Re: trouble with encoded filename
--- Hrvoje Niksic [EMAIL PROTECTED] wrote: Maybe the site expressly disallows Wget to access it? Maybe the site has anti-leech protection that requires sending the `Referer' header to thwart it? Maybe you need to have a cookie to access the site that Opera sends? Well, I found out a little bit more about the real reason for the problem. Opera has a very convenient option called Encode International Web Addresses with UTF-8. When I had this option checked, it could retrieve the file without problems. Without this option enabled, I get the same forbidden response that I received when I used wget. In my never-humble opinion, wget needs this ability also. I had hoped that using the option --restrict-file-names=nocontrol would have disabled encoding of the URL, but apparently, it does not. __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/
Re: trouble with encoded filename
[EMAIL PROTECTED] wrote: Well, I found out a little bit more about the real reason for the problem. Opera has a very convenient option called Encode International Web Addresses with UTF-8. When I had this option checked, it could retrieve the file without problems. Without this option enabled, I get the same forbidden response that I received when I used wget. In my never-humble opinion, wget needs this ability also. I had hoped that using the option --restrict-file-names=nocontrol would have disabled encoding of the URL, but apparently, it does not. Huh? Opera is doing special encoding for some types of web addresses and you hoped that disabling ALL encoding would somehow make wget do the same thing? If special encoding is required: 1) someone has to write the code in wget to perform that encoding and 2) it has to be ENabled (not DISabled). Tony
Re: trouble with encoded filename
--- Tony Lewis [EMAIL PROTECTED] wrote: Opera is doing special encoding for some types of web addresses and you hoped that disabling ALL encoding would somehow make wget do the same thing? Disabling escaped octet encoding would work. It does not seem unreasonable to HOPE (but not expect) that disabling encoding for file names would do that. If special encoding is required: 1) someone has to write the code in wget to perform that encoding I agree with that (I could write it in Perl, but I doubt that that would help much :) ). and 2) it has to be ENabled (not DISabled). No, escaped octet encoding (the default for most user agents) needs to be DISABLED (and the actual character sent). I looked at Opera some more, and in this case, it still only sent one byte for this character (UTF8 uses two bytes for anything not seven bit ASCII). I can't argue with the fact that it worked in Opera though, and I just want wget to send (if I disable escaped ASCII encoding of high bit characters) the unencoded URL. (exactly as I type or paste it). __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/
Re: trouble with encoded filename
none none [EMAIL PROTECTED] writes: (URL changed for privacy) $ wget http://1.2.3.4/?.file --00:00:00-- http://1.2.3.4/%E9.file = `?.file' Connecting to 1.2.3.4:80... connected. HTTP request sent, awaiting response... 403 Forbidden 00:00:00 ERROR 403: Forbidden. $ wget -V GNU Wget 1.9 Gentoo GNU/Linux Opera (for example) has no problem retrieving the URL. Maybe the site expressly disallows Wget to access it? Maybe the site has anti-leech protection that requires sending the `Referer' header to thwart it? Maybe you need to have a cookie to access the site that Opera sends? Try `wget --user-agent=Mozilla --referer=http://1.2.3.4 ...' That gets you past 90% of anti-leech code.
Re: trouble with encoded filename
--- Hrvoje Niksic [EMAIL PROTECTED] wrote: (URL changed for privacy) $ wget http://1.2.3.4/?.file --00:00:00-- http://1.2.3.4/%E9.file = `?.file' Connecting to 1.2.3.4:80... connected. HTTP request sent, awaiting response... 403 Forbidden 00:00:00 ERROR 403: Forbidden. $ wget -V GNU Wget 1.9 Gentoo GNU/Linux Opera (for example) has no problem retrieving the URL. Maybe the site expressly disallows Wget to access it? Maybe the site has anti-leech protection that requires sending the `Referer' header to thwart it? Maybe you need to have a cookie to access the site that Opera sends? No, I retrieved more than fifty other files from this site without any problem (and Opera gave the same access forbidden message when I, as a test, substituted %E9 for that one character in the URL). None of the other filenames had any characters in the upper half of the ascii table. Try `wget --user-agent=Mozilla --referer=http://1.2.3.4 ...' That gets you past 90% of anti-leech code. If you insist. $ wget -S --referer=http://5.6.7.8/index.htm \ --user-agent=Mozilla \ http://1.2.3.4/?.file --00:00:00-- http://1.2.3.4/%E9.file = `?.file' Connecting to 1.2.3.4:80... connected. HTTP request sent, awaiting response... 1 HTTP/1.1 403 Forbidden 2 Date: Wed, 07 Apr 2004 00:00:00 GMT 3 Server: Apache/2.0.48 (Win32) 4 Content-Length: 323 5 Keep-Alive: timeout=15, max=100 6 Connection: Keep-Alive 7 Content-Type: text/html; charset=iso-8859-1 00:00:00 ERROR 403: Forbidden. __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/
Re: trouble with encoded filename
none none wrote: $ wget -S --referer=http://5.6.7.8/index.htm \ --user-agent=Mozilla \ http://1.2.3.4/?.file --00:00:00-- http://1.2.3.4/%E9.file = `?.file' Connecting to 1.2.3.4:80... connected. HTTP request sent, awaiting response... 1 HTTP/1.1 403 Forbidden 2 Date: Wed, 07 Apr 2004 00:00:00 GMT 3 Server: Apache/2.0.48 (Win32) 4 Content-Length: 323 5 Keep-Alive: timeout=15, max=100 6 Connection: Keep-Alive 7 Content-Type: text/html; charset=iso-8859-1 00:00:00 ERROR 403: Forbidden. Hi, Are you really sure that the file you want has the name ?.file with a question mark character?? I see from the headers that the server involved runs under Windows and the question mark is most certainly a character which is not possible in a filename under DOS/Windows. Regards, J.Roderburg
Re: trouble with encoded filename
--- Jochen Roderburg [EMAIL PROTECTED] wrote: Are you really sure that the file you want has the name ?.file with a question mark character?? Not a question mark in wget, or when I sent the mail - it seems that Yahoo mangles it on the way out (another case of characters in the upper half of the ascii table - which I DID mention - not handled correctly). You can see the actual character here (E9): http://bbsinc.com/pic/fnt-mswin.gif Wget did send (what looks like) the correct encoding (which usually works, but did not in this case). Since Opera can retrieve the file, some way does exist to send the name in a way that the server understands it (the forbidden message does seem a little odd, though - I would have expected the server to return a not found. __ Do you Yahoo!? Yahoo! Small Business $15K Web Design Giveaway http://promotions.yahoo.com/design_giveaway/