# Copyright (C) all contributors
# License: AGPL-3.0+
#
# used by PublicInbox::View
# This adds CSS spans for diff highlighting.
# It also generates links for ViewVCS + SolverGit to show
# (or reconstruct) blobs.
package PublicInbox::ViewDiff;
use v5.12;
use parent qw(Exporter);
our @EXPORT_OK = qw(flush_diff uri_escape_path);
use URI::Escape qw(uri_escape_utf8);
use PublicInbox::Hval qw(ascii_html to_attr);
use PublicInbox::Git qw(git_unquote);
my $OID_NULL = '0{7,}';
my $OID_BLOB = '[a-f0-9]{7,}';
my $LF = qr!\n!;
my $ANY = qr![^\n]!;
my $FN = qr!(?:"?[^/\n]+/[^\n]+|/dev/null)!;
# cf. git diff.c :: get_compact_summary
my $DIFFSTAT_COMMENT =
qr/(?: *\((?:new|gone|(?:(?:new|mode) [\+\-][lx]))\))? *\z/s;
my $NULL_TO_BLOB = qr/^(index $OID_NULL\.\.)($OID_BLOB)\b/ms;
my $BLOB_TO_NULL = qr/^index ($OID_BLOB)(\.\.$OID_NULL)\b/ms;
my $BLOB_TO_BLOB = qr/^index ($OID_BLOB)\.\.($OID_BLOB)/ms;
our $EXTRACT_DIFFS = qr/(
(?: # begin header stuff, don't capture filenames, here,
# but instead wait for the --- and +++ lines.
(?:^diff\x20--git\x20$FN\x20$FN$LF)
# old mode || new mode || copy|rename|deleted|...
(?:^[a-z]$ANY+$LF)*
)? # end of optional stuff, everything below is required
^index\x20($OID_BLOB)\.\.($OID_BLOB)$ANY*$LF
^---\x20($FN)$LF
^\+{3}\x20($FN)$LF)/msx;
our $IS_OID = qr/\A$OID_BLOB\z/s;
sub uri_escape_path {
# '/' + $URI::Escape::Unsafe{RFC3986}
uri_escape_utf8($_[0], "^A-Za-z0-9\-\._~/");
}
# link to line numbers in blobs
sub diff_hunk ($$$) {
my ($dctx, $ca, $cb) = @_;
my ($oid_a, $oid_b, $spfx) = @$dctx{qw(oid_a oid_b spfx)};
if (defined($spfx) && defined($oid_a) && defined($oid_b)) {
my $n = ($ca =~ /^-([0-9]+)/) ? "#n$1" : '';
my $x = qq(@@ {Q}$n">$ca);
$n = ($cb =~ /^\+([0-9]+)/) ? "#n$1" : '';
$x .= qq( {Q}$n">$cb @@);
} else {
"@@ $ca $cb @@";
}
}
sub oid ($$$) {
my ($dctx, $spfx, $oid) = @_;
defined($spfx) ? qq({Q}">$oid) : $oid;
}
# returns true if diffstat anchor written, false otherwise
sub anchor0 ($$$) {
my ($ctx, $fn, $rest) = @_;
my $orig = $fn;
# normal git diffstat output is impossible to parse reliably
# without --numstat, and that isn't the default for format-patch.
# So only do best-effort handling of renames for common cases;
# which works well in practice. If projects put "=>", or trailing
# spaces in filenames, oh well :P
$fn =~ s/$DIFFSTAT_COMMENT//;
$fn =~ s/\{(?:.+) => (.+)\}/$1/ or $fn =~ s/.* => (.+)/$1/;
$fn = git_unquote($fn);
# long filenames will require us to check in anchor1()
push(@{$ctx->{-long_path}}, $fn) if $fn =~ s!\A\.\.\./?!!;
my $attr = to_attr($ctx->{-apfx}.$fn) // return;
$ctx->{-anchors}->{$attr} = 1;
my $spaces = ($orig =~ s/( +)\z//) ? $1 : '';
print { $ctx->{zfh} } " ",
ascii_html($orig), '', $spaces,
$ctx->{-linkify}->to_html($rest);
}
# returns "diff --git" anchor destination, undef otherwise
sub anchor1 ($$) {
my ($ctx, $pb) = @_;
my $attr = to_attr($ctx->{-apfx}.$pb) // return;
my $ok = delete $ctx->{-anchors}->{$attr};
# unlikely, check the end of long path names we captured,
# assume diffstat and diff output follow the same order,
# and ignore different ordering (could be malicious input)
unless ($ok) {
my $fn = shift(@{$ctx->{-long_path}}) // return;
$pb =~ /\Q$fn\E\z/s or return;
$attr = to_attr($ctx->{-apfx}.$fn) // return;
$ok = delete $ctx->{-anchors}->{$attr} // return;
}
$ok ? "diff --git" : undef
}
sub diff_header ($$$) {
my ($x, $ctx, $top) = @_;
my (undef, undef, $pa, $pb) = splice(@$top, 0, 4); # ignore oid_{a,b}
my $spfx = $ctx->{-spfx};
my $dctx = { spfx => $spfx };
# get rid of leading "a/" or "b/" (or whatever --{src,dst}-prefix are)
$pa = (split(m'/', git_unquote($pa), 2))[1] if $pa ne '/dev/null';
$pb = (split(m'/', git_unquote($pb), 2))[1] if $pb ne '/dev/null';
if ($pa eq $pb && $pb ne '/dev/null') {
$dctx->{Q} = '?b='.uri_escape_path($pb);
} else {
my @q;
push @q, 'b='.uri_escape_path($pb) if $pb ne '/dev/null';
push @q, 'a='.uri_escape_path($pa) if $pa ne '/dev/null';
$dctx->{Q} = '?'.join('&', @q);
}
# linkify early and all at once, since we know the following
# subst ops on $$x won't need further escaping:
$$x = $ctx->{-linkify}->to_html($$x);
# no need to capture oid_a and oid_b on add/delete,
# we just linkify OIDs directly via s///e in conditional
if ($$x =~ s/$NULL_TO_BLOB/$1 . oid($dctx, $spfx, $2)/e) {
push @{$ctx->{-qry}->{dfpost}}, $2;
} elsif ($$x =~ s/$BLOB_TO_NULL/'index '.oid($dctx, $spfx, $1).$2/e) {
push @{$ctx->{-qry}->{dfpre}}, $1;
} elsif ($$x =~ $BLOB_TO_BLOB) {
# modification-only, not add/delete:
# linkify hunk headers later using oid_a and oid_b
@$dctx{qw(oid_a oid_b)} = ($1, $2);
push @{$ctx->{-qry}->{dfpre}}, $1;
push @{$ctx->{-qry}->{dfpost}}, $2;
} else {
warn "BUG? <$$x> had no ^index line";
}
$$x =~ s!^diff --git!anchor1($ctx, $pb) // 'diff --git'!ems;
print { $ctx->{zfh} } qq(), $$x, '';
$dctx;
}
sub diff_before_or_after ($$) {
my ($ctx, $x) = @_;
if (exists $ctx->{-anchors} && $$x =~ # diffstat lines:
/((?:^\x20(?:[^\n]+?)(?:\x20+\|\x20[^\n]*\n))+)
(\x20[0-9]+\x20files?\x20)changed,/msx) {
my $pre = substr($$x, 0, $-[0]); # (likely) short prefix
substr($$x, 0, $+[0], ''); # sv_chop on $$x ($$x may be long)
my @x = ($2, $1);
my $lnk = $ctx->{-linkify};
my $zfh = $ctx->{zfh};
# uninteresting prefix
print $zfh $lnk->to_html($pre);
for my $l (split(/^/m, pop(@x))) { # $2 per-file stat lines
$l =~ /^ (.+)( +\| .*\z)/s and
anchor0($ctx, $1, $2) and next;
print $zfh $lnk->to_html($l);
}
my $ch = $ctx->{changed_href} // '#related';
print $zfh pop(@x), # $3 /^ \d+ files? /
qq(changed,),
# insertions/deletions, notes, commit message, etc:
$lnk->to_html($$x);
} else {
print { $ctx->{zfh} } $ctx->{-linkify}->to_html($$x);
}
}
# callers must do CRLF => LF conversion before calling this
sub flush_diff ($$) {
my ($ctx, $cur) = @_;
my @top = split($EXTRACT_DIFFS, $$cur);
undef $$cur; # free memory
my $lnk = $ctx->{-linkify};
my $dctx; # {}, keys: Q, oid_a, oid_b
my $zfh = $ctx->zfh;
while (defined(my $x = shift @top)) {
if (scalar(@top) >= 4 &&
$top[1] =~ $IS_OID &&
$top[0] =~ $IS_OID) {
$dctx = diff_header(\$x, $ctx, \@top);
} elsif ($dctx) {
open(my $afh, '>>:utf8', \(my $after='')) or
die "open: $!";
# Quiet "Complex regular subexpression recursion limit"
# warning. Perl will truncate matches upon hitting
# that limit, giving us more (and shorter) scalars than
# would be ideal, but otherwise it's harmless.
#
# We could replace the `+' metacharacter with `{1,100}'
# to limit the matches ourselves to 100, but we can
# let Perl do it for us, quietly.
no warnings 'regexp';
for my $s (split(/((?:(?:^\+[^\n]*\n)+)|
(?:(?:^-[^\n]*\n)+)|
(?:^@@ [^\n]+\n))/xsm, $x)) {
undef $x;
if (!defined($dctx)) {
print $afh $s;
} elsif ($s =~ s/\A@@ (\S+) (\S+) @@//) {
print $zfh qq(),
diff_hunk($dctx, $1, $2),
$lnk->to_html($s),
'';
} elsif ($s =~ /\A\+/) { # $s may be huge
print $zfh qq(),
$lnk->to_html($s),
'';
} elsif ($s =~ /\A-- $/sm) { # email sig starts
$dctx = undef;
print $afh $s;
} elsif ($s =~ /\A-/) { # $s may be huge
print $zfh qq(),
$lnk->to_html($s),
'';
} else { # $s may be huge
print $zfh $lnk->to_html($s);
}
}
if (!$dctx) {
utf8::decode($after);
diff_before_or_after($ctx, \$after);
}
} else {
diff_before_or_after($ctx, \$x);
}
}
}
1;