Use the XPA programs to send/receive data to/from XPA servers from the command line or from scripts.
<data> | xpaset [-h] [-i nsinet] [-m method] [-n] [-p] [-s] [-t sval,lval] [-u users] [-v] <template> [paramlist] xpaget [-h] [-i nsinet] [-m method] [-s] [-t sval,lval] [-u users] <template> [paramlist] xpainfo [-h] [-i nsinet] [-m method] [-n] [-s] [-t sval,lval] [-u users] <template> [paramlist] xpaaccess [-c] [-h] [-i nsinet] [-m method] [-n] [-u users] [-v|-V] <template> [type]
<data> | xpaset [-h] [-i nsinet] [-m method] [-n] [-p] [-s] [-t sval,lval] [-u users] [-v] <template|host:port> [paramlist]
-h print help message -i access XPA point on different machine (override XPA_NSINET) -m override XPA_METHOD environment variable -n don't wait for the status message after server completes -p don't read (or send) buf data from stdin -s enter server mode -t [s,l] set short and long timeouts (override XPA_[SHORT,LONG]_TIMEOUT) -u [users] XPA points can be from specified users (override XPA_NSUSERS) -v verify message to stdout --version display version and exit
Data read from stdin will be sent to access points matching the template or host:port. A set of qualifying parameters can be appended.
Normally, xpaset reads data input from stdin until EOF and sends those data to the XPA target, along with parameters entered on the command line. For example to send a FITS file to the ds9 image display:
cat foo.fits | xpaset ds9 fits
Sometimes, however, it is desirable to send only parameters to an XPA access point, without sending data. For such cases, use the -p switch to indicate that there is no data being send to stdin. For example, to change the colormap used by the ds9 image display program, use:
csh> xpaset -p ds9 cmap HeatOf course, this also can be accomplished by sending EOF to stdin in any of the usual ways:
csh> echo "" | xpaset ds9 cmap Heat csh> xpaget ds9 cmap Heat < /dev/null csh> xpaset ds9 cmap Heat ^D # Ctl-D signals EOF
The -s switch puts xpaset into server mode, in which commands and data can be sent to access points without having to run xpaset multiple times. (Its not clear if this buys you much!) The syntax for sending commands in server mode is:
csh> xpaset -s xpaset ds9 colormap I8 ^D xpaset ds9 regions circle 200 300 40 circle 300 400 50 ^D etc.After the required "xpaset" command is specified, optional ASCII data can be appended (as in the region example). A single data/command set is delimited by ^D. Note that typing ^D when a command is expected terminates the program.
NB: server mode only works from the terminal and only ASCII data can be sent in this way.
Examples:
csh> xpaset ds9 file < foo.fits csh> echo "stop" | xpaset myhost:12345
xpaget [-h] [-i nsinet] [-m method] [-s] [-t sval,lval] [-u users] <template|host:port> [paramlist]
-h print help message -i access XPA point on different machine (override XPA_NSINET) -m override XPA_METHOD environment variable -n don't wait for the status message after server completes -s enter server mode -t [s,l] set short and long timeouts (override XPA_[SHORT,LONG]_TIMEOUT) -u [users] XPA points can be from specified users (override XPA_NSUSERS) --version display version and exit
Data will be retrieved from access points matching the template or host:port. A set of qualifying parameters can be appended.
Examples:
csh> xpaget ds9 images csh> xpaget myhost.harvard.edu:12345
xpainfo [-h] [-i nsinet] [-m method] [-n] [-s] [-t sval,lval] [-u users] <template|host:port> [paramlist]
-h print help message -i access XPA point on different machine (override XPA_NSINET) -m override XPA_METHOD environment variable -n don't wait for the status message after server completes -s enter server mode -t [s,l] set short and long timeouts (override XPA_[SHORT,LONG]_TIMEOUT) -u [users] XPA points can be from specified users (override XPA_NSUSERS) --version display version and exit
Info will be sent to access points matching the template or host:port. A set of qualifying parameters can be appended.
Examples:
csh> xpainfo IMAGE ds9 image
xpaaccess [-c] [-h] [-i nsinet] [-m method] [-n] [-t sval,lval] [-u users] -v <template> [type]
-c contact each access point individually -h print help message -i access XPA point on different machine (override XPA_NSINET) -m override XPA_METHOD environment variable -n return number of matches instead of "yes" or "no" -t [s,l] set short and long timeouts (override XPA_[SHORT,LONG]_TIMEOUT) -u [users] XPA points can be from specified users (override XPA_NSUSERS) -v print info about each successful access point -V print info or error about each access point --version display version and exit
xpaaccess returns "yes" to stdout (with a return error code if 1) if there are existing XPA access points that match the template (and optional access type: g,i,s). Otherwise, it returns "no" (with a return error code of 0). If -n is specified, the number of matches is returned instead (both to stdout and in the returned error code). If -v is specified, each access point is displayed to stdout instead of the number of matches.
By default, xpaaccess simply contacts the xpans name server to find the list of registered access points that match the specified template. It also checks to make sure the specified types are supported by that access point. This is the fastest way to determine available access points. However, an access point might registered but not yet available, if, for example, the server program has not entered its event loop to process XPA requests. To find access points that are guaranteed to be available for processing, use the -c (contact) switch. With this switch, xpaaccess contacts each matching XPA server (rather than the name server) to make sure the registered access point really is ready for processing. In this mode, if an access point is registered but not available, xpaaccess will pause for a period of time equal to the XPA_LONG_TIMEOUT, in order to give the server a chance to ready itself. By default, this timeout is 30 seconds. You can shorten the time of delay using the -t "short,long" switch. For example, to shorten the delay time to 2 seconds, use:
xpaaccess -c -t "2,2" ds9The first argument is the short delay value, and is ignored in this operation. The second is the long delay timeout.
Note also that the default xpaaccess method (no -c switch) does not check access control (acls) but rather only checks whether the access point is both registered with the xpans name server and provides the specified type of access. In other words, the default xpaaccess could return 'yes' when you might not actually have access. This mode also always returns 'yes' for the xpans name server itself, regardless of whether the name server is active. The -c (contact) switch, which contacts the access point directly, can and does check the access control (only for servers using version 2.1 and above) and also returns the real status of xpans.