Add combine --keep (#1110)

* Add combine --keep

Related to https://github.com/nedbat/coveragepy/issues/1108

* Fix unit tests

* Fix line too long

* Fix line too long

8 files changed, 51 insertions, 14 deletions

diff --git a/coverage/cmdline.py b/coverage/cmdline.py
index 9c9ae868..11bc5d99 100644
--- a/coverage/cmdline.py
+++ b/coverage/cmdline.py
@@ -31,6 +31,10 @@ class Opts(object):
         '-a', '--append', action='store_true',
         help="Append coverage data to .coverage, otherwise it starts clean each time.",
     )
+    keep = optparse.make_option(
+        '', '--keep', action='store_true',
+        help="Keep combined coverage files, otherwise they are deleted.",
+    )
     branch = optparse.make_option(
         '', '--branch', action='store_true',
         help="Measure branch coverage in addition to statement coverage.",
@@ -215,6 +219,7 @@ class CoverageOptionParser(optparse.OptionParser, object):
             help=None,
             ignore_errors=None,
             include=None,
+            keep=None,
             module=None,
             omit=None,
             contexts=None,
@@ -333,6 +338,7 @@ CMDS = {
         "combine",
         [
             Opts.append,
+            Opts.keep,
             ] + GLOBAL_ARGS,
         usage="[options] <path1> <path2> ... <pathN>",
         description=(
@@ -585,7 +591,7 @@ class CoverageScript(object):
             if options.append:
                 self.coverage.load()
             data_dirs = args or None
-            self.coverage.combine(data_dirs, strict=True)
+            self.coverage.combine(data_dirs, strict=True, keep=bool(options.keep))
             self.coverage.save()
             return OK
 
diff --git a/coverage/control.py b/coverage/control.py
index 8d129bcb..c952afcd 100644
--- a/coverage/control.py
+++ b/coverage/control.py
@@ -659,7 +659,7 @@ class Coverage(object):
         data = self.get_data()
         data.write()
 
-    def combine(self, data_paths=None, strict=False):
+    def combine(self, data_paths=None, strict=False, keep=False):
         """Combine together a number of similarly-named coverage data files.
 
         All coverage data files whose name starts with `data_file` (from the
@@ -674,6 +674,8 @@ class Coverage(object):
         If `strict` is true, then it is an error to attempt to combine when
         there are no data files to combine.
 
+        If `keep` is true, then combined data files won't be deleted.
+
         .. versionadded:: 4.0
             The `data_paths` parameter.
 
@@ -694,7 +696,8 @@ class Coverage(object):
                 for pattern in paths[1:]:
                     aliases.add(pattern, result)
 
-        combine_parallel_data(self._data, aliases=aliases, data_paths=data_paths, strict=strict)
+        combine_parallel_data(self._data,
+                        aliases=aliases, data_paths=data_paths, strict=strict, keep=keep)
 
     def get_data(self):
         """Get the collected data.
diff --git a/coverage/data.py b/coverage/data.py
index 82bf1d41..5dd1dfe3 100644
--- a/coverage/data.py
+++ b/coverage/data.py
@@ -52,7 +52,7 @@ def add_data_to_hash(data, filename, hasher):
     hasher.update(data.file_tracer(filename))
 
 
-def combine_parallel_data(data, aliases=None, data_paths=None, strict=False):
+def combine_parallel_data(data, aliases=None, data_paths=None, strict=False, keep=False):
     """Combine a number of data files together.
 
     Treat `data.filename` as a file prefix, and combine the data from all
@@ -68,7 +68,7 @@ def combine_parallel_data(data, aliases=None, data_paths=None, strict=False):
     If `data_paths` is not provided, then the directory portion of
     `data.filename` is used as the directory to search for data files.
 
-    Every data file found and combined is then deleted from disk. If a file
+    Unless `keep` is True every data file found and combined is then deleted from disk. If a file
     cannot be read, a warning will be issued, and the file will not be
     deleted.
 
@@ -116,9 +116,10 @@ def combine_parallel_data(data, aliases=None, data_paths=None, strict=False):
         else:
             data.update(new_data, aliases=aliases)
             files_combined += 1
-            if data._debug.should('dataio'):
-                data._debug.write("Deleting combined data file %r" % (f,))
-            file_be_gone(f)
+            if not keep:
+                if data._debug.should('dataio'):
+                    data._debug.write("Deleting combined data file %r" % (f,))
+                file_be_gone(f)
 
     if strict and not files_combined:
         raise CoverageException("No usable data files")
diff --git a/doc/help/combine.rst b/doc/help/combine.rst
index 35180cdd..c35d5b93 100644
--- a/doc/help/combine.rst
+++ b/doc/help/combine.rst
@@ -13,6 +13,7 @@
     Options:
       -a, --append     Append coverage data to .coverage, otherwise it starts
                        clean each time.
+      --keep           Keep combined coverage files, otherwise they are deleted.
       --debug=OPTS     Debug options, separated by commas. [env: COVERAGE_DEBUG]
       -h, --help       Get help on this command.
       --rcfile=RCFILE  Specify configuration file. By default '.coveragerc',
diff --git a/doc/python-coverage.1.txt b/doc/python-coverage.1.txt
index 0bbd44d0..10a8927a 100644
--- a/doc/python-coverage.1.txt
+++ b/doc/python-coverage.1.txt
@@ -109,6 +109,7 @@ COMMAND REFERENCE
     Combine data from multiple coverage files collected with ``run -p``.
     The combined results are written to a single file representing the
     union of the data.
+    Unless --keep is provided the combined coverage files are deleted.
 
     If `PATH` is specified, they are files or directories containing data to
     be combined.
@@ -119,6 +120,9 @@ COMMAND REFERENCE
         Append coverage data to .coverage, otherwise it starts clean each
         time.
 
+    \--keep
+        Keep combined coverage file.
+
 **debug** `TOPIC` ...
 
     Display information about the internals of coverage.py, for diagnosing
diff --git a/tests/test_api.py b/tests/test_api.py
index f8b7b4b2..62fd9ebc 100644
--- a/tests/test_api.py
+++ b/tests/test_api.py
@@ -442,7 +442,7 @@ class ApiTest(CoverageTest):
 
         cov2 = coverage.Coverage()
         with self.assertRaisesRegex(CoverageException, r"No data to combine"):
-            cov2.combine(strict=True)
+            cov2.combine(strict=True, keep=False)
 
         cov3 = coverage.Coverage()
         cov3.combine()
diff --git a/tests/test_cmdline.py b/tests/test_cmdline.py
index 374adb0d..3b11881b 100644
--- a/tests/test_cmdline.py
+++ b/tests/test_cmdline.py
@@ -218,20 +218,20 @@ class CmdLineTest(BaseCmdLineTest):
         # coverage combine with args
         self.cmd_executes("combine datadir1", """\
             cov = Coverage()
-            cov.combine(["datadir1"], strict=True)
+            cov.combine(["datadir1"], strict=True, keep=False)
             cov.save()
             """)
         # coverage combine, appending
         self.cmd_executes("combine --append datadir1", """\
             cov = Coverage()
             cov.load()
-            cov.combine(["datadir1"], strict=True)
+            cov.combine(["datadir1"], strict=True, keep=False)
             cov.save()
             """)
         # coverage combine without args
         self.cmd_executes("combine", """\
             cov = Coverage()
-            cov.combine(None, strict=True)
+            cov.combine(None, strict=True, keep=False)
             cov.save()
             """)
 
@@ -239,12 +239,12 @@ class CmdLineTest(BaseCmdLineTest):
         # https://github.com/nedbat/coveragepy/issues/385
         self.cmd_executes("combine --rcfile cov.ini", """\
             cov = Coverage(config_file='cov.ini')
-            cov.combine(None, strict=True)
+            cov.combine(None, strict=True, keep=False)
             cov.save()
             """)
         self.cmd_executes("combine --rcfile cov.ini data1 data2/more", """\
             cov = Coverage(config_file='cov.ini')
-            cov.combine(["data1", "data2/more"], strict=True)
+            cov.combine(["data1", "data2/more"], strict=True, keep=False)
             cov.save()
             """)
 
diff --git a/tests/test_process.py b/tests/test_process.py
index e4886156..b44147a3 100644
--- a/tests/test_process.py
+++ b/tests/test_process.py
@@ -244,6 +244,28 @@ class ProcessTest(CoverageTest):
         data.read()
         self.assertEqual(line_counts(data)['b_or_c.py'], 7)
 
+    def test_combine_parallel_data_keep(self):
+        self.make_b_or_c_py()
+        out = self.run_command("coverage run -p b_or_c.py b")
+        self.assertEqual(out, 'done\n')
+        self.assert_doesnt_exist(".coverage")
+        self.assert_file_count(".coverage.*", 1)
+
+        out = self.run_command("coverage run -p b_or_c.py c")
+        self.assertEqual(out, 'done\n')
+        self.assert_doesnt_exist(".coverage")
+
+        # After two -p runs, there should be two .coverage.machine.123 files.
+        self.assert_file_count(".coverage.*", 2)
+
+        # Combine the parallel coverage data files into .coverage with the keep flag.
+        self.run_command("coverage combine --keep")
+
+        # After combining, the .coverage file & the original combined file should still be there.
+        self.assert_exists(".coverage")
+        self.assert_file_count(".coverage.*", 2)
+
+
     def test_append_data(self):
         self.make_b_or_c_py()