Using bash, I tried to do something like this:

---
curl -l ftp://ftp.gtk.org/pub/gtk/perl/ | while read FILENAME
do
  curl -sSO ftp://ftp.gtk.org/pub/gtk/perl/$FILENAME
done
---
And I get this:
curl: (3) URL using bad/illegal format or missing URL
curl: (3) URL using bad/illegal format or missing URL
curl: (3) URL using bad/illegal format or missing URL
etc...
The problem is that the the ftp LIST response is in CR/LF format,
and the bash 'read' strips the LF, but not the CR.
I suppose that's my mistake, but it sure took me a while to
figure out what I was doing wrong! Is there any reason why
curl shouldn't strip the CR if it finds one?
This patch may be a bit too simplistic,
but it seems to solve the problem:
--- main.c.OLD  2006-04-10 06:35:53.133455408 -0500
+++ main.c.NEW  2006-04-10 06:35:43.724885728 -0500
@@ -3793,6 +3793,7 @@
         /* size of uploaded file: */
         curl_easy_setopt(curl, CURLOPT_INFILESIZE_LARGE, uploadfilesize);
+        if ( strchr(url, 0xd) ) *strchr(url, 0xd)=0x0;
         curl_easy_setopt(curl, CURLOPT_URL, url);     /* what to fetch */
         curl_easy_setopt(curl, CURLOPT_PROXY, config->proxy); /*
proxy to use */
         curl_easy_setopt(curl, CURLOPT_HEADER, config->conf&CONF_HEADER);
 - Jeff