Merge pull request #88 from pitthsls/beda42-master

Beda42 master

pycounter/constants.py

@@ -174,6 +174,12 @@
         u"Access denied category",
         u"Reporting Period Total",
     ),
+    "PR1": (
+        u"Platform",
+        u"Publisher",
+        u"User Activity",
+        u"Reporting Period Total",
+    ),
     # FIXME: this is outputting counter 5 reports in 4 format for... reasons.
     "TR_J1": (
         u"Journal",

pycounter/report.py

@@ -548,6 +548,45 @@ def as_generic(self):
         return data_line
 
 
+class CounterPlatform(CounterEresource):
+    """a COUNTER platform report line."""
+
+    def __init__(
+        self,
+        period=None,
+        metric=None,
+        month_data=None,
+        platform="",
+        publisher="",
+    ):
+        super(CounterPlatform, self).__init__(
+            period=period,
+            metric=metric,
+            month_data=month_data,
+            title='',  # no title for platform report
+            platform=platform,
+            publisher=publisher
+        )
+        self.isbn = None
+
+    def as_generic(self):
+        """Return data for this line as list of COUNTER report cells."""
+        self._fill_months()
+
+        data_line = [self.platform, self.publisher, self.metric]
+        total_usage = 0
+        month_data = []
+
+        for data in self:
+            total_usage += data[2]
+            month_data.append(six.text_type(data[2]))
+
+        data_line.append(six.text_type(total_usage))
+        data_line.extend(month_data)
+
+        return data_line
+
+
 def parse(filename, filetype=None, encoding="utf-8", fallback_encoding="latin-1"):
     """Parse a COUNTER file, first attempting to determine type.
 
@@ -719,10 +758,12 @@ def parse_generic(report_reader):
         end_date = last_day(convert_date_column(header[last_col - 1]))
         report.period = (start_date, end_date)
 
-    if report.report_type != "DB1" and report.report_version != 5:
+    if report.report_type not in ("DB1", "PR1") and report.report_version != 5:
+        # these reports do not have line with totals
         six.next(report_reader)
 
     if report.report_type == "DB2":
+        # this report has two lines of totals
         six.next(report_reader)
 
     for line in report_reader:
@@ -784,7 +825,8 @@ def _parse_line(line, report, last_col):
     }
     month_data = []
     curr_month = report.period[0]
-    for data in line[5:]:
+    months_start_idx = 5 if report.report_type != 'PR1' else 4
+    for data in line[months_start_idx:]:
         month_data.append((curr_month, format_stat(data)))
         curr_month = next_month(curr_month)
     if report.report_type.startswith("JR") or report.report_type == "TR_J1":
@@ -811,6 +853,10 @@ def _parse_line(line, report, last_col):
         )
     elif report.report_type.startswith("DB"):
         return CounterDatabase(metric=line[3], month_data=month_data, **common_args)
+    elif report.report_type == "PR1":
+        # there is no title in the PR1 report
+        return CounterPlatform(metric=line[2], month_data=month_data, platform=line[0],
+                               publisher=line[1], period=report.period)
     raise PycounterException("Should be unreachable")  # pragma: no cover
 
 
@@ -829,7 +875,7 @@ def _get_type_and_version(specifier):
         report_version = int(rt_match.group(3))
     else:
         raise UnknownReportTypeError("No match in line: %s" % specifier)
-    if not any(report_type.startswith(x) for x in ("JR", "BR", "DB")):
+    if not any(report_type.startswith(x) for x in ("JR", "BR", "DB", "PR1")):
         raise UnknownReportTypeError(report_type)
 
     return report_type, report_version
@@ -851,6 +897,8 @@ def _year_from_header(header, report):
         first_date_col = 8
     elif report.report_type in ("DB1", "DB2") and report.report_version == 4:
         first_date_col = 5
+    elif report.report_type == "PR1" and report.report_version == 4:
+        first_date_col = 4
     year = int(header[first_date_col].split("-")[1])
     if year < 100:
         year += 2000

pycounter/sushi.py

@@ -289,7 +289,8 @@ def raw_to_full(raw_report):
                         pdf_usage += int(inst.Count)
                     elif inst.MetricType == "ft_html":
                         html_usage += int(inst.Count)
-                    elif report.report_type.startswith("DB"):
+                    elif report.report_type.startswith("DB") or \
+                            report.report_type == "PR1":
                         metrics_for_db[inst.MetricType].append(
                             (item_date, int(inst.Count))
                         )
@@ -343,5 +344,16 @@ def raw_to_full(raw_report):
                             month_data=month_data,
                         )
                     )
-
+            elif report.report_type == "PR1":
+                for metric_code, month_data in six.iteritems(metrics_for_db):
+                    metric = pycounter.constants.DB_METRIC_MAP[metric_code]
+                    report.pubs.append(
+                        pycounter.report.CounterPlatform(
+                            platform=platform,
+                            publisher=publisher_name,
+                            period=report.period,
+                            metric=metric,
+                            month_data=month_data,
+                        )
+                    )
     return report

pycounter/test/data/PR1.tsv

@@ -0,0 +1,12 @@
+Platform Report 1 (R4)	Total Searches, Result Clicks, and Record Views by Month and Platform
+A fake institution used for testing
+00000001
+Period covered by Report:
+2018-01-01 to 2018-12-31
+Date run:
+2019-08-01
+Platform	Publisher	User Activity	Reporting Period Total	Jan-2018	Feb-2018	Mar-2018	Apr-2018	May-2018	Jun-2018	Jul-2018	Aug-2018	Sep-2018	Oct-2018	Nov-2018	Dec-2018
+search.alexanderstreet.com	Alexander Street Press	Regular Searches	299	91	41	13	21	44	8	0	0	36	36	7	2
+search.alexanderstreet.com	Alexander Street Press	Searches-federated and automated	0	0	0	0	0	0	0	0	0	0	0	0	0
+search.alexanderstreet.com	Alexander Street Press	Result Clicks	87	19	18	1	8	13	5	0	0	13	6	2	2
+search.alexanderstreet.com	Alexander Street Press	Record Views	100	22	19	1	9	19	5	0	0	15	6	2	2

pycounter/test/test_bad_reports.py

@@ -10,7 +10,7 @@
 
 @pytest.mark.parametrize(
     "report_type",
-    [u"Bogus Report 7 (R4)", u"Platform Report 1 (R4)"],  # unsupported but valid
+    [u"Bogus Report 7 (R4)"],
 )
 def test_report_type(report_type):
     """Report type doesn't exist."""

pycounter/test/test_pr1.py

@@ -0,0 +1,33 @@
+"""Test COUNTER DB1 database report"""
+
+from __future__ import absolute_import
+
+import os
+import unittest
+
+import pycounter.report
+
+
+class ParseCounter4Example(unittest.TestCase):
+    """Tests for parsing C4 PR1"""
+
+    def setUp(self):
+        self.report = pycounter.report.parse(
+            os.path.join(os.path.dirname(__file__), "data/PR1.tsv")
+        )
+
+    def test_reportname(self):
+        self.assertEqual(self.report.report_type, u"PR1")
+
+    def test_stats(self):
+        publication = self.report.pubs[0]
+        self.assertEqual([x[2] for x in publication],
+                         [91, 41, 13, 21, 44, 8, 0, 0, 36, 36, 7, 2])
+
+    def test_row_metric(self):
+        # test metric of the first row
+        jan_data = next(iter(self.report.pubs[0]))
+        self.assertEqual(jan_data[1], "Regular Searches")
+        # test metric of the second row
+        jan_data = next(iter(self.report.pubs[1]))
+        self.assertEqual(jan_data[1], "Searches-federated and automated")

pylintrc

@@ -41,3 +41,6 @@ ignored-argument-names=_.*|^ignored_|^unused_|.*_unused
 defining-attr-methods=__init__,
                       __new__,
                       setUp
+
+[DESIGN]
+max-locals=16