rda-python-metrics 1.0.4__py3-none-any.whl

rda_python_metrics/viewcodusage.py

@@ -0,0 +1,314 @@
+#!/usr/bin/env python3
+#
+###############################################################################
+#
+#     Title : viewcodusage
+#    Author : Zaihua Ji,  zji@ucar.edu
+#      Date : 03/15/2022
+#             2025-03-28 transferred to package rda_python_metrics from
+#             https://github.com/NCAR/rda-database.git
+#   Purpose : python program to view custom OPeNDAP usage information
+#
+#    Github : https://github.com/NCAR/rda-python-metrics.git
+#
+###############################################################################
+#
+import os
+import re
+import sys
+from rda_python_common import PgLOG
+from rda_python_common import PgUtil
+from rda_python_common import PgDBI
+from . import PgView
+
+VUSG = {
+   'SNMS' : "ABCDEGHIJMNOPQRSTUVY",          # cod available short field names in FLDS
+   'OPTS' : 'AabcCdDeEghHiLmnoOqStTUwyz',    # cod available options, used for params
+   'NOPT' : 'abhwz',                         # stand alone option without inputs
+   'ACND' : 'cdegiImoqSty',                 # available array condition options
+   'RCND' : 'DET',                           # available range condition options
+   'CNDS' : 'acdDeEgiImMnoqStTy',           # condition options, ACND, RCND and 'a'
+   'ECND' : 'my',                            # condition options need evaluating
+   'SFLD' : 'DEGINOTU',                      # string fields, to be quoted in condition
+   'UFLD' : 'NO',                            # string fields must be in upper case
+   'LFLD' : 'EMPT'                           # string fields must be in lower case
+}
+
+# keys FLDS - short field names
+# column 0   - column title showing in usage view
+# column 1   - field name in format as shown in select clauses
+# column 2   - field name shown in where condition query string
+# column 3   - table name that the field belongs to 
+# column 4   - output field length, the longer one of data size and comlun title, determine
+#              dynamiccody if it is 0. Negative values indicate right justification
+# column 5   - precision for floating point value if positive and show total value if not zero
+# column 6   - field flag to indicate it is a group, distinct or sum field
+FLDS = {
+# SHRTNM COLUMNNANE   FIELDNAME                         CNDNAME       TBLNAM        Size Prc Grp/Sum
+   'D' : ['DATE',      "date",                           'date',       'codusage',  10,   0,  'G'],
+   'E' : ['EMAIL',     "codusage.email",        'codusage.email',      'codusage',   0,   0,  'G'],
+   'G' : ['AGGREID',   "aid",                           'aid',        'codusage',    0,   0,  'G'],
+   'I' : ['IP',        "ip",                             'ip',         'codusage',   0,   0,  'G'],
+   'M' : ['MONTH',     PgDBI.fmtym("date"),              'date',       'codusage',   7,   0,  'G'],
+   'N' : ['COUNTRY',   "country",                        'country',    'codusage',   0,   0,  'G'],
+   'O' : ['ORGTYPE',   "org_type",                       'org_type',   'codusage',   7,   0,  'G'],
+   'P' : ['DSOWNER',   "specialist",                     'specialist', 'dsowner',    8,   0,  'G'],
+   'Q' : ['QUARTER',   "quarter",                        'quarter',    'codusage',   7,   0,  'G'],
+   'R' : ['DSTITLE',   "search.datasets.title", 'search.datasets.title',   'search.datasets',   0,   0,  'G'],
+   'S' : ['BYTESIZE',  "size",                           'size',       'codusage', -14,  -1,  'G'],
+   'T' : ['DATASET',   "codusage.dsid",         'codusage.dsid',       'codusage',   0,   0,  'G'],
+   'V' : ['PERIOD',    "perod",                          'period',     'codusage',   0,   0,  'G'],
+   'Y' : ['YEAR',      PgDBI.fmtyr("date"),              'date',       'codusage',   4,   0,  'G'],
+   'A' : ['DSCOUNT',   "codusage.dsid",                  'A',          'codusage',  -7,  -1,  'D'],
+   'B' : ['MBYTEREAD', "round(sum(size)/(1000000), 4)",  'B',          'codusage', -14,   3,  'S'],
+   'C' : ['#AGGRE',    "aid",                            'G',          'codusage',  -8,  -1,  'D'],
+   'H' : ['#ACCESS',   "sum(count)",                     'H',          'codusage',  -8,  -1,  'S'],
+   'J' : ['#UNIQUSER', "codusage.email",                 'C',          'codusage',  -9,  -1,  'D'],
+   'U' : ['#UNIQIP',   "codusage.ip",                    'U',          'codusage',  -7,  -1,  'D'],
+   'X' : ['INDEX',     "",                               'X',          '',          -6,   0,  ' ']
+}
+
+# keys EXPAND - short field names allow zero usage
+# column 0   - expand ID for group of fields
+# column 1   - field name shown in where condition query string
+# column 2   - field name in format as shown in select clauses
+# column 3   - table name that the field belongs to 
+EXPAND = {
+# SHRTNM EXPID     CNDSTR    FIELDNAME       TBLNAM
+   'D' : ["TIME",   "dDmy"],
+   'M' : ["TIME",   "dDmy"],
+   'Q' : ["TIME",   "dDmy"],
+   'Y' : ["TIME",   "dDmy"],
+
+   'E' : ["USER",   "eco",  "email",        "wuser",  "user"],
+   'O' : ["USER",   "eco",  "org_type",     "wuser",  "user"],
+   'N' : ["USER",   "eco",  "country",      "wuser",  "user"],
+
+   'R' : ["DSID",   "fFsStT", "search.datasets.title", "search.datasets"],
+   'T' : ["DSID",   "fFsStT", "dataset.dsid",   "dataset"],
+   'P' : ["DSID",   "fFsStT", "specialist",     "dsowner"]
+}
+
+# valid options for params, a hash array of command line parameters
+#   a -- 1 to view cod usage info available
+#   A -- number or records to return
+#   c -- array of specified country codes
+#   C -- a string of short field names for viewing usages
+#   d -- array of specified dates 
+#   D -- dates range, array of 1 or 2 dates in format of YYYY-MM-DD
+#   e -- array of specified email addresses
+#   E -- use given date or date range for email notice of data update
+#   g -- array of specified aggregation IDs
+#   h -- for give emails, include their histical emails registered before
+#   H -- a string of report title to replace the default one
+#   i -- array of specified IP addresses
+#   I -- use given email IDs for email notice of data update
+#   L -- column delimiter for output
+#   m -- array of specified months 
+#   o -- array of specified orginization types
+#   O -- a string of short field names for sorting on
+#   q -- array of the specified quarters, normcody combined with years
+#   S -- array of login names of specialists who owns the datasets
+#   t -- array of specified dataset names
+#   T -- dataset range, array of 1 or 2 dataset names
+#   U -- given unit for file or data sizes
+#   w -- generate view without totals
+#   y -- array of specified years 
+#   z -- generate view including entries without usage
+
+params = {}
+
+# relationship between parameter options and short field names, A option is not
+# related to a field name if it is not in keys SNS 
+SNS = {
+   'c' : 'N', 'd' : 'D', 'D' : 'D', 'e' : 'E', 'g' : 'G', 'i' : 'I', 'm' : 'M',
+   'o' : 'O', 'q' : 'Q', 'S' : 'P', 't' : 'T', 'T' : 'T', 'y' : 'Y'
+}
+
+tablenames = fieldnames = condition = ''
+sfields = []
+gfields = []
+dfields = []
+pgname = 'viewcodusage'
+
+#
+# main function to run this program
+#
+def main():
+
+   PgDBI.view_dbinfo()
+   argv = sys.argv[1:]
+   inputs = []
+   option = 'C'   # default option
+
+   for arg in argv:
+      if re.match(r'^-.*$', arg):
+         curopt = arg[1:2]
+         if curopt and VUSG['OPTS'].find(curopt) > -1:
+            if VUSG['NOPT'].find(option) > -1:
+               params[option] = 1
+            elif inputs:
+               params[option]= inputs   # record input array
+               inputs = []      # empty input array
+            option = curopt     # start a new option
+         else:
+            PgLOG.pglog(arg + ": Unknown Option", PgLOG.LGWNEX)
+      else:
+         val = arg
+         if val != '!':
+            if option == 's':
+               val = int(val)*1000000    # convert MBytes to Bytes
+            elif option in SNS:
+               sfld = SNS[option]
+               if VUSG['SFLD'].find(sfld) > -1:
+                  if VUSG['UFLD'].find(sfld) > -1:
+                     val = arg.upper()     # in case not in upper case
+                  elif VUSG['LFLD'].find(sfld) > -1:
+                     val = arg.lower()     # in case not in lower case
+                  if option == 'c':
+                     val = PgView.get_country_name(val)
+                  elif option == 't' or option == 'T':
+                     val = PgUtil.format_dataset_id(val)   # add 'ds' if only numbers
+                  val = "'{}'".format(val)
+         inputs.append(val)
+   
+   # record the last option
+   if VUSG['NOPT'].find(option) > -1:
+      params[option] = 1
+   elif inputs:
+      params[option] = inputs   # record input array
+   
+   if not params:
+      PgLOG.show_usage(pgname)
+   else:
+      check_enough_options()
+   
+   if 'o' not in params:
+      if 'e' not in params:
+         params['o'] = ['!', "'DSS'"]   # default to exclude 'DSS' for organization
+   elif params['o'][0] == "'ALL'":
+      del params['o']
+
+   usgtable = 'codusage'
+   build_query_strings(usgtable)   # build tablenames, fieldnames, and condtions
+   records = PgDBI.pgmget(tablenames, fieldnames, condition, PgLOG.UCLWEX)
+   if not records: PgLOG.pglog("No Usage Found For Given Conditions", PgLOG.LGWNEX)
+   totals = None if 'w' in params else {}
+   if dfields or totals != None:
+      records = PgView.compact_hash_groups(records, gfields, sfields, dfields, totals)
+   if 'z' in params: records = expand_records(records)
+   ostr = params['O'][0] if 'O' in params else params['C'][0]
+   records = PgView.order_records(records, ostr.replace('X', ''))
+   PgView.simple_output(params, FLDS, records, totals)
+
+   PgLOG.pgexit(0)
+
+#
+# cehck if enough information entered on command line for generate view/report, exit if not
+#
+def check_enough_options():
+
+   cols = params['C'][0] if 'C' in params else 'X'
+   if cols == 'X': PgLOG.pglog("{}: miss field names '{}'".format(pgname, VUSG['SNMS']), PgLOG.LGWNEX)
+
+   if cols.find('Q') > -1 and cols.find('Y') < 0:   # add Y if Q included
+      cols = re.sub('Q', 'YQ', cols)
+      params['C'][0] = cols
+
+   for sn in cols:
+      if sn == 'X': continue  # do not process INDEX field
+      if VUSG['SNMS'].find(sn) < 0:
+         PgLOG.pglog("{}: Field {} must be in field names '{}X'".format(pgname, sn, VUSG['SNMS']), PgLOG.LGWNEX)
+      if 'z' not in params or sn in EXPAND: continue
+      fld = FLDS[sn]
+      if fld[6] != 'G': continue
+      PgLOG.pglog("{}: cannot show zero usage for unexpandable field {} - {}".formt(pgname, sn, fld[0]), PgLOG.LGWNEX)
+
+   if 'E' in params or 'I' in params:
+      if 'z' in params:
+         PgLOG.pglog(pgname + ": option -z and -E/-I can not be present at the same time", PgLOG.LGWNEX)
+      elif 't' not in params or len(params['t']) > 1:
+         PgLOG.pglog(pgname + ": specify one dataset for viewing usage of notified users", PgLOG.LGWNEX)
+      elif 'E' in params and 'I' in params:
+         PgLOG.pglog(pgname + ": option -E and -I can not be present at the same time", PgLOG.LGWNEX)
+
+   for opt in params:
+      if VUSG['CNDS'].find(opt) > -1: return
+   PgLOG.pglog("{}: miss condition options '{}'".format(pgname, VUSG['CNDS']), PgLOG.LGWNEX)
+
+#
+# process parameter options to build cod query strings
+# global variables are used directly and nothing passes in and returns back
+#
+def build_query_strings(usgtable):
+
+   # initialize query strings
+   global condition, fieldnames, tablenames, gfields, sfields
+   joins = having = groupnames = ''
+   tablenames = usgtable
+   cols = params['C'][0]
+
+   if 'U' in params:    # reset units for file and read sizes
+      if cols.find('B') > -1: FLDS['B'] = PgView.set_data_unit(FLDS['B'], params['U'][0], "sum(size)")
+      if cols.find('S') > -1: FLDS['S'] = PgView.set_data_unit(FLDS['S'], params['U'][0], "size")      
+
+   if 'e' in params and 'h' in params: params['e'] = PgView.include_historic_emails(params['e'], 3)
+
+   for opt in params:
+      if opt == 'C':   # build field, table and group names
+         for sn in cols:
+            if sn == 'X': continue  # do not process INDEX field
+            fld = FLDS[sn]
+            if fieldnames: fieldnames += ', '
+            fieldnames += "{} {}".format(fld[1], sn)   # add to field name string
+            (tablenames, joins) = PgView.join_query_tables(fld[3], tablenames, joins, usgtable)
+            if fld[6] == 'S':
+               sfields.append(sn)
+            else:
+               if groupnames: groupnames += ', '
+               groupnames += sn     # add to group name string
+               if fld[6] == 'D':
+                  dfields.append(sn)
+               else:
+                  gfields.append(sn)
+      elif opt == 'O':
+         continue   # order records later
+      elif VUSG['CNDS'].find(opt) > -1:
+         if VUSG['NOPT'].find(opt) > -1: continue
+         sn = SNS[opt]
+         fld = FLDS[sn]
+         # build where conditon strings
+         cnd = PgView.get_view_condition(opt, sn, fld, params, VUSG)
+         if cnd:
+            if condition: condition += ' AND '
+            condition += cnd
+            (tablenames, joins) = PgView.join_query_tables(fld[3], tablenames, joins, usgtable)
+
+   # append joins, group by, order by, and having strings to condition string
+   if 'E' in params or 'I' in params:
+      (tablenames, joins) = PgView.join_query_tables("emreceive", tablenames, joins, usgtable)
+   if joins:
+      if condition:
+         condition = "{} AND {}".format(joins, condition)
+      else:
+         condition = joins
+   if 'E' in params or 'I' in params:
+      condition += PgView.notice_condition(params['E'], params['I'], params['t'][0])
+   if groupnames and sfields: condition += " GROUP BY " + groupnames
+
+def expand_records(records):
+
+   recs = PgView.expand_query("TIME", records, params, EXPAND)
+
+   trecs = PgView.expand_query("USER", records, params, EXPAND, VUSG, SNS, FLDS)
+   if trecs: PgUtil.crosshash(recs, trecs)
+
+   trecs = PgView.expand_query("DSID", records, params, EXPAND, VUSG, SNS, FLDS)
+   if trecs: recs = PgUtil.crosshash(recs, trecs)
+
+   return PgUtil.joinhash(records, recs, 0, 1)
+
+#
+# call main() to start program
+#
+if __name__ == "__main__": main()

rda_python_metrics/viewcodusage.usg

@@ -0,0 +1,184 @@
+
+ View usage information of custom OPeNDAP aggregation Data Services from
+ information stored in MySQL database 'RDADB'.
+
+ Usage: viewcodusage [-C] ColumnNames [-O OrderColumnNames] [-a] \
+                     [-A RowLimit] [-c CountryCodes] [-d DateList] \
+                     [-D StartDate [EndDate]] [-e EMailList] -h \
+                     [-E StartNoticeDate [EndNoticeDate]] \
+                     [-g AggregationIDs] [-i IPAddresses] \
+                     [-I EmailIDList] [-m MonthList] \
+                     [-N MinNumberRead [MaxNumberRead]] \
+                     [-o OrganizationTypes] [-q QuaterList] \
+                     [-S SpecialistLoginNames] [-t DatasetList] \
+                     [-T MinDataset [MaxDataset]] [-y YearList] \
+                     [-H Title] [-L Delimiter] [-U SizeUnit] \
+                     [-w] [-z] [> OutputFileName] [| lp -d PrinterName] 
+
+      Specify [-C] ColumnNames, refer to Option -C section for detail
+      description, and choose at least one of the condition options, -a, -c,
+      -d, -D, -e, -E, -g, -i, -I, -m, -N, -o, -q, -S -t, -T, and -y, to
+      run this application.
+
+      For all condition options, except option -a, an '!' sign can be added
+      between an option flag and its option values to get an excluding
+      condition. For example, choose '-o ! OrganizationTypes' to gather order
+      data usage by users from organization types other than the ones given in
+      OrganizationTypes.  Refer to the example given at the end of this help
+      document for how to select excluding condition.
+
+      String condition options, -c, -e, -g, -i, -o, -S, and -t, allow wildcard
+      inputs. '%' matches any number of characters and '_' matches any one
+      character.  Refer to the example given at the end of this help document
+      for how to use wildcard for string condition options.
+
+      Output of this application is defaulted to page format with a page
+      header on each page. A page header includes main title, sub titles and
+      column titles according to which column names and options are selected,
+      as well as page number and report date. If the output is used directly
+      for input of other applications, add option -w to remove page header
+      and show only the column titles and the usage information.
+
+
+      Column Options:
+      - Option -C, the ColumnNames must be present to run this application.
+           The flag -C can be omitted if it is the first parameter option on
+           the command line. The ColumnNames is a string that includes column
+           names listed below:
+
+         COLUMN - COLUMN       - COLUMN
+           NAME - TITLE        - DESCRIPTION 
+          GroupColumns:
+              D*- DATE         - format as YYYY-MM-DD, for example 2004-04-25
+              E*- EMAIL        - user email address
+              G*- AGGREID      - Aggrenation IDs
+              I*- IP           - user IP address
+              M*- MONTH        - format as YYYY-MM, for example 2004-04
+              N*- COUNTRY      - country codes users from
+              O*- ORGTYPE      - organization types (DSS, NCAR, UNIV and OTHER)
+              P*- DSOWNER      - login names of specialists who own the datasets
+              Q*- QUARTER      - quarter of year, 1, 2, 3, or 4
+              R*- DSTITLE      - dataset titles
+              S - BSIZE        - size of data read each time, default to Bytes
+              T*- DATASET      - format as dsnnn.n, for example d540001
+              Y*- YEAR         - format as YYYY, for example 2004
+
+                * - field names can processed with zero usages
+          SummaryColumns:
+              A - DSCOUNT      - number of datasets in given GroupColumns
+              B - MBREAD       - data sizes, default MB, read by given GroupColumns
+              C - #AGGRRE      - number of unique aggregation IDs in given GroupColumns
+              H - #ACCESS      - number of accesses by given GroupColumns
+              J - #UNIQUSER    - number of unique users in in given GroupColumns
+              U - #UNIQIP      - number of unique users in in given GroupColumns
+ 
+          IndexColumn:
+              X - INDEX        - index of line, it should be the first column
+
+           The column names are used to build up string of ColumnNames, while
+           their associated column titles are shown in view/report output of
+           this application. The display order of the column titles is
+           determined by the order of the column names in the ColumnNames
+           string. At least one of the group and summary columns must be
+           selected, in the ColumnNames string, to generate all usage
+           view/report;
+
+           For example, choose '-C EMB' to display column titles of EMAIL,
+           MONTH and MBREAD, in the first, second and third columns
+           respectively, for numbers of MBytes of data read by each user
+           in each month;
+
+      - Option -O, sort data usage information in ascending or descending
+           order based on the column names specified in OrderColumnNames
+           string. These column names must be in the selected [-C]
+           ColumnNames string. If an column name is in upper case, its
+           associated column is sorted in ascending order, and a lower
+           case means sorting in descending order;
+
+
+      Condition Options:
+      - Option -a, for all usage in table 'codusage';
+
+      - Option -A, gives a row limit for querrying; 
+
+      - Option -c, for files read by users from given country codes;
+
+      - Option -d, for data read on given dates, in format YYYY-MM-DD;
+
+      - Option -D, for data read between two given dates, each date
+           is in format YYYY-MM-DD. Omit EndDate for no upper limit;
+
+      - Option -e, for data read by users with given email addresses;
+
+      - Option -E, for data read by users who have been notified 
+           data update of a specified dataset between two given dates,
+           each date is in format YYYY-MM-DD. Omit EndNoticeDate for
+           no upper limit;
+
+      - Option -g, for data read from specified Aggrenation IDs;
+
+      - Option -h, works with Option -e to include historical user emails
+           registered before;
+
+      - Option -i, for data read from machines with given IP addresses;
+ 
+      - Option -m, for data read in given months, in format YYYY-MM;
+ 
+      - Option -N, for files for numbers of read by each group between
+           MinNumberRead and MaxNumberRead. Omit MaxNumberRead for no
+           upper limit;
+
+      - Option -o, for data read by users from given orgnization types.
+           It defaults to -o ! DSS to exclude usage from DSS specialists;
+           Set it to ALL to include all orgnization types;
+
+      - Option -q, for data read in given quarters;
+
+      - Option -S, for login names of specialsts who own the datasets;
+
+      - Option -t, for data associating to given dataset names;
+
+      - Option -T, for data associating to datasets between 
+           MinDataset and MaxDataset. Omit MaxDataset for no upper limit.
+           For example, -T d540000 d550009, for datasets numbers d540000-d550009;
+
+      - Option -y, for data read in given years in format YYYY;
+
+
+      Miscellaneous Options:
+      - Option -w, view data usage in simple format without totals;
+
+      - Option -z, include datasets without without usage
+
+      - Option -H, use given report title to replace the default one;
+
+      - Option -L, use given delimiter for output, instead of defaulted spaces;
+
+      - Option -U, show data sizes in given unit SizeUnit [BKMG].
+           B - Byte, K - KiloBytes, M  - MegaByte, and G - GigaByte;
+
+      - Option > OutputFilename, redirect output into an output file,
+           for example, ordusage.out, instead of viewing on screen directly;
+
+      - Option | lp -d PrinterName, redirect output to printer of PrinterName.
+           Replace PrinterName with lj100 to print through DSS LaserJet printer.
+
+
+      For example:
+      To view annual data usage in year 2005 with columns, INDEX(X),
+      EMAIL(E), ORGTYPE(O), #AGGRE(C), and MBREAD(B); ordered by ORGTYPE as
+      ascending and MBREAD(B) as descending; the command line should be:
+
+         viewcodusage XEOCB -y 2005 -O Ob
+
+      For usage by users not in Organization 'DDS', out of the file usage
+      gathered above, the command line should be:
+
+         viewcodusage XEOCB -y 2005 -o ! DSS -O Ob
+
+      To redirect the previous output to a file named codusage.out:
+
+         viewcodusage XEOCB -y 2005 ! DSS -O Ob > codusage.out
+
+      Then you can view the file or print it as a report.
+