egghelp/eggheads community

http get

2003-07-04T16:03:08-04:00

TnX for answering ok , one more question what would be a good way to set the first char then ?

and what is lindex used for .... or when hehe

so i probably should do this

set arg [split $arg]
set out [lindex $arg 0]
set url [join [lrange $arg 1 end]]

or am i wrong again :/ lol

so what is the difference between lrange lindex lsearch i think is there as well so i mean when do i use one or an other to me they all look the same just named different

i only understand i can do this instead of using set

lindex {a, b}

and use them as vars .. instead of using set

Statistics: Posted by Ofloo — Fri Jul 04, 2003 4:03 pm

http get

2003-07-04T15:20:58-04:00

You don't have to set a filname each time... Just use the file command on the url and name the file to the same as the one beeing downloaded

Statistics: Posted by Papillon — Fri Jul 04, 2003 3:20 pm

http get

2003-07-04T15:17:29-04:00

Good
just to answer your questions... to let others know ....
the -timeout option is there to set a certain time the connection should try to get a successfull reply, if it doesn't get it by the given time you will get an error.

Code:

set out [lindex $arg 0]

this is still wrong...
lindex is a list command and shall ONLY be used on valid lists. $arg is not a valid list so you will have to make it one before using lindex on it

Statistics: Posted by Papillon — Fri Jul 04, 2003 3:17 pm

http get

2003-07-04T15:06:45-04:00

ok figured it out

Code:

source lib/tcl8.4/http2.4/http.tcl bind pub m !update update:pub proc update:pub {nick uhost hand chan arg} {   set out [lindex $arg 0]   set url [lindex [split $arg] 1]   set fid [open $out w+]   fconfigure $fid -translation binary  set body [http::data [set token [http::geturl $url -binary 1 -timeout 9999]]]   puts $fid $body   close $fid   http::cleanup $token   putquick "NOTICE $nick :Download complete." }

TnX for the help REALY appriciate it hehe

Statistics: Posted by Ofloo — Fri Jul 04, 2003 3:06 pm

http get

2003-07-04T14:27:44-04:00

First of all tnx verry mutch.

hmm i think i start to understand

ah so i need to say what the file will be called hmm ok sounds logic didn't think of that, tought the lib handled that and time out aswell :/ hehe

the time-out i know what it is for just wondering why like this:

you sad before for time out use cleanup, but you do it on the end while the time out is defined in the url

i am not critising just trying to understand it. beter make it logic

to understand beter what you mean whould this bepossible i mean cause a list command is just like arg and all that is after it so if i would wana name the file it would need to be able to name it and i could do this by setting the first arg to an normal not list command if i did it just would it include in the url write ??? or am i wrong.

Code:

source lib/tcl8.4/http2.4/http.tclbind pub m !update update:pub proc update:pub {nick uhost hand chan arg} {   set out [lindex $arg 0]  set url [lindex [split $arg] 1]   set fid [open $out w+]   set body [http::data [set token [http::geturl $url -binary 1 -timeout 9999]]]   puts $fid $body  close $fid   http::cleanup $token   putquick "NOTICE $nick :Download complete."}

i removed the catch cause it kept the file open (couldn't be executed or anything) but i got an other problem, if i check byte size the file is the same, but if i compare them with a hex the new file is little bigger. meaning it has crc errors, i also tryed binary 0 and left it out but no results, ive tryed wiht catch and without.. i added the notice cause like this i could see if all was done so it didn't stop somewhere. i tought the binary was there to prevent crc in compressed or executable files :/

ok like i tought it downloaded it in ascii cause when i download it with or without the -binary 1 it doesn't make a difference and strange enough these apear to be equal so ..., this means to me that it is downloading it in ascii even tho you define it to be binary.

Statistics: Posted by Ofloo — Fri Jul 04, 2003 2:27 pm

http get

2003-07-04T12:12:26-04:00

I use it mainly as a error prevention. No hurt in removing it... just old habit I guess

Statistics: Posted by Papillon — Fri Jul 04, 2003 12:12 pm

http get

2003-07-04T06:32:38-04:00

Code:

catch {close $fid}

There's no need to catch close. Do you have any particular reason for doing it? (I've seen you do it in other posts too iirc)

Statistics: Posted by user — Fri Jul 04, 2003 6:32 am

http get

2003-07-04T06:19:44-04:00

Code:

set url [lindex $arg 0]

first line of bad coding... ALWAYS make sure it is actually a LIST you are doing the list command on

Code:

puts [http::data [http::geturl $url -binary boolean]]

second line
first of as ppslim said.. you need to put it somewhere, open a file to put it in..

-binary boolean <--- change that to -binary 1
also it might be a good idea to do http::cleanup at the end ...and a timeout on the connection to maybe?

so the final code would look something like this:

Code:

bind pub m !update update:pub proc down:pub {nick uhost hand chan arg} {   set url [lindex [split $arg] 0]   set fid [open somefile w+]  set body [http::data [set token [http::geturl $url -binary 1 -timeout 9999]]]  puts $fid $body  catch {close $fid}  http::cleanup $token}

Statistics: Posted by Papillon — Fri Jul 04, 2003 6:19 am

http get

2003-07-04T02:03:10-04:00

Damn.. Ignore my previous reply. I think I'll take a lil vacation or something, I've had some way to screwd up days this week..

Statistics: Posted by caesar — Fri Jul 04, 2003 2:03 am

http get

2003-07-04T01:09:35-04:00

Doh! use ::http::geturl not http::geturl

That's not it. Read the error message (wrong # args)

Prefixing the namespace name with :: is only needed if you're executing code in a differen namespace than the parent and a namespace by that same name exists but is not the one you want.
...uhm Was that even remotely clear?

Statistics: Posted by user — Fri Jul 04, 2003 1:09 am

http get

2003-07-04T00:47:50-04:00

Doh! use ::http::geturl not http::geturl ..

Statistics: Posted by caesar — Fri Jul 04, 2003 12:47 am

http get

2003-07-03T17:51:08-04:00

Sorry i don't know it and like your quote used to say that it is best xplained by ignorence well i have no clue if you know it suggestions what i read ive read ive tryed, maybe i missed something , maybe i misunderstood, but your answer doesn't make me any wizer then i allready was :/

Statistics: Posted by Ofloo — Thu Jul 03, 2003 5:51 pm

http get

2003-07-03T02:30:14-04:00

Are you blind or plain stupid? The *error* says what you should do.. seesh!

Statistics: Posted by caesar — Thu Jul 03, 2003 2:30 am

http get

2003-07-02T19:39:57-04:00

anny one ??

Statistics: Posted by Ofloo — Wed Jul 02, 2003 7:39 pm

http get

2003-06-28T19:39:40-04:00

if i read well i don't need to load http 2.4 cause it is a thing that is loaded in the tcl core i checked lib and it was there.. i checked syntax manual at http://mini.net/tcl/1475 and it showed the syntax i was using but i still get an error

[01:30] Tcl error [update:pub]: wrong # args: should be "::http::geturl url args"

so what am i doing wrong ?

it just needs to download a file that is it an binairy file

Code:

bind pub m !update update:pubproc down:pub {nick uhost hand chan arg} {set url [lindex $arg 0]puts [http::data [http::geturl $url -binary boolean]]}

i understand what you mean by outputting the file now you mention it it sounds logic is there an other way on doing this ?

what do you recommend but easy and simple so i can understand.

i also tryed "source path/http.tcl"

Statistics: Posted by Ofloo — Sat Jun 28, 2003 7:39 pm