=================================================================== RCS file: /cvs/texi2mdoc/main.c,v retrieving revision 1.32 retrieving revision 1.45 diff -u -p -r1.32 -r1.45 --- texi2mdoc/main.c 2015/02/23 12:28:20 1.32 +++ texi2mdoc/main.c 2015/02/25 10:01:54 1.45 @@ -1,4 +1,4 @@ -/* $Id: main.c,v 1.32 2015/02/23 12:28:20 kristaps Exp $ */ +/* $Id: main.c,v 1.45 2015/02/25 10:01:54 kristaps Exp $ */ /* * Copyright (c) 2015 Kristaps Dzonsons * @@ -14,12 +14,11 @@ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. */ -#include -#include - +#if defined(__linux__) || defined(__MINT__) +# define _GNU_SOURCE /* memmem */ +#endif #include #include -#include #include #include #include @@ -28,7 +27,6 @@ #include #include #include -#include #include "extern.h" @@ -47,6 +45,7 @@ static void dobye(struct texi *, enum texicmd, const c 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 doend(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 *); @@ -74,13 +73,18 @@ static void dotab(struct texi *, enum texicmd, const c 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 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 */ @@ -94,6 +98,8 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dosymbol, "!", 1 }, /* TEXICMD_BANG */ { dosymbol, "bullet", 6 }, /* TEXICMD_BULLET */ { dobye, "bye", 3 }, /* TEXICMD_BYE */ + { doblock, "cartouche", 9 }, /* TEXICMD_CARTOUCHE */ + { doaccent, ",", 1 }, /* TEXICMD_CEDILLA */ { doignline, "center", 6 }, /* TEXICMD_CENTER */ { dosection, "chapter", 7 }, /* TEXICMD_CHAPTER */ { doignline, "cindex", 6 }, /* TEXICMD_CINDEX */ @@ -103,6 +109,7 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { doinline, "code", 4 }, /* TEXICMD_CODE */ { dosymbol, ":", 1 }, /* TEXICMD_COLON */ { NULL, "columnfractions", 15 }, /* TEXICMD_COLUMNFRACTIONS */ + { dosymbol, "comma", 5 }, /* TEXICMD_COMMA */ { doinline, "command", 7 }, /* TEXICMD_COMMAND */ { doignline, "c", 1 }, /* TEXICMD_COMMENT */ { doignline, "comment", 7 }, /* TEXICMD_COMMENT_LONG */ @@ -121,6 +128,8 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dodefn, "deftypefnx", 10 }, /* TEXICMD_DEFTYPEFNX */ { dodefn, "deftypefun", 10 }, /* TEXICMD_DEFTYPEFUN */ { dodefn, "deftypefunx", 11 }, /* TEXICMD_DEFTYPEFUNX */ + { dodefn, "deftypemethod", 13 }, /* TEXICMD_DEFTYPEMETHOD */ + { dodefn, "deftypemethodx", 14 }, /* TEXICMD_DEFTYPEMETHODX */ { dodefn, "deftypevar", 10 }, /* TEXICMD_DEFTYPEVAR */ { dodefn, "deftypevarx", 11 }, /* TEXICMD_DEFTYPEVARX */ { dodefn, "deftypevr", 9 }, /* TEXICMD_DEFTYPEVR */ @@ -133,29 +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 */ @@ -180,7 +211,8 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { doignbracket, "image", 5 }, /* TEXICMD_IMAGE */ { doinclude, "include", 7 }, /* TEXICMD_INCLUDE */ { dodisplay, "indentblock", 11 }, /* TEXICMD_INDENTBLOCK */ - { doignline, "", 0 }, /* TEXICMD_USER_INDEX */ + { 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 */ @@ -188,9 +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 */ @@ -199,30 +235,49 @@ 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 */ { dosection, "section", 7 }, /* TEXICMD_SECTION */ { dovalue, "set", 3 }, /* TEXICMD_SET */ { doignline, "setchapternewpage", 17 }, /* TEXICMD_SETCHAPNEWPAGE */ + { doignline, "setcontentsaftertitlepage", 25 }, /* TEXICMD_SETCONTENTSAFTER */ { doignline, "setfilename", 11 }, /* TEXICMD_SETFILENAME */ { dotitle, "settitle", 8 }, /* TEXICMD_SETTITLE */ { doignline, "shortcontents", 13 }, /* TEXICMD_SHORTCONTENTS */ { doinline, "slanted", 7 }, /* TEXICMD_SLANTED */ + { dosymbol, "/", 1 }, /* TEXICMD_SLASH */ { dosp, "sp", 2 }, /* TEXICMD_SP */ { dosymbol, " ", 1 }, /* TEXICMD_SPACE */ { doignline, "smallbook", 9 }, /* TEXICMD_SMALLBOOK */ @@ -232,6 +287,7 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dodisplay, "smallindentblock", 16 }, /* TEXICMD_SMALLINDENTBLOCK */ { dosymbol, "{", 1 }, /* TEXICMD_SQUIGGLE_LEFT */ { dosymbol, "}", 1 }, /* TEXICMD_SQUIGGLE_RIGHT */ + { dosymbol, "ss", 2 }, /* TEXICMD_SS */ { doinline, "strong", 6 }, /* TEXICMD_STRONG */ { dosubsection, "subheading", 10 }, /* TEXICMD_SUBHEADING */ { dosubsection, "subsection", 10 }, /* TEXICMD_SUBSECTION */ @@ -247,12 +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 */ @@ -260,9 +324,12 @@ static const struct texitok __texitoks[TEXICMD__MAX] = { dosubsubsection, "unnumberedsubsubsec", 19 }, /* TEXICMD_UNNUMBEREDSUBSUBSEC */ { 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, "verbatim", 8 }, /* TEXICMD_VERBATIM */ + { doverb, "verb", 4 }, /* TEXICMD_VERB */ + { doverbatim, "verbatim", 8 }, /* TEXICMD_VERBATIM */ { doverbinclude, "verbatiminclude", 15 }, /* TEXICMD_VERBATIMINCLUDE */ { doignline, "vindex", 6 }, /* TEXICMD_VINDEX */ { dosp, "vskip", 5 }, /* TEXICMD_VSKIP */ @@ -314,9 +381,11 @@ dodefn(struct texi *p, enum texicmd cmd, blk = NULL; switch (cmd) { case (TEXICMD_DEFFN): + case (TEXICMD_DEFMAC): case (TEXICMD_DEFTP): case (TEXICMD_DEFTYPEFN): case (TEXICMD_DEFTYPEFUN): + case (TEXICMD_DEFTYPEMETHOD): case (TEXICMD_DEFTYPEVAR): case (TEXICMD_DEFTYPEVR): case (TEXICMD_DEFUN): @@ -339,6 +408,10 @@ dodefn(struct texi *p, enum texicmd cmd, texivspace(p); switch (cmd) { + case (TEXICMD_DEFTYPEMETHOD): + case (TEXICMD_DEFTYPEMETHODX): + texiputchars(p, "Method"); + break; case (TEXICMD_DEFMAC): case (TEXICMD_DEFMACX): texiputchars(p, "Macro"); @@ -387,6 +460,8 @@ dodefn(struct texi *p, enum texicmd cmd, case (TEXICMD_DEFTYPEFUNX): case (TEXICMD_DEFTYPEFN): case (TEXICMD_DEFTYPEFNX): + case (TEXICMD_DEFTYPEMETHOD): + case (TEXICMD_DEFTYPEMETHODX): teximacroopen(p, "Ft"); parselinearg(p, buf, sz, pos); teximacroclose(p); @@ -462,7 +537,7 @@ domacro(struct texi *p, enum texicmd cmd, start = *pos; endtok = "\n@end macro\n"; endtoksz = strlen(endtok); - blk = memmem(&buf[start], sz, endtok, endtoksz); + blk = memmem(&buf[start], sz - start, endtok, endtoksz); if (NULL == blk) texierr(p, "unterminated macro body"); while (&buf[*pos] != blk) @@ -491,34 +566,64 @@ static void doignblock(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, 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 + term - &buf[*pos]; - assert(endpos <= sz); - while (*pos < endpos) - advance(p, buf, pos); + while (stack > 0 && *pos < sz) { + if (stack > 10) + abort(); + endt = memmem(&buf[*pos], sz - *pos, end, esz); + startt = memmem(&buf[*pos], sz - *pos, start, ssz); + if (NULL == endt) { + texiwarn(p, "unterminated \"%s\" " + "block", texitoks[cmd].tok); + *pos = sz; + 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[*pos]); + stack--; + } else { + newpos += ssz + (size_t)(startt - &buf[*pos]); + stack++; + } + + assert(newpos <= sz); + while (*pos < newpos) + advance(p, buf, pos); + } } static void @@ -588,41 +693,75 @@ static void doverb(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, size_t *pos) { + char delim; + size_t start; + + while (*pos < sz && isws(buf[*pos])) + advance(p, buf, pos); + if (*pos == sz || '{' != buf[*pos]) + return; + advance(p, buf, pos); + if (*pos == sz) + return; + + delim = buf[*pos]; + advance(p, buf, 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] == '}') + break; + advance(p, buf, pos); + } + if (*pos == sz - 1) + return; + texiputbuf(p, buf, 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); +} + +static void +doverbatim(struct texi *p, enum texicmd cmd, + const char *buf, size_t sz, 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, buf, sz, pos, 0); + if (*pos == sz) { + 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 + term - &buf[*pos]; + term = memmem(&buf[*pos], sz - *pos, end, endsz); + if (NULL == term) { + texiwarn(p, "unterminated verbatim block"); + endpos = sz; + } else + endpos = *pos + (size_t)(term - &buf[*pos]); - teximacro(p, "Bd -literal -offset indent"); assert(endpos <= sz); + assert('\n' == buf[*pos]); + advance(p, buf, pos); + teximacro(p, "Bd -literal -offset indent"); while (*pos < endpos) { - if (buf[*pos] == '\n') - p->outcol = 0; - else - p->outcol++; - if (*pos > 0 && '.' == buf[*pos]) - if ('\n' == buf[*pos - 1]) - fputs("\\&", stdout); - putchar(buf[*pos]); - if ('\\' == buf[*pos]) - putchar('e'); + texiputchar(p, buf[*pos]); advance(p, buf, pos); } teximacro(p, "Ed"); - advanceto(p, buf, pos, endpos + endsz); + if (*pos < sz) + advanceto(p, buf, pos, endpos + endsz); } static void @@ -826,10 +965,38 @@ static void doaccent(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, size_t *pos) { + int brace = 0; - if (*pos == sz) + if (*pos == sz) { + texiwarn(p, "truncated: @%s", texitoks[cmd].tok); return; - advance(p, buf, pos); + } + + /* Pad us with space, if necessary. */ + if (p->seenws && p->outcol && 0 == p->literal) { + texiputchar(p, ' '); + 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[*pos]) { + brace = 1; + advance(p, buf, pos); + } else if (isalpha(texitoks[cmd].tok[0])) + while (*pos < sz && isws(buf[*pos])) + advance(p, buf, pos); + + if (*pos == sz) { + texiwarn(p, "truncated: @%s", texitoks[cmd].tok); + return; + } + switch (cmd) { case (TEXICMD_ACUTE): switch (buf[*pos]) { @@ -839,12 +1006,22 @@ doaccent(struct texi *p, enum texicmd cmd, 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[*pos]); + advance(p, buf, pos); break; + case (TEXICMD_CEDILLA): + if ('c' == buf[*pos] || 'C' == buf[*pos]) + texiputchars(p, "\\(,"); + else + texiwarn(p, "ignoring accent"); + texiputchar(p, buf[*pos]); + advance(p, buf, pos); + break; case (TEXICMD_CIRCUMFLEX): switch (buf[*pos]) { case ('a'): case ('A'): @@ -853,12 +1030,22 @@ doaccent(struct texi *p, enum texicmd cmd, 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[*pos]); + advance(p, buf, pos); break; + case (TEXICMD_DOTLESS): + if ('i' == buf[*pos] || 'j' == buf[*pos]) + texiputchars(p, "\\(."); + else + texiwarn(p, "ignoring accent"); + texiputchar(p, buf[*pos]); + advance(p, buf, pos); + break; case (TEXICMD_GRAVE): switch (buf[*pos]) { case ('a'): case ('A'): @@ -867,11 +1054,13 @@ doaccent(struct texi *p, enum texicmd cmd, 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[*pos]); + advance(p, buf, pos); break; case (TEXICMD_TILDE): switch (buf[*pos]) { @@ -879,11 +1068,13 @@ doaccent(struct texi *p, enum texicmd cmd, 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[*pos]); + advance(p, buf, pos); break; case (TEXICMD_UMLAUT): switch (buf[*pos]) { @@ -894,14 +1085,38 @@ doaccent(struct texi *p, enum texicmd cmd, case ('u'): case ('U'): case ('y'): texiputchars(p, "\\(:"); - texiputchar(p, buf[*pos]); break; default: + texiwarn(p, "ignoring accent"); + break; + } + texiputchar(p, buf[*pos]); + advance(p, buf, pos); + break; + default: + texiputchar(p, buf[*pos]); + advance(p, buf, pos); + break; + } + + if (brace) { + while (*pos < sz && '}' != buf[*pos]) { texiputchar(p, buf[*pos]); + advance(p, buf, pos); } + if (*pos < sz) + advance(p, buf, pos); + } + + switch (cmd) { + case (TEXICMD_TIEACCENT): + texiputchar(p, ']'); break; + case (TEXICMD_DOTACCENT): + texiputchar(p, '*'); + break; default: - abort(); + break; } } @@ -910,12 +1125,25 @@ dosymbol(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, 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): @@ -931,10 +1159,20 @@ dosymbol(struct texi *p, enum texicmd cmd, case (TEXICMD_BULLET): texiputchars(p, "\\(bu"); break; + case (TEXICMD_COMMA): + texiputchar(p, ','); + break; 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, "..."); break; case (TEXICMD_EQUIV): @@ -943,24 +1181,107 @@ 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; + case (TEXICMD_SLASH): + texiputchar(p, '/'); + break; + case (TEXICMD_SS): + texiputchars(p, "\\(ss"); + break; case (TEXICMD_SQUIGGLE_LEFT): texiputchars(p, "{"); break; @@ -970,6 +1291,18 @@ 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; case (TEXICMD_COLON): case (TEXICMD_HYPHEN): break; @@ -977,7 +1310,8 @@ dosymbol(struct texi *p, enum texicmd cmd, abort(); } - if (texitoks[cmd].len > 1) + /* Alphabetic commands have braces we ignore. */ + if (isalpha(texitoks[cmd].tok[0])) doignbracket(p, cmd, buf, sz, pos); } @@ -995,7 +1329,7 @@ static void domath(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, size_t *pos) { - size_t nest; + size_t nest, start; /* * Math handling is different from everything else. @@ -1011,18 +1345,18 @@ domath(struct texi *p, enum texicmd cmd, if (p->seenws && p->outcol && 0 == p->literal) texiputchar(p, ' '); p->seenws = 0; - for (nest = 1; *pos < sz && nest > 0; ) { + for (nest = 1, start = *pos; *pos < sz && nest > 0; ) { if ('{' == buf[*pos]) nest++; else if ('}' == buf[*pos]) if (0 == --nest) continue; - texiputchar(p, buf[*pos]); advance(p, buf, pos); } if (*pos == sz) return; assert('}' == buf[*pos]); + texiputbuf(p, buf, start, *pos); advance(p, buf, pos); } @@ -1108,16 +1442,21 @@ dolink(struct texi *p, enum texicmd cmd, break; case (TEXICMD_UREF): case (TEXICMD_URL): + case (TEXICMD_INDICATEURL): teximacroopen(p, "Lk"); break; case (TEXICMD_XREF): texiputchars(p, "See Section"); - teximacroopen(p, "Qq"); + teximacroopen(p, "Dq"); break; case (TEXICMD_PXREF): texiputchars(p, "see Section"); - teximacroopen(p, "Qq"); + teximacroopen(p, "Dq"); break; + case (TEXICMD_INFOREF): + texiputchars(p, "See Info file node"); + teximacroopen(p, "Dq"); + break; default: abort(); } @@ -1195,11 +1534,13 @@ dosubsection(struct texi *p, enum texicmd cmd, texierr(p, "\"%s\" in a literal scope!?", sects[sec]); /* We don't have a subsubsection, so make one up. */ - texivspace(p); + if (sec > 1) + texivspace(p); teximacroopen(p, sects[sec]); parseeoln(p, buf, sz, pos); teximacroclose(p); - texivspace(p); + if (sec > 1) + texivspace(p); } static void @@ -1252,7 +1593,10 @@ dosp(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, size_t *pos) { - texivspace(p); + if (p->literal) + texiputchar(p, '\n'); + else + texivspace(p); /* FIXME: ignore and parseeoln. */ advanceeoln(p, buf, sz, pos, 1); } @@ -1289,11 +1633,15 @@ dotop(struct texi *p, enum texicmd cmd, teximacro(p, "Os"); teximacro(p, "Sh NAME"); teximacroopen(p, "Nm"); - texiputchars(p, p->title); + for (cp = p->title; '\0' != *cp; cp++) + texiputchar(p, *cp); teximacroclose(p); teximacroopen(p, "Nd"); - texiputchars(p, NULL != p->subtitle ? - p->subtitle : "Unknown description"); + if (NULL != p->subtitle) + for (cp = p->subtitle; '\0' != *cp; cp++) + texiputchar(p, *cp); + else + texiputchars(p, "Unknown description"); teximacroclose(p); p->seenvs = 1; dosection(p, cmd, buf, sz, pos); @@ -1435,6 +1783,22 @@ dotable(struct texi *p, enum texicmd cmd, } static void +doend(struct texi *p, enum texicmd cmd, + const char *buf, size_t sz, size_t *pos) +{ + size_t start; + + while (*pos < sz && isws(buf[*pos])) + advance(p, buf, pos); + start = *pos; + while (*pos < sz && '\n' != buf[*pos]) + advance(p, buf, pos); + + texiwarn(p, "unexpected \"end\": %.*s", (int)(*pos - start), &buf[start]); + advanceeoln(p, buf, sz, pos, 1); +} + +static void doenumerate(struct texi *p, enum texicmd cmd, const char *buf, size_t sz, size_t *pos) { @@ -1488,7 +1852,7 @@ doignline(struct texi *p, enum texicmd cmd, /* * Parse colon-separated directories from "cp" (if not NULL) and returns * the array of pointers. - * Prepends "base" to the array. + * Prepends "base" to the array, if found. * This does NOT sanitise the directories! */ static char ** @@ -1496,26 +1860,27 @@ parsedirs(struct texi *p, const char *base, const char { char *tok, *str, *tofree; const char *cpp; - size_t i; + size_t i = 0; char **dirs; - *sz = NULL != (cpp = cp) ? 2 : 1; - if (*sz > 1) - for ( ; NULL != (cpp = strchr(cpp, ':')); (*sz)++) + /* Count up our expected arguments. */ + *sz = NULL != base; + if (NULL != (cpp = cp)) + for ((*sz)++; NULL != (cpp = strchr(cpp, ':')); (*sz)++) cpp++; + if (0 == *sz) + return(NULL); if (NULL == (dirs = calloc(*sz, sizeof(char *)))) texiabort(p, NULL); - else if (NULL == (dirs[0] = strdup(base))) + if (NULL != base && NULL == (dirs[i++] = strdup(base))) texiabort(p, NULL); - if (NULL == cp) return(dirs); - if (NULL == (tofree = tok = str = strdup(cp))) texiabort(p, NULL); - for (i = 1; NULL != (tok = strsep(&str, ":")); i++) + for ( ; NULL != (tok = strsep(&str, ":")); i++) if (NULL == (dirs[i] = strdup(tok))) texiabort(p, NULL); @@ -1528,7 +1893,7 @@ main(int argc, char *argv[]) { struct texi texi; int c; - char *path, *dir; + char *dirpath, *dir, *ccp; const char *progname, *Idir, *cp; progname = strrchr(argv[0], '/'); @@ -1538,6 +1903,7 @@ main(int argc, char *argv[]) ++progname; memset(&texi, 0, sizeof(struct texi)); + texi.ign = 1; Idir = NULL; while (-1 != (c = getopt(argc, argv, "I:"))) @@ -1550,33 +1916,33 @@ main(int argc, char *argv[]) } argv += optind; - if (0 == (argc -= optind)) - goto usage; + argc -= optind; - if (NULL == (path = strdup(argv[0]))) - texiabort(&texi, NULL); - else if (NULL == (dir = dirname(path))) - texiabort(&texi, NULL); + if (argc > 0) { + if (NULL == (dirpath = strdup(argv[0]))) + texiabort(&texi, NULL); + if (NULL == (dir = dirname(dirpath))) + texiabort(&texi, NULL); + if (NULL != (cp = strrchr(argv[0], '/'))) + texi.title = strdup(cp + 1); + else + texi.title = strdup(argv[0]); + if (NULL == texi.title) + texiabort(&texi, NULL); + else if (NULL != (ccp = strchr(texi.title, '.'))) + *ccp = '\0'; + texi.dirs = parsedirs(&texi, dir, Idir, &texi.dirsz); + free(dirpath); + parsefile(&texi, argv[0], 1); + } else { + texi.title = strdup("Unknown Manual"); + texi.dirs = parsedirs(&texi, NULL, Idir, &texi.dirsz); + parsestdin(&texi); + } - free(path); - - if (NULL != (cp = strrchr(argv[0], '/'))) - texi.title = strdup(cp + 1); - else - texi.title = strdup(argv[0]); - - if (NULL == texi.title) - texiabort(&texi, NULL); - else if (NULL != (path = strchr(texi.title, '.'))) - *path = '\0'; - - texi.ign = 1; - texi.dirs = parsedirs(&texi, dir, Idir, &texi.dirsz); - parsefile(&texi, argv[0], 1); - /* We shouldn't get here. */ texiexit(&texi); return(EXIT_FAILURE); usage: - fprintf(stderr, "usage: %s [-Idirs] file\n", progname); + fprintf(stderr, "usage: %s [-Idirs] [file]\n", progname); return(EXIT_FAILURE); }