Merge branch 'run-id' into '0.7.x'

docs/packing.rst

@@ -45,7 +45,8 @@ The first part of the configuration file gives general information with respect
     version: <reprozip-version>
     runs:
     # Run 0
-    - architecture: <machine-architecture>
+    - id: <run-id>
+      architecture: <machine-architecture>
       argv: <command-line-arguments>
       binary: <command-line-binary>
       distribution: <linux-distribution>
@@ -58,9 +59,10 @@ The first part of the configuration file gives general information with respect
       workingdir: <working-directory>
       
     # Run 1
-    - ...
+    - id: ...
+    ...
 
-If necessary, users may change command line parameters by editing ``argv``, and add or remove environment variables by editing ``environ``. Other attributes should not be changed in general.
+If necessary, users may change command line parameters by editing ``argv``, and add or remove environment variables by editing ``environ``. Users may also give a more meaningful and user-friendly identifier for a run by changing ``id``. Other attributes should not be changed in general.
 
 The next section brings information about input and output files, including their original paths and which runs read and/or wrote them. These are the files that `reprozip` identified as the main input or result of the experiment, which `reprounzip` will later be able to replace and extract from the experiment when reproducing it. You may add, remove, or edit these files in case `reprozip` fails in recognizing any important information, as well as give meaningful names to them by editing ``name``::
 

reprounzip/reprounzip/common.py

@@ -114,6 +114,7 @@ def read_packages(packages, File=File, Package=Package):
 # 0.5: no change
 # 0.6: no change
 # 0.7: moves input_files and output_files from run to global scope
+# 0.8: adds 'id' field to run
 
 
 Config = optional_return_type(['runs', 'packages', 'other_files'],
@@ -214,8 +215,8 @@ def load_config(filename, canonical, File=File, Package=Package):
     keys_ = set(config)
     if 'version' not in keys_:
         raise InvalidConfig("Missing version")
-    # Accepts versions from 0.2 to 0.7 inclusive
-    elif not LooseVersion('0.2') <= ver < LooseVersion('0.8'):
+    # Accepts versions from 0.2 to 0.8 inclusive
+    elif not LooseVersion('0.2') <= ver < LooseVersion('0.9'):
         pkgname = (__package__ or __name__).split('.', 1)[0]
         raise InvalidConfig("Loading configuration file in unknown format %s; "
                             "this probably means that you should upgrade "
@@ -245,6 +246,11 @@ def load_config(filename, canonical, File=File, Package=Package):
                                    for n, f in iteritems(inputs_outputs)
                                    if i in f.write_runs)
 
+    # reprozip < 0.8 compatibility: assign IDs to runs
+    for i, run in enumerate(runs):
+        if run.get('id') is None:
+            run['id'] = "run%d" % i
+
     record_usage_package(runs, packages, other_files,
                          inputs_outputs,
                          pack_id=config.get('pack_id'))
@@ -312,7 +318,7 @@ def save_config(filename, runs, packages, other_files, reprozip_version,
 """.format(pack_id=(('\npack_id: "%s"' % pack_id) if pack_id is not None
                     else ''),
            version=escape(reprozip_version),
-           format='0.7',
+           format='0.8',
            date=datetime.now().isoformat(),
            what=("# It was generated by the packer and you shouldn't need to "
                  "edit it" if canonical

reprounzip/reprounzip/pack_info.py

@@ -129,10 +129,10 @@ def print_info(args):
         print("Runs (%d):" % len(runs))
         for i, run in enumerate(runs):
             cmdline = ' '.join(shell_escape(a) for a in run['argv'])
-            if len(runs) > 1:
-                print("    %d: %s" % (i, cmdline))
-            else:
+            if len(runs) == 1 and run['id'] == "run0":
                 print("    %s" % cmdline)
+            else:
+                print("    %s: %s" % (run['id'], cmdline))
             if args.verbosity >= 2:
                 print("        wd: %s" % run['workingdir'])
                 if 'signal' in run:

reprounzip/reprounzip/unpackers/common/misc.py

@@ -333,6 +333,9 @@ def get_runs(runs, selected_runs, cmdline):
     Will return an iterable of run numbers. Might also fail loudly or exit
     after printing the original command-line.
     """
+    name_map = dict((r['id'], i) for i, r in enumerate(runs) if 'id' in r)
+    run_list = []
+
     if selected_runs is None:
         if len(runs) == 1:
             selected_runs = '0'
@@ -345,41 +348,47 @@ def parse_run(s):
         try:
             r = int(s)
         except ValueError:
-            logging.critical("Error: Run is not a number")
+            logging.critical("Error: Unknown run %s", s)
             raise UsageError
         if r < 0 or r >= len(runs):
             logging.critical("Error: Expected 0 <= run <= %d, got %d",
                              len(runs) - 1, r)
             sys.exit(1)
         return r
 
-    sep = selected_runs.find('-')
-    if sep == -1:
-        selected_runs = parse_run(selected_runs),
-    else:
-        if sep > 0:
-            first = parse_run(selected_runs[:sep])
-        else:
-            first = 0
-        if sep + 1 < len(selected_runs):
-            last = parse_run(selected_runs[sep + 1:])
+    for run_item in selected_runs.split(','):
+        run_item = run_item.strip()
+        if run_item in name_map:
+            run_list.append(name_map[run_item])
+            continue
+
+        sep = run_item.find('-')
+        if sep == -1:
+            run_list.append(parse_run(run_item))
         else:
-            last = len(runs) - 1
-        if last <= first:
-            logging.critical("Error: Last run number should be greater than "
-                             "the first")
-            sys.exit(1)
-        selected_runs = irange(first, last + 1)
+            if sep > 0:
+                first = parse_run(run_item[:sep])
+            else:
+                first = 0
+            if sep + 1 < len(run_item):
+                last = parse_run(run_item[sep + 1:])
+            else:
+                last = len(runs) - 1
+            if last <= first:
+                logging.critical("Error: Last run number should be greater "
+                                 "than the first")
+                sys.exit(1)
+            run_list.extend(irange(first, last + 1))
 
     # --cmdline without arguments: display the original command-line
     if cmdline == []:
         print("Original command-lines:")
-        for run in selected_runs:
+        for run in run_list:
             print(' '.join(shell_escape(arg)
                            for arg in runs[run]['argv']))
         sys.exit(0)
 
-    return selected_runs
+    return run_list
 
 
 def interruptible_call(*args, **kwargs):

reprozip/reprozip/common.py

@@ -114,6 +114,7 @@ def read_packages(packages, File=File, Package=Package):
 # 0.5: no change
 # 0.6: no change
 # 0.7: moves input_files and output_files from run to global scope
+# 0.8: adds 'id' field to run
 
 
 Config = optional_return_type(['runs', 'packages', 'other_files'],
@@ -214,8 +215,8 @@ def load_config(filename, canonical, File=File, Package=Package):
     keys_ = set(config)
     if 'version' not in keys_:
         raise InvalidConfig("Missing version")
-    # Accepts versions from 0.2 to 0.7 inclusive
-    elif not LooseVersion('0.2') <= ver < LooseVersion('0.8'):
+    # Accepts versions from 0.2 to 0.8 inclusive
+    elif not LooseVersion('0.2') <= ver < LooseVersion('0.9'):
         pkgname = (__package__ or __name__).split('.', 1)[0]
         raise InvalidConfig("Loading configuration file in unknown format %s; "
                             "this probably means that you should upgrade "
@@ -245,6 +246,11 @@ def load_config(filename, canonical, File=File, Package=Package):
                                    for n, f in iteritems(inputs_outputs)
                                    if i in f.write_runs)
 
+    # reprozip < 0.8 compatibility: assign IDs to runs
+    for i, run in enumerate(runs):
+        if run.get('id') is None:
+            run['id'] = "run%d" % i
+
     record_usage_package(runs, packages, other_files,
                          inputs_outputs,
                          pack_id=config.get('pack_id'))
@@ -312,7 +318,7 @@ def save_config(filename, runs, packages, other_files, reprozip_version,
 """.format(pack_id=(('\npack_id: "%s"' % pack_id) if pack_id is not None
                     else ''),
            version=escape(reprozip_version),
-           format='0.7',
+           format='0.8',
            date=datetime.now().isoformat(),
            what=("# It was generated by the packer and you shouldn't need to "
                  "edit it" if canonical

reprozip/reprozip/tracer/trace.py

@@ -360,7 +360,8 @@ def write_configuration(directory, sort_packages, find_inputs_outputs,
             envp = envp[:-1]
         environ = dict(v.split('=', 1) for v in envp)
 
-        runs.append({'binary': r_name, 'argv': argv,
+        runs.append({'id': "run%d" % len(runs),
+                     'binary': r_name, 'argv': argv,
                      'workingdir': unicode_(Path(r_workingdir)),
                      'architecture': platform.machine().lower(),
                      'distribution': distribution,

tests/test_unpackers_common.py

@@ -32,7 +32,9 @@ def test_make_unique_name(self):
 class TestMisc(unittest.TestCase):
     def do_ok(self, arg, expected):
         try:
-            runs = get_runs([{}, {}, {}, {}], arg, None)
+            runs = get_runs(
+                [{'id': 'one'}, {'id': 'two-heh'}, {'id': 'three'}, {}],
+                arg, None)
         except (SystemExit, UsageError):
             self.fail("get_runs(<4 runs>, %r) raised" % arg)
         self.assertEqual(list(runs), expected)
@@ -60,5 +62,10 @@ def test_get_runs(self):
             self.do_fail('0-8')
             self.do_fail('0-4')
             self.do_ok('0-3', [0, 1, 2, 3])
+            self.do_ok('one', [0])
+            self.do_ok('two-heh', [1]),
+            self.do_ok('one,three', [0, 2])
+            self.do_ok('1,three', [1, 2])
+            self.do_ok('2-3,two-heh', [2, 3, 1])
         finally:
             print(">>>>> get_runs tests", file=sys.stderr)