version 1.7, 2014/03/23 23:35:59 |
version 1.11, 2014/04/01 08:31:00 |
|
|
#include <string.h> |
#include <string.h> |
#include <unistd.h> |
#include <unistd.h> |
|
|
|
/* |
|
* In what section can we find Perl manuals? |
|
*/ |
|
#define PERL_SECTION "3p" |
|
|
struct args { |
struct args { |
const char *title; /* override "Dt" title */ |
const char *title; /* override "Dt" title */ |
const char *date; /* override "Dd" date */ |
const char *date; /* override "Dd" date */ |
|
|
LIST__MAX |
LIST__MAX |
}; |
}; |
|
|
|
enum sect { |
|
SECT_NONE = 0, |
|
SECT_NAME, /* NAME section */ |
|
SECT_SYNOPSIS, /* SYNOPSIS section */ |
|
}; |
|
|
struct state { |
struct state { |
int parsing; /* after =cut of before command */ |
int parsing; /* after =cut of before command */ |
int paused; /* in =begin and before =end */ |
int paused; /* in =begin and before =end */ |
int haspar; /* in paragraph: do we need Pp? */ |
int haspar; /* in paragraph: do we need Pp? */ |
int isname; /* are we the NAME section? */ |
enum sect sect; /* which section are we in? */ |
const char *fname; /* file being parsed */ |
const char *fname; /* file being parsed */ |
#define LIST_STACKSZ 128 |
#define LIST_STACKSZ 128 |
enum list lstack[LIST_STACKSZ]; /* open lists */ |
enum list lstack[LIST_STACKSZ]; /* open lists */ |
Line 158 formatescape(const char *buf, size_t *start, size_t en |
|
Line 169 formatescape(const char *buf, size_t *start, size_t en |
|
} |
} |
|
|
/* |
/* |
* Skip space characters. |
* Run some heuristics to intuit a link format. |
|
* I recognise L<foo::bar> as a Perl manpage, printing it in section 3p; |
|
* or a general UNIX foo(5) manpage. |
|
* If I recognise one, I set "start" to be the end of the sequence so |
|
* that the caller can safely just continue processing. |
|
* Otherwise, I don't touch "start". |
*/ |
*/ |
static int |
static int |
skipspace(const char *buf, size_t *start, size_t end) |
trylink(const char *buf, size_t *start, size_t end, size_t dsz) |
{ |
{ |
size_t sv = *start; |
size_t sv, nstart, nend, i, j; |
|
int hasdouble; |
|
|
while (*start < end && ' ' == buf[*start]) |
/* |
(*start)++; |
* Scan to the start of the terminus. |
|
* This function is more or less replicated in the formatcode() |
|
* for null or index formatting codes. |
|
*/ |
|
hasdouble = 0; |
|
for (sv = nstart = *start; nstart < end; nstart++) { |
|
/* Do we have a double-colon? */ |
|
if (':' == buf[nstart] && |
|
nstart > sv && |
|
':' == buf[nstart - 1]) |
|
hasdouble = 1; |
|
if ('>' != buf[nstart]) |
|
continue; |
|
else if (dsz == 1) |
|
break; |
|
assert(nstart > 0); |
|
if (' ' != buf[nstart - 1]) |
|
continue; |
|
i = nstart; |
|
for (j = 0; i < end && j < dsz; j++) |
|
if ('>' != buf[i++]) |
|
break; |
|
if (dsz == j) |
|
break; |
|
} |
|
|
|
/* We don't care about stubs. */ |
|
if (nstart == end || nstart == *start) |
|
return(0); |
|
|
return(*start > sv); |
/* Set nend to the end of content. */ |
|
nend = nstart; |
|
if (dsz > 1) |
|
nend--; |
|
|
|
/* |
|
* Provide for some common invocations of the link primitive. |
|
* First, allow us to link to other Perl manuals. |
|
*/ |
|
if (hasdouble) |
|
printf("Xr %.*s " PERL_SECTION, |
|
(int)(nend - sv), &buf[sv]); |
|
else if (nend - sv > 3 && isalnum(buf[sv]) && |
|
')' == buf[nend - 1] && |
|
isdigit((int)buf[nend - 2]) && |
|
'(' == buf[nend - 3]) |
|
printf("Xr %.*s %c", |
|
(int)(nend - 3 - sv), |
|
&buf[sv], buf[nend - 2]); |
|
else |
|
return(0); |
|
|
|
*start = nstart; |
|
return(1); |
} |
} |
|
|
/* |
/* |
Line 185 skipspace(const char *buf, size_t *start, size_t end) |
|
Line 253 skipspace(const char *buf, size_t *start, size_t end) |
|
* reentrant). |
* reentrant). |
*/ |
*/ |
static int |
static int |
formatcode(const char *buf, size_t *start, |
formatcode(struct state *st, const char *buf, |
size_t end, int reentrant, int nomacro) |
size_t *start, size_t end, int reentrant, int nomacro) |
{ |
{ |
enum fmt fmt; |
enum fmt fmt; |
size_t i, j, dsz; |
size_t i, j, dsz; |
Line 204 formatcode(const char *buf, size_t *start, |
|
Line 272 formatcode(const char *buf, size_t *start, |
|
|
|
if (FMT__MAX == fmt) { |
if (FMT__MAX == fmt) { |
putchar(last = buf[(*start)++]); |
putchar(last = buf[(*start)++]); |
|
if ('\\' == last) |
|
putchar('e'); |
return(0); |
return(0); |
} |
} |
|
|
Line 290 formatcode(const char *buf, size_t *start, |
|
Line 360 formatcode(const char *buf, size_t *start, |
|
printf("Em "); |
printf("Em "); |
break; |
break; |
case (FMT_BOLD): |
case (FMT_BOLD): |
|
/* |
|
* Doclifting: if we're a bold "-xx" and we're |
|
* in the SYNOPSIS section, then it's likely |
|
* that we're a flag. |
|
* Be really strict: only do this when the dash |
|
* is followed by alnums til the end marker, |
|
* which mustn't be a custom. |
|
*/ |
|
if (SECT_SYNOPSIS == st->sect && |
|
end - *start > 1 && |
|
'-' == buf[*start] && |
|
(isalnum((int)buf[*start + 1]) || |
|
'?' == buf[*start + 1])) { |
|
for (i = *start + 1; i < end; i++) |
|
if ( ! isalnum((int)buf[i])) |
|
break; |
|
if (i < end && '>' == buf[i]) { |
|
(*start)++; |
|
printf("Fl "); |
|
break; |
|
} |
|
} |
printf("Sy "); |
printf("Sy "); |
break; |
break; |
case (FMT_CODE): |
case (FMT_CODE): |
printf("Qo Li "); |
printf("Qo Li "); |
break; |
break; |
case (FMT_LINK): |
case (FMT_LINK): |
printf("Lk "); |
if ( ! trylink(buf, start, end, dsz)) |
|
printf("No "); |
break; |
break; |
case (FMT_FILE): |
case (FMT_FILE): |
printf("Pa "); |
printf("Pa "); |
break; |
break; |
case (FMT_NBSP): |
case (FMT_NBSP): |
/* TODO. */ |
|
printf("No "); |
printf("No "); |
break; |
break; |
default: |
default: |
Line 338 formatcode(const char *buf, size_t *start, |
|
Line 430 formatcode(const char *buf, size_t *start, |
|
} |
} |
} |
} |
if (*start + 1 < end && '<' == buf[*start + 1]) { |
if (*start + 1 < end && '<' == buf[*start + 1]) { |
formatcode(buf, start, end, 1, nomacro); |
formatcode(st, buf, start, end, 1, nomacro); |
continue; |
continue; |
} |
} |
|
|
Line 363 formatcode(const char *buf, size_t *start, |
|
Line 455 formatcode(const char *buf, size_t *start, |
|
else |
else |
putchar(last = buf[*start]); |
putchar(last = buf[*start]); |
|
|
|
/* Protect against character escapes. */ |
|
if ('\\' == last) |
|
putchar('e'); |
|
|
(*start)++; |
(*start)++; |
|
|
if (' ' == last) |
if (' ' == last) |
Line 389 formatcode(const char *buf, size_t *start, |
|
Line 485 formatcode(const char *buf, size_t *start, |
|
* Calls formatcode() til the end of a paragraph. |
* Calls formatcode() til the end of a paragraph. |
*/ |
*/ |
static void |
static void |
formatcodeln(const char *buf, size_t *start, size_t end, int nomacro) |
formatcodeln(struct state *st, const char *buf, |
|
size_t *start, size_t end, int nomacro) |
{ |
{ |
|
|
last = ' '; |
last = ' '; |
while (*start < end) { |
while (*start < end) { |
if (*start + 1 < end && '<' == buf[*start + 1]) { |
if (*start + 1 < end && '<' == buf[*start + 1]) { |
formatcode(buf, start, end, 1, nomacro); |
formatcode(st, buf, start, end, 1, nomacro); |
continue; |
continue; |
} |
} |
/* |
/* |
Line 412 formatcodeln(const char *buf, size_t *start, size_t en |
|
Line 509 formatcodeln(const char *buf, size_t *start, size_t en |
|
' ' == buf[*start + 2])) |
' ' == buf[*start + 2])) |
printf("\\&"); |
printf("\\&"); |
|
|
if ('\n' != buf[*start]) |
if ('\n' == buf[*start]) |
putchar(last = buf[*start]); |
|
else |
|
putchar(last = ' '); |
putchar(last = ' '); |
|
else |
|
putchar(last = buf[*start]); |
|
|
|
/* Protect against character escapes. */ |
|
if ('\\' == last) |
|
putchar('e'); |
|
|
(*start)++; |
(*start)++; |
} |
} |
} |
} |
Line 471 command(struct state *st, const char *buf, size_t star |
|
Line 573 command(struct state *st, const char *buf, size_t star |
|
return; |
return; |
|
|
start += csz; |
start += csz; |
skipspace(buf, &start, end); |
while (start < end && ' ' == buf[start]) |
|
start++; |
|
|
len = end - start; |
len = end - start; |
|
|
if (st->paused) { |
if (st->paused) { |
Line 488 command(struct state *st, const char *buf, size_t star |
|
Line 592 command(struct state *st, const char *buf, size_t star |
|
* how pod2man handles it. |
* how pod2man handles it. |
*/ |
*/ |
printf(".Sh "); |
printf(".Sh "); |
st->isname = 0; |
st->sect = SECT_NONE; |
if (end - start == 4) |
if (end - start == 4) { |
if (0 == memcmp(&buf[start], "NAME", 4)) |
if (0 == memcmp(&buf[start], "NAME", 4)) |
st->isname = 1; |
st->sect = SECT_NAME; |
formatcodeln(buf, &start, end, 1); |
} else if (end - start == 8) { |
|
if (0 == memcmp(&buf[start], "SYNOPSIS", 8)) |
|
st->sect = SECT_SYNOPSIS; |
|
} |
|
formatcodeln(st, buf, &start, end, 1); |
putchar('\n'); |
putchar('\n'); |
st->haspar = 1; |
st->haspar = 1; |
break; |
break; |
case (CMD_HEAD2): |
case (CMD_HEAD2): |
printf(".Ss "); |
printf(".Ss "); |
formatcodeln(buf, &start, end, 1); |
formatcodeln(st, buf, &start, end, 1); |
putchar('\n'); |
putchar('\n'); |
st->haspar = 1; |
st->haspar = 1; |
break; |
break; |
case (CMD_HEAD3): |
case (CMD_HEAD3): |
puts(".Pp"); |
puts(".Pp"); |
printf(".Em "); |
printf(".Em "); |
formatcodeln(buf, &start, end, 0); |
formatcodeln(st, buf, &start, end, 0); |
putchar('\n'); |
putchar('\n'); |
puts(".Pp"); |
puts(".Pp"); |
st->haspar = 1; |
st->haspar = 1; |
Line 513 command(struct state *st, const char *buf, size_t star |
|
Line 621 command(struct state *st, const char *buf, size_t star |
|
case (CMD_HEAD4): |
case (CMD_HEAD4): |
puts(".Pp"); |
puts(".Pp"); |
printf(".No "); |
printf(".No "); |
formatcodeln(buf, &start, end, 0); |
formatcodeln(st, buf, &start, end, 0); |
putchar('\n'); |
putchar('\n'); |
puts(".Pp"); |
puts(".Pp"); |
st->haspar = 1; |
st->haspar = 1; |
Line 567 command(struct state *st, const char *buf, size_t star |
|
Line 675 command(struct state *st, const char *buf, size_t star |
|
switch (st->lstack[st->lpos - 1]) { |
switch (st->lstack[st->lpos - 1]) { |
case (LIST_TAG): |
case (LIST_TAG): |
printf(".It "); |
printf(".It "); |
formatcodeln(buf, &start, end, 0); |
formatcodeln(st, buf, &start, end, 0); |
putchar('\n'); |
putchar('\n'); |
break; |
break; |
case (LIST_ENUM): |
case (LIST_ENUM): |
Line 624 command(struct state *st, const char *buf, size_t star |
|
Line 732 command(struct state *st, const char *buf, size_t star |
|
static void |
static void |
verbatim(struct state *st, const char *buf, size_t start, size_t end) |
verbatim(struct state *st, const char *buf, size_t start, size_t end) |
{ |
{ |
size_t sv = start; |
int last; |
|
|
if ( ! st->parsing || st->paused) |
if ( ! st->parsing || st->paused) |
return; |
return; |
|
|
puts(".Bd -literal"); |
puts(".Bd -literal"); |
while (start < end) { |
for (last = ' '; start < end; start++) { |
if (start > sv && '\n' == buf[start - 1]) |
/* |
|
* Handle accidental macros (newline starting with |
|
* control character) and escapes. |
|
*/ |
|
if ('\n' == last) |
if ('.' == buf[start] || '\'' == buf[start]) |
if ('.' == buf[start] || '\'' == buf[start]) |
printf("\\&"); |
printf("\\&"); |
putchar(buf[start++]); |
putchar(last = buf[start]); |
|
if ('\\' == buf[start]) |
|
printf("e"); |
} |
} |
putchar('\n'); |
putchar('\n'); |
puts(".Ed"); |
puts(".Ed"); |
Line 662 ordinary(struct state *st, const char *buf, size_t sta |
|
Line 776 ordinary(struct state *st, const char *buf, size_t sta |
|
* we're in "name - description" format. |
* we're in "name - description" format. |
* To wit, print out a "Nm" and "Nd" in that format. |
* To wit, print out a "Nm" and "Nd" in that format. |
*/ |
*/ |
if (st->isname) { |
if (SECT_NAME == st->sect) { |
for (i = end - 1; i > start; i--) |
for (i = end - 1; i > start; i--) |
if ('-' == buf[i]) |
if ('-' == buf[i]) |
break; |
break; |
Line 673 ordinary(struct state *st, const char *buf, size_t sta |
|
Line 787 ordinary(struct state *st, const char *buf, size_t sta |
|
if ('-' != buf[i]) |
if ('-' != buf[i]) |
break; |
break; |
printf(".Nm "); |
printf(".Nm "); |
formatcodeln(buf, &start, i + 1, 1); |
formatcodeln(st, buf, &start, i + 1, 1); |
putchar('\n'); |
putchar('\n'); |
start = j + 1; |
start = j + 1; |
printf(".Nd "); |
printf(".Nd "); |
formatcodeln(buf, &start, end, 1); |
formatcodeln(st, buf, &start, end, 1); |
putchar('\n'); |
putchar('\n'); |
return; |
return; |
} |
} |
Line 704 ordinary(struct state *st, const char *buf, size_t sta |
|
Line 818 ordinary(struct state *st, const char *buf, size_t sta |
|
else if ('\n' == last && '\'' == buf[start]) |
else if ('\n' == last && '\'' == buf[start]) |
printf("\\&"); |
printf("\\&"); |
putchar(last = buf[start++]); |
putchar(last = buf[start++]); |
|
if ('\\' == last) |
|
putchar('e'); |
} |
} |
|
|
if (start < end - 1 && '<' == buf[start + 1]) { |
if (start < end - 1 && '<' == buf[start + 1]) { |
Line 713 ordinary(struct state *st, const char *buf, size_t sta |
|
Line 829 ordinary(struct state *st, const char *buf, size_t sta |
|
* what, so print a newline now. |
* what, so print a newline now. |
* Then print the (possibly nested) macros and |
* Then print the (possibly nested) macros and |
* following that, a newline. |
* following that, a newline. |
|
* Consume all whitespace so we don't |
|
* accidentally start an implicit literal line. |
*/ |
*/ |
if (formatcode(buf, &start, end, 0, 0)) { |
if (formatcode(st, buf, &start, end, 0, 0)) { |
putchar(last = '\n'); |
putchar(last = '\n'); |
while (start < end && ' ' == buf[start]) |
while (start < end && ' ' == buf[start]) |
start++; |
start++; |
Line 753 ordinary(struct state *st, const char *buf, size_t sta |
|
Line 871 ordinary(struct state *st, const char *buf, size_t sta |
|
else if ('\n' == last && '\'' == buf[start]) |
else if ('\n' == last && '\'' == buf[start]) |
printf("\\&"); |
printf("\\&"); |
putchar(last = buf[start++]); |
putchar(last = buf[start++]); |
|
if ('\\' == last) |
|
putchar('e'); |
} |
} |
} |
} |
|
|
Line 817 dofile(const struct args *args, const char *fname, |
|
Line 937 dofile(const struct args *args, const char *fname, |
|
if (NULL != (cp = strrchr(title, '.'))) { |
if (NULL != (cp = strrchr(title, '.'))) { |
*cp++ = '\0'; |
*cp++ = '\0'; |
if (0 == strcmp(cp, "pm")) |
if (0 == strcmp(cp, "pm")) |
section = "3p"; |
section = PERL_SECTION; |
} |
} |
} |
} |
|
|