cg-log

#!/usr/bin/env bash
#
# Show the list of changes
# Copyright (c) Petr Baudis, 2005.
# Copyright (c) David Woodhouse, 2005.
#
# Display log of changes on a given branch, within a given range of commits,
# and/or concerning given set of files. The log can be further filtered to
# e.g. only changes touching a given string or done by a given user. Several
# output formats are available.
#
# The output will automatically be displayed in a pager unless it is piped to
# a program.
#
# OPTIONS
# -------
# Arguments not interpreted as options will be interpreted as filenames;
# cg-log then displays only changes in those files.
#
# -c::	Colorize
#	Colorize the output. You can customize the colors using the
#	$CG_COLORS environment variable (see below).
#
# -d::	Show diffs against previous commits
#	Accompany each commit with a diff against previous commit.
#	Turns off '-f'.
#
# --diffcore ARGS:: Diffcore arguments to pass the Git diff command
#	Pass the given diffcore arguments the called Git diff command.
#	See e.g. git-diff-tree(1) documentation for the list of possible
#	arguments; '-R', '-B', and '-C' might be of particular interest
#	('-M' is sometimes passed automatically, but not always). This
#	is mostly only relevant in conjunction with the '-d' option.
#
# -f::	List affected files
#	List affected files. (No effect when passed along '-s'.)
#	Turns off '-d'.
#
# -r FROM_ID[..TO_ID]:: Limit to a set of revisions
#	Limit the log information to a set of revisions using either
#	'-r FROM_ID[..TO_ID]' or '-r FROM_ID -r TO_ID'. In both cases the
#	option expects IDs which resolve to commits and will include the
#	specified IDs. If 'TO_ID' is omitted all commits from 'FROM_ID'
#	to the initial commit is shown. If no revisions is specified,
#	the log information starting from 'HEAD' will be shown.
#
# -D DATE:: Limit to revisions newer than given DATE
#	Limit the log information to revisions newer than given DATE,
#	and on second time further restrain it to revisions older than
#	given date. Therefore, '-D "2 days ago" -D "yesterday"' will
#	show all the commits from the day before yesterday.
#
# -m::	End the log at the merge base of the revision set
#	End the log listing at the merge base of the -r arguments
#	to HEAD and 'origin' or the current branch's default remote
#	branch, see `cg-fetch` for details).
#
# -M, --merges:: Show merge commits
#	Display merge commits in the log.
#
# -R, --no-renames:: Do not follow renames
#	This flag is currently no-op. `cg-log` will not follow file history
#	across renames.
#
# -s::	Short output format of the log entries
#	Show the log entries one per line. The entry summary contains
#	information about the commit date, the author, the first line
#	of the commit log and the commit ID. Long author names and commit
#	IDs are trimmed and marked with an ending tilde (~).
#
# --summary:: Group commits by author
#	Generate the changes summary, listing the commit titles grouped
#	by their author. This is also known as a "shortlog", suitable
#	e.g. for contribution summaries of announcements.
#
# -S, --pickaxe STRING:: Limit to changes touching STRING ("pick-axe")
#	List only commits with changes concerning STRING (also known as
#	pick-axe). In other words, only commits where the parent contains
#	STRING and the child does not contain it at the same place in
#	a file or vice versa are shown. The STRING may contain any
#	special characters or even newlines (but you might need to quote
#	it properly when calling `cg-log` from a shell). It is matched
#	verbatim.
#
# -u USERNAME:: Limit to commit where author/committer matches USERNAME
#	List only commits where author or committer contains 'USERNAME'.
#	The search for 'USERNAME' is case-insensitive.
#
# -v::	Verbose header listing
#	By default, only the 'commit' and 'author' headers are shown. This
#	makes `cg-log` show even the other commit headers - 'tree', 'parent',
#	and 'committer'.
#
# ENVIRONMENT VARIABLES
# ---------------------
# PAGER::
#	The pager to display log information in, defaults to `less`.
#
# PAGER_FLAGS::
#	Flags to pass to the pager.
#
# CG_COLORS::
#	Colon-separated list of 'name=color' pairs, where name is
#	one of logcommit, logheader, logauthor, logcommitter,
#	logfilemod, logfileadd, logfiledel, logfileren, logsignoff,
#	logsumauthor, logsumtrim, logsumcommit, logsumdate, default,
#	and value is an ECMA-48 SGR sequence (see e.g. console_codes(4)).
#	You can also customize the diff colors; see `cg-diff` documentation
#	for the appropriate color names.
#
# CG_COLORS_AUTO::
#	Even if -c was passed or specified in ~/.cgrc, if this option
#	is set, use colors only when the output is a terminal and it
#	supports colors.
#
# CG_LESS::
#	This is what the $LESS environment variable value will be set
#	to before invoking $PAGER. It defaults to $LESS concatenated
#	with the `R` and `S` flags to allow displaying of colorized output
#	and to avoid long lines from wrapping when using `-s`.
#
# CONFIGURATION VARIABLES
# -----------------------
# The following GIT configuration file variables are recognized:
#
# log.usecolor::
#	If enabled, colorify the output like with -c if the output
#	is a terminal.
#
# EXAMPLE USAGE
# -------------
# To show a log of changes between two releases tagged as 'releasetag-0.9'
# and 'releasetag-0.10' do:
#
#	$ cg-log -r releasetag-0.9..releasetag-0.10
#
# Similarily, to see which commits are in branch A but not in branch B,
#
#	$ cg-log -r B..A
#
# (meaning "all the commits which newly appear along the way from B to A").
#
# If you see a dubious "if (current->uid = 0)" test in a file and wonder
# about its genesis, you can run
#
#	$ cg-log -d -S "if (current->uid = 0)" filename
#
# to show the commits adding, removing or modifying that string, together
# with the relevant patches (you can obviously refrain from limiting
# the pick-axe to a particular file, but it will make it significantly
# slower).
#
# NOTES
# -----
# The ':' is equivalent to '..' in revisions range specification (to make
# things more comfortable to SVN users). See cogito(7) for more details
# about revision specification.

# Testsuite: TODO

USAGE="cg-log [-D DATE] [-r FROM_ID[..TO_ID]] [-d] [-s | --summary] [OTHER_OPTIONS] [FILE]..."
_git_wc_unneeded=1

. "${COGITO_LIB}"cg-Xlib || exit 1
# Try to fix the annoying "Broken pipe" output. May not help, but apparently
# at least somewhere it does. Bash is broken.
trap exit SIGPIPE

	

setup_colors()
{
	local C="logcommit=32:logheader=32"
	C="$C:logauthor=36:logcommitter=35:logsignoff=33"
	C="$C:logfilemod=34:logfileadd=36:logfiledel=31:logfileren=33"
	C="$C:logsumauthor=36:logsumtrim=35"
	C="$C:logsumcommit=34:logsumdate=32"
	C="$C:default=0"
	[ "$show_diffs" ] && C="$C:logcommit=32;1"
	# Remove bolds from diffcolors - they tended to overshadow cg-log's
	# highlighting, making it less obvious where one commit ends and
	# another appears.
	colorify_setup "$C:${colorify_diffcolors//1;/}"
	collogfile_M0="$collogfilemod"
	collogfile_A0="$collogfileadd"
	collogfile_D0="$collogfiledel"
	collogfile_R0="$collogfileren"
	collogfile_R1="$collogfileadd"
}

print_oneline()
{
	commit="${commit%:*}"
	author="${author% <*}"
	[ "${#author}" -gt 15 ] && author="${author:0:14}$collogsumtrim~"
	sumcommit="${commit:0:12}$collogsumtrim~"

	# We want wordsplitting in the $date here, to get
	# TZ as separate argument.
	date=(${committer#*> })
	showdate ${date[*]} '+%F %H:%M'; date="$_showdate"

	read -r title

	printf "$collogsumcommit%s $collogsumauthor%-15s $collogsumdate%s $coldefault%s\n" \
		"$sumcommit" "$author" "$date" "${title:3}"
}

list_commit_files()
{
	if [ ${#files[@]} -eq 0 ]; then
		echo "    * no changes:"
		echo
		return
	fi
	tree1="$1"
	tree2="$2"
	line=
	sep="    * "
	for (( i=0; i < ${#files[@]}; i++ )); do
		echo -n "$sep"
		sep="$collogfilemod, "
		line="$line$sep${files[$i]}"
		if [ ${#line} -le 74 ]; then
			echo -n "${filecols[$i]}${files[$i]}"
		else
			line="      ${files[$i]}"
			echo "$coldefault"
			echo -n "      ${filecols[$i]}${files[$i]}"
		fi
	done
	echo "$coldefault:"
	echo
}

print_commit_contents()
{
	[ "$list_files" ] && list_commit_files
	echo "$msg"
}

reset_commit_info()
{
	commit=
	tree=
	parents=()
	author=
	committer=
	msg=
	files=()
	filecols=()

	# The $state variable mostly describes what should happen on the next
	# empty line.
	state=printhdr
}

process_commit_line()
{
	if [ "$key" = "%" ] || [ "$key" = "%$collogsignoff" ]; then
		# The fast common case
		[ "$state" = silent ] || msg="$msg    ${rest#?}
"
		return
	fi
	case "$key" in
	"commit"|"diff-tree")
		oldcommit="$commit"
		reset_commit_info
		commit="${rest:0:40}"
		# If we've just seen a commit, we are seeing multiple
		# instances of the same commit caused by git-diff-tree --stdin
		# hitting a merge. We show only the diff against the first
		# parent since heuristically, this is the interesting one.
		# In some cases, this might not be true, but this is hopefully
		# a good general strategy (always except in the change-here-
		# -merge-there-fastforward-here case and doing anything else
		# results in unusably huge file lists etc).
		[ "$commit" = "$oldcommit" ] && state=silent
		;;
	"tree")
		tree="$rest"
		;;
	"parent")
		parents[${#parents[@]}]="$rest"
		;;
	"committer")
		committer="$rest"
		;;
	"author")
		author="$rest"
		;;
	:)
		orest="$rest";
		rest="${orest#*	}"
		crest="${orest%%	*}"
		crest="${crest%[0-9][0-9][0-9]}" # rename similarity
		i=0
		while [ x"$rest" != x"$orest" ]; do
			local l="collogfile_${crest:(-1):1}$i"
			filecols[${#files[@]}]="${!l}"
			files[${#files[@]}]="${rest%%	*}"
			# Multiple tab-separated filenames are present in case
			# of rename entries.
			orest="$rest"
			rest="${rest#*	}"
			i=$((i+1))
		done
		;;
	"")
		if [ "$state" = silent ]; then
			return
		fi
		if [ "$state" = waitdiff ]; then
			# We cannot hook this to ^: since the diff may be empty
			[ "$show_diffs" ] && msg="$msg
"
			state=showcommit
			return
		fi
		if [ "$state" = showcommit ]; then
			print_commit_contents
			[ "$show_diffs" ] && echo
			state=randomjunk
			return
		fi
		if [ "$state" != printhdr ]; then
			die "internal error - state '$state'"
		fi

		if [ "$user" ]; then
			if ! [[ "$author" == *"$user"* || "$committer" == *"$user"* ]]; then
				state=silent
				return
			fi
		fi
		if [ "$oneline" ]; then
			print_oneline
			state=silent
			return
		fi

		merge=
		[ ! "$verbose" -a ${#parents[@]} -gt 1 ] && merge=" (merge)"
		echo "${collogcommit}Commit: ${commit%:*}$merge $coldefault"

		if [ "$verbose" ]; then
			echo "${collogheader}Tree: $tree $coldefault"

			for parent in "${parents[@]}"; do
				echo "${collogheader}Parent: $parent $coldefault"
			done
		fi

		# We want wordsplitting in the $date here, to get
		# TZ as separate argument.
		date=(${author#*> })
		showdate ${date[*]}; pdate="$_showdate"
		[ "$pdate" ] && author="${author%> *}> $pdate"
		echo "${collogauthor}Author: $author $coldefault"

		if [ "$verbose" ]; then
			date=(${committer#*> })
			showdate ${date[*]}; pdate="$_showdate"
			[ "$pdate" ] && committer="${committer%> *}> $pdate"
			echo "${collogcommitter}Committer: $committer $coldefault"
		fi

		echo
		if [ "$difffilter" ]; then
			state=waitdiff
		else
			state=showcommit
		fi
		;;
	esac
}

print_commit_log()
{
	[ "$show_diffs" ] || colorify_diffsed=
	reset_commit_info
	sed -e '
		s/^:/: /
		s/^    \(.*\)/% @\1/
		/^% *@[Ss]igned-[Oo]ff-[Bb]y:.*/ s/^% @\(.*\)/% @'$collogsignoff'\1'$coldefault'/
		/^% *@[Aa][Cc][Kk]ed-[Bb]y:.*/ s/^% @\(.*\)/% @'$collogsignoff'\1'$coldefault'/
	' -e "$colorify_diffsed" | { while IFS=$'\n' read -r line; do
		trap exit SIGPIPE
		if [ "$state" = "showcommit" -a "$show_diffs" -a -n "$line" ]; then
			[ x"${line#% @}" = x"$line" ] || line="    ${line#% @}" # undo sed damage
			[ "$state" = silent ] || msg="$msg$line
"
			continue
		fi
		key="${line%%[ 	]*}"
		rest="${line#*[ 	]}"
		process_commit_line
	done; [ "$state" = "showcommit" ] && print_commit_contents # the last commit
	}
}


colors=
collogheader=
collogauthor=
collogcommitter=
collogfiles=
collogsignoff=
collogsumauthor=
collogsumcommit=
collogsumdate=
collogsumtrim=
coldefault=
difffilter=
followrenames=
neverfollowrenames=
verbose=

list_files=
show_diffs=
diffcore=
id1=
id2=
oneline=
shortlog=
user=
mergebase=
date_from=
date_to=
no_merges=--no-merges
always=--always
diffmerges=
diffpatches=
pickaxe=()

while optparse; do
	if optparse -c; then
		colors=1
	elif optparse -d; then
		show_diffs=1
		difffilter=showdiffs
		diffpatches=-p
		list_files=
	elif optparse -f; then
		list_files=1
		difffilter=listfiles
		show_diffs=
		diffpatches=
	elif optparse -u=; then
		user="$OPTARG"
	elif optparse -r=; then
		if echo "$OPTARG" | fgrep -q '..'; then
			id1="${OPTARG%..*}"
			id2="${OPTARG#*..}"
			# id2 was specified as empty commit, that is HEAD;
			# but leaving it empty now would give the code below
			# wrong idea.
			[ "$id2" ] || id2="HEAD"
		elif echo "$OPTARG" | grep -q ':'; then
			id1="${OPTARG%:*}"
			id2="${OPTARG#*:}"
			[ "$id2" ] || id2="HEAD"
		elif [ -z "$id1" ]; then
			id1="$OPTARG"
		elif [ -z "$id2" ]; then
			id2="$OPTARG"
		else
			die "too many revisions"
		fi
	elif optparse -D=; then
		if [ -z "$date_from" ]; then
			date_from="--max-age=$(date -d "$OPTARG" +%s)" || exit 1
		else
			date_to="--min-age=$(date -d "$OPTARG" +%s)" || exit 1
		fi
	elif optparse -d=; then
		die "the -d option was renamed to -D"
	elif optparse -m; then
		mergebase=1
	elif optparse -M || optparse --merges; then
		no_merges=
		diffmerges=-m
	elif optparse -R || optparse --no-renames; then
		neverfollowrenames=1
	elif optparse -s; then
		oneline=1
	elif optparse -S= || optparse --pickaxe=; then
		always=
		pickaxe=(-S"$OPTARG")
		# The trouble with this is that less behaves really strange.
		# It withholds the output until it reads all the input, for
		# example. So this didn't work out very well so far. :-(
		# pickaxe_less=$'+/\013\022'"${OPTARG}"
		# pickaxe_less="${pickaxe_less%$'\n'*}" # stupid less!
		difffilter=pickaxe
	elif optparse --diffcore=; then
		diffcore="$OPTARG"
	elif optparse --summary; then
		shortlog=1
	elif optparse -v; then
		verbose=1
	else
		optfail
	fi
done


# [ "$pickaxe_less" -a "$show_diffs" ] && _local_CG_LESS="$pickaxe_less"
colorify_detect "$colors" log && setup_colors
if [ "$show_diffs" -a "${ARGS[*]}" ]; then
	#warn "-d is buggy and cannot follow renames yet; implying --no-renames"
	neverfollowrenames=1
fi


# Word splitting is ok here and we want to auto-drop empty dates.
revls="$no_merges $date_from $date_to"


if [ "$mergebase" ]; then
	[ "$id1" ] || id1="HEAD"
	[ "$id2" ] || { id2="$(choose_origin refs/heads "what to log against?")" || exit 1; }

	id1="$(cg-object-id -c "$id1")" || exit 1
	id2="$(cg-object-id -c "$id2")" || exit 1
	conservative_merge_base "$id1" "$id2" || exit 1
	[ "$_cg_base_conservative" ] &&
		warn -b "multiple merge bases, picking the most conservative one"
	id1="$_cg_baselist"

else
	id1="$(cg-object-id -c "$id1")" || exit 1
fi

if [ "$id2" ]; then
	id2="$(cg-object-id -c "$id2")" || exit 1
	revls="$revls ^$id1"
	revlsstart="$id2"
else
	revlsstart="$id1"
fi


if [ "$shortlog" ]; then
	fmt="--pretty=short"
else
	fmt="--pretty=raw"
fi

sep=
if [ "${ARGS[*]}" ]; then
	[ "$neverfollowrenames" ] || followrenames=1
	sep=--
fi

# Translate arguments to relpath:
if [ "$_git_relpath" ]; then
	for (( i=0; i<${#ARGS[@]}; i++ )); do
		ARGS[$i]="$_git_relpath${ARGS[$i]}"
	done
fi

[ "$followrenames" ] && difffilter=followrenames


# A curious pipeline:
rev_extract()
{
	# XXX: Following renames is broken and turns out to be massive
	# performance hog.
#	if [ "$followrenames" ]; then
#		[ "${ARGS[*]}" ] || die "internal error: no files to follow renames on"
#		# We ignore $fmt but that's no biggie, shortlog
#		# will actually work anyway.
#		"${COGITO_LIB}"cg-Xfollowrenames $revls -- \
#			--root --pickaxe-all $diffmerges $diffpatches \
#				$always $diffcore "${pickaxe[@]}" -- \
#			$revlsstart -- "${ARGS[@]}"
#	el
	if [ "$difffilter" ]; then
		git-rev-list $revls $revlsstart $sep "${ARGS[@]}" | \
		git-diff-tree -r --stdin --root --pickaxe-all \
			$diffmerges $diffpatches $always $diffcore $fmt \
			"${pickaxe[@]}"
	else
		git-rev-list $revls $revlsstart $fmt $sep "${ARGS[@]}"
	fi
}

rev_show()
{
	if [ "$shortlog" ]; then
		git-shortlog | pager
	else
		# LESS="S" will prevent less to wrap too long titles
		# to multiple lines; you can scroll horizontally.
		print_commit_log | _local_CG_LESS="S $_local_CG_LESS" pager
	fi
}

rev_extract | rev_show

exit 0