Load the database for this run which is for example
     ./db/parity03_1441.db for run = 1441
 The flag 'usectrl' can be set via the '-D control.db' command
 line option; this forces us to use 'control.db' as the database.
 See also the comments in ChkDbCommand()

 Command line over-ride of database
 Logic does not allow infinite combinations, NOTE:
*  If  '-D control.db' we read from control.db.  Other
     '-D' command subsequently over-ride table(s).
  *  If  '-D useroot filename' we read database from ROOT
     file 'filename' and ignore all other commands.
  *  If  '-D mysql' we read database from MYSQL. Other '-D'
     commands subsequently over-ride table(s).
  *  It is an error to specify both mysql and useroot. If
     you try that, you'll get no data at all (ERROR).
  *  '-D table values'  will over-write the table with those
values, provided the line has proper syntax.  NOTE:  The
     table is dropped; therefore, you cannot add to a table,
     instead you must add all lines of a table when you start one.
     Normally this is ok for trivial tables like 'lobeam'.

Use the '-D' database command to over-ride table(s)
 of database.  See comments for ChkDbCommand()

 Load the database from a ROOT file 'filename'.
 This is done with '-D useroot filename' command line option.
 This choice, if chosen, takes precedence, i.e. all other source
 of data including command line is therefore ignored.

 Human-readable printout for end-of-run summary
 This tells the user what variables were used for this run.
 Print() should be called BEFORE any Put() method is used
 because it shows the state of database at start of analysis.

Returns kTRUE if the database makes good sense.
 This enforces some rules.

 Thorough debug checkout of database.  Ok, this is
 partly redundant with Print()

 Technical printout of entire database

 Technical printout of data map.

Get analysis type, e.g. anatype = 'beam'

 Get timestamp in format "yyyy-mm-dd hh:mm:ss"

 Get timestamp as a TDatime object

 Return the blinding parameters -- sign, mean, and offscale.
 Return defaults if absent from database.  Length of returned
 vector is always 3.

 Return the Cavity monitor calibration parameters-
    xcal, ycal.
 Return defaults if absent from database.

 Return the Cavity monitor calibration parameters-
    xcal, ycal.
 Return defaults if absent from database.

 Return the Cavity monitor calibration parameters-
    xcal, ycal.
 Return defaults if absent from database.

 returns pair type (pair or quad) for this run.

 get the feedback switch state corresponding to feedback type fdbktype.

 get the timescale of feedback type fdbktype

 get the timescale of feedback type fdbktype

 Read or Write the database from Mysql server.
If reading, the ASCII database will be overwritten, so it is a
 good idea to keep backups.
 If writing, it is assumed the ASCII database was already written.
 Syntax of the Mysql script command is:
     $MYSQL_SCRIPT_DIR/mysql.pl  action  dbfile  run
 where
   $MYSQL_SCRIPT_DIR = environment variable pointing to script
   mysql.pl = Perl script that does all the work.
   action = "read" or "write"
   dbfile = name of ASCII database file to read or write
   run = run number

 Write the database for this run (if data was "Put").
 Output goes to both an ASCII file and to MYSQL.
 This method should be called AFTER all Put() methods are used.

 Return single value from table "table".  This assumes the data
 are in a pair  "table   value" where table is a unique string and
 value the single Double_t that belongs to it.

 Generic get method if you know the 'table' and 'keys' you want.
 Structure is 'table (key, data) (key, data)' i.e. a series of
 pairs of (string key, double data) after the string table
 The entire list of keys must be found, not a parital match.

 Return key'd data at index in dtype vector.

 Get Dac noise parameters for adc,chan with key = 'slope'

Get Pedestals for adc, chan

Get Pedestals for scaler, chan

 Get Headers for decoding

Get Mask for decoding

Get a cut value for 'value'.  e.g. value = 'lobeam', 'burpcut'

Get a vector of values for 'value'.  e.g. value = 'posburp'

 Get number of cuts "ncuts" in the database.

Get cut extensions, low and high

Get cut evhi event intervals

 Get detector weights

 Get blumi weights

 Get flumi weights

 Get number of bad event intervals

For bad event intervals, get formatted results
 First element of map goes from 0 to GetNumBadEv(), second is a vector of
 results in prescribed order: (evlo, evhi, cut num, cut value)

 returns ROOT file compression factor

 returns number of events to process

 returns helicity delay (in windows)

 returns oversample factor

 returns pair type (pair or quad) for this run.

 Get current monitor to use for cuts and normalizations for this run

 Get current monitor to use for Hall C cuts for this run

 Return true if calibration variables (not pedestal subtracted) are
 to be included in the raw data tree (default=false)

 Return single value from table "table".  This assumes the data
 are in a pair  "table   value" where table is a unique string and
 value the single Double_t that belongs to it.

 Return "string" from table "table".  This assumes the data
 are in a pair  "table  string" where table is unique and
 value the single string that belongs to it.

Put Dac noise parameters slope  for adc,chan

Put Pedestals for adc, chan

Put Pedestals for scaler, chan

 Put a single bad event interval into the database.
 The user must call this for each cut interval.
The vector 'myevint' is a prescribed order:
   (evlo, evhi, cut num, cut value)

 To put data into the database.
 PRIVATE method used by the public "Put" methods.
WARNINGS: PutData() will REPLACE the table.
 And, no, you do not want to set 'didput' here.
 Performance is poor but its ok because you will not call this often.

 Define the database structure.
 Table names correspond to first column in Ascii DB.
 Careful, an error here can lead to incomprehensible core dump.

 Load the database for a 'table'
 The first call will set up the structure but load empty data.
 The 2nd call (loadflg==1) must overwrite the empty data of 1st call.
 Subsequent calls load data.

 Return 'case insensitive' table name if table exists

Copy database to ROOT representation.

 To reset the datamap iterator.  Must be done before using
 NextDataMap() the first time.

 Find the location in datamap of the first key, according to the rules
 in ./doc/DATABASE.TXT

 From the device name, recognize the suffix "_r" or "_ur"
 if it exists, and return an integer flag:
      1  -->  device is rotated
      0  -->  device is not rotated
     -1  -->  unknown

 Get random or toggle mode
 Get blinding parameters