.\" generated by cd2nroff 0.1 from CURLOPT_FTP_FILEMETHOD.md
.TH CURLOPT_FTP_FILEMETHOD 3 "March 22 2024" libcurl
.SH NAME
CURLOPT_FTP_FILEMETHOD \- select directory traversing method for FTP
.SH SYNOPSIS
.nf
#include <curl/curl.h>

CURLcode curl_easy_setopt(CURL *handle, CURLOPT_FTP_FILEMETHOD,
                          long method);
.fi
.SH DESCRIPTION
Pass a long telling libcurl which \fImethod\fP to use to reach a file on a
FTP(S) server.

This option exists because some server implementations are not compliant to
what the standards say should work.

The argument should be one of the following alternatives:
.IP CURLFTPMETHOD_MULTICWD
libcurl does a single CWD operation for each path part in the given URL. For
deep hierarchies this means many commands. This is how RFC 1738 says it should
be done. This is the default but the slowest behavior.
.IP CURLFTPMETHOD_NOCWD
libcurl makes no CWD at all. libcurl does SIZE, RETR, STOR etc and gives a
full path to the server for all these commands. This is the fastest behavior
since it skips having to change directories.
.IP CURLFTPMETHOD_SINGLECWD
libcurl does one CWD with the full target directory and then operates on the
file &"normally" (like in the multicwd case). This is somewhat more standards
compliant than \(aqnocwd\(aq but without the full penalty of \(aqmulticwd\(aq.
.SH DEFAULT
CURLFTPMETHOD_MULTICWD
.SH PROTOCOLS
FTP
.SH EXAMPLE
.nf
int main(void)
{
  CURL *curl = curl_easy_init();
  if(curl) {
    CURLcode res;
    curl_easy_setopt(curl, CURLOPT_URL, "ftp://example.com/1/2/3/4/new.txt");
    curl_easy_setopt(curl, CURLOPT_FTP_FILEMETHOD,
                     (long)CURLFTPMETHOD_SINGLECWD);

    res = curl_easy_perform(curl);

    curl_easy_cleanup(curl);
  }
}
.fi
.SH AVAILABILITY
Added in 7.15.1
.SH RETURN VALUE
Returns CURLE_OK if the option is supported, and CURLE_UNKNOWN_OPTION if not.
.SH SEE ALSO
.BR CURLOPT_DIRLISTONLY (3),
.BR CURLOPT_FTP_SKIP_PASV_IP (3)