#!/bin/sh -ef
export LC_ALL=C

cd "${workdir:?}"
[ -n "$*" ] || set -- -ap
rpmelfsym "$@" >sym
awk -F'\t' '$3~/[DT]/' sym >def
sort -t$'\t' -k4 -o def def

# avoid most frequent symbols
nelf=`sort -u -k2,2 def |wc -l`
cut -f4 def |uniq -cd |sort -n >freq
awk -v n=$nelf '$1<sqrt(3*n){print$NF}' freq |grep -vf /dev/fd/3 >defsym 3<<'__EOF__'
__cxa_
__cxxabiv1
__gnu_cxx
__gxx_personality
^__dynamic_cast$
^_Znwj$
^_ZdlPv$
^_Znaj$
^_ZdaPv$
^_ZNSt[0-9]
^_ZNKSt[0-9]
^_ZSt[0-9]
^_Unwind_
^__register_frame
^yy[a-z_]
__EOF__
sort -o defsym -u defsym
join -t$'\t' -14 -21 -o '1.1 1.2 1.3 1.4' def defsym >xdef

# product ::= pkg1 elf1 pkg2 elf2 symtype symname
join -t$'\t' -j 4 -o '1.1 1.2 1.3 2.1 2.2 2.3 0' xdef xdef |
	awk -F'\t' 'function prio(f) {
			if (f~"^/lib/lib[^/]+[.]so[^/]*$") return 1
			if (f~"^/usr/lib/lib[^/]+[.]so[^/]*$") return 2
			if (f~"^/usr/X11R./lib/lib[^/]+[.]so[^/]*$") return 3
			if (f~"lib/lib[^/]+[.]so[^/]*$") return 4
			if (f~"/lib[^/]+[.]so[^/]*$") return 5
			if (f~"[^/][.]so[^/]*$") return 6
			return 9
		}
		($2<$5 && $3==$6) {
			OFS="\t"
			pd = prio($2)-prio($5)
			if (pd<0 || pd==0 && length($2)<=length($5))
				print $1,$2,$4,$5,$6,$7
			else
				print $4,$5,$1,$2,$6,$7 }' >product

# suspect ::= pkg1 elf1 pkg2 elf2 log2(nsym)
# I need logarithmic scale here to facilitate diffing
#	3-5 symbols (approx. 4)		log2 = 2
#	6-11 symbols (approx. 8)	log2 = 3
#	12-22 symbols (approx. 16)	log2 = 4
#	23-45 symbols (approx. 32)	log2 = 5
#	46-90 symbols (approx. 64)	log2 = 6
#	91-181 symbols (approx. 128)	log2 = 7
#	182-362 symbols (approx. 256)	log2 = 8
cut -f-4 product |sort |uniq -cd |sed 's/^ *\([0-9]\+\) \+\(.\+\)$/\2\t\1/' |
	awk -F'\t' 'BEGIN{OFS=FS}($NF=int(log($NF)/log(2)+0.5))>1' >suspect
sort -o suspect -u suspect
cat suspect
cd - >/dev/null