gitweb: make hash size independent
Gitweb has several hard-coded 40 values throughout it to check for values that are passed in or acquired from Git. To simplify the code, introduce a regex variable that matches either exactly 40 or exactly 64 hex characters, and use this variable anywhere we would have previously hard-coded a 40 in a regex. Add some helper functions which allow us to write tighter regexes that match exactly the number of hex characters we're expecting. Similarly, switch the code that looks for deleted diffinfo information to look for either 40 or 64 zeros, and update one piece of code to use this function. Finally, when formatting a log line, allow an abbreviated describe output to contain up to 64 characters. Helped-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com> Signed-off-by: brian m. carlson <sandals@crustytoothpaste.net> Signed-off-by: Junio C Hamano <gitster@pobox.com>
brian m. carlson committed
Feb 19, 2019 at 00:05 UTC
cfb049119c9e8b8f66a2e1faf1621ac48f714412
1 file changed
+67
-30
gitweb/gitweb.perl
+67
-30
@@ -788,6 +788,38 @@ sub check_loadavg {
788
# ======================================================================
789
# input validation and dispatch
790
791
+# Various hash size-related values.
792
+my $sha1_len = 40;
793
+my $sha256_extra_len = 24;
794
+my $sha256_len = $sha1_len + $sha256_extra_len;
795
+
796
+# A regex matching $len hex characters. $len may be a range (e.g. 7,64).
797
+sub oid_nlen_regex {
798
+ my $len = shift;
799
+ my $hchr = qr/[0-9a-fA-F]/;
800
+ return qr/(?:(?:$hchr){$len})/;
801
+}
802
+
803
+# A regex matching two sets of $nlen hex characters, prefixed by the literal
804
+# string $prefix and with the literal string $infix between them.
805
+sub oid_nlen_prefix_infix_regex {
806
+ my $nlen = shift;
807
+ my $prefix = shift;
808
+ my $infix = shift;
809
+
810
+ my $rx = oid_nlen_regex($nlen);
811
+
812
+ return qr/^\Q$prefix\E$rx\Q$infix\E$rx$/;
813
+}
814
+
815
+# A regex matching a valid object ID.
816
+our $oid_regex;
817
+{
818
+ my $x = oid_nlen_regex($sha1_len);
819
+ my $y = oid_nlen_regex($sha256_extra_len);
820
+ $oid_regex = qr/(?:$x(?:$y)?)/;
821
+}
822
+
823
# input parameters can be collected from a variety of sources (presently, CGI
824
# and PATH_INFO), so we define an %input_params hash that collects them all
825
# together during validation: this allows subsequent uses (e.g. href()) to be
@@ -1516,7 +1548,7 @@ sub is_valid_refname {
1548
1549
return undef unless defined $input;
1550
# textual hashes are O.K.
1519
- if ($input =~ m/^[0-9a-fA-F]{40}$/) {
1551
+ if ($input =~ m/^$oid_regex$/) {
1552
return 1;
1553
}
1554
# it must be correct pathname
@@ -2028,6 +2060,9 @@ sub file_type_long {
2060
sub format_log_line_html {
2061
my $line = shift;
2062
2063
+ # Potentially abbreviated OID.
2064
+ my $regex = oid_nlen_regex("7,64");
2065
+
2066
$line = esc_html($line, -nbsp=>1);
2067
$line =~ s{
2068
\b
@@ -2037,10 +2072,10 @@ sub format_log_line_html {
2072
(?<!-) # see strbuf_check_tag_ref(). Tags can't start with -
2073
[A-Za-z0-9.-]+
2074
(?!\.) # refs can't end with ".", see check_refname_format()
2040
- -g[0-9a-fA-F]{7,40}
2075
+ -g$regex
2076
|
2077
# Just a normal looking Git SHA1
2043
- [0-9a-fA-F]{7,40}
2078
+ $regex
2079
)
2080
\b
2081
}{
@@ -2286,7 +2321,8 @@ sub format_extended_diff_header_line {
2321
')</span>';
2322
}
2323
# match <hash>
2289
- if ($line =~ m/^index [0-9a-fA-F]{40},[0-9a-fA-F]{40}/) {
2324
+ if ($line =~ oid_nlen_prefix_infix_regex($sha1_len, "index ", ",") |
2325
+ $line =~ oid_nlen_prefix_infix_regex($sha256_len, "index ", ",")) {
2326
# can match only for combined diff
2327
$line = 'index ';
2328
for (my $i = 0; $i < $diffinfo->{'nparents'}; $i++) {
@@ -2308,7 +2344,8 @@ sub format_extended_diff_header_line {
2344
$line .= '0' x 7;
2345
}
2346
2311
- } elsif ($line =~ m/^index [0-9a-fA-F]{40}..[0-9a-fA-F]{40}/) {
2347
+ } elsif ($line =~ oid_nlen_prefix_infix_regex($sha1_len, "index ", "..") |
2348
+ $line =~ oid_nlen_prefix_infix_regex($sha256_len, "index ", "..")) {
2349
# can match only for ordinary diff
2350
my ($from_link, $to_link);
2351
if ($from->{'href'}) {
@@ -2834,7 +2871,7 @@ sub git_get_hash_by_path {
2871
}
2872
2873
#'100644 blob 0fa3f3a66fb6a137f6ec2c19351ed4d807070ffa panic.c'
2837
- $line =~ m/^([0-9]+) (.+) ([0-9a-fA-F]{40})\t/;
2874
+ $line =~ m/^([0-9]+) (.+) ($oid_regex)\t/;
2875
if (defined $type && $type ne $2) {
2876
# type doesn't match
2877
return undef;
@@ -3333,7 +3370,7 @@ sub git_get_references {
3370
3371
while (my $line = <$fd>) {
3372
chomp $line;
3336
- if ($line =~ m!^([0-9a-fA-F]{40})\srefs/($type.*)$!) {
3373
+ if ($line =~ m!^($oid_regex)\srefs/($type.*)$!) {
3374
if (defined $refs{$1}) {
3375
push @{$refs{$1}}, $2;
3376
} else {
@@ -3407,7 +3444,7 @@ sub parse_tag {
3444
$tag{'id'} = $tag_id;
3445
while (my $line = <$fd>) {
3446
chomp $line;
3410
- if ($line =~ m/^object ([0-9a-fA-F]{40})$/) {
3447
+ if ($line =~ m/^object ($oid_regex)$/) {
3448
$tag{'object'} = $1;
3449
} elsif ($line =~ m/^type (.+)$/) {
3450
$tag{'type'} = $1;
@@ -3451,15 +3488,15 @@ sub parse_commit_text {
3488
}
3489
3490
my $header = shift @commit_lines;
3454
- if ($header !~ m/^[0-9a-fA-F]{40}/) {
3491
+ if ($header !~ m/^$oid_regex/) {
3492
return;
3493
}
3494
($co{'id'}, my @parents) = split ' ', $header;
3495
while (my $line = shift @commit_lines) {
3496
last if $line eq "\n";
3460
- if ($line =~ m/^tree ([0-9a-fA-F]{40})$/) {
3497
+ if ($line =~ m/^tree ($oid_regex)$/) {
3498
$co{'tree'} = $1;
3462
- } elsif ((!defined $withparents) && ($line =~ m/^parent ([0-9a-fA-F]{40})$/)) {
3499
+ } elsif ((!defined $withparents) && ($line =~ m/^parent ($oid_regex)$/)) {
3500
push @parents, $1;
3501
} elsif ($line =~ m/^author (.*) ([0-9]+) (.*)$/) {
3502
$co{'author'} = to_utf8($1);
@@ -3591,7 +3628,7 @@ sub parse_difftree_raw_line {
3628
3629
# ':100644 100644 03b218260e99b78c6df0ed378e59ed9205ccc96d 3b93d5e7cc7f7dd4ebed13a5cc1a4ad976fc94d8 M ls-files.c'
3630
# ':100644 100644 7f9281985086971d3877aca27704f2aaf9c448ce bc190ebc71bbd923f2b728e505408f5e54bd073a M rev-tree.c'
3594
- if ($line =~ m/^:([0-7]{6}) ([0-7]{6}) ([0-9a-fA-F]{40}) ([0-9a-fA-F]{40}) (.)([0-9]{0,3})\t(.*)$/) {
3631
+ if ($line =~ m/^:([0-7]{6}) ([0-7]{6}) ($oid_regex) ($oid_regex) (.)([0-9]{0,3})\t(.*)$/) {
3632
$res{'from_mode'} = $1;
3633
$res{'to_mode'} = $2;
3634
$res{'from_id'} = $3;
@@ -3606,7 +3643,7 @@ sub parse_difftree_raw_line {
3643
}
3644
# '::100755 100755 100755 60e79ca1b01bc8b057abe17ddab484699a7f5fdb 94067cc5f73388f33722d52ae02f44692bc07490 94067cc5f73388f33722d52ae02f44692bc07490 MR git-gui/git-gui.sh'
3645
# combined diff (for merge commit)
3609
- elsif ($line =~ s/^(::+)((?:[0-7]{6} )+)((?:[0-9a-fA-F]{40} )+)([a-zA-Z]+)\t(.*)$//) {
3646
+ elsif ($line =~ s/^(::+)((?:[0-7]{6} )+)((?:$oid_regex )+)([a-zA-Z]+)\t(.*)$//) {
3647
$res{'nparents'} = length($1);
3648
$res{'from_mode'} = [ split(' ', $2) ];
3649
$res{'to_mode'} = pop @{$res{'from_mode'}};
@@ -3616,7 +3653,7 @@ sub parse_difftree_raw_line {
3653
$res{'to_file'} = unquote($5);
3654
}
3655
# 'c512b523472485aef4fff9e57b229d9d243c967f'
3619
- elsif ($line =~ m/^([0-9a-fA-F]{40})$/) {
3656
+ elsif ($line =~ m/^($oid_regex)$/) {
3657
$res{'commit'} = $1;
3658
}
3659
@@ -3644,7 +3681,7 @@ sub parse_ls_tree_line {
3681
3682
if ($opts{'-l'}) {
3683
#'100644 blob 0fa3f3a66fb6a137f6ec2c19351ed4d807070ffa 16717 panic.c'
3647
- $line =~ m/^([0-9]+) (.+) ([0-9a-fA-F]{40}) +(-|[0-9]+)\t(.+)$/s;
3684
+ $line =~ m/^([0-9]+) (.+) ($oid_regex) +(-|[0-9]+)\t(.+)$/s;
3685
3686
$res{'mode'} = $1;
3687
$res{'type'} = $2;
@@ -3657,7 +3694,7 @@ sub parse_ls_tree_line {
3694
}
3695
} else {
3696
#'100644 blob 0fa3f3a66fb6a137f6ec2c19351ed4d807070ffa panic.c'
3660
- $line =~ m/^([0-9]+) (.+) ([0-9a-fA-F]{40})\t(.+)$/s;
3697
+ $line =~ m/^([0-9]+) (.+) ($oid_regex)\t(.+)$/s;
3698
3699
$res{'mode'} = $1;
3700
$res{'type'} = $2;
@@ -4799,7 +4836,7 @@ sub fill_from_file_info {
4836
sub is_deleted {
4837
my $diffinfo = shift;
4838
4802
- return $diffinfo->{'to_id'} eq ('0' x 40);
4839
+ return $diffinfo->{'to_id'} eq ('0' x 40) || $diffinfo->{'to_id'} eq ('0' x 64);
4840
}
4841
4842
# does patch correspond to [previous] difftree raw line
@@ -6285,7 +6322,7 @@ sub git_search_changes {
6322
-class => "list subject"},
6323
chop_and_escape_str($co{'title'}, 50) . "<br/>");
6324
} elsif (defined $set{'to_id'}) {
6288
- next if ($set{'to_id'} =~ m/^0{40}$/);
6325
+ next if is_deleted(\%set);
6326
6327
print $cgi->a({-href => href(action=>"blob", hash_base=>$co{'id'},
6328
hash=>$set{'to_id'}, file_name=>$set{'to_file'}),
@@ -6829,7 +6866,7 @@ sub git_blame_common {
6866
# the header: <SHA-1> <src lineno> <dst lineno> [<lines in group>]
6867
# no <lines in group> for subsequent lines in group of lines
6868
my ($full_rev, $orig_lineno, $lineno, $group_size) =
6832
- ($line =~ /^([0-9a-f]{40}) (\d+) (\d+)(?: (\d+))?$/);
6869
+ ($line =~ /^($oid_regex) (\d+) (\d+)(?: (\d+))?$/);
6870
if (!exists $metainfo{$full_rev}) {
6871
$metainfo{$full_rev} = { 'nprevious' => 0 };
6872
}
@@ -6879,7 +6916,7 @@ sub git_blame_common {
6916
}
6917
# 'previous' <sha1 of parent commit> <filename at commit>
6918
if (exists $meta->{'previous'} &&
6882
- $meta->{'previous'} =~ /^([a-fA-F0-9]{40}) (.*)$/) {
6919
+ $meta->{'previous'} =~ /^($oid_regex) (.*)$/) {
6920
$meta->{'parent'} = $1;
6921
$meta->{'file_parent'} = unquote($2);
6922
}
@@ -6996,7 +7033,7 @@ sub git_blob_plain {
7033
} else {
7034
die_error(400, "No file name defined");
7035
}
6999
- } elsif ($hash =~ m/^[0-9a-fA-F]{40}$/) {
7036
+ } elsif ($hash =~ m/^$oid_regex$/) {
7037
# blobs defined by non-textual hash id's can be cached
7038
$expires = "+1d";
7039
}
@@ -7057,7 +7094,7 @@ sub git_blob {
7094
} else {
7095
die_error(400, "No file name defined");
7096
}
7060
- } elsif ($hash =~ m/^[0-9a-fA-F]{40}$/) {
7097
+ } elsif ($hash =~ m/^$oid_regex$/) {
7098
# blobs defined by non-textual hash id's can be cached
7099
$expires = "+1d";
7100
}
@@ -7515,7 +7552,7 @@ sub git_commit {
7552
7553
# non-textual hash id's can be cached
7554
my $expires;
7518
- if ($hash =~ m/^[0-9a-fA-F]{40}$/) {
7555
+ if ($hash =~ m/^$oid_regex$/) {
7556
$expires = "+1d";
7557
}
7558
my $refs = git_get_references();
@@ -7609,7 +7646,7 @@ sub git_object {
7646
close $fd;
7647
7648
#'100644 blob 0fa3f3a66fb6a137f6ec2c19351ed4d807070ffa panic.c'
7612
- unless ($line && $line =~ m/^([0-9]+) (.+) ([0-9a-fA-F]{40})\t/) {
7649
+ unless ($line && $line =~ m/^([0-9]+) (.+) ($oid_regex)\t/) {
7650
die_error(404, "File or directory for given base does not exist");
7651
}
7652
$type = $2;
@@ -7649,7 +7686,7 @@ sub git_blobdiff {
7686
or die_error(404, "Blob diff not found");
7687
7688
} elsif (defined $hash &&
7652
- $hash =~ /[0-9a-fA-F]{40}/) {
7689
+ $hash =~ $oid_regex) {
7690
# try to find filename from $hash
7691
7692
# read filtered raw output
@@ -7659,7 +7696,7 @@ sub git_blobdiff {
7696
@difftree =
7697
# ':100644 100644 03b21826... 3b93d5e7... M ls-files.c'
7698
# $hash == to_id
7662
- grep { /^:[0-7]{6} [0-7]{6} [0-9a-fA-F]{40} $hash/ }
7699
+ grep { /^:[0-7]{6} [0-7]{6} $oid_regex $hash/ }
7700
map { chomp; $_ } <$fd>;
7701
close $fd
7702
or die_error(404, "Reading git-diff-tree failed");
@@ -7682,8 +7719,8 @@ sub git_blobdiff {
7719
$hash ||= $diffinfo{'to_id'};
7720
7721
# non-textual hash id's can be cached
7685
- if ($hash_base =~ m/^[0-9a-fA-F]{40}$/ &&
7686
- $hash_parent_base =~ m/^[0-9a-fA-F]{40}$/) {
7722
+ if ($hash_base =~ m/^$oid_regex$/ &&
7723
+ $hash_parent_base =~ m/^$oid_regex$/) {
7724
$expires = '+1d';
7725
}
7726
@@ -7819,7 +7856,7 @@ sub git_commitdiff {
7856
$hash_parent ne '-c' && $hash_parent ne '--cc') {
7857
# commitdiff with two commits given
7858
my $hash_parent_short = $hash_parent;
7822
- if ($hash_parent =~ m/^[0-9a-fA-F]{40}$/) {
7859
+ if ($hash_parent =~ m/^$oid_regex$/) {
7860
$hash_parent_short = substr($hash_parent, 0, 7);
7861
}
7862
$formats_nav .=
@@ -7928,7 +7965,7 @@ sub git_commitdiff {
7965
7966
# non-textual hash id's can be cached
7967
my $expires;
7931
- if ($hash =~ m/^[0-9a-fA-F]{40}$/) {
7968
+ if ($hash =~ m/^$oid_regex$/) {
7969
$expires = "+1d";
7970
}
7971