current version uses strstr() to check if the file has the expected ascii
text extension ... this is buggy becouse this way a file named
".txt_hidden_file" or "foo.config.dat" are both seen as .txt files.

the following patch fixes the problem, plz apply to cvs:



#######################################################################
--- elogd_orig.c        2005-02-03 16:46:10.000000000 +0100
+++ elogd_extchk_fix.c  2005-02-08 17:32:21.000000000 +0100
@@ -1160,6 +1160,28 @@

 #define my_toupper(_c)    ( ((_c)>='a' && (_c)<='z') ? ((_c)-'a'+'A') : (_c) )

+static BOOL chkext(const char *str, const char *ext)
+{
+   int extl, strl;
+   char c1, c2;
+
+   if (ext == NULL || str == NULL)
+      return FALSE;
+
+   extl = strlen(ext);
+   strl = strlen(str);
+   if (extl >= strl)
+      return FALSE;
+   str = str+strl-extl;
+   while (*str) {
+      c1 = *str++;
+      c2 = *ext++;
+      if (my_toupper(c1) != my_toupper(c2))
+         return FALSE;
+   }
+   return TRUE;
+}
+
 BOOL strieq(const char *str1, const char *str2)
 {
    char c1, c2;
@@ -1168,6 +1190,8 @@
       return TRUE;
    if (str1 == NULL || str2 == NULL)
       return FALSE;
+   if (strlen(str1)!=strlen(str2))
+      return FALSE;

    while (*str1) {
       c1 = *str1++;
@@ -13698,8 +13722,8 @@
                       ("<tr><td colspan=%d class=\"attachment\">%s %d: <a
href=\"%s\">%s</a>\n",
                        colspan, loc("Attachment"), index + 1, ref,
attachment[index] + 14);

-                  if ((strstr(str, ".TXT") || strstr(str, ".ASC") ||
strstr(str, ".CFG")
-                       || strstr(str, ".CONF")
+                  if ((chkext(str, ".TXT") || chkext(str, ".ASC") ||
chkext(str, ".CFG")
+                       || chkext(str, ".CONF")
                        || strchr(str, '.') == NULL) && show_attachments) {
                      /* display attachment */
                      rsprintf("</td></tr><tr><td colspan=%d
class=\"messagelist\"><pre>", colspan);
@@ -14779,7 +14803,7 @@
    regex_t re_buf[MAX_N_ATTR + 1];
    regmatch_t pmatch[10];

-   /* redirect if enpty parameters */
+   /* redirect if empty parameters */
    if (strstr(_cmdline, "=&")) {
       while ((pt1 = strstr(_cmdline, "=&")) != NULL) {
          pt2 = pt1;
#######################################################################

> 
> Sorry, I just double-checked, this userdatabase, I was talking about is the Password-file...
> 
> Please try the following snippet :
> 
> allow password change = 0
> Authentication = Webserver, File
> Preset Author = $short_name <$long_name>
> Locked Attributes = Author
> Password file = PASSWD.file
> 
> Christof

Many thanks, that is what I was looking for.

best,
  Stefano

   If you have only one entry in the logbook and have it set up in threaded 
mode like: ?cmd=Search&mode=threaded&reverse=1, when you click on the icon 
of the entry or the number of the entry to see it, it does not allow you 
to.  If there is more than one entry, it works fine.  
   Also I wanted to make a submit page from where you could either go back 
to the logbook which you just entered a message for or type a new entry for 
that logbook.  First of all, is there a way to have only one file called 
submitpage.html in which I can define this for all logbooks or do i need it 
for ever logbook.  And also, is there a way I can even set a link that will 
allow to create a new entry.  I tried: 
    <a href="demo/?cmd=new">Enter</a> another message<p>

but for some reason it does not work.  Is there a way to implement this?

Thanks,
Mo

FYI.

Removing "wrap=hard" on the line #11461 in the elogd.cxx file resolves my problem.

- rsprintf("<textarea rows=%d cols=%d wrap=hard name=\"Text\">\n", height, width);*/
+ rsprintf("<textarea rows=%d cols=%d name=\"Text\">\n", height, width);

- rsprintf("<textarea rows=%d cols=%d wrap=hard name=\"Text\">\n", height, width);
+ rsprintf("<textarea rows=%d cols=%d name=\"Text\">\n", height, width);

my vote is to remove "wrap=hard":

1) I try to read the specs and my head explodes: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/textarea
2) textarea should just accept input typed by user, should not try to "neatify" it. if user wants long lines, we should let them.
3) this bug (introduced in recent safari, the best I can tell)

K.O.

> - rsprintf("<textarea rows=%d cols=%d wrap=hard name=\"Text\">\n", height, width);
> + rsprintf("<textarea rows=%d cols=%d name=\"Text\">\n", height, width);
> 
> my vote is to remove "wrap=hard":
> 
> 1) I try to read the specs and my head explodes: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/textarea
> 2) textarea should just accept input typed by user, should not try to "neatify" it. if user wants long lines, we should let them.
> 3) this bug (introduced in recent safari, the best I can tell)
> 
> K.O.

I agree with K.O. Does anybody see a problem in removing "wrap=hard"? 

It was there more for historical reasons. In the old days screens were not so wide and wrapping was more of an issue.
People tended to write longer lines and complained that the long lines got reformatted differently for different screen
sizes. So by adding hard CRLF the formatting looked the same on different screens. These days this is not such an issue
any more and I agree with 2) above. If the user wants a long line, the user should get it.

Stefan

Ahh, now I remember. Well, the I put that in like 25 years ago ;-)

Let's assume the user write a very long line and relies on the wrapping of the text box. So the input might look like the 
first attachment. Then the user hits submit and gets just one long line (second attachment) and has to scroll one kilometre
to the right to see the full line. So there is an inconsistency between the entry form and what the user sees after the
submission. Having "wrap=hard" tells the browser to put CRLF where the wrapping in the textarea happens, so the text looks
the same during entry and after submission. If we remove the "wrap=hard", we would be back to the situation below in the two
attachments.

Opinions?

Stefan