Reduce memory and time consumption on certain malformed input files

[mandoc.git] / demandoc.c

diff --git a/demandoc.c b/demandoc.c
index 63c7b7bbb11f078733ef3495d9bd1ce65a91f8f6..e9aa62afb2ba9b68ca0d7772f27e893632278ad2 100644 (file)
--- a/demandoc.c
+++ b/demandoc.c
@@ -1,4 +1,4 @@
-/*     $Id: demandoc.c,v 1.4 2011/09/01 20:55:50 kristaps Exp $ */
+/*     $Id: demandoc.c,v 1.12 2014/10/28 17:36:19 schwarze Exp $ */
 /*
  * Copyright (c) 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  *
@@ -14,9 +14,9 @@
  * 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"
-#endif
+
+#include <sys/types.h>
 
 #include <assert.h>
 #include <ctype.h>
@@ -43,6 +43,7 @@ int
 main(int argc, char *argv[])
 {
        struct mparse   *mp;
+       struct mchars   *mchars;
        int              ch, i, list;
        extern int       optind;
 
@@ -76,7 +77,8 @@ main(int argc, char *argv[])
        argc -= optind;
        argv += optind;
 
-       mp = mparse_alloc(MPARSE_AUTO, MANDOCLEVEL_FATAL, NULL, NULL);
+       mchars = mchars_alloc();
+       mp = mparse_alloc(MPARSE_SO, MANDOCLEVEL_FATAL, NULL, mchars, NULL);
        assert(mp);
 
        if (0 == argc)
@@ -88,6 +90,7 @@ main(int argc, char *argv[])
        }
 
        mparse_free(mp);
+       mchars_free(mchars);
        return((int)MANDOCLEVEL_OK);
 }
 
@@ -110,7 +113,7 @@ pmandoc(struct mparse *mp, int fd, const char *fn, int list)
                return;
        }
 
-       mparse_result(mp, &mdoc, &man);
+       mparse_result(mp, &mdoc, &man, NULL);
        line = 1;
        col = 0;
 
@@ -121,7 +124,8 @@ pmandoc(struct mparse *mp, int fd, const char *fn, int list)
        else
                return;
 
-       putchar('\n');
+       if ( ! list)
+               putchar('\n');
 }
 
 /*
@@ -131,12 +135,70 @@ static void
 pstring(const char *p, int col, int *colp, int list)
 {
        enum mandoc_esc  esc;
+       const char      *start, *end;
+       int              emit;
+
+       /*
+        * Print as many column spaces til we achieve parity with the
+        * input document.
+        */
+
+again:
+       if (list && '\0' != *p) {
+               while (isspace((unsigned char)*p))
+                       p++;
+
+               while ('\'' == *p || '(' == *p || '"' == *p)
+                       p++;
+
+               emit = isalpha((unsigned char)p[0]) &&
+                       isalpha((unsigned char)p[1]);
+
+               for (start = p; '\0' != *p; p++)
+                       if ('\\' == *p) {
+                               p++;
+                               esc = mandoc_escape(&p, NULL, NULL);
+                               if (ESCAPE_ERROR == esc)
+                                       return;
+                               emit = 0;
+                       } else if (isspace((unsigned char)*p))
+                               break;
+
+               end = p - 1;
+
+               while (end > start)
+                       if ('.' == *end || ',' == *end || 
+                                       '\'' == *end || '"' == *end ||
+                                       ')' == *end || '!' == *end ||
+                                       '?' == *end || ':' == *end ||
+                                       ';' == *end)
+                               end--;
+                       else
+                               break;
+
+               if (emit && end - start >= 1) {
+                       for ( ; start <= end; start++)
+                               if (ASCII_HYPH == *start)
+                                       putchar('-');
+                               else
+                                       putchar((unsigned char)*start);
+                       putchar('\n');
+               }
+
+               if (isspace((unsigned char)*p))
+                       goto again;
+
+               return;
+       }
 
        while (*colp < col) {
                putchar(' ');
                (*colp)++;
        }
 
+       /*
+        * Print the input word, skipping any special characters.
+        */
        while ('\0' != *p) 
                if ('\\' == *p) {
                        p++;
@@ -153,6 +215,14 @@ static void
 pline(int line, int *linep, int *col, int list)
 {
 
+       if (list)
+               return;
+
+       /*
+        * Print out as many lines as needed to reach parity with the
+        * original input. 
+        */
+
        while (*linep < line) {
                putchar('\n');
                (*linep)++;