The primary use of D0DAD is to replicate standard DØ data streams streams
in a manner which does not require splitting data for different streams
into separate Zebra files. The data in a given stream is read by
specifying a D0DAD file as the input file to an analysis job in the
place of a (set of) Zebra file(s). The standard set of D0DAD files is created
for the streams defined in omni_filter and stored in the directory
D0$DATA$D0DAD. Table 
lists the standard stream names and the
corresponding D0DAD filenames for the run 1b data. The run 1a data will have
the D0DAD file names as for run 1b, with the string RUN1B replaced by the
string RUN1A. An example of reading
D0DAD-specified data within the D0USER frame is given in figure .
D0DAD files are also created for all triggers and filters defined in normal physics running. These files are in the directory D0$DATA$D0DAD$TRIGFILT. The file names are RUN1x_cccc_trigger/filter.D0DADF. Here x is either A or B for runs Ia and Ib respectively, cccc is either TRIG or FILT for level 1 and level 2 respectivley and trigger/filter is the name defined in the trigger list.
In order to be able to read only runs added after a certain
date, the file NEW_RUNS_YYMMDD.LIS will be created
in the directory D0$DATA$D0DAD each day that
the master catalog is updated. This file will list all runs entered into the
catalog on the date MM/DD/YY in which YY, MM and DD
are the year, month and day respectively. These files can be used with the
RCP parameters described in section to select specific runs
either by run number or via a time stamp.
Table: Standard streams and the corresponding D0DAD filenames. This set
is maintained by the production group. The filenames shown are
for the run 1b data. The run 1a filenames have
the string RUN1B replaced by RUN1A with the rest of the file name
unchanged.
All files are in the directory D0$DATA$D0DAD.
Figure: An example of the input file specification portion of a D0USER job
which reads the NP_MET stream data using a D0DAD file. The user's
typing appears in bold.
For each Zebra file successfully opened by D0DAD, a message of the form
D0DAD: ProcessingFile: <filename>is printed. If a Zebra data file cannot be opened by d0dad, an error message of the form:
D0DAD_READ_EVENT-W :OPEN_FAILURE
FailedFile: D0$DATA$DST:all_072814m01.x_mds01reu1210_all41_nonex00_4070812
is printed using the standard DØ error message utility. All such occurances
are printed, so the log file can be searched to find the name of files to be
excluded from the luminosity calculation.
In addition, the error-handling RCP parameters
described in section allow an analysis job to be stopped with
a fatal error message if any unreadable files are encountered
(See parameter FILE_ERROR_IS_FATAL)
and allow the writing of a d0dad file containing references to all unreadable
events (See parameter ERROR_LIST_FILE).