Grrr, patch(1) merged the new function to the wrong place in this file.

[mandoc.git] / mandoc.c

diff --git a/mandoc.c b/mandoc.c
index 0462134a08c8b359bb922db2b12ef67dff7f42d4..e82093b935dbbe55c524f31ff1b29d61183a2671 100644 (file)
--- a/mandoc.c
+++ b/mandoc.c
@@ -1,4 +1,4 @@
-/*     $Id: mandoc.c,v 1.78 2014/04/08 01:37:27 schwarze Exp $ */
+/*     $Id: mandoc.c,v 1.87 2014/10/13 17:17:45 schwarze Exp $ */

 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2011, 2012, 2013, 2014 Ingo Schwarze <schwarze@openbsd.org>
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2011, 2012, 2013, 2014 Ingo Schwarze <schwarze@openbsd.org>


@@ -15,9 +15,7 @@
  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  */
  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
  */

-#ifdef HAVE_CONFIG_H

 #include "config.h"
 #include "config.h"

-#endif

 
 #include <sys/types.h>
 
 
 #include <sys/types.h>
 


@@ -46,7 +44,7 @@ mandoc_escape(const char **end, const char **start, int *sz)
        const char      *local_start;
        int              local_sz;
        char             term;
        const char      *local_start;
        int              local_sz;
        char             term;

-       enum mandoc_esc  gly; 
+       enum mandoc_esc  gly;

 
        /*
         * When the caller doesn't provide return storage,
 
        /*
         * When the caller doesn't provide return storage,


@@ -75,80 +73,69 @@ mandoc_escape(const char **end, const char **start, int *sz)
         * these, but each eventually returns a substring of the glyph
         * name.
         */
         * these, but each eventually returns a substring of the glyph
         * name.
         */

-       case ('('):
+       case '(':

                gly = ESCAPE_SPECIAL;
                *sz = 2;
                break;
                gly = ESCAPE_SPECIAL;
                *sz = 2;
                break;

-       case ('['):
+       case '[':

                gly = ESCAPE_SPECIAL;
                gly = ESCAPE_SPECIAL;

-               /*
-                * Unicode escapes are defined in groff as \[uXXXX] to
-                * \[u10FFFF], where the contained value must be a valid
-                * Unicode codepoint.  Here, however, only check whether
-                * it's not a zero-width escape.
-                */
-               if ('u' == (*start)[0] && ']' != (*start)[1])
-                       gly = ESCAPE_UNICODE;

                term = ']';
                break;
                term = ']';
                break;

-       case ('C'):
+       case 'C':

                if ('\'' != **start)
                        return(ESCAPE_ERROR);
                *start = ++*end;
                if ('\'' != **start)
                        return(ESCAPE_ERROR);
                *start = ++*end;

-               if ('u' == (*start)[0] && '\'' != (*start)[1])
-                       gly = ESCAPE_UNICODE;
-               else
-                       gly = ESCAPE_SPECIAL;
+               gly = ESCAPE_SPECIAL;

                term = '\'';
                break;
 
        /*
         * Escapes taking no arguments at all.
         */
                term = '\'';
                break;
 
        /*
         * Escapes taking no arguments at all.
         */

-       case ('d'):
+       case 'd':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('u'):
+       case 'u':

                return(ESCAPE_IGNORE);
 
        /*
         * The \z escape is supposed to output the following
                return(ESCAPE_IGNORE);
 
        /*
         * The \z escape is supposed to output the following

-        * character without advancing the cursor position.  
+        * character without advancing the cursor position.

         * Since we are mostly dealing with terminal mode,
         * let us just skip the next character.
         */
         * Since we are mostly dealing with terminal mode,
         * let us just skip the next character.
         */

-       case ('z'):
+       case 'z':

                return(ESCAPE_SKIPCHAR);
 
        /*
         * Handle all triggers matching \X(xy, \Xx, and \X[xxxx], where
         * 'X' is the trigger.  These have opaque sub-strings.
         */
                return(ESCAPE_SKIPCHAR);
 
        /*
         * Handle all triggers matching \X(xy, \Xx, and \X[xxxx], where
         * 'X' is the trigger.  These have opaque sub-strings.
         */

-       case ('F'):
+       case 'F':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('g'):
+       case 'g':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('k'):
+       case 'k':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('M'):
+       case 'M':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('m'):
+       case 'm':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('n'):
+       case 'n':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('V'):
+       case 'V':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('Y'):
+       case 'Y':

                gly = ESCAPE_IGNORE;
                /* FALLTHROUGH */
                gly = ESCAPE_IGNORE;
                /* FALLTHROUGH */

-       case ('f'):
+       case 'f':

                if (ESCAPE_ERROR == gly)
                        gly = ESCAPE_FONT;
                switch (**start) {
                if (ESCAPE_ERROR == gly)
                        gly = ESCAPE_FONT;
                switch (**start) {

-               case ('('):
+               case '(':

                        *start = ++*end;
                        *sz = 2;
                        break;
                        *start = ++*end;
                        *sz = 2;
                        break;

-               case ('['):
+               case '[':

                        *start = ++*end;
                        term = ']';
                        break;
                        *start = ++*end;
                        term = ']';
                        break;


@@ -163,19 +150,19 @@ mandoc_escape(const char **end, const char **start, int *sz)
         * and 'Y' is any string.  These have opaque sub-strings.
         * The \B and \w escapes are handled in roff.c, roff_res().
         */
         * and 'Y' is any string.  These have opaque sub-strings.
         * The \B and \w escapes are handled in roff.c, roff_res().
         */

-       case ('A'):
+       case 'A':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('b'):
+       case 'b':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('D'):
+       case 'D':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('o'):
+       case 'o':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('R'):
+       case 'R':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('X'):
+       case 'X':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('Z'):
+       case 'Z':

                if ('\0' == **start)
                        return(ESCAPE_ERROR);
                gly = ESCAPE_IGNORE;
                if ('\0' == **start)
                        return(ESCAPE_ERROR);
                gly = ESCAPE_IGNORE;


@@ -187,21 +174,24 @@ mandoc_escape(const char **end, const char **start, int *sz)
         * These escapes are of the form \X'N', where 'X' is the trigger
         * and 'N' resolves to a numerical expression.
         */
         * These escapes are of the form \X'N', where 'X' is the trigger
         * and 'N' resolves to a numerical expression.
         */

-       case ('h'):
+       case 'h':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('H'):
+       case 'H':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('L'):
+       case 'L':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('l'):
+       case 'l':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('S'):
+       case 'S':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('v'):
+       case 'v':

                /* FALLTHROUGH */
                /* FALLTHROUGH */

-       case ('x'):
-               if (strchr("\0 %&()*+-./0123456789:<=>", **start))
+       case 'x':
+               if (strchr(" %&()*+-./0123456789:<=>", **start)) {
+                       if ('\0' != **start)
+                               ++*end;

                        return(ESCAPE_ERROR);
                        return(ESCAPE_ERROR);

+               }

                gly = ESCAPE_IGNORE;
                term = **start;
                *start = ++*end;
                gly = ESCAPE_IGNORE;
                term = **start;
                *start = ++*end;


@@ -211,7 +201,7 @@ mandoc_escape(const char **end, const char **start, int *sz)
         * Special handling for the numbered character escape.
         * XXX Do any other escapes need similar handling?
         */
         * Special handling for the numbered character escape.
         * XXX Do any other escapes need similar handling?
         */

-       case ('N'):
+       case 'N':

                if ('\0' == **start)
                        return(ESCAPE_ERROR);
                (*end)++;
                if ('\0' == **start)
                        return(ESCAPE_ERROR);
                (*end)++;


@@ -227,10 +217,10 @@ mandoc_escape(const char **end, const char **start, int *sz)
                        (*end)++;
                return(ESCAPE_NUMBERED);
 
                        (*end)++;
                return(ESCAPE_NUMBERED);
 

-       /* 
+       /*

         * Sizes get a special category of their own.
         */
         * Sizes get a special category of their own.
         */

-       case ('s'):
+       case 's':

                gly = ESCAPE_IGNORE;
 
                /* See +/- counts as a sign. */
                gly = ESCAPE_IGNORE;
 
                /* See +/- counts as a sign. */


@@ -238,15 +228,15 @@ mandoc_escape(const char **end, const char **start, int *sz)
                        (*end)++;
 
                switch (**end) {
                        (*end)++;
 
                switch (**end) {

-               case ('('):
+               case '(':

                        *start = ++*end;
                        *sz = 2;
                        break;
                        *start = ++*end;
                        *sz = 2;
                        break;

-               case ('['):
+               case '[':

                        *start = ++*end;
                        term = ']';
                        break;
                        *start = ++*end;
                        term = ']';
                        break;

-               case ('\''):
+               case '\'':

                        *start = ++*end;
                        term = '\'';
                        break;
                        *start = ++*end;
                        term = '\'';
                        break;


@@ -278,9 +268,9 @@ mandoc_escape(const char **end, const char **start, int *sz)
        if ('\0' != term) {
                while (**end != term) {
                        switch (**end) {
        if ('\0' != term) {
                while (**end != term) {
                        switch (**end) {

-                       case ('\0'):
+                       case '\0':

                                return(ESCAPE_ERROR);
                                return(ESCAPE_ERROR);

-                       case ('\\'):
+                       case '\\':

                                (*end)++;
                                if (ESCAPE_ERROR ==
                                    mandoc_escape(end, NULL, NULL))
                                (*end)++;
                                if (ESCAPE_ERROR ==
                                    mandoc_escape(end, NULL, NULL))


@@ -302,7 +292,7 @@ mandoc_escape(const char **end, const char **start, int *sz)
        /* Run post-processors. */
 
        switch (gly) {
        /* Run post-processors. */
 
        switch (gly) {

-       case (ESCAPE_FONT):
+       case ESCAPE_FONT:

                if (2 == *sz) {
                        if ('C' == **start) {
                                /*
                if (2 == *sz) {
                        if ('C' == **start) {
                                /*


@@ -320,29 +310,39 @@ mandoc_escape(const char **end, const char **start, int *sz)
                        break;
 
                switch (**start) {
                        break;
 
                switch (**start) {

-               case ('3'):
+               case '3':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('B'):
+               case 'B':

                        gly = ESCAPE_FONTBOLD;
                        break;
                        gly = ESCAPE_FONTBOLD;
                        break;

-               case ('2'):
+               case '2':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('I'):
+               case 'I':

                        gly = ESCAPE_FONTITALIC;
                        break;
                        gly = ESCAPE_FONTITALIC;
                        break;

-               case ('P'):
+               case 'P':

                        gly = ESCAPE_FONTPREV;
                        break;
                        gly = ESCAPE_FONTPREV;
                        break;

-               case ('1'):
+               case '1':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('R'):
+               case 'R':

                        gly = ESCAPE_FONTROMAN;
                        break;
                }
                break;
                        gly = ESCAPE_FONTROMAN;
                        break;
                }
                break;

-       case (ESCAPE_SPECIAL):
+       case ESCAPE_SPECIAL:

                if (1 == *sz && 'c' == **start)
                        gly = ESCAPE_NOSPACE;
                if (1 == *sz && 'c' == **start)
                        gly = ESCAPE_NOSPACE;

+               /*
+                * Unicode escapes are defined in groff as \[uXXXX]
+                * to \[u10FFFF], where the contained value must be
+                * a valid Unicode codepoint.  Here, however, only
+                * check the length and the validity of all digits.
+                */
+               else if (*sz > 4 && *sz < 8 && **start == 'u' &&
+                   (int)strspn(*start + 1, "0123456789ABCDEFabcdef")
+                   + 1 == *sz)
+                       gly = ESCAPE_UNICODE;

                break;
        default:
                break;
                break;
        default:
                break;


@@ -372,7 +372,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
        if ('"' == *start) {
                quoted = 1;
                start++;
        if ('"' == *start) {
                quoted = 1;
                start++;

-       } 
+       }

 
        pairs = 0;
        white = 0;
 
        pairs = 0;
        white = 0;


@@ -391,14 +391,14 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
                         * backslashes and backslash-t to literal tabs.
                         */
                        switch (cp[1]) {
                         * backslashes and backslash-t to literal tabs.
                         */
                        switch (cp[1]) {

-                       case ('t'):
+                       case 't':

                                cp[0] = '\t';
                                /* FALLTHROUGH */
                                cp[0] = '\t';
                                /* FALLTHROUGH */

-                       case ('\\'):
+                       case '\\':

                                pairs++;
                                cp++;
                                break;
                                pairs++;
                                cp++;
                                break;

-                       case (' '):
+                       case ' ':

                                /* Skip escaped blanks. */
                                if (0 == quoted)
                                        cp++;
                                /* Skip escaped blanks. */
                                if (0 == quoted)
                                        cp++;


@@ -427,7 +427,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
 
        /* Quoted argument without a closing quote. */
        if (1 == quoted)
 
        /* Quoted argument without a closing quote. */
        if (1 == quoted)

-               mandoc_msg(MANDOCERR_BADQUOTE, parse, ln, *pos, NULL);
+               mandoc_msg(MANDOCERR_ARG_QUOTE, parse, ln, *pos, NULL);

 
        /* NUL-terminate this argument and move to the next one. */
        if (pairs)
 
        /* NUL-terminate this argument and move to the next one. */
        if (pairs)


@@ -441,7 +441,7 @@ mandoc_getarg(struct mparse *parse, char **cpp, int ln, int *pos)
        *cpp = cp;
 
        if ('\0' == *cp && (white || ' ' == cp[-1]))
        *cpp = cp;
 
        if ('\0' == *cp && (white || ' ' == cp[-1]))

-               mandoc_msg(MANDOCERR_EOLNSPACE, parse, ln, *pos, NULL);
+               mandoc_msg(MANDOCERR_SPACE_EOL, parse, ln, *pos, NULL);

 
        return(start);
 }
 
        return(start);
 }


@@ -455,7 +455,7 @@ a2time(time_t *t, const char *fmt, const char *p)
        memset(&tm, 0, sizeof(struct tm));
 
        pp = NULL;
        memset(&tm, 0, sizeof(struct tm));
 
        pp = NULL;

-#ifdef HAVE_STRPTIME
+#if HAVE_STRPTIME

        pp = strptime(p, fmt, &tm);
 #endif
        if (NULL != pp && '\0' == *pp) {
        pp = strptime(p, fmt, &tm);
 #endif
        if (NULL != pp && '\0' == *pp) {


@@ -509,14 +509,14 @@ mandoc_normdate(struct mparse *parse, char *in, int ln, int pos)
 
        if (NULL == in || '\0' == *in ||
            0 == strcmp(in, "$" "Mdocdate$")) {
 
        if (NULL == in || '\0' == *in ||
            0 == strcmp(in, "$" "Mdocdate$")) {

-               mandoc_msg(MANDOCERR_NODATE, parse, ln, pos, NULL);
+               mandoc_msg(MANDOCERR_DATE_MISSING, parse, ln, pos, NULL);

                time(&t);
        }
        else if (a2time(&t, "%Y-%m-%d", in))
                t = 0;
        else if (!a2time(&t, "$" "Mdocdate: %b %d %Y $", in) &&
            !a2time(&t, "%b %d, %Y", in)) {
                time(&t);
        }
        else if (a2time(&t, "%Y-%m-%d", in))
                t = 0;
        else if (!a2time(&t, "$" "Mdocdate: %b %d %Y $", in) &&
            !a2time(&t, "%b %d, %Y", in)) {

-               mandoc_msg(MANDOCERR_BADDATE, parse, ln, pos, NULL);
+               mandoc_msg(MANDOCERR_DATE_BAD, parse, ln, pos, in);

                t = 0;
        }
        out = t ? time2a(t) : NULL;
                t = 0;
        }
        out = t ? time2a(t) : NULL;


@@ -541,21 +541,21 @@ mandoc_eos(const char *p, size_t sz)
        enclosed = found = 0;
        for (q = p + (int)sz - 1; q >= p; q--) {
                switch (*q) {
        enclosed = found = 0;
        for (q = p + (int)sz - 1; q >= p; q--) {
                switch (*q) {

-               case ('\"'):
+               case '\"':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('\''):
+               case '\'':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case (']'):
+               case ']':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case (')'):
+               case ')':

                        if (0 == found)
                                enclosed = 1;
                        break;
                        if (0 == found)
                                enclosed = 1;
                        break;

-               case ('.'):
+               case '.':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('!'):
+               case '!':

                        /* FALLTHROUGH */
                        /* FALLTHROUGH */

-               case ('?'):
+               case '?':

                        found = 1;
                        break;
                default:
                        found = 1;
                        break;
                default: