/* Copy the characters in the file (up to the next newline) into line
* and advance s to the next line */
static NEOERR *_copy_line_advance(const char **s, STRING *line)
{
NEOERR *err;
int x = 0;
const char *st = *s;
const char *nl;
nl = strchr(st, '\n');
if (nl == NULL)
{
x = strlen(st);
err = string_appendn(line, st, x);
if (err) return nerr_pass(err);
*s = st + x;
}
else
{
x = nl - st;
err = string_appendn(line, st, x);
if (err) return nerr_pass(err);
*s = nl + 1;
}
return STATUS_OK;
}
NEOERR *ne_net_read_line(NSOCK *sock, char **buf)
{
NEOERR *err;
STRING str;
UINT8 *nl;
int l;
string_init(&str);
while (1)
{
if (sock->il - sock->ib > 0)
{
nl = memchr(sock->ibuf + sock->ib, '\n', sock->il - sock->ib);
if (nl == NULL)
{
l = sock->il - sock->ib;
err = string_appendn(&str, (char *)(sock->ibuf + sock->ib), l);
sock->ib += l;
if (err) break;
}
else
{
l = nl - (sock->ibuf + sock->ib);
err = string_appendn(&str, (char *)(sock->ibuf + sock->ib), l);
sock->ib += l;
if (err) break;
*buf = str.buf;
return STATUS_OK;
}
}
else
{
err = ne_net_fill(sock);
if (err) break;
if (sock->il == 0) return STATUS_OK;
}
}
string_clear(&str);
return nerr_pass(err);
}
static PyObject * p_html_ws_strip (PyObject *self, PyObject *args)
{
char *s;
NEOERR *err;
PyObject *rv;
int len;
int lvl;
STRING html;
if (!PyArg_ParseTuple(args, "s#i:htmlStrip(str, level)", &s, &len, &lvl))
return NULL;
string_init (&html);
err = string_appendn (&html, s, len);
if (err) return p_neo_error (err);
cgi_html_ws_strip (&html, lvl);
rv = Py_BuildValue ("s", html.buf);
string_clear (&html);
return rv;
}
static NEOERR *split_and_convert (const char *src, int slen,
STRING *out, HTML_CONVERT_OPTS *opts)
{
NEOERR *err = STATUS_OK;
regmatch_t email_match, url_match;
int errcode;
char *ptr, *esc;
char errbuf[256];
struct _parts *parts;
int part_count;
int part;
int x, i;
int spaces = 0;
if (!CompiledRe)
{
#ifdef HAVE_PTHREADS
/* In threaded environments, we have to mutex lock to do this regcomp, but
* we don't want to use a mutex every time to check that it was regcomp.
* So, we only lock if our first test of compiled was false */
err = mLock(&InitLock);
if (err != STATUS_OK) return nerr_pass(err);
if (CompiledRe == 0) {
#endif
if ((errcode = regcomp (&EmailRegex, EmailRe, REG_ICASE | REG_EXTENDED)))
{
regerror (errcode, &EmailRegex, errbuf, sizeof(errbuf));
err = nerr_raise (NERR_PARSE, "Unable to compile EmailRE: %s", errbuf);
}
if ((errcode = regcomp (&UrlRegex, URLRe, REG_ICASE | REG_EXTENDED)))
{
regerror (errcode, &UrlRegex, errbuf, sizeof(errbuf));
err = nerr_raise (NERR_PARSE, "Unable to compile URLRe: %s", errbuf);
}
CompiledRe = 1;
#ifdef HAVE_PTHREADS
}
if (err) {
mUnlock(&InitLock);
return err;
}
err = mUnlock(&InitLock);
if (err != STATUS_OK) return nerr_pass(err);
#else
if (err) {
return err;
}
#endif
}
part_count = 20;
parts = (struct _parts *) malloc (sizeof(struct _parts) * part_count);
part = 0;
x = 0;
if (regexec (&EmailRegex, src+x, 1, &email_match, 0) != 0)
{
email_match.rm_so = -1;
email_match.rm_eo = -1;
}
else
{
email_match.rm_so += x;
email_match.rm_eo += x;
}
if (regexec (&UrlRegex, src+x, 1, &url_match, 0) != 0)
{
url_match.rm_so = -1;
url_match.rm_eo = -1;
}
else
{
url_match.rm_so += x;
url_match.rm_eo += x;
}
while ((x < slen) && !((email_match.rm_so == -1) && (url_match.rm_so == -1)))
{
if (part >= part_count)
{
void *new_ptr;
part_count *= 2;
new_ptr = realloc (parts, sizeof(struct _parts) * part_count);
if (new_ptr == NULL) {
free(parts);
return nerr_raise (NERR_NOMEM,
"Unable to increase url matcher to %d urls",
part_count);
}
parts = (struct _parts *) new_ptr;
}
if ((url_match.rm_so != -1) && ((email_match.rm_so == -1) || (url_match.rm_so <= email_match.rm_so)))
{
parts[part].begin = url_match.rm_so;
parts[part].end = url_match.rm_eo;
parts[part].type = SC_TYPE_URL;
x = parts[part].end + 1;
part++;
if (x < slen)
{
if (regexec (&UrlRegex, src+x, 1, &url_match, 0) != 0)
{
url_match.rm_so = -1;
url_match.rm_eo = -1;
}
else
{
url_match.rm_so += x;
url_match.rm_eo += x;
}
if ((email_match.rm_so != -1) && (x > email_match.rm_so))
{
if (regexec (&EmailRegex, src+x, 1, &email_match, 0) != 0)
{
email_match.rm_so = -1;
email_match.rm_eo = -1;
}
else
{
email_match.rm_so += x;
email_match.rm_eo += x;
}
}
}
}
else
{
parts[part].begin = email_match.rm_so;
parts[part].end = email_match.rm_eo;
parts[part].type = SC_TYPE_EMAIL;
x = parts[part].end + 1;
part++;
if (x < slen)
{
if (regexec (&EmailRegex, src+x, 1, &email_match, 0) != 0)
{
email_match.rm_so = -1;
email_match.rm_eo = -1;
}
else
{
email_match.rm_so += x;
email_match.rm_eo += x;
}
if ((url_match.rm_so != -1) && (x > url_match.rm_so))
{
if (regexec (&UrlRegex, src+x, 1, &url_match, 0) != 0)
{
url_match.rm_so = -1;
url_match.rm_eo = -1;
}
else
{
url_match.rm_so += x;
url_match.rm_eo += x;
}
}
}
}
}
i = 0;
x = 0;
while (x < slen)
{
if ((i >= part) || (x < parts[i].begin))
{
ptr = strpbrk(src + x, "&<>\r\n ");
if (ptr == NULL)
{
if (spaces)
{
int sp;
for (sp = 0; sp < spaces - 1; sp++)
{
err = string_append (out, " ");
if (err != STATUS_OK) break;
}
if (err != STATUS_OK) break;
err = string_append_char (out, ' ');
}
spaces = 0;
if (i < part)
{
err = string_appendn (out, src + x, parts[i].begin - x);
x = parts[i].begin;
}
else
{
err = string_append (out, src + x);
x = slen;
}
}
else
{
if ((i >= part) || ((ptr - src) < parts[i].begin))
{
if (spaces)
{
int sp;
for (sp = 0; sp < spaces - 1; sp++)
{
err = string_append (out, " ");
if (err != STATUS_OK) break;
}
if (err != STATUS_OK) break;
err = string_append_char (out, ' ');
}
spaces = 0;
err = string_appendn (out, src + x, (ptr - src) - x);
if (err != STATUS_OK) break;
x = ptr - src;
if (src[x] == ' ')
{
if (opts->space_convert)
{
spaces++;
}
else
err = string_append_char (out, ' ');
}
else
{
if (src[x] != '\n' && spaces)
{
int sp;
for (sp = 0; sp < spaces - 1; sp++)
{
err = string_append (out, " ");
if (err != STATUS_OK) break;
}
if (err != STATUS_OK) break;
err = string_append_char (out, ' ');
}
spaces = 0;
if (src[x] == '&')
err = string_append (out, "&");
else if (src[x] == '<')
err = string_append (out, "<");
else if (src[x] == '>')
err = string_append (out, ">");
else if (src[x] == '\n')
if (opts->newlines_convert)
err = string_append (out, "<br/>\n");
else if (x && src[x-1] == '\n')
err = string_append (out, "<p/>\n");
else
err = string_append_char (out, '\n');
else if (src[x] != '\r')
err = nerr_raise (NERR_ASSERT, "src[x] == '%c'", src[x]);
}
x++;
}
else
{
if (spaces)
{
int sp;
for (sp = 0; sp < spaces - 1; sp++)
{
err = string_append (out, " ");
if (err != STATUS_OK) break;
}
if (err != STATUS_OK) break;
err = string_append_char (out, ' ');
}
spaces = 0;
err = string_appendn (out, src + x, parts[i].begin - x);
x = parts[i].begin;
}
}
}
else
{
if (spaces)
{
int sp;
for (sp = 0; sp < spaces - 1; sp++)
{
err = string_append (out, " ");
if (err != STATUS_OK) break;
}
if (err != STATUS_OK) break;
err = string_append_char (out, ' ');
}
spaces = 0;
if (parts[i].type == SC_TYPE_URL)
{
char last_char = src[parts[i].end-1];
int suffix=0;
if (last_char == '.' || last_char == ',') { suffix=1; }
err = string_append (out, " <a ");
if (err != STATUS_OK) break;
if (opts->url_class)
{
err = string_appendf (out, "class=%s ", opts->url_class);
if (err) break;
}
if (opts->url_target)
{
err = string_appendf (out, "target=\"%s\" ", opts->url_target);
if (err) break;
}
err = string_append(out, "href=\"");
if (err) break;
if (opts->bounce_url)
{
char *url, *esc_url, *new_url;
int url_len;
if (!strncasecmp(src + x, "www.", 4))
{
url_len = 7 + parts[i].end - x - suffix;
url = (char *) malloc(url_len+1);
if (url == NULL)
{
err = nerr_raise(NERR_NOMEM,
"Unable to allocate memory to convert url");
break;
}
strcpy(url, "http://");
strncat(url, src + x, parts[i].end - x - suffix);
}
else
{
url_len = parts[i].end - x - suffix;
url = (char *) malloc(url_len+1);
if (url == NULL)
{
err = nerr_raise(NERR_NOMEM,
"Unable to allocate memory to convert url");
break;
}
strncpy(url, src + x, parts[i].end - x - suffix);
url[url_len] = '\0';
}
err = cgi_url_escape(url, &esc_url);
free(url);
if (err) {
free(esc_url);
break;
}
new_url = sprintf_alloc(opts->bounce_url, esc_url);
free(esc_url);
if (new_url == NULL)
{
err = nerr_raise(NERR_NOMEM, "Unable to allocate memory to convert url");
break;
}
err = string_append (out, new_url);
free(new_url);
if (err) break;
}
else
{
if (!strncasecmp(src + x, "www.", 4))
{
err = string_append (out, "http://");
if (err != STATUS_OK) break;
}
err = string_appendn (out, src + x, parts[i].end - x - suffix);
if (err != STATUS_OK) break;
}
err = string_append (out, "\">");
if (err != STATUS_OK) break;
if (opts->link_name) {
err = html_escape_alloc((opts->link_name),
strlen(opts->link_name), &esc);
} else {
err = html_escape_alloc((src + x), parts[i].end - x - suffix, &esc);
}
if (err != STATUS_OK) break;
err = string_append (out, esc);
free(esc);
if (err != STATUS_OK) break;
err = string_append (out, "</a>");
if (suffix) {
err = string_appendn(out,src + parts[i].end - 1,1);
if (err != STATUS_OK) break;
}
}
else /* type == SC_TYPE_EMAIL */
{
err = string_append (out, "<a ");
if (err != STATUS_OK) break;
if (opts->mailto_class)
{
err = string_appendf (out, "class=%s ", opts->mailto_class);
if (err) break;
}
err = string_append(out, "href=\"mailto:");
if (err) break;
err = string_appendn (out, src + x, parts[i].end - x);
if (err != STATUS_OK) break;
err = string_append (out, "\">");
if (err != STATUS_OK) break;
err = html_escape_alloc(src + x, parts[i].end - x, &esc);
if (err != STATUS_OK) break;
err = string_append (out, esc);
free(esc);
if (err != STATUS_OK) break;
err = string_append (out, "</a>");
}
x = parts[i].end;
i++;
}
if (err != STATUS_OK) break;
}
free (parts);
return err;
}