get_maintainer: add commit author information to --rolestats

get_maintainer currently uses "Signed-off-by" style lines to find interested parties to send patches to when the MAINTAINERS file does not have a specific section entry with a matching file pattern. Add statistics for commit authors and lines added and deleted to the information provided by --rolestats. These statistics are also emitted whenever --rolestats and --git are selected even when there is a specified maintainer. This can have the effect of expanding the number of people that are shown as possible "maintainers" of a particular file because "authors", "added_lines", and "removed_lines" are also used as criterion for the --max-maintainers option separate from the "commit_signers". The first "--git-max-maintainers" values of each criterion are emitted. Any "ties" are not shown. For example: (forcedeth does not have a named maintainer) Old output: $ ./scripts/get_maintainer.pl -f drivers/net/ethernet/nvidia/forcedeth.c "David S. Miller" <davem@davemloft.net> (commit_signer:8/10=80%) Jiri Pirko <jiri@resnulli.us> (commit_signer:2/10=20%) Patrick McHardy <kaber@trash.net> (commit_signer:2/10=20%) Larry Finger <Larry.Finger@lwfinger.net> (commit_signer:1/10=10%) Peter Zijlstra <peterz@infradead.org> (commit_signer:1/10=10%) netdev@vger.kernel.org (open list:NETWORKING DRIVERS) linux-kernel@vger.kernel.org (open list) New output: $ ./scripts/get_maintainer.pl -f drivers/net/ethernet/nvidia/forcedeth.c "David S. Miller" <davem@davemloft.net> (commit_signer:8/10=80%) Jiri Pirko <jiri@resnulli.us> (commit_signer:2/10=20%,authored:2/10=20%,removed_lines:3/33=9%) Patrick McHardy <kaber@trash.net> (commit_signer:2/10=20%,authored:2/10=20%,added_lines:12/95=13%,removed_lines:10/33=30%) Larry Finger <Larry.Finger@lwfinger.net> (commit_signer:1/10=10%,authored:1/10=10%,added_lines:35/95=37%) Peter Zijlstra <peterz@infradead.org> (commit_signer:1/10=10%) "Peter Hüwe" <PeterHuewe@gmx.de> (authored:1/10=10%,removed_lines:15/33=45%) Joe Perches <joe@perches.com> (authored:1/10=10%) Neil Horman <nhorman@tuxdriver.com> (added_lines:40/95=42%) Bill Pemberton <wfp5p@virginia.edu> (removed_lines:3/33=9%) netdev@vger.kernel.org (open list:NETWORKING DRIVERS) linux-kernel@vger.kernel.org (open list) Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

get_maintainer: add commit author information to --rolestats
get_maintainer currently uses "Signed-off-by" style lines to find interested parties to send patches to when the MAINTAINERS file does not have a specific section entry with a matching file pattern. Add statistics for commit authors and lines added and deleted to the information provided by --rolestats. These statistics are also emitted whenever --rolestats and --git are selected even when there is a specified maintainer. This can have the effect of expanding the number of people that are shown as possible "maintainers" of a particular file because "authors", "added_lines", and "removed_lines" are also used as criterion for the --max-maintainers option separate from the "commit_signers". The first "--git-max-maintainers" values of each criterion are emitted. Any "ties" are not shown. For example: (forcedeth does not have a named maintainer) Old output: $ ./scripts/get_maintainer.pl -f drivers/net/ethernet/nvidia/forcedeth.c "David S. Miller" <davem@davemloft.net> (commit_signer:8/10=80%) Jiri Pirko <jiri@resnulli.us> (commit_signer:2/10=20%) Patrick McHardy <kaber@trash.net> (commit_signer:2/10=20%) Larry Finger <Larry.Finger@lwfinger.net> (commit_signer:1/10=10%) Peter Zijlstra <peterz@infradead.org> (commit_signer:1/10=10%) netdev@vger.kernel.org (open list:NETWORKING DRIVERS) linux-kernel@vger.kernel.org (open list) New output: $ ./scripts/get_maintainer.pl -f drivers/net/ethernet/nvidia/forcedeth.c "David S. Miller" <davem@davemloft.net> (commit_signer:8/10=80%) Jiri Pirko <jiri@resnulli.us> (commit_signer:2/10=20%,authored:2/10=20%,removed_lines:3/33=9%) Patrick McHardy <kaber@trash.net> (commit_signer:2/10=20%,authored:2/10=20%,added_lines:12/95=13%,removed_lines:10/33=30%) Larry Finger <Larry.Finger@lwfinger.net> (commit_signer:1/10=10%,authored:1/10=10%,added_lines:35/95=37%) Peter Zijlstra <peterz@infradead.org> (commit_signer:1/10=10%) "Peter Hüwe" <PeterHuewe@gmx.de> (authored:1/10=10%,removed_lines:15/33=45%) Joe Perches <joe@perches.com> (authored:1/10=10%) Neil Horman <nhorman@tuxdriver.com> (added_lines:40/95=42%) Bill Pemberton <wfp5p@virginia.edu> (removed_lines:3/33=9%) netdev@vger.kernel.org (open list:NETWORKING DRIVERS) linux-kernel@vger.kernel.org (open list) Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Joe Perches · Linus Torvalds
1 parent 1d3fa37034
Showing 1 changed file with 85 additions and 6 deletions Side-by-side Diff
scripts/get_maintainer.pl
@@ -98,6 +98,7 @@
     "available" => '(which("git") ne "") && (-d ".git")',
     "find_signers_cmd" =>
 	"git log --no-color --follow --since=\$email_git_since " .
+	    '--numstat --no-merges ' .
 	    '--format="GitCommit: %H%n' .
 		      'GitAuthor: %an <%ae>%n' .
 		      'GitDate: %aD%n' .
@@ -106,6 +107,7 @@
 	    " -- \$file",
     "find_commit_signers_cmd" =>
 	"git log --no-color " .
+	    '--numstat ' .
 	    '--format="GitCommit: %H%n' .
 		      'GitAuthor: %an <%ae>%n' .
 		      'GitDate: %aD%n' .
@@ -114,6 +116,7 @@
 	    " -1 \$commit",
     "find_commit_author_cmd" =>
 	"git log --no-color " .
+	    '--numstat ' .
 	    '--format="GitCommit: %H%n' .
 		      'GitAuthor: %an <%ae>%n' .
 		      'GitDate: %aD%n' .
@@ -125,6 +128,7 @@
     "blame_commit_pattern" => "^([0-9a-f]+) ",
     "author_pattern" => "^GitAuthor: (.*)",
     "subject_pattern" => "^GitSubject: (.*)",
+    "stat_pattern" => "^(\\d+)\\t(\\d+)\\t\$file\$",
 );
  
 my %VCS_cmds_hg = (
@@ -152,6 +156,7 @@
     "blame_commit_pattern" => "^([ 0-9a-f]+):",
     "author_pattern" => "^HgAuthor: (.*)",
     "subject_pattern" => "^HgSubject: (.*)",
+    "stat_pattern" => "^(\\d+)\t(\\d+)\t\$file\$",
 );
  
 my $conf = which_conf(".get_maintainer.conf");
  
  
  
  
  
  
  
@@ -1269,21 +1274,31 @@
 }
  
 sub vcs_find_signers {
-    my ($cmd) = @_;
+    my ($cmd, $file) = @_;
     my $commits;
     my @lines = ();
     my @signatures = ();
+    my @authors = ();
+    my @stats = ();
  
     @lines = &{$VCS_cmds{"execute_cmd"}}($cmd);
  
     my $pattern = $VCS_cmds{"commit_pattern"};
+    my $author_pattern = $VCS_cmds{"author_pattern"};
+    my $stat_pattern = $VCS_cmds{"stat_pattern"};
  
+    $stat_pattern =~ s/(\$\w+)/$1/eeg;		#interpolate $stat_pattern
+
     $commits = grep(/$pattern/, @lines);	# of commits
  
+    @authors = grep(/$author_pattern/, @lines);
     @signatures = grep(/^[ \t]*${signature_pattern}.*\@.*$/, @lines);
+    @stats = grep(/$stat_pattern/, @lines);
  
-    return (0, @signatures) if !@signatures;
+#    print("stats: <@stats>\n");
  
+    return (0, \@signatures, \@authors, \@stats) if !@signatures;
+
     save_commits_by_author(@lines) if ($interactive);
     save_commits_by_signer(@lines) if ($interactive);
  
  
@@ -1291,9 +1306,10 @@
 	@signatures = grep(!/${penguin_chiefs}/i, @signatures);
     }
  
+    my ($author_ref, $authors_ref) = extract_formatted_signatures(@authors);
     my ($types_ref, $signers_ref) = extract_formatted_signatures(@signatures);
  
-    return ($commits, @$signers_ref);
+    return ($commits, $signers_ref, $authors_ref, \@stats);
 }
  
 sub vcs_find_author {
  
@@ -1849,7 +1865,12 @@
 sub vcs_file_signoffs {
     my ($file) = @_;
  
+    my $authors_ref;
+    my $signers_ref;
+    my $stats_ref;
+    my @authors = ();
     my @signers = ();
+    my @stats = ();
     my $commits;
  
     $vcs_used = vcs_exists();
  
  
@@ -1858,13 +1879,59 @@
     my $cmd = $VCS_cmds{"find_signers_cmd"};
     $cmd =~ s/(\$\w+)/$1/eeg;		# interpolate $cmd
  
-    ($commits, @signers) = vcs_find_signers($cmd);
+    ($commits, $signers_ref, $authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
  
+    @signers = @{$signers_ref} if defined $signers_ref;
+    @authors = @{$authors_ref} if defined $authors_ref;
+    @stats = @{$stats_ref} if defined $stats_ref;
+
+#    print("commits: <$commits>\nsigners:<@signers>\nauthors: <@authors>\nstats: <@stats>\n");
+
     foreach my $signer (@signers) {
 	$signer = deduplicate_email($signer);
     }
  
     vcs_assign("commit_signer", $commits, @signers);
+    vcs_assign("authored", $commits, @authors);
+    if ($#authors == $#stats) {
+	my $stat_pattern = $VCS_cmds{"stat_pattern"};
+	$stat_pattern =~ s/(\$\w+)/$1/eeg;	#interpolate $stat_pattern
+
+	my $added = 0;
+	my $deleted = 0;
+	for (my $i = 0; $i <= $#stats; $i++) {
+	    if ($stats[$i] =~ /$stat_pattern/) {
+		$added += $1;
+		$deleted += $2;
+	    }
+	}
+	my @tmp_authors = uniq(@authors);
+	foreach my $author (@tmp_authors) {
+	    $author = deduplicate_email($author);
+	}
+	@tmp_authors = uniq(@tmp_authors);
+	my @list_added = ();
+	my @list_deleted = ();
+	foreach my $author (@tmp_authors) {
+	    my $auth_added = 0;
+	    my $auth_deleted = 0;
+	    for (my $i = 0; $i <= $#stats; $i++) {
+		if ($author eq deduplicate_email($authors[$i]) &&
+		    $stats[$i] =~ /$stat_pattern/) {
+		    $auth_added += $1;
+		    $auth_deleted += $2;
+		}
+	    }
+	    for (my $i = 0; $i < $auth_added; $i++) {
+		push(@list_added, $author);
+	    }
+	    for (my $i = 0; $i < $auth_deleted; $i++) {
+		push(@list_deleted, $author);
+	    }
+	}
+	vcs_assign("added_lines", $added, @list_added);
+	vcs_assign("removed_lines", $deleted, @list_deleted);
+    }
 }
  
 sub vcs_file_blame {
@@ -1887,6 +1954,10 @@
     if ($email_git_blame_signatures) {
 	if (vcs_is_hg()) {
 	    my $commit_count;
+	    my $commit_authors_ref;
+	    my $commit_signers_ref;
+	    my $stats_ref;
+	    my @commit_authors = ();
 	    my @commit_signers = ();
 	    my $commit = join(" -r ", @commits);
 	    my $cmd;
  
  
@@ -1894,19 +1965,27 @@
 	    $cmd = $VCS_cmds{"find_commit_signers_cmd"};
 	    $cmd =~ s/(\$\w+)/$1/eeg;	#substitute variables in $cmd
  
-	    ($commit_count, @commit_signers) = vcs_find_signers($cmd);
+	    ($commit_count, $commit_signers_ref, $commit_authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
+	    @commit_authors = @{$commit_authors_ref} if defined $commit_authors_ref;
+	    @commit_signers = @{$commit_signers_ref} if defined $commit_signers_ref;
  
 	    push(@signers, @commit_signers);
 	} else {
 	    foreach my $commit (@commits) {
 		my $commit_count;
+		my $commit_authors_ref;
+		my $commit_signers_ref;
+		my $stats_ref;
+		my @commit_authors = ();
 		my @commit_signers = ();
 		my $cmd;
  
 		$cmd = $VCS_cmds{"find_commit_signers_cmd"};
 		$cmd =~ s/(\$\w+)/$1/eeg;	#substitute variables in $cmd
  
-		($commit_count, @commit_signers) = vcs_find_signers($cmd);
+		($commit_count, $commit_signers_ref, $commit_authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
+		@commit_authors = @{$commit_authors_ref} if defined $commit_authors_ref;
+		@commit_signers = @{$commit_signers_ref} if defined $commit_signers_ref;
  
 		push(@signers, @commit_signers);
 	    }
...	...	@@ -98,6 +98,7 @@
98	98	"available" => '(which("git") ne "") && (-d ".git")',
99	99	"find_signers_cmd" =>
100	100	"git log --no-color --follow --since=\$email_git_since " .
	101	+ '--numstat --no-merges ' .
101	102	'--format="GitCommit: %H%n' .
102	103	'GitAuthor: %an <%ae>%n' .
103	104	'GitDate: %aD%n' .
...	...	@@ -106,6 +107,7 @@
106	107	" -- \$file",
107	108	"find_commit_signers_cmd" =>
108	109	"git log --no-color " .
	110	+ '--numstat ' .
109	111	'--format="GitCommit: %H%n' .
110	112	'GitAuthor: %an <%ae>%n' .
111	113	'GitDate: %aD%n' .
...	...	@@ -114,6 +116,7 @@
114	116	" -1 \$commit",
115	117	"find_commit_author_cmd" =>
116	118	"git log --no-color " .
	119	+ '--numstat ' .
117	120	'--format="GitCommit: %H%n' .
118	121	'GitAuthor: %an <%ae>%n' .
119	122	'GitDate: %aD%n' .
...	...	@@ -125,6 +128,7 @@
125	128	"blame_commit_pattern" => "^([0-9a-f]+) ",
126	129	"author_pattern" => "^GitAuthor: (.*)",
127	130	"subject_pattern" => "^GitSubject: (.*)",
	131	+ "stat_pattern" => "^(\\d+)\\t(\\d+)\\t\$file\$",
128	132	);
129	133
130	134	my %VCS_cmds_hg = (
...	...	@@ -152,6 +156,7 @@
152	156	"blame_commit_pattern" => "^([ 0-9a-f]+):",
153	157	"author_pattern" => "^HgAuthor: (.*)",
154	158	"subject_pattern" => "^HgSubject: (.*)",
	159	+ "stat_pattern" => "^(\\d+)\t(\\d+)\t\$file\$",
155	160	);
156	161
157	162	my $conf = which_conf(".get_maintainer.conf");
158	163
159	164
160	165
161	166
162	167
163	168
164	169
...	...	@@ -1269,21 +1274,31 @@
1269	1274	}
1270	1275
1271	1276	sub vcs_find_signers {
1272		- my ($cmd) = @_;
	1277	+ my ($cmd, $file) = @_;
1273	1278	my $commits;
1274	1279	my @lines = ();
1275	1280	my @signatures = ();
	1281	+ my @authors = ();
	1282	+ my @stats = ();
1276	1283
1277	1284	@lines = &{$VCS_cmds{"execute_cmd"}}($cmd);
1278	1285
1279	1286	my $pattern = $VCS_cmds{"commit_pattern"};
	1287	+ my $author_pattern = $VCS_cmds{"author_pattern"};
	1288	+ my $stat_pattern = $VCS_cmds{"stat_pattern"};
1280	1289
	1290	+ $stat_pattern =~ s/(\$\w+)/$1/eeg; #interpolate $stat_pattern
	1291	+
1281	1292	$commits = grep(/$pattern/, @lines); # of commits
1282	1293
	1294	+ @authors = grep(/$author_pattern/, @lines);
1283	1295	@signatures = grep(/^[ \t]${signature_pattern}.\@.*$/, @lines);
	1296	+ @stats = grep(/$stat_pattern/, @lines);
1284	1297
1285		- return (0, @signatures) if !@signatures;
	1298	+# print("stats: <@stats>\n");
1286	1299
	1300	+ return (0, \@signatures, \@authors, \@stats) if !@signatures;
	1301	+
1287	1302	save_commits_by_author(@lines) if ($interactive);
1288	1303	save_commits_by_signer(@lines) if ($interactive);
1289	1304
1290	1305
...	...	@@ -1291,9 +1306,10 @@
1291	1306	@signatures = grep(!/${penguin_chiefs}/i, @signatures);
1292	1307	}
1293	1308
	1309	+ my ($author_ref, $authors_ref) = extract_formatted_signatures(@authors);
1294	1310	my ($types_ref, $signers_ref) = extract_formatted_signatures(@signatures);
1295	1311
1296		- return ($commits, @$signers_ref);
	1312	+ return ($commits, $signers_ref, $authors_ref, \@stats);
1297	1313	}
1298	1314
1299	1315	sub vcs_find_author {
1300	1316
...	...	@@ -1849,7 +1865,12 @@
1849	1865	sub vcs_file_signoffs {
1850	1866	my ($file) = @_;
1851	1867
	1868	+ my $authors_ref;
	1869	+ my $signers_ref;
	1870	+ my $stats_ref;
	1871	+ my @authors = ();
1852	1872	my @signers = ();
	1873	+ my @stats = ();
1853	1874	my $commits;
1854	1875
1855	1876	$vcs_used = vcs_exists();
1856	1877
1857	1878
...	...	@@ -1858,13 +1879,59 @@
1858	1879	my $cmd = $VCS_cmds{"find_signers_cmd"};
1859	1880	$cmd =~ s/(\$\w+)/$1/eeg; # interpolate $cmd
1860	1881
1861		- ($commits, @signers) = vcs_find_signers($cmd);
	1882	+ ($commits, $signers_ref, $authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
1862	1883
	1884	+ @signers = @{$signers_ref} if defined $signers_ref;
	1885	+ @authors = @{$authors_ref} if defined $authors_ref;
	1886	+ @stats = @{$stats_ref} if defined $stats_ref;
	1887	+
	1888	+# print("commits: <$commits>\nsigners:<@signers>\nauthors: <@authors>\nstats: <@stats>\n");
	1889	+
1863	1890	foreach my $signer (@signers) {
1864	1891	$signer = deduplicate_email($signer);
1865	1892	}
1866	1893
1867	1894	vcs_assign("commit_signer", $commits, @signers);
	1895	+ vcs_assign("authored", $commits, @authors);
	1896	+ if ($#authors == $#stats) {
	1897	+ my $stat_pattern = $VCS_cmds{"stat_pattern"};
	1898	+ $stat_pattern =~ s/(\$\w+)/$1/eeg; #interpolate $stat_pattern
	1899	+
	1900	+ my $added = 0;
	1901	+ my $deleted = 0;
	1902	+ for (my $i = 0; $i <= $#stats; $i++) {
	1903	+ if ($stats[$i] =~ /$stat_pattern/) {
	1904	+ $added += $1;
	1905	+ $deleted += $2;
	1906	+ }
	1907	+ }
	1908	+ my @tmp_authors = uniq(@authors);
	1909	+ foreach my $author (@tmp_authors) {
	1910	+ $author = deduplicate_email($author);
	1911	+ }
	1912	+ @tmp_authors = uniq(@tmp_authors);
	1913	+ my @list_added = ();
	1914	+ my @list_deleted = ();
	1915	+ foreach my $author (@tmp_authors) {
	1916	+ my $auth_added = 0;
	1917	+ my $auth_deleted = 0;
	1918	+ for (my $i = 0; $i <= $#stats; $i++) {
	1919	+ if ($author eq deduplicate_email($authors[$i]) &&
	1920	+ $stats[$i] =~ /$stat_pattern/) {
	1921	+ $auth_added += $1;
	1922	+ $auth_deleted += $2;
	1923	+ }
	1924	+ }
	1925	+ for (my $i = 0; $i < $auth_added; $i++) {
	1926	+ push(@list_added, $author);
	1927	+ }
	1928	+ for (my $i = 0; $i < $auth_deleted; $i++) {
	1929	+ push(@list_deleted, $author);
	1930	+ }
	1931	+ }
	1932	+ vcs_assign("added_lines", $added, @list_added);
	1933	+ vcs_assign("removed_lines", $deleted, @list_deleted);
	1934	+ }
1868	1935	}
1869	1936
1870	1937	sub vcs_file_blame {
...	...	@@ -1887,6 +1954,10 @@
1887	1954	if ($email_git_blame_signatures) {
1888	1955	if (vcs_is_hg()) {
1889	1956	my $commit_count;
	1957	+ my $commit_authors_ref;
	1958	+ my $commit_signers_ref;
	1959	+ my $stats_ref;
	1960	+ my @commit_authors = ();
1890	1961	my @commit_signers = ();
1891	1962	my $commit = join(" -r ", @commits);
1892	1963	my $cmd;
1893	1964
1894	1965
...	...	@@ -1894,19 +1965,27 @@
1894	1965	$cmd = $VCS_cmds{"find_commit_signers_cmd"};
1895	1966	$cmd =~ s/(\$\w+)/$1/eeg; #substitute variables in $cmd
1896	1967
1897		- ($commit_count, @commit_signers) = vcs_find_signers($cmd);
	1968	+ ($commit_count, $commit_signers_ref, $commit_authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
	1969	+ @commit_authors = @{$commit_authors_ref} if defined $commit_authors_ref;
	1970	+ @commit_signers = @{$commit_signers_ref} if defined $commit_signers_ref;
1898	1971
1899	1972	push(@signers, @commit_signers);
1900	1973	} else {
1901	1974	foreach my $commit (@commits) {
1902	1975	my $commit_count;
	1976	+ my $commit_authors_ref;
	1977	+ my $commit_signers_ref;
	1978	+ my $stats_ref;
	1979	+ my @commit_authors = ();
1903	1980	my @commit_signers = ();
1904	1981	my $cmd;
1905	1982
1906	1983	$cmd = $VCS_cmds{"find_commit_signers_cmd"};
1907	1984	$cmd =~ s/(\$\w+)/$1/eeg; #substitute variables in $cmd
1908	1985
1909		- ($commit_count, @commit_signers) = vcs_find_signers($cmd);
	1986	+ ($commit_count, $commit_signers_ref, $commit_authors_ref, $stats_ref) = vcs_find_signers($cmd, $file);
	1987	+ @commit_authors = @{$commit_authors_ref} if defined $commit_authors_ref;
	1988	+ @commit_signers = @{$commit_signers_ref} if defined $commit_signers_ref;
1910	1989
1911	1990	push(@signers, @commit_signers);
1912	1991	}