CSIT-1078: Optimize input data files download and processing

Change-Id: Id1b0af996c4b77878621cfc8c262795f51e537fc
Signed-off-by: Tibor Frank <tifrank@cisco.com>

1 files changed, 46 insertions, 29 deletions

diff --git a/resources/tools/presentation/input_data_parser.py b/resources/tools/presentation/input_data_parser.py
index 0ad07a95db..e12e2fb8df 100644
--- a/resources/tools/presentation/input_data_parser.py
+++ b/resources/tools/presentation/input_data_parser.py
@@ -29,6 +29,8 @@ from collections import OrderedDict
 from string import replace
 from os import remove
 
+from input_data_files import download_and_unzip_data_file
+
 
 class ExecutionChecker(ResultVisitor):
     """Class to traverse through the test suite structure.
@@ -179,7 +181,7 @@ class ExecutionChecker(ResultVisitor):
     REGEX_MRR = re.compile(r'MaxReceivedRate_Results\s\[pkts/(\d*)sec\]:\s'
                            r'tx\s(\d*),\srx\s(\d*)')
 
-    def __init__(self, **metadata):
+    def __init__(self, metadata):
         """Initialisation.
 
         :param metadata: Key-value pairs to be included in "metadata" part of
@@ -251,8 +253,6 @@ class ExecutionChecker(ResultVisitor):
             self._data["metadata"]["version"] = self._version
             self._msg_type = None
 
-            logging.info("    VPP version: {0}".format(self._version))
-
     def _get_vat_history(self, msg):
         """Called when extraction of VAT command history is required.
 
@@ -748,21 +748,29 @@ class InputData(object):
         return self.data[job][build]["tests"]
 
     @staticmethod
-    def _parse_tests(job, build):
+    def _parse_tests(job, build, get_timestamp=False):
         """Process data from robot output.xml file and return JSON structured
         data.
 
         :param job: The name of job which build output data will be processed.
         :param build: The build which output data will be processed.
+        :param get_timestamp: If True, timestamp is read form the xml source
+            file.
         :type job: str
         :type build: dict
+        :type get_timestamp: bool
         :returns: JSON data structure.
         :rtype: dict
         """
 
-        tree = ET.parse(build["file-name"])
-        root = tree.getroot()
-        generated = root.attrib["generated"]
+        metadata = {
+            "job": job,
+            "build": build
+        }
+        if get_timestamp:
+            tree = ET.parse(build["file-name"])
+            root = tree.getroot()
+            metadata["generated"] = root.attrib["generated"]
 
         with open(build["file-name"], 'r') as data_file:
             try:
@@ -771,46 +779,55 @@ class InputData(object):
                 logging.error("Error occurred while parsing output.xml: {0}".
                               format(err))
                 return None
-        checker = ExecutionChecker(job=job, build=build, generated=generated)
+        checker = ExecutionChecker(metadata)
         result.visit(checker)
 
         return checker.data
 
-    def read_data(self):
-        """Parse input data from input files and store in pandas' Series.
+    def download_and_parse_data(self, get_timestamp=False):
+        """Download the input data files, parse input data from input files and
+        store in pandas' Series.
+
+        :param get_timestamp: If True, timestamp is read form the xml source
+            file.
+        :type get_timestamp: bool
         """
 
-        logging.info("Parsing input files ...")
+        logging.info("Downloading and parsing input files ...")
 
         job_data = dict()
         for job, builds in self._cfg.builds.items():
-            logging.info("  Extracting data from the job '{0}' ...'".
+            logging.info("  Processing data from the job '{0}' ...'".
                          format(job))
             builds_data = dict()
             for build in builds:
-                if build["status"] == "failed" \
-                        or build["status"] == "not found":
+                logging.info("    Processing the build '{0}'".
+                             format(build["build"]))
+                self._cfg.set_input_state(job, build["build"], "failed")
+                if not download_and_unzip_data_file(self._cfg, job, build):
+                    logging.error("It is not possible to download the input "
+                                  "data file from the job '{job}', build "
+                                  "'{build}', or it is damaged. Skipped.".
+                                  format(job=job, build=build["build"]))
                     continue
-                logging.info("    Extracting data from the build '{0}'".
+
+                logging.info("      Processing data from the build '{0}' ...".
                              format(build["build"]))
-                logging.info("    Processing the file '{0}'".
-                             format(build["file-name"]))
+                data = InputData._parse_tests(job, build,
+                                              get_timestamp=get_timestamp)
+                if data is None:
+                    logging.error("Input data file from the job '{job}', build "
+                                  "'{build}' is damaged. Skipped.".
+                                  format(job=job, build=build["build"]))
+                    continue
 
-                data = InputData._parse_tests(job, build)
+                self._cfg.set_input_state(job, build["build"], "processed")
 
-                logging.info("    Removing the file '{0}'".
-                             format(build["file-name"]))
                 try:
                     remove(build["file-name"])
-                    build["status"] = "processed"
                 except OSError as err:
-                    logging.error("   Cannot remove the file '{0}': {1}".
+                    logging.error("Cannot remove the file '{0}': {1}".
                                   format(build["file-name"], err))
-                if data is None:
-                    logging.error("Input data file from the job '{job}', build "
-                                  "'{build}' is damaged. Skipped.".
-                                  format(job=job, build=build["build"]))
-                    continue
 
                 build_data = pd.Series({
                     "metadata": pd.Series(data["metadata"].values(),
@@ -818,9 +835,9 @@ class InputData(object):
                     "suites": pd.Series(data["suites"].values(),
                                         index=data["suites"].keys()),
                     "tests": pd.Series(data["tests"].values(),
-                                       index=data["tests"].keys()),
-                    })
+                                       index=data["tests"].keys())})
                 builds_data[str(build["build"])] = build_data
+                build["status"] = "processed"
                 logging.info("    Done.")
 
             job_data[job] = pd.Series(builds_data.values(),