X-Git-Url: http://www.git.stargrave.org/?a=blobdiff_plain;f=lib%2FPublicInbox%2FViewVCS.pm;h=3b4fa393ed3ff16436db04a0dc3259d62825edbf;hb=35b7adee1f01a7cce720d944e1666aa44717d024;hp=ead8c2b45fe1fd8e1791227b13872f480e2560aa;hpb=b316f7541ec263fd04ea50530a1d500f43773299;p=public-inbox.git diff --git a/lib/PublicInbox/ViewVCS.pm b/lib/PublicInbox/ViewVCS.pm index ead8c2b4..3b4fa393 100644 --- a/lib/PublicInbox/ViewVCS.pm +++ b/lib/PublicInbox/ViewVCS.pm @@ -1,8 +1,7 @@ -# Copyright (C) 2019 all contributors +# Copyright (C) all contributors # License: AGPL-3.0+ # show any VCS object, similar to "git show" -# FIXME: we only show blobs for now # # This can use a "solver" to reconstruct blobs based on git # patches (with abbreviated OIDs in the header). However, the @@ -15,62 +14,86 @@ package PublicInbox::ViewVCS; use strict; -use warnings; -use bytes (); # only for bytes::length +use v5.10.1; +use File::Temp 0.19 (); # newdir use PublicInbox::SolverGit; -use PublicInbox::WwwStream; +use PublicInbox::GitAsyncCat; +use PublicInbox::WwwStream qw(html_oneshot); use PublicInbox::Linkify; use PublicInbox::Tmpfile; +use PublicInbox::ViewDiff qw(flush_diff uri_escape_path); +use PublicInbox::View; +use PublicInbox::Eml; +use Text::Wrap qw(wrap); use PublicInbox::Hval qw(ascii_html to_filename); my $hl = eval { require PublicInbox::HlMod; PublicInbox::HlMod->new; }; -my %QP_MAP = ( A => 'oid_a', B => 'oid_b', a => 'path_a', b => 'path_b' ); +my %QP_MAP = ( A => 'oid_a', a => 'path_a', b => 'path_b' ); our $MAX_SIZE = 1024 * 1024; # TODO: configurable my $BIN_DETECT = 8000; # same as git +my $SHOW_FMT = '--pretty=format:'.join('%n', '%P', '%p', '%H', '%T', '%s', '%f', + '%an <%ae> %ai', '%cn <%ce> %ci', '%b%x00'); -sub html_i { # WwwStream::getline callback - my ($nr, $ctx) = @_; - $nr == 1 ? ${delete $ctx->{rv}} : undef; -} +my %GIT_MODE = ( + '100644' => ' ', # blob + '100755' => 'x', # executable blob + '040000' => 'd', # tree + '120000' => 'l', # symlink + '160000' => 'g', # commit (gitlink) +); -sub html_page ($$$) { - my ($ctx, $code, $strref) = @_; +sub html_page ($$;@) { + my ($ctx, $code) = @_[0, 1]; my $wcb = delete $ctx->{-wcb}; $ctx->{-upfx} = '../../'; # from "/$INBOX/$OID/s/" - $ctx->{rv} = $strref; - my $res = PublicInbox::WwwStream->response($ctx, $code, \&html_i); + my $res = html_oneshot($ctx, $code, @_[2..$#_]); $wcb ? $wcb->($res) : $res; } +sub dbg_log ($) { + my ($ctx) = @_; + my $log = delete $ctx->{lh} // die 'BUG: already captured debug log'; + if (!seek($log, 0, 0)) { + warn "seek(log): $!"; + return '
debug log seek error
'; + } + $log = do { local $/; <$log> } // do { + warn "readline(log): $!"; + return '
debug log read error
'; + }; + $ctx->{-linkify} //= PublicInbox::Linkify->new; + "
debug log:\n\n".
+		$ctx->{-linkify}->to_html($log).'
'; +} + sub stream_blob_parse_hdr { # {parse_hdr} for Qspawn my ($r, $bref, $ctx) = @_; - my ($res, $logref) = delete @$ctx{qw(-res -logref)}; - my ($git, $oid, $type, $size, $di) = @$res; + my ($git, $oid, $type, $size, $di) = @{$ctx->{-res}}; my @cl = ('Content-Length', $size); - if (!defined $r) { # error - html_page($ctx, 500, $logref); + if (!defined $r) { # sysread error + html_page($ctx, 500, dbg_log($ctx)); } elsif (index($$bref, "\0") >= 0) { [200, [qw(Content-Type application/octet-stream), @cl] ]; } else { - my $n = bytes::length($$bref); + my $n = length($$bref); if ($n >= $BIN_DETECT || $n == $size) { return [200, [ 'Content-Type', 'text/plain; charset=UTF-8', @cl ] ]; } if ($r == 0) { - warn "premature EOF on $oid $$logref\n"; - return html_page($ctx, 500, $logref); + my $log = dbg_log($ctx); + warn "premature EOF on $oid $log"; + return html_page($ctx, 500, $log); } undef; # bref keeps growing } } -sub stream_large_blob ($$$$) { - my ($ctx, $res, $logref, $fn) = @_; - $ctx->{-logref} = $logref; +sub stream_large_blob ($$) { + my ($ctx, $res) = @_; $ctx->{-res} = $res; my ($git, $oid, $type, $size, $di) = @$res; my $cmd = ['git', "--git-dir=$git->{git_dir}", 'cat-file', $type, $oid]; @@ -80,96 +103,346 @@ sub stream_large_blob ($$$$) { $qsp->psgi_return($env, undef, \&stream_blob_parse_hdr, $ctx); } -sub show_other_result ($$) { +sub show_other_result ($$) { # tag my ($bref, $ctx) = @_; - my ($qsp, $logref) = delete @$ctx{qw(-qsp -logref)}; - if (my $err = $qsp->{err}) { - utf8::decode($$err); - $$logref .= "git show error: $err"; - return html_page($ctx, 500, $logref); + if (my $qsp_err = delete $ctx->{-qsp_err}) { + return html_page($ctx, 500, dbg_log($ctx) . + "git show error:$qsp_err"); } my $l = PublicInbox::Linkify->new; utf8::decode($$bref); - $l->linkify_1($$bref); - $$bref = '
'. $l->linkify_2(ascii_html($$bref));
-	$$bref .= '
' . $$logref; - html_page($ctx, 200, $bref); + html_page($ctx, 200, '
', $l->to_html($$bref), '
', + dbg_log($ctx)); } -sub show_other ($$$$) { - my ($ctx, $res, $logref, $fn) = @_; - my ($git, $oid, $type, $size) = @$res; - if ($size > $MAX_SIZE) { - $$logref = "$oid is too big to show\n" . $$logref; - return html_page($ctx, 200, $logref); +sub cmt_title { # git->cat_async callback + my ($bref, $oid, $type, $size, $ctx) = @_; + utf8::decode($$bref); + my $title = $$bref =~ /\r?\n\r?\n([^\r\n]+)\r?\n?/ ? $1 : ''; + push(@{$ctx->{-cmt_pt}} , ascii_html($title)) == @{$ctx->{-cmt_P}} and + cmt_finalize($ctx); +} + +sub show_commit_start { # ->psgi_qx callback + my ($bref, $ctx) = @_; + if (my $qsp_err = delete $ctx->{-qsp_err}) { + return html_page($ctx, 500, dbg_log($ctx) . + "git show/patch-id error:$qsp_err"); + } + my $patchid = (split(/ /, $$bref))[0]; # ignore commit + $ctx->{-q_value_html} = "patchid:$patchid" if defined $patchid; + open my $fh, '<:utf8', "$ctx->{-tmp}/h" or + die "open $ctx->{-tmp}/h: $!"; + chop(my $buf = do { local $/ = "\0"; <$fh> }); + chomp $buf; + my ($P, $p); + ($P, $p, @{$ctx->{cmt_info}}) = split(/\n/, $buf, 9); + return cmt_finalize($ctx) if !$P; + @{$ctx->{-cmt_P}} = split(/ /, $P); + @{$ctx->{-cmt_p}} = split(/ /, $p); # abbreviated + if ($ctx->{env}->{'pi-httpd.async'}) { + for (@{$ctx->{-cmt_P}}) { + ibx_async_cat($ctx, $_, \&cmt_title, $ctx); + } + } else { # synchronous + for (@{$ctx->{-cmt_P}}) { + $ctx->{git}->cat_async($_, \&cmt_title, $ctx); + } + $ctx->{git}->cat_async_wait; + } +} + +sub cmt_finalize { + my ($ctx) = @_; + $ctx->{-linkify} //= PublicInbox::Linkify->new; + my $upfx = $ctx->{-upfx} = '../../'; # from "/$INBOX/$OID/s/" + my ($H, $T, $s, $f, $au, $co, $bdy) = @{delete $ctx->{cmt_info}}; + # try to keep author and committer dates lined up + my $x = length($au) - length($co); + if ($x > 0) { + $x = ' ' x $x; + $co =~ s/>/>$x/; + } elsif ($x < 0) { + $x = ' ' x (-$x); + $au =~ s/>/>$x/; + } + $_ = ascii_html($_) for ($au, $co); + $au =~ s!(> +)([0-9]{4,}-\S+ \S+)! + my ($gt, $t) = ($1, $2); + $t =~ tr/ :-//d; + qq($gt$2) + !e; + $ctx->{-title_html} = $s = $ctx->{-linkify}->to_html($s); + my ($P, $p, $pt) = delete @$ctx{qw(-cmt_P -cmt_p -cmt_pt)}; + $_ = qq().shift(@$p).' '.shift(@$pt) for @$P; + if (@$P == 1) { + $x = qq{ (patch)\n parent $P->[0]}; + } elsif (@$P > 1) { + $x = qq(\n parents $P->[0]\n); + shift @$P; + $x .= qq( $_\n) for @$P; + chop $x; + } else { + $x = ' (root commit)'; + } + PublicInbox::WwwStream::html_init($ctx); + $ctx->zmore(< commit $H$x + tree $T + author $au +committer $co + +$s +EOM + $ctx->zmore("\n", $ctx->{-linkify}->to_html($bdy)) if length($bdy); + $bdy = ''; + open my $fh, '<:utf8', "$ctx->{-tmp}/p" or + die "open $ctx->{-tmp}/p: $!"; + if (-s $fh > $MAX_SIZE) { + $ctx->zmore("---\n patch is too large to show\n"); + } else { # prepare flush_diff: + read($fh, $x, -s _); + $ctx->{obuf} = \$bdy; + $ctx->{-apfx} = $ctx->{-spfx} = $upfx; + $x =~ s/\r?\n/\n/gs; + $ctx->{-anchors} = {} if $x =~ /^diff --git /sm; + flush_diff($ctx, \$x); # undefs $x + $ctx->zmore($bdy); + undef $bdy; + # TODO: should there be another textarea which attempts to + # search for the exact email which was applied to make this + # commit? + if (my $qry = delete $ctx->{-qry}) { + my $q = ''; + for (@{$qry->{dfpost}}, @{$qry->{dfpre}}) { + # keep blobs as short as reasonable, emails + # are going to be older than what's in git + substr($_, 7, 64, ''); + $q .= "dfblob:$_ "; + } + chop $q; # no trailing SP + local $Text::Wrap::columns = PublicInbox::View::COLS; + local $Text::Wrap::huge = 'overflow'; + $q = wrap('', '', $q); + my $rows = ($q =~ tr/\n/\n/) + 1; + $q = ascii_html($q); + $ctx->zmore(<
find related emails, including ancestors/descendants/conflicts
+
+\t(help)
+EOM + } + } + chop($x = <
glossary
+--------
+Commit objects reference one tree, and zero or more parents.
+
+Single parent commits can typically generate a patch in
+unified diff format via `git format-patch'.
+
+Multiple parents means the commit is a merge.
+
+Root commits have no ancestor.  Note that it is
+possible to have multiple root commits when merging independent histories.
+
+Every commit references one top-level tree object.
+EOM + $x = $ctx->zflush($x, $ctx->_html_end); + my $res_hdr = delete $ctx->{-res_hdr}; + push @$res_hdr, 'Content-Length', length($x); + delete($ctx->{env}->{'qspawn.wcb'})->([200, $res_hdr, [$x]]); +} + +sub stream_patch_parse_hdr { # {parse_hdr} for Qspawn + my ($r, $bref, $ctx) = @_; + if (!defined $r) { # sysread error + html_page($ctx, 500, dbg_log($ctx)); + } elsif (index($$bref, "\n\n") >= 0) { + my $eml = bless { hdr => $bref }, 'PublicInbox::Eml'; + my $fn = to_filename($eml->header('Subject') // ''); + $fn = substr($fn // 'PATCH-no-subject', 6); # drop "PATCH-" + return [ 200, [ 'Content-Type', 'text/plain; charset=UTF-8', + 'Content-Disposition', + qq(inline; filename=$fn.patch) ] ]; + } elsif ($r == 0) { + my $log = dbg_log($ctx); + warn "premature EOF on $ctx->{patch_oid} $log"; + return html_page($ctx, 500, $log); + } else { + undef; # bref keeps growing until "\n\n" } +} + +sub show_patch ($$) { + my ($ctx, $res) = @_; + my ($git, $oid) = @$res; + my @cmd = ('git', "--git-dir=$git->{git_dir}", + qw(format-patch -1 --stdout -C), + "--signature=git format-patch -1 --stdout -C $oid", $oid); + my $qsp = PublicInbox::Qspawn->new(\@cmd); + $ctx->{env}->{'qspawn.wcb'} = delete $ctx->{-wcb}; + $ctx->{patch_oid} = $oid; + $qsp->psgi_return($ctx->{env}, undef, \&stream_patch_parse_hdr, $ctx); +} + +sub show_commit ($$) { + my ($ctx, $res) = @_; + return show_patch($ctx, $res) if ($ctx->{fn} // '') =~ /\.patch\z/; + my ($git, $oid) = @$res; + # patch-id needs two passes, and we use the initial show to ensure + # a patch embedded inside the commit message body doesn't get fed + # to patch-id: + my $cmd = [ '/bin/sh', '-c', + "git show --encoding=UTF-8 '$SHOW_FMT'". + " -z --no-notes --no-patch $oid >h && ". + 'git show --encoding=UTF-8 --pretty=format:%n -M'. + " --stat -p $oid >p && ". + "git patch-id --stable $git->{git_dir} }; + my $qsp = PublicInbox::Qspawn->new($cmd, $e, { -C => "$ctx->{-tmp}" }); + $qsp->{qsp_err} = \($ctx->{-qsp_err} = ''); + $ctx->{env}->{'qspawn.wcb'} = delete $ctx->{-wcb}; + $ctx->{git} = $git; + $qsp->psgi_qx($ctx->{env}, undef, \&show_commit_start, $ctx); +} + +sub show_other ($$) { + my ($ctx, $res) = @_; + my ($git, $oid, $type, $size) = @$res; + $size > $MAX_SIZE and return html_page($ctx, 200, + ascii_html($type)." $oid is too big to show\n". dbg_log($ctx)); my $cmd = ['git', "--git-dir=$git->{git_dir}", qw(show --encoding=UTF-8 --no-color --no-abbrev), $oid ]; my $qsp = PublicInbox::Qspawn->new($cmd); - my $env = $ctx->{env}; - $ctx->{-qsp} = $qsp; - $ctx->{-logref} = $logref; - $qsp->psgi_qx($env, undef, \&show_other_result, $ctx); + $qsp->{qsp_err} = \($ctx->{-qsp_err} = ''); + $qsp->psgi_qx($ctx->{env}, undef, \&show_other_result, $ctx); } -# user_cb for SolverGit, called as: user_cb->($result_or_error, $uarg) -sub solve_result { - my ($res, $ctx) = @_; - my ($log, $hints, $fn) = delete @$ctx{qw(log hints fn)}; - - unless (seek($log, 0, 0)) { - $ctx->{env}->{'psgi.errors'}->print("seek(log): $!\n"); - return html_page($ctx, 500, \'seek error'); +sub show_tree_result ($$) { + my ($bref, $ctx) = @_; + if (my $qsp_err = delete $ctx->{-qsp_err}) { + return html_page($ctx, 500, dbg_log($ctx) . + "git ls-tree -z error:$qsp_err"); + } + my @ent = split(/\0/, $$bref); + my $qp = delete $ctx->{qp}; + my $l = $ctx->{-linkify} //= PublicInbox::Linkify->new; + my $pfx = $qp->{b}; + $$bref = "
tree $ctx->{tree_oid}";
+	if (defined $pfx) {
+		my $x = ascii_html($pfx);
+		$pfx .= '/';
+		$$bref .= qq(  path: $x\n);
+	} else {
+		$pfx = '';
+		$$bref .= qq[  (path unknown)\n];
 	}
-	$log = do { local $/; <$log> };
+	my ($x, $m, $t, $oid, $sz, $f, $n);
+	$$bref .= "\n	size	name";
+	for (@ent) {
+		($x, $f) = split(/\t/, $_, 2);
+		undef $_;
+		($m, $t, $oid, $sz) = split(/ +/, $x, 4);
+		$m = $GIT_MODE{$m} // '?';
+		utf8::decode($f);
+		$n = ascii_html($f);
+		if ($m eq 'g') { # gitlink submodule commit
+			$$bref .= "\ng\t\t$n @ commit$oid";
+			next;
+		}
+		my $q = 'b='.ascii_html(uri_escape_path($pfx.$f));
+		if ($m eq 'd') { $n .= '/' }
+		elsif ($m eq 'x') { $n = "$n" }
+		elsif ($m eq 'l') { $n = "$n" }
+		$$bref .= qq(\n$m\t$sz\t$n);
+	}
+	$$bref .= dbg_log($ctx);
+	$$bref .= <glossary
+--------
+Tree objects belong to commits or other tree objects.  Trees may
+reference blobs, sub-trees, or commits of submodules.
 
-	my $ref = ref($res);
-	my $l = PublicInbox::Linkify->new;
-	$l->linkify_1($log);
-	$log = '
debug log:
' .
-		$l->linkify_2(ascii_html($log)) . '
';
+Path names are stored in tree objects, but trees do not know
+their own path name.  A tree's path name comes from their parent tree,
+or it is the root tree referenced by a commit object.  Thus, this web UI
+relies on the `b=' URI parameter as a hint to display the path name.
+
+Commit objects may be stored in trees to reference submodules.
+EOM + chop $$bref; + html_page($ctx, 200, $$bref); +} - $res or return html_page($ctx, 404, \$log); - $ref eq 'ARRAY' or return html_page($ctx, 500, \$log); +sub show_tree ($$) { + my ($ctx, $res) = @_; + my ($git, $oid, undef, $size) = @$res; + $size > $MAX_SIZE and return html_page($ctx, 200, + "tree $oid is too big to show\n". dbg_log($ctx)); + my $cmd = [ 'git', "--git-dir=$git->{git_dir}", + qw(ls-tree -z -l --no-abbrev), $oid ]; + my $qsp = PublicInbox::Qspawn->new($cmd); + $ctx->{tree_oid} = $oid; + $qsp->{qsp_err} = \($ctx->{-qsp_err} = ''); + $qsp->psgi_qx($ctx->{env}, undef, \&show_tree_result, $ctx); +} + +# user_cb for SolverGit, called as: user_cb->($result_or_error, $uarg) +sub solve_result { + my ($res, $ctx) = @_; + my $hints = delete $ctx->{hints}; + $res or return html_page($ctx, 404, dbg_log($ctx)); + ref($res) eq 'ARRAY' or return html_page($ctx, 500, dbg_log($ctx)); my ($git, $oid, $type, $size, $di) = @$res; - return show_other($ctx, $res, \$log, $fn) if $type ne 'blob'; - my $path = to_filename($di->{path_b} || $hints->{path_b} || 'blob'); + return show_commit($ctx, $res) if $type eq 'commit'; + return show_tree($ctx, $res) if $type eq 'tree'; + return show_other($ctx, $res) if $type ne 'blob'; + my $path = to_filename($di->{path_b} // $hints->{path_b} // 'blob'); my $raw_link = "(raw)"; if ($size > $MAX_SIZE) { - return stream_large_blob($ctx, $res, \$log, $fn) if defined $fn; - $log = "
Too big to show, download available\n" .
-			"$oid $type $size bytes $raw_link
" . $log; - return html_page($ctx, 200, \$log); + return stream_large_blob($ctx, $res) if defined $ctx->{fn}; + return html_page($ctx, 200, <Too big to show, download available +blob $oid $size bytes $raw_link +EOM } my $blob = $git->cat_file($oid); if (!$blob) { # WTF? my $e = "Failed to retrieve generated blob ($oid)"; - $ctx->{env}->{'psgi.errors'}->print("$e ($git->{git_dir})\n"); - $log = "
$e
" . $log; - return html_page($ctx, 500, \$log); + warn "$e ($git->{git_dir})"; + return html_page($ctx, 500, "
$e
".dbg_log($ctx)) } my $bin = index(substr($$blob, 0, $BIN_DETECT), "\0") >= 0; - if (defined $fn) { + if (defined $ctx->{fn}) { my $h = [ 'Content-Length', $size, 'Content-Type' ]; push(@$h, ($bin ? 'application/octet-stream' : 'text/plain')); return delete($ctx->{-wcb})->([200, $h, [ $$blob ]]); } - if ($bin) { - $log = "
$oid $type $size bytes (binary)" .
-			" $raw_link
" . $log; - return html_page($ctx, 200, \$log); - } + $bin and return html_page($ctx, 200, + "
blob $oid $size bytes (binary)" .
+				" $raw_link
".dbg_log($ctx)); # TODO: detect + convert to ensure validity utf8::decode($$blob); - my $nl = ($$blob =~ tr/\n/\n/); + my $nl = ($$blob =~ s/\r?\n/\n/sg); my $pad = length($nl); - $l->linkify_1($$blob); + ($ctx->{-linkify} //= PublicInbox::Linkify->new)->linkify_1($$blob); my $ok = $hl->do_hl($blob, $path) if $hl; if ($ok) { $blob = $ok; @@ -177,37 +450,38 @@ sub solve_result { $$blob = ascii_html($$blob); } - # using some of the same CSS class names and ids as cgit - $log = "
$oid $type $size bytes $raw_link
" . + my $x = "
blob $oid $size bytes $raw_link
" . "
". - "
" . join('', map {
-			sprintf("% ${pad}u\n", $_)
-		} (1..$nl)) . '
' . - '
'. # pad for non-CSS users - "" . - $l->linkify_2($$blob) . - '' . $log; + "
";
+	$x .= sprintf("% ${pad}u\n", $_) for (1..$nl);
+	$x .= '
'. # pad for non-CSS users + ""; - html_page($ctx, 200, \$log); + # using some of the same CSS class names and ids as cgit + html_page($ctx, 200, $x, $ctx->{-linkify}->linkify_2($$blob), + ''.dbg_log($ctx)); } +# GET /$INBOX/$GIT_OBJECT_ID/s/ +# GET /$INBOX/$GIT_OBJECT_ID/s/$FILENAME sub show ($$;$) { my ($ctx, $oid_b, $fn) = @_; my $qp = $ctx->{qp}; my $hints = $ctx->{hints} = {}; while (my ($from, $to) = each %QP_MAP) { defined(my $v = $qp->{$from}) or next; - $hints->{$to} = $v; + $hints->{$to} = $v if $v ne ''; } - - $ctx->{'log'} = tmpfile("solve.$oid_b"); $ctx->{fn} = $fn; - my $solver = PublicInbox::SolverGit->new($ctx->{-inbox}, + $ctx->{-tmp} = File::Temp->newdir("solver.$oid_b-XXXX", TMPDIR => 1); + open $ctx->{lh}, '+>>', "$ctx->{-tmp}/solve.log" or die "open: $!"; + my $solver = PublicInbox::SolverGit->new($ctx->{ibx}, \&solve_result, $ctx); + $solver->{tmp} = $ctx->{-tmp}; # share tmpdir # PSGI server will call this immediately and give us a callback (-wcb) sub { $ctx->{-wcb} = $_[0]; # HTTP write callback - $solver->solve($ctx->{env}, $ctx->{log}, $oid_b, $hints); + $solver->solve($ctx->{env}, $ctx->{lh}, $oid_b, $hints); }; }