=================================================================== RCS file: /cvs/texi2mdoc/main.c,v retrieving revision 1.41 retrieving revision 1.46 diff -u -p -r1.41 -r1.46 --- texi2mdoc/main.c 2015/02/24 14:35:40 1.41 +++ texi2mdoc/main.c 2015/02/25 14:37:17 1.46 @@ -1,4 +1,4 @@ -/* $Id: main.c,v 1.41 2015/02/24 14:35:40 kristaps Exp $ */ +/* $Id: main.c,v 1.46 2015/02/25 14:37:17 kristaps Exp $ */ /* * Copyright (c) 2015 Kristaps Dzonsons * @@ -38,48 +38,53 @@ static const char *const sects[SECTSZ] = { "No", }; -static void doaccent(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doblock(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dobracket(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dobye(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dodefindex(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dodefn(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dodisplay(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doenumerate(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doexample(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doignargn(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doignblock(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doignbracket(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doignline(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doinline(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doinclude(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doitem(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doitemize(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dolink(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void domacro(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void domath(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void domultitable(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doquotation(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dotable(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dotop(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosecoffs(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosection(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosp(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosubsection(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosubsubsection(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dosymbol(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dotab(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dotitle(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void dovalue(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doverb(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doverbatim(struct texi *, enum texicmd, const char *, size_t, size_t *); -static void doverbinclude(struct texi *, enum texicmd, const char *, size_t, size_t *); +static void doaccent(struct texi *, enum texicmd, size_t *); +static void doblock(struct texi *, enum texicmd, size_t *); +static void dobracket(struct texi *, enum texicmd, size_t *); +static void dobye(struct texi *, enum texicmd, size_t *); +static void dodefindex(struct texi *, enum texicmd, size_t *); +static void dodefn(struct texi *, enum texicmd, size_t *); +static void dodisplay(struct texi *, enum texicmd, size_t *); +static void doend(struct texi *, enum texicmd, size_t *); +static void doenumerate(struct texi *, enum texicmd, size_t *); +static void doexample(struct texi *, enum texicmd, size_t *); +static void doignargn(struct texi *, enum texicmd, size_t *); +static void doignblock(struct texi *, enum texicmd, size_t *); +static void doignbracket(struct texi *, enum texicmd, size_t *); +static void doignline(struct texi *, enum texicmd, size_t *); +static void doinline(struct texi *, enum texicmd, size_t *); +static void doinclude(struct texi *, enum texicmd, size_t *); +static void doitem(struct texi *, enum texicmd, size_t *); +static void doitemize(struct texi *, enum texicmd, size_t *); +static void dolink(struct texi *, enum texicmd, size_t *); +static void domacro(struct texi *, enum texicmd, size_t *); +static void domath(struct texi *, enum texicmd, size_t *); +static void domultitable(struct texi *, enum texicmd, size_t *); +static void doquotation(struct texi *, enum texicmd, size_t *); +static void dotable(struct texi *, enum texicmd, size_t *); +static void dotop(struct texi *, enum texicmd, size_t *); +static void dosecoffs(struct texi *, enum texicmd, size_t *); +static void dosection(struct texi *, enum texicmd, size_t *); +static void dosp(struct texi *, enum texicmd, size_t *); +static void dosubsection(struct texi *, enum texicmd, size_t *); +static void dosubsubsection(struct texi *, enum texicmd, size_t *); +static void dosymbol(struct texi *, enum texicmd, size_t *); +static void dotab(struct texi *, enum texicmd, size_t *); +static void dotitle(struct texi *, enum texicmd, size_t *); +static void dovalue(struct texi *, enum texicmd, size_t *); +static void doverb(struct texi *, enum texicmd, size_t *); +static void doverbatim(struct texi *, enum texicmd, size_t *); +static void doverbinclude(struct texi *, enum texicmd, size_t *); static const struct texitok __texitoks[TEXICMD__MAX] = { /* TEXICMD__BEGIN */ + { doignline, "afourpaper", 10 }, /* TEXICMD_A4PAPER */ + { dosymbol, "AA", 2 }, /* TEXICMD_AA */ + { dosymbol, "aa", 2 }, /* TEXICMD_AASMALL */ { doignargn, "acronym", 7 }, /* TEXICMD_ACRONYM */ { doaccent, "'", 1 }, /* TEXICMD_ACUTE */ - { doignline, "afourpaper", 10 }, /* TEXICMD_A4PAPER */ + { dosymbol, "AE", 2 }, /* TEXICMD_AE */ + { dosymbol, "ae", 2 }, /* TEXICMD_AESMALL */ { doignbracket, "anchor", 6 }, /* TEXICMD_ANCHOR */ { dosection, "appendix", 8 }, /* TEXICMD_APPENDIX */ { dosection, "appendixsec", 11 }, /* TEXICMD_APPENDIXSEC */ @@ -137,32 +142,51 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dodefn, "defvrx", 6 }, /* TEXICMD_DEFVRX */ { doignblock, "detailmenu", 10 }, /* TEXICMD_DETAILMENU */ { doinline, "dfn", 3 }, /* TEXICMD_DFN */ + { dosymbol, "DH", 2 }, /* TEXICMD_DH */ + { dosymbol, "dh", 2 }, /* TEXICMD_DHSMALL */ { doignline, "dircategory", 11 }, /* TEXICMD_DIRCATEGORY */ { doignblock, "direntry", 8 }, /* TEXICMD_DIRENTRY */ { dodisplay, "display", 7 }, /* TEXICMD_DISPLAY */ { doignbracket, "dmn", 3 }, /* TEXICMD_DMN */ { doignblock, "documentdescription", 19 }, /* TEXICMD_DOCUMENTDESCRIPTION */ { doignline, "documentencoding", 16 }, /* TEXICMD_DOCUMENTENCODING */ + { doignline, "documentlanguage", 16 }, /* TEXICMD_DOCUMENTLANGUAGE */ + { doaccent, "dotaccent", 9 }, /* TEXICMD_DOTACCENT */ + { doaccent, "dotless", 7 }, /* TEXICMD_DOTLESS */ { dosymbol, "dots", 4 }, /* TEXICMD_DOTS */ { dolink, "email", 5 }, /* TEXICMD_EMAIL */ { doinline, "emph", 4 }, /* TEXICMD_EMPH */ - { NULL, "end", 3 }, /* TEXICMD_END */ + { doend, "end", 3 }, /* TEXICMD_END */ { dosymbol, "enddots", 7 }, /* TEXICMD_ENDDOTS */ { doenumerate, "enumerate", 9 }, /* TEXICMD_ENUMERATE */ { doinline, "env", 3 }, /* TEXICMD_ENV */ { dosymbol, "equiv", 5 }, /* TEXICMD_EQUIV */ { dosymbol, "error", 5 }, /* TEXICMD_ERROR */ + { dosymbol, "euro", 4 }, /* TEXICMD_EURO */ { doexample, "example", 7 }, /* TEXICMD_EXAMPLE */ + { dosymbol, "exclamdown", 10 }, /* TEXICMD_EXCLAMDOWN */ { doignline, "exdent", 6 }, /* TEXICMD_EXDENT */ { dosymbol, "expansion", 9 }, /* TEXICMD_EXPANSION */ { doinline, "file", 4 }, /* TEXICMD_FILE */ { doignline, "finalout", 8 }, /* TEXICMD_FINALOUT */ { doignline, "findex", 6 }, /* TEXICMD_FINDEX */ + { doblock, "flushleft", 9 }, /* TEXICMD_FLUSHLEFT */ + { doblock, "flushright", 10 }, /* TEXICMD_FLUSHRIGHT */ + { doignline, "firstparagraphindent", 20 }, /* TEXICMD_FIRSTPARAGRAPHINDENT */ { doignbracket, "footnote", 8 }, /* TEXICMD_FOOTNOTE */ + { doignline, "footnotestyle", 13 }, /* TEXICMD_FOOTNOTESTYLE */ { dotable, "ftable", 6 }, /* TEXICMD_FTABLE */ { dodisplay, "format", 6 }, /* TEXICMD_FORMAT */ + { dosymbol, "geq", 3 }, /* TEXICMD_GEQ */ { doaccent, "`", 1 }, /* TEXICMD_GRAVE */ { doblock, "group", 5 }, /* TEXICMD_GROUP */ + { dosymbol, "guillemetleft", 13 }, /* TEXICMD_GUILLEMETLEFT */ + { dosymbol, "guillemetright", 14 }, /* TEXICMD_GUILLEMETRIGHT */ + { dosymbol, "guillemotleft", 13 }, /* TEXICMD_GUILLEMOTLEFT */ + { dosymbol, "guillemotright", 14 }, /* TEXICMD_GUILLEMOTRIGHT */ + { dosymbol, "guilsinglleft", 13 }, /* TEXICMD_GUILSINGLLEFT */ + { dosymbol, "guilsinglright", 14 }, /* TEXICMD_GUILSINGLRIGHT */ + { doaccent, "H", 1 }, /* TEXICMD_H */ { dosection, "heading", 7 }, /* TEXICMD_HEADING */ { doignline, "headings", 8 }, /* TEXICMD_HEADINGS */ { doitem, "headitem", 8 }, /* TEXICMD_HEADITEM */ @@ -187,7 +211,8 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { doignbracket, "image", 5 }, /* TEXICMD_IMAGE */ { doinclude, "include", 7 }, /* TEXICMD_INCLUDE */ { dodisplay, "indentblock", 11 }, /* TEXICMD_INDENTBLOCK */ - { dolink, "inforef", 7 }, /* TEXICMD_INDENTBLOCK */ + { dolink, "indicateurl", 11 }, /* TEXICMD_INDICATEURL */ + { dolink, "inforef", 7 }, /* TEXICMD_INFOREF */ { doignline, "insertcopying", 13 }, /* TEXICMD_INSERTCOPYING */ { doitem, "item", 4 }, /* TEXICMD_ITEM */ { doitemize, "itemize", 7 }, /* TEXICMD_ITEMIZE */ @@ -195,9 +220,13 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { doinline, "kbd", 3 }, /* TEXICMD_KBD */ { dobracket, "key", 3 }, /* TEXICMD_KEY */ { doignline, "kindex", 6 }, /* TEXICMD_KINDEX */ + { dosymbol, "L", 1 }, /* TEXICMD_L */ { dosymbol, "LaTeX", 5 }, /* TEXICMD_LATEX */ + { dosymbol, "leq", 3 }, /* TEXICMD_LEQ */ { dosecoffs, "lowersections", 13 }, /* TEXICMD_LOWERSECTIONS */ + { dosymbol, "l", 1 }, /* TEXICMD_LSMALL */ { domacro, "macro", 5 }, /* TEXICMD_MACRO */ + { doaccent, "=", 1 }, /* TEXICMD_MACRON */ { domath, "math", 4 }, /* TEXICMD_MATH */ { doignblock, "menu", 4 }, /* TEXICMD_MENU */ { dosymbol, "minus", 5 }, /* TEXICMD_MINUS */ @@ -206,20 +235,37 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dosymbol, "\n", 1 }, /* TEXICMD_NEWLINE */ { doignline, "node", 4 }, /* TEXICMD_NODE */ { doignline, "noindent", 8 }, /* TEXICMD_NOINDENT */ + { dosymbol, "O", 1 }, /* TEXICMD_O */ + { dosymbol, "OE", 2 }, /* TEXICMD_OE */ + { dosymbol, "oe", 2 }, /* TEXICMD_OESMALL */ + { doaccent, "ogonek", 6 }, /* TEXICMD_OGONEK */ { doinline, "option", 6 }, /* TEXICMD_OPTION */ - { dolink, "pxref", 5 }, /* TEXICMD_PXREF */ - { dosymbol, "?", 1 }, /* TEXICMD_QUESTIONMARK */ - { doquotation, "quotation", 9 }, /* TEXICMD_QUOTATION */ + { dosymbol, "ordf", 4 }, /* TEXICMD_ORDF */ + { dosymbol, "ordm", 4 }, /* TEXICMD_ORDM */ + { dosymbol, "o", 1 }, /* TEXICMD_OSMALL */ { doignline, "page", 4 }, /* TEXICMD_PAGE */ { doignline, "paragraphindent", 15 }, /* TEXICMD_PARINDENT */ { dosymbol, ".", 1 }, /* TEXICMD_PERIOD */ { doignline, "pindex", 6 }, /* TEXICMD_PINDEX */ + { dosymbol, "pounds", 6 }, /* TEXICMD_POUNDS */ { doignline, "printindex", 10 }, /* TEXICMD_PRINTINDEX */ + { dolink, "pxref", 5 }, /* TEXICMD_PXREF */ + { dosymbol, "questiondown", 12 }, /* TEXICMD_QUESTIONDOWN */ + { dosymbol, "?", 1 }, /* TEXICMD_QUESTIONMARK */ + { doquotation, "quotation", 9 }, /* TEXICMD_QUOTATION */ + { dosymbol, "quotedblbase", 12 }, /* TEXICMD_QUOTEDBLBASE */ + { dosymbol, "quotedblleft", 12 }, /* TEXICMD_QUOTEDBLLEFT */ + { dosymbol, "quotedblright", 13 }, /* TEXICMD_QUOTEDBLRIGHT */ + { dosymbol, "quotesinglbase", 14 }, /* TEXICMD_QUOTESINGLBASE */ + { dosymbol, "quoteleft", 9 }, /* TEXICMD_QUOTELEFT */ + { dosymbol, "quoteright", 10 }, /* TEXICMD_QUOTERIGHT */ { doinline, "r", 1 }, /* TEXICMD_R */ { dosecoffs, "raisesections", 13 }, /* TEXICMD_RAISESECTIONS */ { dobracket, "ref", 3 }, /* TEXICMD_REF */ { doignline, "refill", 6 }, /* TEXICMD_REFILL */ + { dosymbol, "registeredsymbol", 16 }, /* TEXICMD_REGISTEREDSYMBOL */ { dosymbol, "result", 6 }, /* TEXICMD_RESULT */ + { doaccent, "ringaccent", 10 }, /* TEXICMD_RINGACCENT */ { doinline, "samp", 4 }, /* TEXICMD_SAMP */ { doinline, "sansserif", 9 }, /* TEXICMD_SANSSERIF */ { dobracket, "sc", 2 }, /* TEXICMD_SC */ @@ -257,13 +303,20 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dotable, "table", 5 }, /* TEXICMD_TABLE */ { doignblock, "tex", 3 }, /* TEXICMD_TEX */ { dosymbol, "TeX", 3 }, /* TEXICMD_TEXSYM */ + { dosymbol, "textdegree", 10 }, /* TEXICMD_TEXTDEGREE */ + { dosymbol, "TH", 2 }, /* TEXICMD_TH */ + { dosymbol, "th", 2 }, /* TEXICMD_THSMALL */ { dosymbol, "tie", 3 }, /* TEXICMD_TIE */ + { doaccent, "tieaccent", 9 }, /* TEXICMD_TIEACCENT */ { doaccent, "~", 1 }, /* TEXICMD_TILDE */ { doignline, "tindex", 6 }, /* TEXICMD_TINDEX */ { doignline, "title", 5 }, /* TEXICMD_TITLE */ { dobracket, "titlefont", 9 }, /* TEXICMD_TITLEFONT */ { doignblock, "titlepage", 9 }, /* TEXICMD_TITLEPAGE */ { dotop, "top", 3 }, /* TEXICMD_TOP */ + { doaccent, "u", 1 }, /* TEXICMD_U */ + { doaccent, "ubaraccent", 10 }, /* TEXICMD_UBARACCENT */ + { doaccent, "udotaccent", 10 }, /* TEXICMD_UDOTACCENT */ { doaccent, "\"", 1 }, /* TEXICMD_UMLAUT */ { dosection, "unnumbered", 10 }, /* TEXICMD_UNNUMBERED */ { dosection, "unnumberedsec", 13 }, /* TEXICMD_UNNUMBEREDSEC */ @@ -272,6 +325,7 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dolink, "uref", 4 }, /* TEXICMD_UREF */ { dolink, "url", 3 }, /* TEXICMD_URL */ { doignline, "", 0 }, /* TEXICMD_USER_INDEX */ + { doaccent, "v", 1 }, /* TEXICMD_V */ { dovalue, "value", 5 }, /* TEXICMD_VALUE */ { doinline, "var", 3 }, /* TEXICMD_VAR */ { doverb, "verb", 4 }, /* TEXICMD_VERB */ @@ -288,26 +342,25 @@ static const struct texitok __texitoks[TEXICMD__MAX] = const struct texitok *const texitoks = __texitoks; static void -dodefindex(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dodefindex(struct texi *p, enum texicmd cmd, size_t *pos) { size_t start, end; char *cp; - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); start = end = *pos; - while (end < sz && ! ismspace(buf[end])) + while (end < BUFSZ(p) && ! ismspace(BUF(p)[end])) end++; if (start == end) { - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); return; } else if (NULL == (cp = malloc(end - start + 1))) texiabort(p, NULL); - memcpy(cp, &buf[start], end - start); + memcpy(cp, &BUF(p)[start], end - start); cp[end - start] = '\0'; p->indexs = realloc(p->indexs, @@ -319,8 +372,7 @@ dodefindex(struct texi *p, enum texicmd cmd, } static void -dodefn(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dodefn(struct texi *p, enum texicmd cmd, size_t *pos) { const char *blk; @@ -345,8 +397,8 @@ dodefn(struct texi *p, enum texicmd cmd, if (p->ign) { NULL != blk ? - parseto(p, buf, sz, pos, blk) : - parseeoln(p, buf, sz, pos); + parseto(p, pos, blk) : + parseeoln(p, pos); return; } @@ -375,7 +427,7 @@ dodefn(struct texi *p, enum texicmd cmd, texiputchars(p, "Function"); break; default: - parselinearg(p, buf, sz, pos); + parselinearg(p, pos); break; } @@ -385,7 +437,7 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFMAC): case (TEXICMD_DEFMACX): teximacroopen(p, "Dv"); - while (parselinearg(p, buf, sz, pos)) + while (parselinearg(p, pos)) /* Spin. */ ; teximacroclose(p); break; @@ -394,10 +446,10 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFUN): case (TEXICMD_DEFUNX): teximacroopen(p, "Fo"); - parselinearg(p, buf, sz, pos); + parselinearg(p, pos); teximacroclose(p); teximacroopen(p, "Fa"); - while (parselinearg(p, buf, sz, pos)) + while (parselinearg(p, pos)) /* Spin. */ ; teximacroclose(p); teximacro(p, "Fc"); @@ -409,13 +461,13 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFTYPEMETHOD): case (TEXICMD_DEFTYPEMETHODX): teximacroopen(p, "Ft"); - parselinearg(p, buf, sz, pos); + parselinearg(p, pos); teximacroclose(p); teximacroopen(p, "Fo"); - parselinearg(p, buf, sz, pos); + parselinearg(p, pos); teximacroclose(p); teximacroopen(p, "Fa"); - while (parselinearg(p, buf, sz, pos)) + while (parselinearg(p, pos)) /* Spin. */ ; teximacroclose(p); teximacro(p, "Fc"); @@ -427,7 +479,7 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFTYPEVR): case (TEXICMD_DEFTYPEVRX): teximacroopen(p, "Vt"); - while (parselinearg(p, buf, sz, pos)) + while (parselinearg(p, pos)) /* Spin. */ ; teximacroclose(p); break; @@ -436,7 +488,7 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFVR): case (TEXICMD_DEFVRX): teximacroopen(p, "Va"); - while (parselinearg(p, buf, sz, pos)) + while (parselinearg(p, pos)) /* Spin. */ ; teximacroclose(p); break; @@ -446,12 +498,11 @@ dodefn(struct texi *p, enum texicmd cmd, texivspace(p); if (NULL != blk) - parseto(p, buf, sz, pos, blk); + parseto(p, pos, blk); } static void -domacro(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +domacro(struct texi *p, enum texicmd cmd, size_t *pos) { size_t start, end, endtoksz, len; struct teximacro m; @@ -459,42 +510,58 @@ domacro(struct texi *p, enum texicmd cmd, memset(&m, 0, sizeof(struct teximacro)); - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); - for (start = end = *pos; end < sz; end++) - if (ismspace(buf[end]) || '{' == buf[end]) + for (start = end = *pos; end < BUFSZ(p); end++) + if (ismspace(BUF(p)[end]) || '{' == BUF(p)[end]) break; if (start == end) texierr(p, "zero-length macro name"); - advanceto(p, buf, pos, end); + advanceto(p, pos, end); m.key = malloc(end - start + 1); if (NULL == m.key) texiabort(p, NULL); - memcpy(m.key, &buf[start], end - start); + memcpy(m.key, &BUF(p)[start], end - start); m.key[end - start] = '\0'; - m.args = argparse(p, buf, sz, pos, &m.argsz, 0); - advanceeoln(p, buf, sz, pos, 0); + m.args = argparse(p, pos, &m.argsz, 0); + /* Note: we advance to the beginning of the macro. */ + advanceeoln(p, pos, 1); + + /* + * According to the Texinfo manual, the macro ends on the + * newline subsequent the @end macro. + * That's COMPLETELY FUCKING WRONG. + * It ends inclusive the newline, which is why so many macros + * say things like @r{hello}@c, where the subsequent @c swallows + * the newline. + * However, it does swallow the leading newline, so look for the + * @end macro without the leading newline else we might look + * past empty macros. + */ start = *pos; - endtok = "\n@end macro\n"; + endtok = "@end macro\n"; endtoksz = strlen(endtok); - blk = memmem(&buf[start], sz, endtok, endtoksz); + blk = memmem(&BUF(p)[start], BUFSZ(p) - start, endtok, endtoksz); if (NULL == blk) texierr(p, "unterminated macro body"); - while (&buf[*pos] != blk) - advance(p, buf, pos); - assert('\n' == buf[*pos]); - advance(p, buf, pos); - len = blk - &buf[start]; + /* Roll us back one character. */ + while (&BUF(p)[*pos] != blk) + advance(p, pos); + assert('@' == BUF(p)[*pos]); + if ('\n' != BUF(p)[*pos - 1]) + texierr(p, "cannot handle @end macro in-line"); + + len = blk - &BUF(p)[start]; m.value = malloc(len + 1); if (NULL == m.value) texiabort(p, NULL); - memcpy(m.value, &buf[start], len); + memcpy(m.value, &BUF(p)[start], len); m.value[len] = '\0'; p->macros = realloc @@ -505,54 +572,81 @@ domacro(struct texi *p, enum texicmd cmd, texiabort(p, NULL); p->macros[p->macrosz++] = m; - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); } static void -doignblock(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doignblock(struct texi *p, enum texicmd cmd, size_t *pos) { - char end[32]; - const char *term; - size_t endsz, endpos; + char end[32], start[32]; + const char *endt, *startt; + size_t esz, ssz, newpos, stack; /* - * We want to completely ignore everything in these blocks, so - * simply jump to the @end block. + * FIXME: this is cheating. + * These tokens are supposed to begin on a newline. + * However, if we do that, then we would need to check within + * the loop for trailer (or leading, as the case may be) + * newline, and that's just a bit too complicated right now. + * This is becasue + * @ifset BAR + * @ifset FOO + * @end ifset + * @end ifset + * won't work right now: we'd read after the first "@end ifset" + * to the next line, then look for the next line after that. */ - endsz = snprintf(end, sizeof(end), - "\n@end %s\n", texitoks[cmd].tok); - assert(endsz < sizeof(end)); + ssz = snprintf(start, sizeof(start), + "@%s", texitoks[cmd].tok); + assert(ssz < sizeof(start)); + esz = snprintf(end, sizeof(end), + "@end %s\n", texitoks[cmd].tok); + assert(esz < sizeof(end)); + stack = 1; - /* - * Look up where our end token occurs. - * Set our end position based on the relative offset of that - * from our current position, or the EOF if we don't have a - * proper ending point. + /* + * Here we look for the end token "end" somewhere in the file in + * front of us. + * It's not that easy, of course: if we have a nested block, + * then there'll be an "end" token of the same kind between us. + * Thus, we keep track of scopes for matching "end" blocks. */ - term = memmem(&buf[*pos], sz, end, endsz); - endpos = NULL == term ? sz : - *pos + (size_t)(term - &buf[*pos]); - assert(endpos <= sz); - while (*pos < endpos) - advance(p, buf, pos); + while (stack > 0 && *pos < BUFSZ(p)) { + if (stack > 10) + abort(); + endt = memmem(&BUF(p)[*pos], BUFSZ(p) - *pos, end, esz); + startt = memmem(&BUF(p)[*pos], BUFSZ(p) - *pos, start, ssz); + if (NULL == endt) { + texiwarn(p, "unterminated \"%s\" " + "block", texitoks[cmd].tok); + *pos = BUFSZ(p); + break; + } - /* Only do this if we're not already at the end. */ - if (endpos < sz) - advanceto(p, buf, pos, endpos + endsz); + newpos = *pos; + if (NULL == startt || startt > endt) { + newpos += esz + (size_t)(endt - &BUF(p)[*pos]); + stack--; + } else { + newpos += ssz + (size_t)(startt - &BUF(p)[*pos]); + stack++; + } + + assert(newpos <= BUFSZ(p)); + while (*pos < newpos) + advance(p, pos); + } } static void -doblock(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doblock(struct texi *p, enum texicmd cmd, size_t *pos) { - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); } static void -doinline(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doinline(struct texi *p, enum texicmd cmd, size_t *pos) { const char *macro = NULL; @@ -594,91 +688,92 @@ doinline(struct texi *p, enum texicmd cmd, } if (NULL == macro || p->literal || TEXILIST_TABLE == p->list) { - parsebracket(p, buf, sz, pos); + parsebracket(p, pos); return; } teximacroopen(p, macro); p->seenws = 0; - parsebracket(p, buf, sz, pos); - texipunctuate(p, buf, sz, pos); + parsebracket(p, pos); + texipunctuate(p, pos); teximacroclose(p); } static void -doverb(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doverb(struct texi *p, enum texicmd cmd, size_t *pos) { char delim; size_t start; - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - if (*pos == sz || '{' != buf[*pos]) + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + if (*pos == BUFSZ(p) || '{' != BUF(p)[*pos]) return; - advance(p, buf, pos); - if (*pos == sz) + advance(p, pos); + if (*pos == BUFSZ(p)) return; - delim = buf[*pos]; - advance(p, buf, pos); + delim = BUF(p)[*pos]; + advance(p, pos); /* Make sure we flush out our initial whitespace... */ if (p->seenws && p->outcol && 0 == p->literal) texiputchar(p, ' '); p->seenws = 0; start = *pos; /* Read until we see the delimiter then end-brace. */ - while (*pos < sz - 1) { - if (buf[*pos] == delim && buf[*pos + 1] == '}') + while (*pos < BUFSZ(p) - 1) { + if (BUF(p)[*pos] == delim && BUF(p)[*pos + 1] == '}') break; - advance(p, buf, pos); + advance(p, pos); } - if (*pos == sz - 1) + if (*pos == BUFSZ(p) - 1) return; - texiputbuf(p, buf, start, *pos); + texiputbuf(p, start, *pos); /* Make sure we read after the end-brace. */ - assert(delim == buf[*pos]); - advance(p, buf, pos); - assert('}' == buf[*pos]); - advance(p, buf, pos); + assert(delim == BUF(p)[*pos]); + advance(p, pos); + assert('}' == BUF(p)[*pos]); + advance(p, pos); } static void -doverbatim(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doverbatim(struct texi *p, enum texicmd cmd, size_t *pos) { const char *end, *term; size_t endsz, endpos; - advanceeoln(p, buf, sz, pos, 1); - - /* We end at exactly this token. */ + /* We read from the @verbatim\n newline inclusive! */ end = "\n@end verbatim\n"; endsz = strlen(end); + advanceeoln(p, pos, 0); + if (*pos == BUFSZ(p)) { + texiwarn(p, "unexpected end of file"); + return; + } - /* - * Look up where our end token occurs. - * Set our end position based on the relative offset of that - * from our current position. - */ - term = memmem(&buf[*pos], sz, end, endsz); - endpos = NULL == term ? sz : - *pos + (size_t)(term - &buf[*pos]); + term = memmem(&BUF(p)[*pos], BUFSZ(p) - *pos, end, endsz); + if (NULL == term) { + texiwarn(p, "unterminated verbatim block"); + endpos = BUFSZ(p); + } else + endpos = *pos + (size_t)(term - &BUF(p)[*pos]); + assert(endpos <= BUFSZ(p)); + assert('\n' == BUF(p)[*pos]); + advance(p, pos); teximacro(p, "Bd -literal -offset indent"); - assert(endpos <= sz); while (*pos < endpos) { - texiputchar(p, buf[*pos]); - advance(p, buf, pos); + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); } teximacro(p, "Ed"); - advanceto(p, buf, pos, endpos + endsz); + if (*pos < BUFSZ(p)) + advanceto(p, pos, endpos + endsz); } static void -doverbinclude(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doverbinclude(struct texi *p, enum texicmd cmd, size_t *pos) { char fname[PATH_MAX], path[PATH_MAX]; int rc; @@ -686,22 +781,22 @@ doverbinclude(struct texi *p, enum texicmd cmd, const char *v; enum texicmd type; - while (*pos < sz && ' ' == buf[*pos]) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && ' ' == BUF(p)[*pos]) + advance(p, pos); - for (i = 0; *pos < sz && '\n' != buf[*pos]; ) { + for (i = 0; *pos < BUFSZ(p) && '\n' != BUF(p)[*pos]; ) { if (i == sizeof(fname) - 1) break; - if ('@' != buf[*pos]) { - fname[i++] = buf[*pos]; - advance(p, buf, pos); + if ('@' != BUF(p)[*pos]) { + fname[i++] = BUF(p)[*pos]; + advance(p, pos); continue; } - type = texicmd(p, buf, *pos, sz, &end, NULL); - advanceto(p, buf, pos, end); + type = texicmd(p, *pos, &end, NULL); + advanceto(p, pos, end); if (TEXICMD_VALUE != type) texierr(p, "unknown verbatiminclude command"); - v = valueblookup(p, buf, sz, pos); + v = valueblookup(p, pos); if (NULL == v) continue; while ('\0' != *v) { @@ -715,7 +810,7 @@ doverbinclude(struct texi *p, enum texicmd cmd, if (i == 0) texierr(p, "path too short"); - else if ('\n' != buf[*pos]) + else if ('\n' != BUF(p)[*pos]) texierr(p, "path too long"); else if ('/' == fname[0]) texierr(p, "no absolute paths"); @@ -735,8 +830,7 @@ doverbinclude(struct texi *p, enum texicmd cmd, } static void -doinclude(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doinclude(struct texi *p, enum texicmd cmd, size_t *pos) { char fname[PATH_MAX], path[PATH_MAX]; size_t i, end; @@ -744,23 +838,23 @@ doinclude(struct texi *p, enum texicmd cmd, const char *v; enum texicmd type; - while (*pos < sz && ' ' == buf[*pos]) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && ' ' == BUF(p)[*pos]) + advance(p, pos); /* Read in the filename. */ - for (i = 0; *pos < sz && '\n' != buf[*pos]; ) { + for (i = 0; *pos < BUFSZ(p) && '\n' != BUF(p)[*pos]; ) { if (i == sizeof(fname) - 1) break; - if ('@' != buf[*pos]) { - fname[i++] = buf[*pos]; - advance(p, buf, pos); + if ('@' != BUF(p)[*pos]) { + fname[i++] = BUF(p)[*pos]; + advance(p, pos); continue; } - type = texicmd(p, buf, *pos, sz, &end, NULL); - advanceto(p, buf, pos, end); + type = texicmd(p, *pos, &end, NULL); + advanceto(p, pos, end); if (TEXICMD_VALUE != type) texierr(p, "unknown include command"); - v = valueblookup(p, buf, sz, pos); + v = valueblookup(p, pos); if (NULL == v) continue; while ('\0' != *v) { @@ -774,7 +868,7 @@ doinclude(struct texi *p, enum texicmd cmd, if (i == 0) texierr(p, "path too short"); - else if ('\n' != buf[*pos]) + else if ('\n' != BUF(p)[*pos]) texierr(p, "path too long"); else if ('/' == fname[0]) texierr(p, "no absolute paths"); @@ -801,16 +895,14 @@ doinclude(struct texi *p, enum texicmd cmd, } static void -dobracket(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dobracket(struct texi *p, enum texicmd cmd, size_t *pos) { - parsebracket(p, buf, sz, pos); + parsebracket(p, pos); } static void -dodisplay(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dodisplay(struct texi *p, enum texicmd cmd, size_t *pos) { switch (cmd) { @@ -825,28 +917,26 @@ dodisplay(struct texi *p, enum texicmd cmd, p->seenvs = 1; /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); - parseto(p, buf, sz, pos, texitoks[cmd].tok); + advanceeoln(p, pos, 1); + parseto(p, pos, texitoks[cmd].tok); teximacro(p, "Ed"); } static void -doexample(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doexample(struct texi *p, enum texicmd cmd, size_t *pos) { teximacro(p, "Bd -literal -offset indent"); /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); p->literal++; - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); p->literal--; teximacro(p, "Ed"); } static void -dobye(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dobye(struct texi *p, enum texicmd cmd, size_t *pos) { texiexit(p); @@ -854,112 +944,140 @@ dobye(struct texi *p, enum texicmd cmd, } static void -dotitle(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dotitle(struct texi *p, enum texicmd cmd, size_t *pos) { size_t start, end; - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); start = end = *pos; - while (end < sz && '\n' != buf[end]) + while (end < BUFSZ(p) && '\n' != BUF(p)[end]) end++; - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); free(p->subtitle); p->subtitle = malloc(end - start + 1); if (NULL == p->subtitle) texiabort(p, NULL); - memcpy(p->subtitle, &buf[start], end - start); + memcpy(p->subtitle, &BUF(p)[start], end - start); p->subtitle[end - start] = '\0'; } static void -doaccent(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doaccent(struct texi *p, enum texicmd cmd, size_t *pos) { + int brace = 0; - if (*pos == sz) + if (*pos == BUFSZ(p)) { + texiwarn(p, "truncated: @%s", texitoks[cmd].tok); return; - if (p->seenws && p->outcol && 0 == p->literal) + } + + /* Pad us with space, if necessary. */ + if (p->seenws && p->outcol && 0 == p->literal) { texiputchar(p, ' '); - p->seenws = 0; + p->seenws = 0; + } + + /* + * If we're braced, then that's that. + * Otherwise, in a special Texinfo case: if we're a non + * alphabetic command of one letter, then the next character is + * the critical one. + * Otherwise, space can sit between us and our argument. + */ + if ('{' == BUF(p)[*pos]) { + brace = 1; + advance(p, pos); + } else if (isalpha(texitoks[cmd].tok[0])) + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + + if (*pos == BUFSZ(p)) { + texiwarn(p, "truncated: @%s", texitoks[cmd].tok); + return; + } + switch (cmd) { - case (TEXICMD_CEDILLA): - /* Strange rules... */ - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - if (*pos == sz || '{' != buf[*pos]) - return; - advance(p, buf, pos); - switch (buf[*pos]) { - case ('c'): case ('C'): - texiputchars(p, "\\(,"); - texiputchar(p, buf[*pos]); - break; - default: - texiputchar(p, buf[*pos]); - break; - } - advance(p, buf, pos); - break; case (TEXICMD_ACUTE): - switch (buf[*pos]) { + switch (BUF(p)[*pos]) { case ('a'): case ('A'): case ('e'): case ('E'): case ('i'): case ('I'): case ('o'): case ('O'): case ('u'): case ('U'): texiputchars(p, "\\(\'"); - texiputchar(p, buf[*pos]); break; default: - texiputchar(p, buf[*pos]); + texiwarn(p, "ignoring accent"); + break; } + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); break; + case (TEXICMD_CEDILLA): + if ('c' == BUF(p)[*pos] || 'C' == BUF(p)[*pos]) + texiputchars(p, "\\(,"); + else + texiwarn(p, "ignoring accent"); + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); + break; case (TEXICMD_CIRCUMFLEX): - switch (buf[*pos]) { + switch (BUF(p)[*pos]) { case ('a'): case ('A'): case ('e'): case ('E'): case ('i'): case ('I'): case ('o'): case ('O'): case ('u'): case ('U'): texiputchars(p, "\\(^"); - texiputchar(p, buf[*pos]); break; default: - texiputchar(p, buf[*pos]); + texiwarn(p, "ignoring accent"); break; } + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); break; + case (TEXICMD_DOTLESS): + if ('i' == BUF(p)[*pos] || 'j' == BUF(p)[*pos]) + texiputchars(p, "\\(."); + else + texiwarn(p, "ignoring accent"); + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); + break; case (TEXICMD_GRAVE): - switch (buf[*pos]) { + switch (BUF(p)[*pos]) { case ('a'): case ('A'): case ('e'): case ('E'): case ('i'): case ('I'): case ('o'): case ('O'): case ('u'): case ('U'): texiputchars(p, "\\(`"); - texiputchar(p, buf[*pos]); break; default: - texiputchar(p, buf[*pos]); + texiwarn(p, "ignoring accent"); + break; } + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); break; case (TEXICMD_TILDE): - switch (buf[*pos]) { + switch (BUF(p)[*pos]) { case ('a'): case ('A'): case ('n'): case ('N'): case ('o'): case ('O'): texiputchars(p, "\\(~"); - texiputchar(p, buf[*pos]); break; default: - texiputchar(p, buf[*pos]); + texiwarn(p, "ignoring accent"); break; } + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); break; case (TEXICMD_UMLAUT): - switch (buf[*pos]) { + switch (BUF(p)[*pos]) { case ('a'): case ('A'): case ('e'): case ('E'): case ('i'): case ('I'): @@ -967,30 +1085,64 @@ doaccent(struct texi *p, enum texicmd cmd, case ('u'): case ('U'): case ('y'): texiputchars(p, "\\(:"); - texiputchar(p, buf[*pos]); break; default: - texiputchar(p, buf[*pos]); + texiwarn(p, "ignoring accent"); break; } + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); break; default: - abort(); + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); + break; } - advance(p, buf, pos); + + if (brace) { + while (*pos < BUFSZ(p) && '}' != BUF(p)[*pos]) { + texiputchar(p, BUF(p)[*pos]); + advance(p, pos); + } + if (*pos < BUFSZ(p)) + advance(p, pos); + } + + switch (cmd) { + case (TEXICMD_TIEACCENT): + texiputchar(p, ']'); + break; + case (TEXICMD_DOTACCENT): + texiputchar(p, '*'); + break; + default: + break; + } } static void -dosymbol(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosymbol(struct texi *p, enum texicmd cmd, size_t *pos) { - if (p->seenws && p->outcol && 0 == p->literal) { + /* Remember to pad us. */ + if (p->seenws && p->outcol && 0 == p->literal) texiputchar(p, ' '); - p->seenws = 0; - } + p->seenws = 0; + switch (cmd) { + case (TEXICMD_AA): + texiputchars(p, "\\(oA"); + break; + case (TEXICMD_AASMALL): + texiputchars(p, "\\(oa"); + break; + case (TEXICMD_AE): + texiputchars(p, "\\(AE"); + break; + case (TEXICMD_AESMALL): + texiputchars(p, "\\(ae"); + break; case (TEXICMD_ASTERISK): case (TEXICMD_NEWLINE): case (TEXICMD_SPACE): @@ -1012,6 +1164,12 @@ dosymbol(struct texi *p, enum texicmd cmd, case (TEXICMD_COPYRIGHT): texiputchars(p, "\\(co"); break; + case (TEXICMD_DH): + texiputchars(p, "\\(-D"); + break; + case (TEXICMD_DHSMALL): + texiputchars(p, "\\(Sd"); + break; case (TEXICMD_DOTS): case (TEXICMD_ENDDOTS): texiputchars(p, "..."); @@ -1022,21 +1180,98 @@ dosymbol(struct texi *p, enum texicmd cmd, case (TEXICMD_ERROR): texiputchars(p, "error\\(->"); break; + case (TEXICMD_EURO): + texiputchars(p, "\\(Eu"); + break; + case (TEXICMD_EXCLAMDOWN): + texiputchars(p, "\\(r!"); + break; case (TEXICMD_EXPANSION): texiputchars(p, "\\(->"); break; + case (TEXICMD_GEQ): + texiputchars(p, "\\(>="); + break; + case (TEXICMD_GUILLEMETLEFT): + case (TEXICMD_GUILLEMOTLEFT): + texiputchars(p, "\\(Fo"); + break; + case (TEXICMD_GUILLEMETRIGHT): + case (TEXICMD_GUILLEMOTRIGHT): + texiputchars(p, "\\(Fc"); + break; + case (TEXICMD_GUILSINGLLEFT): + texiputchars(p, "\\(fo"); + break; + case (TEXICMD_GUILSINGLRIGHT): + texiputchars(p, "\\(fc"); + break; + case (TEXICMD_L): + texiputchars(p, "\\(/L"); + break; case (TEXICMD_LATEX): texiputchars(p, "LaTeX"); break; + case (TEXICMD_LEQ): + texiputchars(p, "\\(<="); + break; + case (TEXICMD_LSMALL): + texiputchars(p, "\\(/l"); + break; case (TEXICMD_MINUS): texiputchars(p, "\\-"); break; + case (TEXICMD_O): + texiputchars(p, "\\(/O"); + break; + case (TEXICMD_OE): + texiputchars(p, "\\(OE"); + break; + case (TEXICMD_OESMALL): + texiputchars(p, "\\(oe"); + break; + case (TEXICMD_ORDF): + texiputchars(p, "a"); + break; + case (TEXICMD_ORDM): + texiputchars(p, "o"); + break; + case (TEXICMD_OSMALL): + texiputchars(p, "\\(/o"); + break; case (TEXICMD_PERIOD): texiputchar(p, '.'); break; + case (TEXICMD_POUNDS): + texiputchars(p, "\\(Po"); + break; + case (TEXICMD_QUESTIONDOWN): + texiputchars(p, "\\(r?"); + break; case (TEXICMD_QUESTIONMARK): texiputchar(p, '?'); break; + case (TEXICMD_QUOTEDBLBASE): + texiputchars(p, "\\(Bq"); + break; + case (TEXICMD_QUOTEDBLLEFT): + texiputchars(p, "\\(lq"); + break; + case (TEXICMD_QUOTEDBLRIGHT): + texiputchars(p, "\\(rq"); + break; + case (TEXICMD_QUOTESINGLBASE): + texiputchars(p, "\\(bq"); + break; + case (TEXICMD_QUOTELEFT): + texiputchars(p, "\\(oq"); + break; + case (TEXICMD_QUOTERIGHT): + texiputchars(p, "\\(cq"); + break; + case (TEXICMD_REGISTEREDSYMBOL): + texiputchars(p, "\\(rg"); + break; case (TEXICMD_RESULT): texiputchars(p, "\\(rA"); break; @@ -1055,6 +1290,15 @@ dosymbol(struct texi *p, enum texicmd cmd, case (TEXICMD_TEXSYM): texiputchars(p, "TeX"); break; + case (TEXICMD_TEXTDEGREE): + texiputchars(p, "\\(de"); + break; + case (TEXICMD_TH): + texiputchars(p, "\\(TP"); + break; + case (TEXICMD_THSMALL): + texiputchars(p, "\\(Tp"); + break; case (TEXICMD_TIE): texiputchars(p, "\\ "); break; @@ -1065,23 +1309,22 @@ dosymbol(struct texi *p, enum texicmd cmd, abort(); } - if (texitoks[cmd].len > 1) - doignbracket(p, cmd, buf, sz, pos); + /* Alphabetic commands have braces we ignore. */ + if (isalpha(texitoks[cmd].tok[0])) + doignbracket(p, cmd, pos); } static void -doquotation(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doquotation(struct texi *p, enum texicmd cmd, size_t *pos) { teximacro(p, "Qo"); - parseto(p, buf, sz, pos, "quotation"); + parseto(p, pos, "quotation"); teximacro(p, "Qc"); } static void -domath(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +domath(struct texi *p, enum texicmd cmd, size_t *pos) { size_t nest, start; @@ -1091,102 +1334,91 @@ domath(struct texi *p, enum texicmd cmd, * terms of @-commands. * This departs from GNU's rules, but whatever. */ - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - if (*pos == sz || '{' != buf[*pos]) + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + if (*pos == BUFSZ(p) || '{' != BUF(p)[*pos]) return; - advance(p, buf, pos); + advance(p, pos); if (p->seenws && p->outcol && 0 == p->literal) texiputchar(p, ' '); p->seenws = 0; - for (nest = 1, start = *pos; *pos < sz && nest > 0; ) { - if ('{' == buf[*pos]) + for (nest = 1, start = *pos; *pos < BUFSZ(p) && nest > 0; ) { + if ('{' == BUF(p)[*pos]) nest++; - else if ('}' == buf[*pos]) + else if ('}' == BUF(p)[*pos]) if (0 == --nest) continue; - advance(p, buf, pos); + advance(p, pos); } - if (*pos == sz) + if (*pos == BUFSZ(p)) return; - assert('}' == buf[*pos]); - texiputbuf(p, buf, start, *pos); - advance(p, buf, pos); + assert('}' == BUF(p)[*pos]); + texiputbuf(p, start, *pos); + advance(p, pos); } static void -dovalue(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dovalue(struct texi *p, enum texicmd cmd, size_t *pos) { - size_t start, end, i; + size_t start, end; char *key, *val; const char *cp; if (TEXICMD_SET == cmd) { - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - for (start = end = *pos; end < sz; end++) - if (ismspace(buf[end])) + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + for (start = end = *pos; end < BUFSZ(p); end++) + if (ismspace(BUF(p)[end])) break; /* We don't allow empty keys. */ if (start == end) return; - advanceto(p, buf, pos, end); + advanceto(p, pos, end); key = malloc(end - start + 1); if (NULL == key) texiabort(p, NULL); - memcpy(key, &buf[start], end - start); + memcpy(key, &BUF(p)[start], end - start); key[end - start] = '\0'; - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - for (start = end = *pos; end < sz; end++) - if ('\n' == buf[end]) + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + for (start = end = *pos; end < BUFSZ(p); end++) + if ('\n' == BUF(p)[end]) break; /* We do allow empty values. */ - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); val = malloc(end - start + 1); if (NULL == val) texiabort(p, NULL); - memcpy(val, &buf[start], end - start); + memcpy(val, &BUF(p)[start], end - start); val[end - start] = '\0'; valueadd(p, key, val); } else if (TEXICMD_VALUE == cmd) { if (p->seenws) texiputchar(p, ' '); p->seenws = 0; - if (NULL != (cp = valueblookup(p, buf, sz, pos))) { - for (i = 0; i < p->valstackpos; i++) - if (cp == p->valstack[i]) - break; - if (i < p->valstackpos) - texierr(p, "recursive value"); - if (64 == p->valstackpos) - texierr(p, "too many nested values"); - p->valstack[p->valstackpos++] = cp; - parsemembuf(p, cp, strlen(cp)); - p->valstackpos--; - } else + if (NULL != (cp = valueblookup(p, pos))) + texisplice(p, cp, strlen(cp), pos); + else texiputchars(p, "{No value}"); } else if (TEXICMD_IFCLEAR == cmd) { - if (NULL != valuellookup(p, buf, sz, pos)) - doignblock(p, cmd, buf, sz, pos); + if (NULL != valuellookup(p, pos)) + doignblock(p, cmd, pos); else - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); } else if (TEXICMD_IFSET == cmd) { - if (NULL == valuellookup(p, buf, sz, pos)) - doignblock(p, cmd, buf, sz, pos); + if (NULL == valuellookup(p, pos)) + doignblock(p, cmd, pos); else - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); } else if (TEXICMD_CLEAR == cmd) - valuelclear(p, buf, sz, pos); + valuelclear(p, pos); } static void -dolink(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dolink(struct texi *p, enum texicmd cmd, size_t *pos) { int c; @@ -1196,6 +1428,7 @@ dolink(struct texi *p, enum texicmd cmd, break; case (TEXICMD_UREF): case (TEXICMD_URL): + case (TEXICMD_INDICATEURL): teximacroopen(p, "Lk"); break; case (TEXICMD_XREF): @@ -1214,26 +1447,25 @@ dolink(struct texi *p, enum texicmd cmd, abort(); } - c = parsearg(p, buf, sz, pos, 0); + c = parsearg(p, pos, 0); p->ign++; while (c > 0) - c = parsearg(p, buf, sz, pos, 1); + c = parsearg(p, pos, 1); p->ign--; - texipunctuate(p, buf, sz, pos); + texipunctuate(p, pos); teximacroclose(p); } static void -doignargn(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doignargn(struct texi *p, enum texicmd cmd, size_t *pos) { int c; - c = parsearg(p, buf, sz, pos, 0); + c = parsearg(p, pos, 0); p->ign++; while (c > 0) - c = parsearg(p, buf, sz, pos, 1); + c = parsearg(p, pos, 1); p->ign--; } @@ -1258,8 +1490,7 @@ sectioner(struct texi *p, int sec) } static void -dosubsubsection(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosubsubsection(struct texi *p, enum texicmd cmd, size_t *pos) { int sec; @@ -1268,14 +1499,13 @@ dosubsubsection(struct texi *p, enum texicmd cmd, /* We don't have a subsubsubsection, so make one up. */ texivspace(p); teximacroopen(p, sects[sec]); - parseeoln(p, buf, sz, pos); + parseeoln(p, pos); teximacroclose(p); texivspace(p); } static void -dosubsection(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosubsection(struct texi *p, enum texicmd cmd, size_t *pos) { int sec; @@ -1290,15 +1520,14 @@ dosubsection(struct texi *p, enum texicmd cmd, if (sec > 1) texivspace(p); teximacroopen(p, sects[sec]); - parseeoln(p, buf, sz, pos); + parseeoln(p, pos); teximacroclose(p); if (sec > 1) texivspace(p); } static void -dosecoffs(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosecoffs(struct texi *p, enum texicmd cmd, size_t *pos) { if (TEXICMD_RAISESECTIONS == cmd) @@ -1308,8 +1537,7 @@ dosecoffs(struct texi *p, enum texicmd cmd, } static void -dosection(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosection(struct texi *p, enum texicmd cmd, size_t *pos) { int sec; @@ -1336,14 +1564,13 @@ dosection(struct texi *p, enum texicmd cmd, texierr(p, "\"%s\" in a literal scope!?", sects[sec]); teximacroopen(p, sects[sec]); - parseeoln(p, buf, sz, pos); + parseeoln(p, pos); teximacroclose(p); p->seenvs = 1; } static void -dosp(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dosp(struct texi *p, enum texicmd cmd, size_t *pos) { if (p->literal) @@ -1351,19 +1578,18 @@ dosp(struct texi *p, enum texicmd cmd, else texivspace(p); /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); } static void -dotop(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dotop(struct texi *p, enum texicmd cmd, size_t *pos) { const char *cp; time_t t; char date[32]; if (--p->ign) - texierr(p, "@top command while ignoring (%d)", p->ign); + texierr(p, "@top command while ignoring"); /* * Here we print our standard mdoc(7) prologue. @@ -1397,12 +1623,11 @@ dotop(struct texi *p, enum texicmd cmd, texiputchars(p, "Unknown description"); teximacroclose(p); p->seenvs = 1; - dosection(p, cmd, buf, sz, pos); + dosection(p, cmd, pos); } static void -doitem(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doitem(struct texi *p, enum texicmd cmd, size_t *pos) { /* Multitable is using raw tbl(7). */ @@ -1430,7 +1655,7 @@ doitem(struct texi *p, enum texicmd cmd, /* Trick so we don't start with Pp. */ p->seenvs = 1; - parseeoln(p, buf, sz, pos); + parseeoln(p, pos); if (TEXILIST_ITEM == p->list) teximacroclose(p); @@ -1439,8 +1664,7 @@ doitem(struct texi *p, enum texicmd cmd, } static void -dotab(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dotab(struct texi *p, enum texicmd cmd, size_t *pos) { /* This command is only useful in @multitable. */ @@ -1449,8 +1673,7 @@ dotab(struct texi *p, enum texicmd cmd, } static void -domultitable(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +domultitable(struct texi *p, enum texicmd cmd, size_t *pos) { enum texilist sv = p->list; int svliteral = p->literal; @@ -1467,29 +1690,29 @@ domultitable(struct texi *p, enum texicmd cmd, columns = 0; /* Advance to the first argument... */ - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); /* Make sure we don't print anything when scanning. */ p->ign++; - if ('@' == buf[*pos]) { + if ('@' == BUF(p)[*pos]) { /* * Look for @columnfractions. * We ignore these, but we do use the number of * arguments to set the number of columns that we'll * have. */ - type = texicmd(p, buf, *pos, sz, &end, NULL); - advanceto(p, buf, pos, end); + type = texicmd(p, *pos, &end, NULL); + advanceto(p, pos, end); if (TEXICMD_COLUMNFRACTIONS != type) texierr(p, "unknown multitable command"); - while (*pos < sz && '\n' != buf[*pos]) { - while (*pos < sz && isws(buf[*pos])) - advance(p, buf, pos); - while (*pos < sz && ! isws(buf[*pos])) { - if ('\n' == buf[*pos]) + while (*pos < BUFSZ(p) && '\n' != BUF(p)[*pos]) { + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + while (*pos < BUFSZ(p) && ! isws(BUF(p)[*pos])) { + if ('\n' == BUF(p)[*pos]) break; - advance(p, buf, pos); + advance(p, pos); } columns++; } @@ -1500,7 +1723,7 @@ domultitable(struct texi *p, enum texicmd cmd, * tbl(7) figure it out. * So use this only to count arguments. */ - while (parselinearg(p, buf, sz, pos) > 0) + while (parselinearg(p, pos) > 0) columns++; p->ign--; @@ -1512,7 +1735,7 @@ domultitable(struct texi *p, enum texicmd cmd, } texiputchars(p, ".\n"); p->outmacro++; - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); p->outmacro--; teximacro(p, "TE"); p->literal = svliteral; @@ -1520,40 +1743,52 @@ domultitable(struct texi *p, enum texicmd cmd, } static void -dotable(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +dotable(struct texi *p, enum texicmd cmd, size_t *pos) { enum texilist sv = p->list; p->list = TEXILIST_ITEM; teximacro(p, "Bl -tag -width Ds"); /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); p->seenvs = 1; - parseto(p, buf, sz, pos, texitoks[cmd].tok); + parseto(p, pos, texitoks[cmd].tok); teximacro(p, "El"); p->list = sv; } static void -doenumerate(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doend(struct texi *p, enum texicmd cmd, size_t *pos) { + size_t start; + + while (*pos < BUFSZ(p) && isws(BUF(p)[*pos])) + advance(p, pos); + start = *pos; + while (*pos < BUFSZ(p) && '\n' != BUF(p)[*pos]) + advance(p, pos); + + texiwarn(p, "unexpected \"end\": %.*s", (int)(*pos - start), &BUF(p)[start]); + advanceeoln(p, pos, 1); +} + +static void +doenumerate(struct texi *p, enum texicmd cmd, size_t *pos) +{ enum texilist sv = p->list; p->list = TEXILIST_NOITEM; teximacro(p, "Bl -enum"); p->seenvs = 1; /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); - parseto(p, buf, sz, pos, "enumerate"); + advanceeoln(p, pos, 1); + parseto(p, pos, "enumerate"); teximacro(p, "El"); p->list = sv; } static void -doitemize(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doitemize(struct texi *p, enum texicmd cmd, size_t *pos) { enum texilist sv = p->list; @@ -1561,29 +1796,27 @@ doitemize(struct texi *p, enum texicmd cmd, teximacro(p, "Bl -bullet"); p->seenvs = 1; /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); - parseto(p, buf, sz, pos, "itemize"); + advanceeoln(p, pos, 1); + parseto(p, pos, "itemize"); teximacro(p, "El"); p->list = sv; } static void -doignbracket(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doignbracket(struct texi *p, enum texicmd cmd, size_t *pos) { p->ign++; - parsebracket(p, buf, sz, pos); + parsebracket(p, pos); p->ign--; } static void -doignline(struct texi *p, enum texicmd cmd, - const char *buf, size_t sz, size_t *pos) +doignline(struct texi *p, enum texicmd cmd, size_t *pos) { /* FIXME: ignore and parseeoln. */ - advanceeoln(p, buf, sz, pos, 1); + advanceeoln(p, pos, 1); } /*