mandoc/mdoc_validate.c - diff

Return to mdoc_validate.c CVS log

Up to [cvsweb.bsd.lv] / mandoc

Diff for /mandoc/mdoc_validate.c between version 1.331 and 1.343

version 1.331, 2017/06/07 23:29:48

version 1.343, 2017/06/25 17:43:45

Line 53 typedef void (*v_post)(POST_ARGS);

static int build_list(struct roff_man *, int);

static void check_text(struct roff_man *, int, int, char *);

static void check_bsd(struct roff_man *, int, int, char *);

static void check_argv(struct roff_man *,

struct roff_node *, struct mdoc_argv *);

static void check_args(struct roff_man *, struct roff_node *);

static void check_toptext(struct roff_man *, int, int, const char *);

static int child_an(const struct roff_node *);

static size_t macro2len(enum roff_tok);

static void rewrite_macro2len(struct roff_man *, char **);

static int similar(const char *, const char *);

static void post_an(POST_ARGS);

static void post_an_norm(POST_ARGS);

Line 75 static void post_bx(POST_ARGS);

Line 76 static void post_bx(POST_ARGS);

static void post_defaults(POST_ARGS);

static void post_display(POST_ARGS);

static void post_dd(POST_ARGS);

static void post_delim(POST_ARGS);

static void post_dt(POST_ARGS);

static void post_en(POST_ARGS);

static void post_es(POST_ARGS);

Line 124 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

Line 126 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

post_bl, /* Bl */

NULL, /* El */

post_it, /* It */

NULL, /* Ad */

post_delim, /* Ad */

post_an, /* An */

NULL, /* Ap */

post_defaults, /* Ar */

NULL, /* Cd */

NULL, /* Cm */

post_delim, /* Cm */

NULL, /* Dv */

post_delim, /* Dv */

NULL, /* Er */

post_delim, /* Er */

NULL, /* Ev */

post_delim, /* Ev */

post_ex, /* Ex */

post_fa, /* Fa */

NULL, /* Fd */

NULL, /* Fl */

post_delim, /* Fl */

post_fn, /* Fn */

NULL, /* Ft */

post_delim, /* Ft */

NULL, /* Ic */

post_delim, /* Ic */

NULL, /* In */

post_delim, /* In */

post_defaults, /* Li */

post_nd, /* Nd */

post_nm, /* Nm */

NULL, /* Op */

post_delim, /* Op */

post_obsolete, /* Ot */

post_defaults, /* Pa */

post_rv, /* Rv */

post_st, /* St */

NULL, /* Va */

post_delim, /* Va */

NULL, /* Vt */

post_delim, /* Vt */

post_xr, /* Xr */

NULL, /* %A */

post_hyph, /* %B */ /* FIXME: can be used outside Rs/Re. */

Line 164 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

Line 166 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

post_hyph, /* %T */ /* FIXME: can be used outside Rs/Re. */

NULL, /* %V */

NULL, /* Ac */

NULL, /* Ao */

post_delim, /* Ao */

NULL, /* Aq */

post_delim, /* Aq */

post_at, /* At */

NULL, /* Bc */

post_bf, /* Bf */

NULL, /* Bo */

post_delim, /* Bo */

NULL, /* Bq */

post_xx, /* Bsx */

post_bx, /* Bx */

Line 179 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

Line 181 static const v_post __mdoc_valids[MDOC_MAX - MDOC_Dd]

NULL, /* Dq */

NULL, /* Ec */

NULL, /* Ef */

NULL, /* Em */

post_delim, /* Em */

NULL, /* Eo */

post_xx, /* Fx */

NULL, /* Ms */

post_delim, /* Ms */

NULL, /* No */

post_ns, /* Ns */

post_xx, /* Nx */

post_xx, /* Ox */

NULL, /* Pc */

NULL, /* Pf */

NULL, /* Po */

post_delim, /* Po */

NULL, /* Pq */

post_delim, /* Pq */

NULL, /* Qc */

NULL, /* Ql */

post_delim, /* Ql */

NULL, /* Qo */

post_delim, /* Qo */

NULL, /* Qq */

post_delim, /* Qq */

NULL, /* Re */

post_rs, /* Rs */

NULL, /* Sc */

NULL, /* So */

post_delim, /* So */

NULL, /* Sq */

post_delim, /* Sq */

post_sm, /* Sm */

post_hyph, /* Sx */

NULL, /* Sy */

post_delim, /* Sy */

post_useless, /* Tn */

post_xx, /* Ux */

NULL, /* Xc */

NULL, /* Xo */

post_fo, /* Fo */

NULL, /* Fc */

NULL, /* Oo */

post_delim, /* Oo */

NULL, /* Oc */

post_bk, /* Bk */

NULL, /* Ek */

post_eoln, /* Bt */

NULL, /* Hf */

post_obsolete, /* Hf */

post_obsolete, /* Fr */

post_eoln, /* Ud */

post_lb, /* Lb */

post_par, /* Lp */

NULL, /* Lk */

post_delim, /* Lk */

post_defaults, /* Mt */

NULL, /* Brq */

post_delim, /* Brq */

NULL, /* Bro */

post_delim, /* Bro */

NULL, /* Brc */

NULL, /* %C */

post_es, /* Es */

Line 303 mdoc_node_validate(struct roff_man *mdoc)

Line 305 mdoc_node_validate(struct roff_man *mdoc)

if (n->sec != SEC_SYNOPSIS ||

(n->parent->tok != MDOC_Cd && n->parent->tok != MDOC_Fd))

check_text(mdoc, n->line, n->pos, n->string);

if (n->parent->tok == MDOC_Sh ||

if (n->parent->tok == MDOC_It ||

n->parent->tok == MDOC_Ss ||

(n->parent->type == ROFFT_BODY &&

n->parent->tok == MDOC_It)

(n->parent->tok == MDOC_Sh ||

check_bsd(mdoc, n->line, n->pos, n->string);

n->parent->tok == MDOC_Ss)))

check_toptext(mdoc, n->line, n->pos, n->string);

break;

case ROFFT_EQN:

case ROFFT_TBL:

Line 389 check_text(struct roff_man *mdoc, int ln, int pos, cha

Line 392 check_text(struct roff_man *mdoc, int ln, int pos, cha

}

static void

check_bsd(struct roff_man *mdoc, int ln, int pos, char *p)

check_toptext(struct roff_man *mdoc, int ln, int pos, const char *p)

{

const char *cp;

const char *cp, *cpr;

if (*p == '\0')

return;

if ((cp = strstr(p, "OpenBSD")) != NULL)

mandoc_msg(MANDOCERR_BX, mdoc->parse,

ln, pos + (cp - p), "Ox");

Line 405 check_bsd(struct roff_man *mdoc, int ln, int pos, char

Line 411 check_bsd(struct roff_man *mdoc, int ln, int pos, char

if ((cp = strstr(p, "DragonFly")) != NULL)

mandoc_msg(MANDOCERR_BX, mdoc->parse,

ln, pos + (cp - p), "Dx");

cp = p;

while ((cp = strstr(cp + 1, "()")) != NULL) {

for (cpr = cp - 1; cpr >= p; cpr--)

if (*cpr != '_' && !isalnum((unsigned char)*cpr))

break;

if ((cpr < p || *cpr == ' ') && cpr + 1 < cp) {

cpr++;

mandoc_vmsg(MANDOCERR_FUNC, mdoc->parse,

ln, pos + (cpr - p),

"%.*s()", (int)(cp - cpr), cpr);

}

static void

post_delim(POST_ARGS)

{

const struct roff_node *nch;

const char *lc, *cp;

int nw;

enum mdelim delim;

enum roff_tok tok;

* Find candidates: at least two bytes,

* the last one a closing or middle delimiter.

tok = mdoc->last->tok;

nch = mdoc->last->last;

if (nch == NULL || nch->type != ROFFT_TEXT)

return;

lc = strchr(nch->string, '\0') - 1;

if (lc <= nch->string)

return;

delim = mdoc_isdelim(lc);

if (delim == DELIM_NONE || delim == DELIM_OPEN)

return;

* Reduce false positives by allowing various cases.

/* Escaped delimiters. */

if (lc > nch->string + 1 && lc[-2] == '\\' &&

(lc[-1] == '&' || lc[-1] == 'e'))

return;

/* Specific byte sequences. */

switch (*lc) {

case ')':

for (cp = lc; cp >= nch->string; cp--)

if (*cp == '(')

return;

break;

case '.':

if (lc > nch->string + 1 && lc[-2] == '.' && lc[-1] == '.')

return;

if (lc[-1] == '.')

return;

break;

case ';':

if (tok == MDOC_Vt)

return;

break;

case '?':

if (lc[-1] == '?')

return;

break;

case ']':

for (cp = lc; cp >= nch->string; cp--)

if (*cp == '[')

return;

break;

case '|':

if (lc == nch->string + 1 && lc[-1] == '|')

return;

default:

break;

}

/* Exactly two non-alphanumeric bytes. */

if (lc == nch->string + 1 && !isalnum((unsigned char)lc[-1]))

return;

/* At least three alphabetic words with a sentence ending. */

if (strchr("!.:?", *lc) != NULL && (tok == MDOC_Em ||

tok == MDOC_Li || tok == MDOC_Po || tok == MDOC_Pq ||

tok == MDOC_Sy)) {

nw = 0;

for (cp = lc - 1; cp >= nch->string; cp--) {

if (*cp == ' ') {

nw++;

if (cp > nch->string && cp[-1] == ',')

cp--;

} else if (isalpha((unsigned int)*cp)) {

if (nw > 1)

return;

} else

break;

}

mandoc_vmsg(MANDOCERR_DELIM, mdoc->parse,

nch->line, nch->pos + (lc - nch->string),

"%s%s %s", roff_name[tok],

nch == mdoc->last->child ? "" : " ...", nch->string);

}

static void

post_bl_norm(POST_ARGS)

{

struct roff_node *n;

Line 782 post_lb(POST_ARGS)

Line 896 post_lb(POST_ARGS)

struct roff_node *n;

const char *p;

post_delim(mdoc);

n = mdoc->last;

assert(n->child->type == ROFFT_TEXT);

mdoc->next = ROFF_NEXT_CHILD;

Line 794 post_lb(POST_ARGS)

Line 910 post_lb(POST_ARGS)

return;

}

mandoc_vmsg(MANDOCERR_LB_BAD, mdoc->parse, n->child->line,

n->child->pos, "Lb %s", n->child->string);

roff_word_alloc(mdoc, n->line, n->pos, "library");

mdoc->last->flags = NODE_NOSRC;

roff_word_alloc(mdoc, n->line, n->pos, "\\(Lq");

Line 1039 post_fa(POST_ARGS)

Line 1158 post_fa(POST_ARGS)

break;

}

post_delim(mdoc);

}

static void

Line 1061 post_nm(POST_ARGS)

Line 1181 post_nm(POST_ARGS)

mandoc_msg(MANDOCERR_NM_NONAME, mdoc->parse,

n->line, n->pos, "Nm");

if (n->type == ROFFT_ELEM)

post_delim(mdoc);

if ((n->type != ROFFT_ELEM && n->type != ROFFT_HEAD) ||

(n->child != NULL && n->child->type == ROFFT_TEXT) ||

mdoc->meta.name == NULL)

Line 1151 post_defaults(POST_ARGS)

Line 1274 post_defaults(POST_ARGS)

{

struct roff_node *nn;

if (mdoc->last->child != NULL) {

post_delim(mdoc);

return;

}

* The `Ar' defaults to "file ..." if no value is provided as an

* argument; the `Mt' and `Pa' macros use "~"; the `Li' just

* gets an empty string.

if (mdoc->last->child != NULL)

return;

nn = mdoc->last;

switch (nn->tok) {

case MDOC_Ar:

mdoc->next = ROFF_NEXT_CHILD;

Line 1225 post_an(POST_ARGS)

Line 1349 post_an(POST_ARGS)

if (nch == NULL)

mandoc_msg(MANDOCERR_MACRO_EMPTY, mdoc->parse,

np->line, np->pos, "An");

else

post_delim(mdoc);

} else if (nch != NULL)

mandoc_vmsg(MANDOCERR_ARG_EXCESS, mdoc->parse,

nch->line, nch->pos, "An ... %s", nch->string);

Line 1253 post_xx(POST_ARGS)

Line 1379 post_xx(POST_ARGS)

struct roff_node *n;

const char *os;

post_delim(mdoc);

n = mdoc->last;

switch (n->tok) {

case MDOC_Bsx:

Line 1582 post_bl(POST_ARGS)

Line 1710 post_bl(POST_ARGS)

nchild = nnext;

}

if (mdoc->meta.os_e != MDOC_OS_NETBSD)

if (mdoc->meta.os_e != MANDOC_OS_NETBSD)

return;

prev_Er = NULL;

Line 1601 post_bl(POST_ARGS)

Line 1729 post_bl(POST_ARGS)

if (order > 0)

mandoc_vmsg(MANDOCERR_ER_ORDER,

mdoc->parse, nnext->line, nnext->pos,

"Er %s %s", prev_Er, nnext->string);

"Er %s %s (NetBSD)",

prev_Er, nnext->string);

else if (order == 0)

mandoc_vmsg(MANDOCERR_ER_REP,

mdoc->parse, nnext->line, nnext->pos,

"Er %s", prev_Er);

"Er %s (NetBSD)", prev_Er);

}

prev_Er = nnext->string;

}

Line 1658 post_sm(POST_ARGS)

Line 1787 post_sm(POST_ARGS)

static void

post_root(POST_ARGS)

{

const char *openbsd_arch[] = {

"alpha", "amd64", "arm64", "armv7", "hppa", "i386",

"landisk", "loongson", "luna88k", "macppc", "mips64",

"octeon", "sgi", "socppc", "sparc64", NULL

};

const char *netbsd_arch[] = {

"acorn26", "acorn32", "algor", "alpha", "amiga",

"arc", "atari",

"bebox", "cats", "cesfic", "cobalt", "dreamcast",

"emips", "evbarm", "evbmips", "evbppc", "evbsh3", "evbsh5",

"hp300", "hpcarm", "hpcmips", "hpcsh", "hppa",

"i386", "ibmnws", "luna68k",

"mac68k", "macppc", "mipsco", "mmeye", "mvme68k", "mvmeppc",

"netwinder", "news68k", "newsmips", "next68k",

"pc532", "playstation2", "pmax", "pmppc", "prep",

"sandpoint", "sbmips", "sgimips", "shark",

"sparc", "sparc64", "sun2", "sun3",

"vax", "walnut", "x68k", "x86", "x86_64", "xen", NULL

};

const char **arches[] = { NULL, netbsd_arch, openbsd_arch };

struct roff_node *n;

const char **arch;

/* Add missing prologue data. */

if (mdoc->meta.date == NULL)

mdoc->meta.date = mdoc->quick ?

mdoc->meta.date = mdoc->quick ? mandoc_strdup("") :

mandoc_strdup("") :

mandoc_normdate(mdoc, NULL, 0, 0);

mandoc_normdate(mdoc->parse, NULL, 0, 0);

if (mdoc->meta.title == NULL) {

mandoc_msg(MANDOCERR_DT_NOTITLE,

Line 1680 post_root(POST_ARGS)

Line 1830 post_root(POST_ARGS)

mandoc_msg(MANDOCERR_OS_MISSING,

mdoc->parse, 0, 0, NULL);

mdoc->meta.os = mandoc_strdup("");

} else if (mdoc->meta.os_e &&

(mdoc->meta.rcsids & (1 << mdoc->meta.os_e)) == 0)

mandoc_msg(MANDOCERR_RCS_MISSING, mdoc->parse, 0, 0,

mdoc->meta.os_e == MANDOC_OS_OPENBSD ?

"(OpenBSD)" : "(NetBSD)");

if (mdoc->meta.arch != NULL &&

(arch = arches[mdoc->meta.os_e]) != NULL) {

while (*arch != NULL && strcmp(*arch, mdoc->meta.arch))

arch++;

if (*arch == NULL) {

n = mdoc->first->child;

while (n->tok != MDOC_Dt)

n = n->next;

n = n->child->next->next;

mandoc_vmsg(MANDOCERR_ARCH_BAD,

mdoc->parse, n->line, n->pos,

"Dt ... %s %s", mdoc->meta.arch,

mdoc->meta.os_e == MANDOC_OS_OPENBSD ?

"(OpenBSD)" : "(NetBSD)");

}

/* Check that we begin with a proper `Sh'. */

Line 1978 post_sh_authors(POST_ARGS)

Line 2149 post_sh_authors(POST_ARGS)

mdoc->last->line, mdoc->last->pos, NULL);

}

* Return an upper bound for the string distance (allowing

* transpositions). Not a full Levenshtein implementation

* because Levenshtein is quadratic in the string length

* and this function is called for every standard name,

* so the check for each custom name would be cubic.

* The following crude heuristics is linear, resulting

* in quadratic behaviour for checking one custom name,

* which does not cause measurable slowdown.

static int

similar(const char *s1, const char *s2)

{

const int maxdist = 3;

int dist = 0;

while (s1[0] != '\0' && s2[0] != '\0') {

if (s1[0] == s2[0]) {

s1++;

s2++;

continue;

}

if (++dist > maxdist)

return INT_MAX;

if (s1[1] == s2[1]) { /* replacement */

s1++;

s2++;

} else if (s1[0] == s2[1] && s1[1] == s2[0]) {

s1 += 2; /* transposition */

s2 += 2;

} else if (s1[0] == s2[1]) /* insertion */

s2++;

else if (s1[1] == s2[0]) /* deletion */

s1++;

else

return INT_MAX;

}

dist += strlen(s1) + strlen(s2);

return dist > maxdist ? INT_MAX : dist;

}

static void

post_sh_head(POST_ARGS)

{

struct roff_node *nch;

const char *goodsec;

const char *const *testsec;

int dist, mindist;

enum roff_sec sec;

Line 2020 post_sh_head(POST_ARGS)

Line 2234 post_sh_head(POST_ARGS)

/* We don't care about custom sections after this. */

if (sec == SEC_CUSTOM)

if (sec == SEC_CUSTOM) {

if ((nch = mdoc->last->child) == NULL ||

nch->type != ROFFT_TEXT || nch->next != NULL)

return;

goodsec = NULL;

mindist = INT_MAX;

for (testsec = secnames + 1; *testsec != NULL; testsec++) {

dist = similar(nch->string, *testsec);

if (dist < mindist) {

goodsec = *testsec;

mindist = dist;

}

if (goodsec != NULL)

mandoc_vmsg(MANDOCERR_SEC_TYPO, mdoc->parse,

nch->line, nch->pos, "Sh %s instead of %s",

nch->string, goodsec);

return;

}

* Check whether our non-custom section is being repeated or is

Line 2087 post_xr(POST_ARGS)

Line 2318 post_xr(POST_ARGS)

if (nch->next == NULL) {

mandoc_vmsg(MANDOCERR_XR_NOSEC, mdoc->parse,

n->line, n->pos, "Xr %s", nch->string);

return;

} else

}

assert(nch->next == n->last);

post_delim(mdoc);

}

static void

Line 2221 post_dd(POST_ARGS)

Line 2452 post_dd(POST_ARGS)

if (n->child == NULL || n->child->string[0] == '\0') {

mdoc->meta.date = mdoc->quick ? mandoc_strdup("") :

mandoc_normdate(mdoc->parse, NULL, n->line, n->pos);

mandoc_normdate(mdoc, NULL, n->line, n->pos);

return;

}

Line 2230 post_dd(POST_ARGS)

Line 2461 post_dd(POST_ARGS)

if (mdoc->quick)

mdoc->meta.date = datestr;

else {

mdoc->meta.date = mandoc_normdate(mdoc->parse,

mdoc->meta.date = mandoc_normdate(mdoc,

datestr, n->line, n->pos);

free(datestr);

}

Line 2338 post_bx(POST_ARGS)

Line 2569 post_bx(POST_ARGS)

struct roff_node *n, *nch;

const char *macro;

post_delim(mdoc);

n = mdoc->last;

nch = n->child;

Line 2417 post_os(POST_ARGS)

Line 2650 post_os(POST_ARGS)

if (mdoc->meta.os)

goto out;

if (mdoc->defos) {

if (mdoc->os_s != NULL) {

mdoc->meta.os = mandoc_strdup(mdoc->defos);

mdoc->meta.os = mandoc_strdup(mdoc->os_s);

goto out;

}

Line 2437 post_os(POST_ARGS)

Line 2670 post_os(POST_ARGS)

mdoc->meta.os = mandoc_strdup(defbuf);

#endif /*!OSNAME*/

out: mdoc->meta.os_e = strstr(mdoc->meta.os, "OpenBSD") != NULL ?

out:

MDOC_OS_OPENBSD : strstr(mdoc->meta.os, "NetBSD") != NULL ?

if (mdoc->meta.os_e == MANDOC_OS_OTHER) {

MDOC_OS_NETBSD : MDOC_OS_OTHER;

if (strstr(mdoc->meta.os, "OpenBSD") != NULL)

mdoc->meta.os_e = MANDOC_OS_OPENBSD;

else if (strstr(mdoc->meta.os, "NetBSD") != NULL)

mdoc->meta.os_e = MANDOC_OS_NETBSD;

}

* This is the earliest point where we can check

* Mdocdate conventions because we don't know

* the operating system earlier.

if (n->child != NULL)

mandoc_vmsg(MANDOCERR_OS_ARG, mdoc->parse,

n->child->line, n->child->pos,

"Os %s (%s)", n->child->string,

mdoc->meta.os_e == MANDOC_OS_OPENBSD ?

"OpenBSD" : "NetBSD");

while (n->tok != MDOC_Dd)

if ((n = n->prev) == NULL)

return;

if ((n = n->child) == NULL)

return;

if (strncmp(n->string, "$" "Mdocdate", 9)) {

if (mdoc->meta.os_e == MANDOC_OS_OPENBSD)

mandoc_vmsg(MANDOCERR_MDOCDATE_MISSING,

mdoc->parse, n->line, n->pos,

"Dd %s (OpenBSD)", n->string);

} else {

if (mdoc->meta.os_e == MANDOC_OS_NETBSD)

mandoc_vmsg(MANDOCERR_MDOCDATE,

mdoc->parse, n->line, n->pos,

"Dd %s (NetBSD)", n->string);

}

enum roff_sec

CVSweb