version 1.163, 2011/07/27 20:55:28 |
version 1.170, 2011/09/18 23:51:31 |
|
|
/* Maximum number of nested if-else conditionals. */ |
/* Maximum number of nested if-else conditionals. */ |
#define RSTACK_MAX 128 |
#define RSTACK_MAX 128 |
|
|
|
/* Maximum number of string expansions per line, to break infinite loops. */ |
|
#define EXPAND_LIMIT 1000 |
|
|
enum rofft { |
enum rofft { |
ROFF_ad, |
ROFF_ad, |
ROFF_am, |
ROFF_am, |
|
|
* Registers are assumed to be unsigned ints for now. |
* Registers are assumed to be unsigned ints for now. |
*/ |
*/ |
struct reg { |
struct reg { |
int set; /* whether set or not */ |
int set; /* whether set or not */ |
unsigned int u; /* unsigned integer */ |
unsigned int u; /* unsigned integer */ |
}; |
}; |
|
|
|
/* |
|
* An incredibly-simple string buffer. |
|
*/ |
struct roffstr { |
struct roffstr { |
char *key; /* key of symbol */ |
char *p; /* nil-terminated buffer */ |
size_t keysz; |
size_t sz; /* saved strlen(p) */ |
char *val; /* current value */ |
|
size_t valsz; |
|
struct roffstr *next; /* next in list */ |
|
}; |
}; |
|
|
|
/* |
|
* A key-value roffstr pair as part of a singly-linked list. |
|
*/ |
|
struct roffkv { |
|
struct roffstr key; |
|
struct roffstr val; |
|
struct roffkv *next; /* next in list */ |
|
}; |
|
|
struct roff { |
struct roff { |
struct mparse *parse; /* parse point */ |
struct mparse *parse; /* parse point */ |
struct roffnode *last; /* leaf of stack */ |
struct roffnode *last; /* leaf of stack */ |
enum roffrule rstack[RSTACK_MAX]; /* stack of !`ie' rules */ |
enum roffrule rstack[RSTACK_MAX]; /* stack of !`ie' rules */ |
int rstackpos; /* position in rstack */ |
int rstackpos; /* position in rstack */ |
struct reg regs[REG__MAX]; |
struct reg regs[REG__MAX]; |
struct roffstr *first_string; /* user-defined strings & macros */ |
struct roffkv *strtab; /* user-defined strings & macros */ |
|
struct roffkv *xmbtab; /* multi-byte trans table (`tr') */ |
|
struct roffstr *xtab; /* single-byte trans table (`tr') */ |
const char *current_string; /* value of last called user macro */ |
const char *current_string; /* value of last called user macro */ |
struct tbl_node *first_tbl; /* first table parsed */ |
struct tbl_node *first_tbl; /* first table parsed */ |
struct tbl_node *last_tbl; /* last table parsed */ |
struct tbl_node *last_tbl; /* last table parsed */ |
Line 162 static enum rofferr roff_cond_sub(ROFF_ARGS); |
|
Line 176 static enum rofferr roff_cond_sub(ROFF_ARGS); |
|
static enum rofferr roff_ds(ROFF_ARGS); |
static enum rofferr roff_ds(ROFF_ARGS); |
static enum roffrule roff_evalcond(const char *, int *); |
static enum roffrule roff_evalcond(const char *, int *); |
static void roff_free1(struct roff *); |
static void roff_free1(struct roff *); |
static void roff_freestr(struct roff *); |
static void roff_freestr(struct roffkv *); |
static char *roff_getname(struct roff *, char **, int, int); |
static char *roff_getname(struct roff *, char **, int, int); |
static const char *roff_getstrn(const struct roff *, |
static const char *roff_getstrn(const struct roff *, |
const char *, size_t); |
const char *, size_t); |
static enum rofferr roff_line_ignore(ROFF_ARGS); |
static enum rofferr roff_line_ignore(ROFF_ARGS); |
static enum rofferr roff_nr(ROFF_ARGS); |
static enum rofferr roff_nr(ROFF_ARGS); |
static void roff_openeqn(struct roff *, const char *, |
static void roff_openeqn(struct roff *, const char *, |
int, int, const char *); |
int, int, const char *); |
static enum rofft roff_parse(struct roff *, const char *, int *); |
static enum rofft roff_parse(struct roff *, const char *, int *); |
static enum rofferr roff_parsetext(char *); |
static enum rofferr roff_parsetext(char *); |
Line 177 static void roff_res(struct roff *, |
|
Line 191 static void roff_res(struct roff *, |
|
static enum rofferr roff_rm(ROFF_ARGS); |
static enum rofferr roff_rm(ROFF_ARGS); |
static void roff_setstr(struct roff *, |
static void roff_setstr(struct roff *, |
const char *, const char *, int); |
const char *, const char *, int); |
|
static void roff_setstrn(struct roffkv **, const char *, |
|
size_t, const char *, size_t, int); |
static enum rofferr roff_so(ROFF_ARGS); |
static enum rofferr roff_so(ROFF_ARGS); |
|
static enum rofferr roff_tr(ROFF_ARGS); |
static enum rofferr roff_TE(ROFF_ARGS); |
static enum rofferr roff_TE(ROFF_ARGS); |
static enum rofferr roff_TS(ROFF_ARGS); |
static enum rofferr roff_TS(ROFF_ARGS); |
static enum rofferr roff_EQ(ROFF_ARGS); |
static enum rofferr roff_EQ(ROFF_ARGS); |
Line 216 static struct roffmac roffs[ROFF_MAX] = { |
|
Line 233 static struct roffmac roffs[ROFF_MAX] = { |
|
{ "rm", roff_rm, NULL, NULL, 0, NULL }, |
{ "rm", roff_rm, NULL, NULL, 0, NULL }, |
{ "so", roff_so, NULL, NULL, 0, NULL }, |
{ "so", roff_so, NULL, NULL, 0, NULL }, |
{ "ta", roff_line_ignore, NULL, NULL, 0, NULL }, |
{ "ta", roff_line_ignore, NULL, NULL, 0, NULL }, |
{ "tr", roff_line_ignore, NULL, NULL, 0, NULL }, |
{ "tr", roff_tr, NULL, NULL, 0, NULL }, |
{ "TS", roff_TS, NULL, NULL, 0, NULL }, |
{ "TS", roff_TS, NULL, NULL, 0, NULL }, |
{ "TE", roff_TE, NULL, NULL, 0, NULL }, |
{ "TE", roff_TE, NULL, NULL, 0, NULL }, |
{ "T&", roff_T_, NULL, NULL, 0, NULL }, |
{ "T&", roff_T_, NULL, NULL, 0, NULL }, |
Line 336 roff_free1(struct roff *r) |
|
Line 353 roff_free1(struct roff *r) |
|
{ |
{ |
struct tbl_node *t; |
struct tbl_node *t; |
struct eqn_node *e; |
struct eqn_node *e; |
|
int i; |
|
|
while (NULL != (t = r->first_tbl)) { |
while (NULL != (t = r->first_tbl)) { |
r->first_tbl = t->next; |
r->first_tbl = t->next; |
Line 354 roff_free1(struct roff *r) |
|
Line 372 roff_free1(struct roff *r) |
|
while (r->last) |
while (r->last) |
roffnode_pop(r); |
roffnode_pop(r); |
|
|
roff_freestr(r); |
roff_freestr(r->strtab); |
} |
roff_freestr(r->xmbtab); |
|
|
|
r->strtab = r->xmbtab = NULL; |
|
|
|
if (r->xtab) |
|
for (i = 0; i < 128; i++) |
|
free(r->xtab[i].p); |
|
|
|
free(r->xtab); |
|
r->xtab = NULL; |
|
} |
|
|
void |
void |
roff_reset(struct roff *r) |
roff_reset(struct roff *r) |
{ |
{ |
Line 413 roff_res(struct roff *r, char **bufp, size_t *szp, int |
|
Line 440 roff_res(struct roff *r, char **bufp, size_t *szp, int |
|
const char *stnam; /* start of the name, after "[(*" */ |
const char *stnam; /* start of the name, after "[(*" */ |
const char *cp; /* end of the name, e.g. before ']' */ |
const char *cp; /* end of the name, e.g. before ']' */ |
const char *res; /* the string to be substituted */ |
const char *res; /* the string to be substituted */ |
int i, maxl; |
int i, maxl, expand_count; |
size_t nsz; |
size_t nsz; |
char *n; |
char *n; |
|
|
|
expand_count = 0; |
|
|
again: |
again: |
cp = *bufp + pos; |
cp = *bufp + pos; |
while (NULL != (cp = strchr(cp, '\\'))) { |
while (NULL != (cp = strchr(cp, '\\'))) { |
|
|
|
|
*bufp = n; |
*bufp = n; |
*szp = nsz; |
*szp = nsz; |
goto again; |
|
|
if (EXPAND_LIMIT >= ++expand_count) |
|
goto again; |
|
|
|
/* Just leave the string unexpanded. */ |
|
mandoc_msg(MANDOCERR_ROFFLOOP, r->parse, ln, pos, NULL); |
|
return; |
} |
} |
} |
} |
|
|
Line 1340 roff_TS(ROFF_ARGS) |
|
Line 1375 roff_TS(ROFF_ARGS) |
|
|
|
/* ARGSUSED */ |
/* ARGSUSED */ |
static enum rofferr |
static enum rofferr |
|
roff_tr(ROFF_ARGS) |
|
{ |
|
const char *p, *first, *second; |
|
size_t fsz, ssz; |
|
enum mandoc_esc esc; |
|
|
|
p = *bufp + pos; |
|
|
|
if ('\0' == *p) { |
|
mandoc_msg(MANDOCERR_ARGCOUNT, r->parse, ln, ppos, NULL); |
|
return(ROFF_IGN); |
|
} |
|
|
|
while ('\0' != *p) { |
|
fsz = ssz = 1; |
|
|
|
first = p++; |
|
if ('\\' == *first) { |
|
esc = mandoc_escape(&p, NULL, NULL); |
|
if (ESCAPE_ERROR == esc) { |
|
mandoc_msg |
|
(MANDOCERR_BADESCAPE, r->parse, |
|
ln, (int)(p - *bufp), NULL); |
|
return(ROFF_IGN); |
|
} |
|
fsz = (size_t)(p - first); |
|
} |
|
|
|
second = p++; |
|
if ('\\' == *second) { |
|
esc = mandoc_escape(&p, NULL, NULL); |
|
if (ESCAPE_ERROR == esc) { |
|
mandoc_msg |
|
(MANDOCERR_BADESCAPE, r->parse, |
|
ln, (int)(p - *bufp), NULL); |
|
return(ROFF_IGN); |
|
} |
|
ssz = (size_t)(p - second); |
|
} else if ('\0' == *second) { |
|
mandoc_msg(MANDOCERR_ARGCOUNT, r->parse, |
|
ln, (int)(p - *bufp), NULL); |
|
second = " "; |
|
p--; |
|
} |
|
|
|
if (fsz > 1) { |
|
roff_setstrn(&r->xmbtab, first, |
|
fsz, second, ssz, 0); |
|
continue; |
|
} |
|
|
|
if (NULL == r->xtab) |
|
r->xtab = mandoc_calloc |
|
(128, sizeof(struct roffstr)); |
|
|
|
free(r->xtab[(int)*first].p); |
|
r->xtab[(int)*first].p = mandoc_strndup(second, ssz); |
|
r->xtab[(int)*first].sz = ssz; |
|
} |
|
|
|
return(ROFF_IGN); |
|
} |
|
|
|
/* ARGSUSED */ |
|
static enum rofferr |
roff_so(ROFF_ARGS) |
roff_so(ROFF_ARGS) |
{ |
{ |
char *name; |
char *name; |
|
|
roff_setstr(struct roff *r, const char *name, const char *string, |
roff_setstr(struct roff *r, const char *name, const char *string, |
int multiline) |
int multiline) |
{ |
{ |
struct roffstr *n; |
|
char *c; |
|
size_t oldch, newch; |
|
|
|
|
roff_setstrn(&r->strtab, name, strlen(name), string, |
|
string ? strlen(string) : 0, multiline); |
|
} |
|
|
|
static void |
|
roff_setstrn(struct roffkv **r, const char *name, size_t namesz, |
|
const char *string, size_t stringsz, int multiline) |
|
{ |
|
struct roffkv *n; |
|
char *c; |
|
int i; |
|
size_t oldch, newch; |
|
|
/* Search for an existing string with the same name. */ |
/* Search for an existing string with the same name. */ |
n = r->first_string; |
n = *r; |
while (n && strcmp(name, n->key)) |
|
|
while (n && strcmp(name, n->key.p)) |
n = n->next; |
n = n->next; |
|
|
if (NULL == n) { |
if (NULL == n) { |
/* Create a new string table entry. */ |
/* Create a new string table entry. */ |
n = mandoc_malloc(sizeof(struct roffstr)); |
n = mandoc_malloc(sizeof(struct roffkv)); |
n->key = mandoc_strdup(name); |
n->key.p = mandoc_strndup(name, namesz); |
n->keysz = strlen(name); |
n->key.sz = namesz; |
n->val = NULL; |
n->val.p = NULL; |
n->valsz = 0; |
n->val.sz = 0; |
n->next = r->first_string; |
n->next = *r; |
r->first_string = n; |
*r = n; |
} else if (0 == multiline) { |
} else if (0 == multiline) { |
/* In multiline mode, append; else replace. */ |
/* In multiline mode, append; else replace. */ |
free(n->val); |
free(n->val.p); |
n->val = NULL; |
n->val.p = NULL; |
n->valsz = 0; |
n->val.sz = 0; |
} |
} |
|
|
if (NULL == string) |
if (NULL == string) |
Line 1493 roff_setstr(struct roff *r, const char *name, const ch |
|
Line 1604 roff_setstr(struct roff *r, const char *name, const ch |
|
* One additional byte for the '\n' in multiline mode, |
* One additional byte for the '\n' in multiline mode, |
* and one for the terminating '\0'. |
* and one for the terminating '\0'. |
*/ |
*/ |
newch = strlen(string) + (multiline ? 2u : 1u); |
newch = stringsz + (multiline ? 2u : 1u); |
if (NULL == n->val) { |
|
n->val = mandoc_malloc(newch); |
if (NULL == n->val.p) { |
*n->val = '\0'; |
n->val.p = mandoc_malloc(newch); |
|
*n->val.p = '\0'; |
oldch = 0; |
oldch = 0; |
} else { |
} else { |
oldch = n->valsz; |
oldch = n->val.sz; |
n->val = mandoc_realloc(n->val, oldch + newch); |
n->val.p = mandoc_realloc(n->val.p, oldch + newch); |
} |
} |
|
|
/* Skip existing content in the destination buffer. */ |
/* Skip existing content in the destination buffer. */ |
c = n->val + (int)oldch; |
c = n->val.p + (int)oldch; |
|
|
/* Append new content to the destination buffer. */ |
/* Append new content to the destination buffer. */ |
while (*string) { |
i = 0; |
|
while (i < (int)stringsz) { |
/* |
/* |
* Rudimentary roff copy mode: |
* Rudimentary roff copy mode: |
* Handle escaped backslashes. |
* Handle escaped backslashes. |
*/ |
*/ |
if ('\\' == *string && '\\' == *(string + 1)) |
if ('\\' == string[i] && '\\' == string[i + 1]) |
string++; |
i++; |
*c++ = *string++; |
*c++ = string[i++]; |
} |
} |
|
|
/* Append terminating bytes. */ |
/* Append terminating bytes. */ |
Line 1522 roff_setstr(struct roff *r, const char *name, const ch |
|
Line 1635 roff_setstr(struct roff *r, const char *name, const ch |
|
*c++ = '\n'; |
*c++ = '\n'; |
|
|
*c = '\0'; |
*c = '\0'; |
n->valsz = (int)(c - n->val); |
n->val.sz = (int)(c - n->val.p); |
} |
} |
|
|
static const char * |
static const char * |
roff_getstrn(const struct roff *r, const char *name, size_t len) |
roff_getstrn(const struct roff *r, const char *name, size_t len) |
{ |
{ |
const struct roffstr *n; |
const struct roffkv *n; |
|
|
for (n = r->first_string; n; n = n->next) |
for (n = r->strtab; n; n = n->next) |
if (0 == strncmp(name, n->key, len) && |
if (0 == strncmp(name, n->key.p, len) && |
'\0' == n->key[(int)len]) |
'\0' == n->key.p[(int)len]) |
return(n->val); |
return(n->val.p); |
|
|
return(NULL); |
return(NULL); |
} |
} |
|
|
static void |
static void |
roff_freestr(struct roff *r) |
roff_freestr(struct roffkv *r) |
{ |
{ |
struct roffstr *n, *nn; |
struct roffkv *n, *nn; |
|
|
for (n = r->first_string; n; n = nn) { |
for (n = r; n; n = nn) { |
free(n->key); |
free(n->key.p); |
free(n->val); |
free(n->val.p); |
nn = n->next; |
nn = n->next; |
free(n); |
free(n); |
} |
} |
|
|
r->first_string = NULL; |
|
} |
} |
|
|
const struct tbl_span * |
const struct tbl_span * |
Line 1567 roff_eqn(const struct roff *r) |
|
Line 1678 roff_eqn(const struct roff *r) |
|
return(r->last_eqn ? &r->last_eqn->eqn : NULL); |
return(r->last_eqn ? &r->last_eqn->eqn : NULL); |
} |
} |
|
|
char |
/* |
roff_eqndelim(const struct roff *r) |
* Duplicate an input string, making the appropriate character |
|
* conversations (as stipulated by `tr') along the way. |
|
* Returns a heap-allocated string with all the replacements made. |
|
*/ |
|
char * |
|
roff_strdup(const struct roff *r, const char *p) |
{ |
{ |
|
const struct roffkv *cp; |
|
char *res; |
|
const char *pp; |
|
size_t ssz, sz; |
|
enum mandoc_esc esc; |
|
|
return('\0'); |
if (NULL == r->xmbtab && NULL == r->xtab) |
|
return(mandoc_strdup(p)); |
|
else if ('\0' == *p) |
|
return(mandoc_strdup("")); |
|
|
|
/* |
|
* Step through each character looking for term matches |
|
* (remember that a `tr' can be invoked with an escape, which is |
|
* a glyph but the escape is multi-character). |
|
* We only do this if the character hash has been initialised |
|
* and the string is >0 length. |
|
*/ |
|
|
|
res = NULL; |
|
ssz = 0; |
|
|
|
while ('\0' != *p) { |
|
if ('\\' != *p && r->xtab && r->xtab[(int)*p].p) { |
|
sz = r->xtab[(int)*p].sz; |
|
res = mandoc_realloc(res, ssz + sz + 1); |
|
memcpy(res + ssz, r->xtab[(int)*p].p, sz); |
|
ssz += sz; |
|
p++; |
|
continue; |
|
} else if ('\\' != *p) { |
|
res = mandoc_realloc(res, ssz + 2); |
|
res[ssz++] = *p++; |
|
continue; |
|
} |
|
|
|
/* Search for term matches. */ |
|
for (cp = r->xmbtab; cp; cp = cp->next) |
|
if (0 == strncmp(p, cp->key.p, cp->key.sz)) |
|
break; |
|
|
|
if (NULL != cp) { |
|
/* |
|
* A match has been found. |
|
* Append the match to the array and move |
|
* forward by its keysize. |
|
*/ |
|
res = mandoc_realloc |
|
(res, ssz + cp->val.sz + 1); |
|
memcpy(res + ssz, cp->val.p, cp->val.sz); |
|
ssz += cp->val.sz; |
|
p += (int)cp->key.sz; |
|
continue; |
|
} |
|
|
|
/* |
|
* Handle escapes carefully: we need to copy |
|
* over just the escape itself, or else we might |
|
* do replacements within the escape itself. |
|
* Make sure to pass along the bogus string. |
|
*/ |
|
pp = p++; |
|
esc = mandoc_escape(&p, NULL, NULL); |
|
if (ESCAPE_ERROR == esc) { |
|
sz = strlen(pp); |
|
res = mandoc_realloc(res, ssz + sz + 1); |
|
memcpy(res + ssz, pp, sz); |
|
break; |
|
} |
|
/* |
|
* We bail out on bad escapes. |
|
* No need to warn: we already did so when |
|
* roff_res() was called. |
|
*/ |
|
sz = (int)(p - pp); |
|
res = mandoc_realloc(res, ssz + sz + 1); |
|
memcpy(res + ssz, pp, sz); |
|
ssz += sz; |
|
} |
|
|
|
res[(int)ssz] = '\0'; |
|
return(res); |
} |
} |