Question: I typically use wget to download files. On some systems, wget is not installed and only curl is available. Can you explain me with a simple example on how I can download a remote file using curl? Are there any difference between curl and wget?
Answer: On a high-level, both wget and curl are command line utilities that do the same thing.
- They both can be used to download files using FTP and HTTP(s).
- You can also send HTTP POST request using curl and wget
- However curl provides APIs that can be used by programmers inside their own code. curl uses libcurl which is a cross-platform library.
- wget is just a command-line tool without any APIs.
- Curl also supports lot more protocols that wget doesn’t support. For example: SCP, SFTP, TFTP, TELNET, LDAP(S), FILE, POP3, IMAP, SMTP, RTMP and RTSP.
- There is a major advantage of using wget. wget supports recursive download, while curl doesn’t.
The following example downloads the file and stores in the same name as the remote server.
The following example download the file and stores in a different name than the remote server. This is helpful when the remote URL doesn’t contain the file name in the url as shown in the example below.
wget -O taglist.zip http://www.vim.org/scripts/download_script.php?src_id=7701
More wget examples: The Ultimate Wget Download Guide With 15 Awesome Examples
$ curl -O http://www.openss7.org/repos/tarballs/strx25-0.9.2.1.tar.bz2 % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 28 3762k 28 1085k 0 0 72771 0 0:00:52 0:00:15 0:00:37 54267
Option -O (upper-case O) is important. Without this, curl will start dumping the downloaded file on the stdout. Using -O, it downloads the files in the same name as the remote server. In the above example, we are downloading strx25-0.9.2.1.tar.bz2, so the downloaded file will also be the same name.
Instead of -O, you an also specify, “–remote-name” as shown below. Both are the same.
$ curl --remote-name http://www.openss7.org/repos/tarballs/strx25-0.9.2.1.tar.bz2
While curl is downloading it gives the following useful information:
- % – The total % of the download that was completed as of now. When it gets to 100% the download is completed. In the above example, it has downloaded only 28% of the file.
- Total – The total size of the file
- Received – The total size of the file that was has been downloaded so far. In the above example, it has downloaded 1085k so far (out of 3762k total)
- Xferd – This will be used when you upload some files to the remote server. During upload, this will indicate the total size of the file that has been uploaded so far. Since we are downloading a file, in this example, this is 0.
- Average Speed Dload – This indicates the average download speed.
- AVerage Speed Upload – While uploading a file, this will indicate the average upload speed
- Time Total – This indicates the total time it will take to download (or upload) the whole file based on the current download (or upload) speed. In this example, it will take approximately a total of 52 seconds to download this file.
- Time Spend – The time curl has spent so far downloading (or uploading) the file. In this example, it has spent 15 seconds so far.
- Time Left – This is caculated based on “Time Total” – “Time Spent”.
- Current Speed – This indicates the current download/upload speed. Compare this with Average Spped Dload/UPload to see how fast or slow your system is downloading currently.
If you want to download the file and store it in a different name than the name of the file in the remote server, use -o (lower-case o) as shown below. This is helpful when the remote URL doesn’t contain the file name in the url as shown in the example below.
$ curl -o taglist.zip http://www.vim.org/scripts/download_script.php?src_id=7701 % Total % Received % Xferd Average Speed Time Time Time Current Dload Upload Total Spent Left Speed 100 50243 100 50243 0 0 170k 0 --:--:-- --:--:-- --:--:-- 400k
In the above example, there is no file name in the remote URL, it just calls a php script that passes some parameter to it. However, the file will be downloaded and saved as taglist.zip on your local system. Instead of -o, you an also specify, “–output”.
More curl examples: 15 Practical Linux cURL Command Examples
Comments on this entry are closed.
thanks, Very nice article
Thanks so muc for the info…….helpful to many guys
You sure know how to make useful posts!
Is there any way to download specific type of file form a site using wget or curl?
Suppose, I want to download just the .jpg files from my site.
Thank you..good info..
I am trying to download a file from GE library. However, we need to login to GE network before downloading. So ,is there any option to give login credentials like my user id and password in the code so that i can download the file.
Please share your comments,
Your help is appreciated.
How can I use this to download the latest version of to date of say a text.txt file from my web site?
Thx a lot!
I want downloading file for video
Its awesome. i love your code. thank you so much… your website is become an angel form because i am tired to find out wget optimum use of code.thanks again.
Thanks you for all the useful posts and do continue the good work.
Thank you for all the useful posts and do continue the good work.
Thank you. Please help me answer that how to resume download file when file was interrupted?
Wish you a good day.
It’s probably a smart move to also use the -L option with curl, which will follow any HTTP redirects.
What if a site requires authentication prior able to download? Eg: mysql binary?
Thanks for the article.
2 quick questions-
(1). When I followed using curl -O example why the file being downloaded has a name of file as “pdf?cycleId=1” where it is picking up the last 2 words from my request ?
(2). I would like to know how can I pass Method type (GET or POST) with curl -O ?
using curl, how do i continue the download from where it ended after it timedout