simple implementation of the roff(7) .als (macro alias) request,

[mandoc.git] / read.c

diff --git a/read.c b/read.c
index 5c8af39778818883220bc6a9f3dec7665301c100..f86d57003a9726b9f4bea2adf1b974f498d5084d 100644 (file)
--- a/read.c
+++ b/read.c
@@ -1,4 +1,4 @@
-/*     $Id: read.c,v 1.159 2017/02/03 17:56:59 schwarze Exp $ */
+/*     $Id: read.c,v 1.176 2017/06/11 19:37:01 schwarze Exp $ */

 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2017 Ingo Schwarze <schwarze@openbsd.org>
 /*
  * Copyright (c) 2008, 2009, 2010, 2011 Kristaps Dzonsons <kristaps@bsd.lv>
  * Copyright (c) 2010-2017 Ingo Schwarze <schwarze@openbsd.org>


@@ -30,7 +30,6 @@
 #include <errno.h>
 #include <fcntl.h>
 #include <stdarg.h>
 #include <errno.h>
 #include <fcntl.h>
 #include <stdarg.h>

-#include <stdint.h>

 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <string.h>


@@ -48,8 +47,8 @@
 #define        REPARSE_LIMIT   1000
 
 struct mparse {
 #define        REPARSE_LIMIT   1000
 
 struct mparse {

-       struct roff_man  *man; /* man parser */

        struct roff      *roff; /* roff parser (!NULL) */
        struct roff      *roff; /* roff parser (!NULL) */

+       struct roff_man  *man; /* man parser */

        char             *sodest; /* filename pointed to by .so */
        const char       *file; /* filename of current input file */
        struct buf       *primary; /* buffer currently being parsed */
        char             *sodest; /* filename pointed to by .so */
        const char       *file; /* filename of current input file */
        struct buf       *primary; /* buffer currently being parsed */


@@ -67,7 +66,7 @@ struct        mparse {
 
 static void      choose_parser(struct mparse *);
 static void      resize_buf(struct buf *, size_t);
 
 static void      choose_parser(struct mparse *);
 static void      resize_buf(struct buf *, size_t);

-static void      mparse_buf_r(struct mparse *, struct buf, size_t, int);
+static int       mparse_buf_r(struct mparse *, struct buf, size_t, int);

 static int       read_whole_file(struct mparse *, const char *, int,
                                struct buf *, int *);
 static void      mparse_end(struct mparse *);
 static int       read_whole_file(struct mparse *, const char *, int,
                                struct buf *, int *);
 static void      mparse_end(struct mparse *);


@@ -76,7 +75,7 @@ static        void      mparse_parse_buffer(struct mparse *, struct buf,
 
 static const enum mandocerr    mandoclimits[MANDOCLEVEL_MAX] = {
        MANDOCERR_OK,
 
 static const enum mandocerr    mandoclimits[MANDOCLEVEL_MAX] = {
        MANDOCERR_OK,

-       MANDOCERR_WARNING,
+       MANDOCERR_STYLE,

        MANDOCERR_WARNING,
        MANDOCERR_ERROR,
        MANDOCERR_UNSUPP,
        MANDOCERR_WARNING,
        MANDOCERR_ERROR,
        MANDOCERR_UNSUPP,


@@ -87,6 +86,19 @@ static       const enum mandocerr    mandoclimits[MANDOCLEVEL_MAX] = {
 static const char * const      mandocerrs[MANDOCERR_MAX] = {
        "ok",
 
 static const char * const      mandocerrs[MANDOCERR_MAX] = {
        "ok",
 

+       "generic style suggestion",
+
+       "Mdocdate found",
+       "Mdocdate missing",
+       "legacy man(7) date format",
+       "useless macro",
+       "consider using OS macro",
+       "errnos out of order",
+       "duplicate errno",
+       "description line ends with a full stop",
+       "no blank before trailing delimiter",
+       "function name without markup",
+

        "generic warning",
 
        /* related to the prologue */
        "generic warning",
 
        /* related to the prologue */


@@ -114,6 +126,7 @@ static      const char * const      mandocerrs[MANDOCERR_MAX] = {
        "bad NAME section content",
        "missing comma before name",
        "missing description line, using \"\"",
        "bad NAME section content",
        "missing comma before name",
        "missing description line, using \"\"",

+       "description line outside NAME section",

        "sections out of conventional order",
        "duplicate section title",
        "unexpected section",
        "sections out of conventional order",
        "duplicate section title",
        "unexpected section",


@@ -133,6 +146,7 @@ static      const char * const      mandocerrs[MANDOCERR_MAX] = {
        "fill mode already enabled, skipping",
        "fill mode already disabled, skipping",
        "line scope broken",
        "fill mode already enabled, skipping",
        "fill mode already disabled, skipping",
        "line scope broken",

+       "skipping blank line in line scope",

 
        /* related to missing macro arguments */
        "skipping empty request",
 
        /* related to missing macro arguments */
        "skipping empty request",


@@ -168,6 +182,7 @@ static      const char * const      mandocerrs[MANDOCERR_MAX] = {
        "unknown AT&T UNIX version",
        "comma in function argument",
        "parenthesis in function name",
        "unknown AT&T UNIX version",
        "comma in function argument",
        "parenthesis in function name",

+       "unknown library name",

        "invalid content in Rs block",
        "invalid Boolean argument",
        "unknown font, skipping request",
        "invalid content in Rs block",
        "invalid Boolean argument",
        "unknown font, skipping request",


@@ -220,6 +235,7 @@ static      const char * const      mandocerrs[MANDOCERR_MAX] = {
        "NOT IMPLEMENTED: Bd -file",
        "skipping display without arguments",
        "missing list type, using -item",
        "NOT IMPLEMENTED: Bd -file",
        "skipping display without arguments",
        "missing list type, using -item",

+       "argument is not numeric, using 1",

        "missing manual name, using \"\"",
        "uname(3) system call failed, using UNKNOWN",
        "unknown standard specifier",
        "missing manual name, using \"\"",
        "uname(3) system call failed, using UNKNOWN",
        "unknown standard specifier",


@@ -241,7 +257,7 @@ static      const char * const      mandocerrs[MANDOCERR_MAX] = {
 
 static const char * const      mandoclevels[MANDOCLEVEL_MAX] = {
        "SUCCESS",
 
 static const char * const      mandoclevels[MANDOCLEVEL_MAX] = {
        "SUCCESS",

-       "RESERVED",
+       "STYLE",

        "WARNING",
        "ERROR",
        "UNSUPP",
        "WARNING",
        "ERROR",
        "UNSUPP",


@@ -293,14 +309,15 @@ choose_parser(struct mparse *curp)
        }
 
        if (format == MPARSE_MDOC) {
        }
 
        if (format == MPARSE_MDOC) {

-               mdoc_hash_init();

                curp->man->macroset = MACROSET_MDOC;
                curp->man->macroset = MACROSET_MDOC;

-               curp->man->first->tok = TOKEN_NONE;
+               if (curp->man->mdocmac == NULL)
+                       curp->man->mdocmac = roffhash_alloc(MDOC_Dd, MDOC_MAX);

        } else {
        } else {

-               man_hash_init();

                curp->man->macroset = MACROSET_MAN;
                curp->man->macroset = MACROSET_MAN;

-               curp->man->first->tok = TOKEN_NONE;
+               if (curp->man->manmac == NULL)
+                       curp->man->manmac = roffhash_alloc(MAN_TH, MAN_MAX);

        }
        }

+       curp->man->first->tok = TOKEN_NONE;

 }
 
 /*
 }
 
 /*


@@ -310,7 +327,7 @@ choose_parser(struct mparse *curp)
  * macros, inline equations, and input line traps)
  * and indirectly (for .so file inclusion).
  */
  * macros, inline equations, and input line traps)
  * and indirectly (for .so file inclusion).
  */

-static void
+static int

 mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
 {
        const struct tbl_span   *span;
 mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
 {
        const struct tbl_span   *span;


@@ -318,7 +335,6 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
        const char      *save_file;
        char            *cp;
        size_t           pos; /* byte number in the ln buffer */
        const char      *save_file;
        char            *cp;
        size_t           pos; /* byte number in the ln buffer */

-       size_t           j;  /* auxiliary byte number in the blk buffer */

        enum rofferr     rr;
        int              of;
        int              lnn; /* line number in the real file */
        enum rofferr     rr;
        int              of;
        int              lnn; /* line number in the real file */


@@ -400,79 +416,14 @@ mparse_buf_r(struct mparse *curp, struct buf blk, size_t i, int start)
                                continue;
                        }
 
                                continue;
                        }
 

-                       /* Trailing backslash = a plain char. */
-
-                       if (blk.buf[i] != '\\' || i + 1 == blk.sz) {
-                               ln.buf[pos++] = blk.buf[i++];
-                               continue;
-                       }
-
-                       /*
-                        * Found escape and at least one other character.
-                        * When it's a newline character, skip it.
-                        * When there is a carriage return in between,
-                        * skip that one as well.
-                        */
-
-                       if ('\r' == blk.buf[i + 1] && i + 2 < blk.sz &&
-                           '\n' == blk.buf[i + 2])
-                               ++i;
-                       if ('\n' == blk.buf[i + 1]) {
-                               i += 2;
-                               ++lnn;
-                               continue;
-                       }
-
-                       if ('"' == blk.buf[i + 1] || '#' == blk.buf[i + 1]) {
-                               j = i;
-                               i += 2;
-                               /* Comment, skip to end of line */
-                               for (; i < blk.sz; ++i) {
-                                       if (blk.buf[i] != '\n')
-                                               continue;
-                                       if (blk.buf[i - 1] == ' ' ||
-                                           blk.buf[i - 1] == '\t')
-                                               mandoc_msg(
-                                                   MANDOCERR_SPACE_EOL,
-                                                   curp, curp->line,
-                                                   pos + i-1 - j, NULL);
-                                       ++i;
-                                       ++lnn;
-                                       break;
-                               }
-
-                               /* Backout trailing whitespaces */
-                               for (; pos > 0; --pos) {
-                                       if (ln.buf[pos - 1] != ' ')
-                                               break;
-                                       if (pos > 2 && ln.buf[pos - 2] == '\\')
-                                               break;
-                               }
-                               break;
-                       }
-
-                       /* Catch escaped bogus characters. */
-
-                       c = (unsigned char) blk.buf[i+1];
-
-                       if ( ! (isascii(c) &&
-                           (isgraph(c) || isblank(c)))) {
-                               mandoc_vmsg(MANDOCERR_CHAR_BAD, curp,
-                                   curp->line, pos, "0x%x", c);
-                               i += 2;
-                               ln.buf[pos++] = '?';
-                               continue;
-                       }
-
-                       /* Some other escape sequence, copy & cont. */
-
-                       ln.buf[pos++] = blk.buf[i++];

                        ln.buf[pos++] = blk.buf[i++];
                }
 
                        ln.buf[pos++] = blk.buf[i++];
                }
 

-               if (pos >= ln.sz)
+               if (pos + 1 >= ln.sz)

                        resize_buf(&ln, 256);
 
                        resize_buf(&ln, 256);
 

+               if (i == blk.sz || blk.buf[i] == '\0')
+                       ln.buf[pos++] = '\n';

                ln.buf[pos] = '\0';
 
                /*
                ln.buf[pos] = '\0';
 
                /*


@@ -511,13 +462,16 @@ rerun:
 
                switch (rr) {
                case ROFF_REPARSE:
 
                switch (rr) {
                case ROFF_REPARSE:

-                       if (REPARSE_LIMIT >= ++curp->reparse_count)
-                               mparse_buf_r(curp, ln, of, 0);
-                       else
+                       if (++curp->reparse_count > REPARSE_LIMIT)

                                mandoc_msg(MANDOCERR_ROFFLOOP, curp,
                                    curp->line, pos, NULL);
                                mandoc_msg(MANDOCERR_ROFFLOOP, curp,
                                    curp->line, pos, NULL);

-                       pos = 0;
-                       continue;
+                       else if (mparse_buf_r(curp, ln, of, 0) == 1 ||
+                           start == 1) {
+                               pos = 0;
+                               continue;
+                       }
+                       free(ln.buf);
+                       return 0;

                case ROFF_APPEND:
                        pos = strlen(ln.buf);
                        continue;
                case ROFF_APPEND:
                        pos = strlen(ln.buf);
                        continue;


@@ -531,7 +485,7 @@ rerun:
                            (i >= blk.sz || blk.buf[i] == '\0')) {
                                curp->sodest = mandoc_strdup(ln.buf + of);
                                free(ln.buf);
                            (i >= blk.sz || blk.buf[i] == '\0')) {
                                curp->sodest = mandoc_strdup(ln.buf + of);
                                free(ln.buf);

-                               return;
+                               return 1;

                        }
                        /*
                         * We remove `so' clauses from our lookaside
                        }
                        /*
                         * We remove `so' clauses from our lookaside


@@ -597,18 +551,18 @@ rerun:
        }
 
        free(ln.buf);
        }
 
        free(ln.buf);

+       return 1;

 }
 
 static int
 read_whole_file(struct mparse *curp, const char *file, int fd,
                struct buf *fb, int *with_mmap)
 {
 }
 
 static int
 read_whole_file(struct mparse *curp, const char *file, int fd,
                struct buf *fb, int *with_mmap)
 {

+       struct stat      st;

        gzFile           gz;
        size_t           off;
        ssize_t          ssz;
 
        gzFile           gz;
        size_t           off;
        ssize_t          ssz;
 

-       struct stat      st;
-

        if (fstat(fd, &st) == -1)
                err((int)MANDOCLEVEL_SYSERR, "%s", file);
 
        if (fstat(fd, &st) == -1)
                err((int)MANDOCLEVEL_SYSERR, "%s", file);
 


@@ -815,11 +769,13 @@ mparse_alloc(int options, enum mandoclevel wlevel, mandocmsg mmsg,
        curp->man = roff_man_alloc( curp->roff, curp, curp->defos,
                curp->options & MPARSE_QUICK ? 1 : 0);
        if (curp->options & MPARSE_MDOC) {
        curp->man = roff_man_alloc( curp->roff, curp, curp->defos,
                curp->options & MPARSE_QUICK ? 1 : 0);
        if (curp->options & MPARSE_MDOC) {

-               mdoc_hash_init();

                curp->man->macroset = MACROSET_MDOC;
                curp->man->macroset = MACROSET_MDOC;

+               if (curp->man->mdocmac == NULL)
+                       curp->man->mdocmac = roffhash_alloc(MDOC_Dd, MDOC_MAX);

        } else if (curp->options & MPARSE_MAN) {
        } else if (curp->options & MPARSE_MAN) {

-               man_hash_init();

                curp->man->macroset = MACROSET_MAN;
                curp->man->macroset = MACROSET_MAN;

+               if (curp->man->manmac == NULL)
+                       curp->man->manmac = roffhash_alloc(MAN_TH, MAN_MAX);

        }
        curp->man->first->tok = TOKEN_NONE;
        return curp;
        }
        curp->man->first->tok = TOKEN_NONE;
        return curp;


@@ -830,13 +786,14 @@ mparse_reset(struct mparse *curp)
 {
        roff_reset(curp->roff);
        roff_man_reset(curp->man);
 {
        roff_reset(curp->roff);
        roff_man_reset(curp->man);

+
+       free(curp->sodest);
+       curp->sodest = NULL;
+

        if (curp->secondary)
                curp->secondary->sz = 0;
 
        curp->file_status = MANDOCLEVEL_OK;
        if (curp->secondary)
                curp->secondary->sz = 0;
 
        curp->file_status = MANDOCLEVEL_OK;

-
-       free(curp->sodest);
-       curp->sodest = NULL;

        curp->gzip = 0;
 }
 
        curp->gzip = 0;
 }
 


@@ -844,9 +801,10 @@ void
 mparse_free(struct mparse *curp)
 {
 
 mparse_free(struct mparse *curp)
 {
 

+       roffhash_free(curp->man->mdocmac);
+       roffhash_free(curp->man->manmac);

        roff_man_free(curp->man);
        roff_man_free(curp->man);

-       if (curp->roff)
-               roff_free(curp->roff);
+       roff_free(curp->roff);

        if (curp->secondary)
                free(curp->secondary->buf);
 
        if (curp->secondary)
                free(curp->secondary->buf);