a sequence
of nodes $p_0p_1\ldots p_m$ where $p_0$ is a glue node, $p_1\ldots p_{m-1}$ are
either character or ligature or whatsit or implicit kern nodes, and $p_m$ is a
glue or penalty or insertion or adjust or mark or whatsit or explicit kern node.
(Therefore hyphenation is disabled by boxes, math formulas, and discretionary
nodes already inserted by the user.) The ligature nodes among $p_1\ldots p_{m-1}$
are effectively expanded into the original non-ligature characters; the kern
nodes and whatsits are ignored. Each character |c| is now classified as either a
nonletter (if |lc_code(c)=0|), a lowercase letter (if |lc_code(c)=c|), or an
uppercase letter (otherwise); an uppercase letter is treated as if it were
|lc_code(c)| for purposes of hyphenation. The characters generated by $p_1\ldots
p_{m-1}$ may begin with nonletters; let $c_1$ be the first letter that is not in
the middle of a ligature. Whatsit nodes preceding $c_1$ are ignored; a whatsit
found after $c_1$ will be the terminating node $p_m$. All characters that do not
have the same font as $c_1$ will be treated as nonletters. The |hyphen_char| for
that font must be between 0 and 255, otherwise hyphenation will not be attempted.
\TeX\ looks ahead for as many consecutive letters $c_1\ldots c_n$ as possible;
however, |n| must be less than 64, so a character that would otherwise be
$c_{64}$ is effectively not a letter. Furthermore $c_n$ must not be in the middle
of a ligature. In this way we obtain a string of letters $c_1\ldots c_n$ that are
generated by nodes $p_a\ldots p_b$, where |1<=a<=b+1<=m|. If |n>=l_hyf+r_hyf|,
this string qualifies for hyphenation; however, |uc_hyph| must be positive, if
$c_1$ is uppercase.

The hyphenation process takes place in three stages. First, the candidate
sequence $c_1\ldots c_n$ is found; then potential positions for hyphens are
determined by referring to hyphenation tables; and finally, the nodes $p_a\ldots
p_b$ are replaced by a new sequence of nodes that includes the discretionary
breaks found.

Fortunately, we do not have to do all this calculation very often, because of the
way it has been taken out of \TeX's inner loop. For example, when the second
edition of the author's 700-page book {\sl Seminumerical Algorithms} was typeset
by \TeX, only about 1.2 hyphenations needed to be @^Knuth, Donald Ervin@> tried
per paragraph, since the line breaking algorithm needed to use two passes on only
about 5 per cent of the paragraphs.

When a word been set up to contain a candidate for hyphenation, \TeX\ first looks
to see if it is in the user's exception dictionary. If not, hyphens are inserted
based on patterns that appear within the given word, using an algorithm due to
Frank~M. Liang. @^Liang, Franklin Mark@>

@ This is incompatible with TEX because the first word of a paragraph can be
hyphenated, but most european users seem to agree that prohibiting hyphenation
there was not the best idea ever.

@c
/*
    More strict: \hyphenationbounds

    0 = not strict
    1 = strict start
    2 = strict end
    3 = strict start and strict end

    \parindent0pt \hsize=1.1cm
    12-34-56 \par
    12-34-\hbox{56} \par
    12-34-\vrule width 1em height 1.5ex \par
    12-\hbox{34}-56 \par
    12-\vrule width 1em height 1.5ex-56 \par
    \hjcode`\1=`\1 \hjcode`\2=`\2 \hjcode`\3=`\3 \hjcode`\4=`\4 \vskip.5cm
    12-34-56 \par
    12-34-\hbox{56} \par
    12-34-\vrule width 1em height 1.5ex \par
    12-\hbox{34}-56 \par
    12-\vrule width 1em height 1.5ex-56 \par

*/

static halfword find_next_wordstart(halfword r, halfword first_language, halfword strict_bound)
{
    register int l;
    register int start_ok = 1;
    int mathlevel = 1;
    int chr ;
    halfword t ;
    while (r != null) {
        switch (type(r)) {
        case boundary_node:
            if (subtype(r) == word_boundary) {
                start_ok = 1;
            }
            break;
        case hlist_node: /* new > 0.95 */
        case vlist_node: /* new > 0.95 */
        case rule_node:  /* new > 0.95 */
        case dir_node:
        case whatsit_node:
            if (strict_bound == 1 || strict_bound == 3) {
                start_ok = 0;
            }
            break;
        case glue_node:
            start_ok = 1;
            break;
        case math_node:
            while (mathlevel > 0) {
                r = vlink(r);
                if (r == null)
                    return r;
                if (type(r) == math_node) {
                    if (subtype(r) == before) {
                        mathlevel++;
                    } else {
                        mathlevel--;
                    }
                }
            }
            break;
        case glyph_node:
            if (is_simple_character(r)) {
                chr = character(r) ;
                if (chr == ex_hyphen_char_par) {
                    /*
                        We only accept an explicit hyphen when there is a preceding glyph and
                        we skip a sequence of explicit hyphens as that normally indicates a
                        -- or --- ligature in which case we can in a worse case usage get bad
                        node lists later on due to messed up ligature building as these dashes
                        are ligatures in base fonts. This is a side effect of the separating the
                        hyphenation, ligaturing and kerning steps. A test is cmr with ------.
                    */
                    t = vlink(r) ;
                    if ((start_ok == 0) && (t!=null) && (type(t) == glyph_node) && (character(t) != ex_hyphen_char_par)) {
                        t = compound_word_break(r, char_lang(r));
                        subtype(t) = automatic_disc;
                        start_ok = 1 ;
                    } else {
                        start_ok = 0;
                    }
                } else if (start_ok && (char_lang(r)>=first_language) && ((l = get_hj_code(char_lang(r),chr)) > 0)) {
                    if (char_uchyph(r) || l == chr || l <= 32) {
                        return r;
                    } else {
                        start_ok = 0;
                    }
                }
            }
            break;
        default:
            start_ok = 0;
            break;
        }
        r = vlink(r);
    }
    return r;
}

@ @c
static int valid_wordend(halfword s, halfword strict_bound)
{
    register halfword r = s;
    register int clang = char_lang(s);
    if (r == null)
        return 1;
    while ( (r != null) &&
           (    (type(r) == glyph_node && is_simple_character(r) && clang == char_lang(r))
             || (type(r) == kern_node && (subtype(r) == normal))
            )
           ) {
        r = vlink(r);
    }
    if (r == null || (type(r) == glyph_node && is_simple_character(r) && clang != char_lang(r))
                  ||  type(r) == glue_node
                  ||  type(r) == penalty_node
                  || (type(r) == kern_node && (subtype(r) == explicit_kern || /* so why not italic correction ? */
                                               subtype(r) == italic_kern   ||
                                               subtype(r) == accent_kern   ))
                  ||  ((type(r) == hlist_node   || /* new > 0.95 */
                        type(r) == vlist_node   || /* new > 0.95 */
                        type(r) == rule_node    || /* new > 0.95 */
                        type(r) == dir_node     || /* new > 0.97 */
                        type(r) == whatsit_node ||
                        type(r) == ins_node     || /* yes or no strict test */
                        type(r) == adjust_node     /* yes or no strict test */
                       ) && ! (strict_bound == 2 || strict_bound == 3))
                  ||  type(r) == boundary_node
        )
        return 1;
    return 0;
}

@ @c
void hnj_hyphenation(halfword head, halfword tail)
{
    int lchar, i;
    struct tex_language *lang;
    lang_variables langdata;
    char utf8word[(4 * MAX_WORD_LEN) + 1] = { 0 };
    int wordlen = 0;
    char *hy = utf8word;
    char *replacement = NULL;
    boolean explicit_hyphen = false;
    halfword first_language = first_valid_language_par;
    halfword strict_bound = hyphenation_bounds_par;
    halfword s, r = head, wordstart = null, save_tail1 = null, left = null, right = null;

    /* this first movement assures two things:
     \item{a)} that we won't waste lots of time on something that has been
      handled already (in that case, none of the glyphs match |simple_character|).
     \item{b)} that the first word can be hyphenated. if the movement was
     not explicit, then the indentation at the start of a paragraph
     list would make |find_next_wordstart()| look too far ahead.
     */

    while (r != null && (type(r) != glyph_node || !is_simple_character(r))) {
        r = vlink(r);
    }
    /* this will make |r| a glyph node with subtype character */
    r = find_next_wordstart(r,first_language,strict_bound);
    if (r == null)
        return;

    assert(tail != null);
    save_tail1 = vlink(tail);
    s = new_penalty(0);
    couple_nodes(tail, s);

    while (r != null) {         /* could be while(1), but let's be paranoid */
        int clang, lhmin, rhmin, hmin;
        halfword hyf_font;
        halfword end_word = r;
        wordstart = r;
        assert(is_simple_character(wordstart));
        hyf_font = font(wordstart);
        if (hyphen_char(hyf_font) < 0)  /* for backward compat */
            hyf_font = 0;
        clang = char_lang(wordstart);
        lhmin = char_lhmin(wordstart);
        rhmin = char_rhmin(wordstart);
        hmin = get_hyphenation_min(clang);
        langdata.pre_hyphen_char = get_pre_hyphen_char(clang);
        langdata.post_hyphen_char = get_post_hyphen_char(clang);
        while (    r != null
                && type(r) == glyph_node
                && is_simple_character(r)
                && clang == char_lang(r)
                && (    (     (clang >= first_language)
                           && (lchar = get_hj_code(clang,character(r))) > 0
                        )
                     || (     character(r) == ex_hyphen_char_par
                           && (lchar = ex_hyphen_char_par)
                        )
                   )
              ) {
            if (character(r) == ex_hyphen_char_par) {
                explicit_hyphen = true;
            }
            wordlen++;
            if (lchar <= 32) {
                if (lchar == 32) {
                    lchar = 0 ;
                }
                if (wordlen <= lhmin) {
                    lhmin = lhmin - lchar + 1 ;
                    if (lhmin < 0)
                        lhmin = 1;
                }
                if (wordlen >= rhmin) {
                    rhmin = rhmin - lchar + 1 ;
                    if (rhmin < 0)
                        rhmin = 1;
                }
                hmin = hmin - lchar + 1 ;
                if (hmin < 0)
                    rhmin = 1;
                lchar = character(r) ;
            }
            hy = uni2string(hy, (unsigned) lchar);
            /* this should not be needed  any more */
            /*if (vlink(r)!=null) alink(vlink(r))=r; */
            end_word = r;
            r = vlink(r);
        }
        if (     valid_wordend(r,strict_bound)
              && clang >= first_language
              && wordlen >= lhmin + rhmin
              && (hmin <= 0 || wordlen >= hmin)
              && (hyf_font != 0)
              && (lang = tex_languages[clang]) != NULL
           ) {
            *hy = 0;
            if (    lang->exceptions != 0
                 && (replacement = hyphenation_exception(lang->exceptions, utf8word)) != NULL
               ) {
#ifdef VERBOSE
                formatted_warning("hyphenation","replacing %s (c=%d) by %s", utf8word, clang, replacement);
#endif
                do_exception(wordstart, r, replacement);
                free(replacement);
            } else if (explicit_hyphen == true) {
                /*
                    insert an explicit discretionary after each of the last in a
                    set of explicit hyphens
                */
                halfword rr = r;
                halfword t = null;
#ifdef VERBOSE
                formatted_warning("hyphenation","explicit hyphen(s) found in %s (c=%d)", utf8word, clang);
#endif
                while (rr != wordstart) {
                if (is_simple_character(rr)) {
                        if (character(rr) == ex_hyphen_char_par) {
                            t = compound_word_break(rr, clang);
                            subtype(t) = automatic_disc;
                            while (character(alink(rr)) == ex_hyphen_char_par)
                                rr = alink(rr);
                            if (rr == wordstart)
                                break;
                        }
                    }
                    rr = alink(rr);
                }
            } else if (lang->patterns != NULL) {
                left = wordstart;
                for (i = lhmin; i > 1; i--) {
                    left = vlink(left);
                    while (!is_simple_character(left)) {
                        left = vlink(left);
                    }
                    /*
                    if (!left)
                        break ;
                    */
                    /* what is left overruns right .. a bit messy */
                }
                right = r;
                for (i = rhmin; i > 0; i--) {
                    right = alink(right);
                    while (!is_simple_character(right)) {
                        right = alink(right);
                    }
                    /*
                    if (!right)
                        break ;
                    */
                    /* what is right overruns left .. a bit messy */
                }
                /* maybe an extra check ... */
                /* if (left && right) { */
#ifdef VERBOSE
                    formatted_warning("hyphenation","hyphenate %s (c=%d,l=%d,r=%d) from %c to %c",
                        utf8word, clang, lhmin, rhmin, character(left), character(right));
#endif
                    (void) hnj_hyphen_hyphenate(lang->patterns, wordstart, end_word, wordlen, left, right, &langdata);
                /* } */
            }
        }
        explicit_hyphen = false;
        wordlen = 0;
        hy = utf8word;
        if (r == null)
            break;
        r = find_next_wordstart(r,first_language,strict_bound);
    }
    flush_node(vlink(tail));
    vlink(tail) = save_tail1;
}

@ @c
void new_hyphenation(halfword head, halfword tail)
{
    register int callback_id = 0;
    if (head == null || vlink(head) == null)
        return;
    fix_node_list(head);
    callback_id = callback_defined(hyphenate_callback);
    if (callback_id > 0) {
        if (!get_callback(Luas, callback_id)) {
            lua_pop(Luas, 2);
            return;
        }
        nodelist_to_lua(Luas, head);
        nodelist_to_lua(Luas, tail);
        if (lua_pcall(Luas, 2, 0, 0) != 0) {
            formatted_warning("hyphenation","bad specification: %s",lua_tostring(Luas, -1));
            lua_pop(Luas, 2);
            lua_error(Luas);
            return;
        }
        lua_pop(Luas, 1);
    } else if (callback_id == 0) {
        hnj_hyphenation(head, tail);
    }
}

@ dumping and undumping languages

@c
#define dump_string(a)                \
  if (a!=NULL) {                      \
      x = (int)strlen(a)+1;           \
    dump_int(x);  dump_things(*a, x); \
  } else {                            \
    x = 0; dump_int(x);               \
  }

static void dump_one_language(int i)
{
    char *s = NULL;
    int x = 0;
    struct tex_language *lang;
    lang = tex_languages[i];
    dump_int(lang->id);
    dump_int(lang->pre_hyphen_char);
    dump_int(lang->post_hyphen_char);
    dump_int(lang->pre_exhyphen_char);
    dump_int(lang->post_exhyphen_char);
    dump_int(lang->hyphenation_min);
    if (lang->patterns != NULL) {
        s = (char *) hnj_serialize(lang->patterns);
    }
    dump_string(s);
    if (s != NULL) {
        free(s);
        s = NULL;
    }
    if (lang->exceptions != 0)
        s = exception_strings(lang);
    dump_string(s);
    if (s != NULL) {
        free(s);
    }
    free(lang);
}

void dump_language_data(void)
{
    int i;
    dump_int(next_lang_id);
    for (i = 0; i < next_lang_id; i++) {
        if (tex_languages[i]) {
            dump_int(1);
            dump_one_language(i);
        } else {
            dump_int(0);
        }
    }
}

static void undump_one_language(int i)
{
    char *s = NULL;
    int x = 0;
    struct tex_language *lang = get_language(i);
    undump_int(x);
    lang->id = x;
    undump_int(x);
    lang->pre_hyphen_char = x;
    undump_int(x);
    lang->post_hyphen_char = x;
    undump_int(x);
    lang->pre_exhyphen_char = x;
    undump_int(x);
    lang->post_exhyphen_char = x;
    undump_int(x);
    lang->hyphenation_min = x;
    /* patterns */
    undump_int(x);
    if (x > 0) {
        s = xmalloc((unsigned) x);
        undump_things(*s, x);
        load_patterns(lang, (unsigned char *) s);
        free(s);
    }
    /* exceptions */
    undump_int(x);
    if (x > 0) {
        s = xmalloc((unsigned) x);
        undump_things(*s, x);
        load_hyphenation(lang, (unsigned char *) s);
        free(s);
    }
}

void undump_language_data(void)
{
    int i, x, numlangs;
    undump_int(numlangs);
    next_lang_id = numlangs;
    for (i = 0; i < numlangs; i++) {
        undump_int(x);
        if (x == 1) {
            undump_one_language(i);
        }
    }
}

@ When \TeX\ has scanned `\.{\\hyphenation}', it calls on a procedure named
|new_hyph_exceptions| to do the right thing.

@c
void new_hyph_exceptions(void)
{                               /* enters new exceptions */
    (void) scan_toks(false, true);
    load_tex_hyphenation(language_par, def_ref);
    flush_list(def_ref);
}

@ Similarly, when \TeX\ has scanned `\.{\\patterns}', it calls on a
procedure named |new_patterns|.

@c
void new_patterns(void)
{                               /* initializes the hyphenation pattern data */
    (void) scan_toks(false, true);
    load_tex_patterns(language_par, def_ref);
    flush_list(def_ref);
}

@ `\.{\\prehyphenchar}', sets the |pre_break| character, and
`\.{\\posthyphenchar}' the |post_break| character. Their respective defaults are
ascii hyphen ("-") and zero (nul).

@c
void new_pre_hyphen_char(void)
{
    scan_optional_equals();
    scan_int();
    set_pre_hyphen_char(language_par, cur_val);
}

void new_post_hyphen_char(void)
{
    scan_optional_equals();
    scan_int();
    set_post_hyphen_char(language_par, cur_val);
}

@ `\.{\\preexhyphenchar}', sets the |pre_break| character, and
`\.{\\postexhyphenchar}' the |post_break| character. Their defaults are both zero
(nul).

@c
void new_pre_exhyphen_char(void)
{
    scan_optional_equals();
    scan_int();
    set_pre_exhyphen_char(language_par, cur_val);
}

void new_post_exhyphen_char(void)
{
    scan_optional_equals();
    scan_int();
    set_post_exhyphen_char(language_par, cur_val);
}

void new_hyphenation_min(void)
{
    scan_optional_equals();
    scan_int();
    set_hyphenation_min(language_par, cur_val);
}

void new_hj_code(void)
{
    int i ;
    scan_int();
    i = cur_val;
    scan_optional_equals();
    scan_int();
    set_hj_code(language_par, i, cur_val, -1);
}