src/third_party/llvm-project/llvm/utils/DSAextract.py

#! /usr/bin/python

#this is a script to extract given named nodes from a dot file, with
#the associated edges.  An edge is kept iff for edge x -> y
# x and y are both nodes specified to be kept.

#known issues: if a line contains '->' and is not an edge line
#problems will occur.  If node labels do not begin with
#Node this also will not work.  Since this is designed to work
#on DSA dot output and not general dot files this is ok.
#If you want to use this on other files rename the node labels
#to Node[.*] with a script or something.  This also relies on
#the length of a node name being 13 characters (as it is in all
#DSA dot output files)

#Note that the name of the node can be any substring of the actual
#name in the dot file.  Thus if you say specify COLLAPSED
#as a parameter this script will pull out all COLLAPSED
#nodes in the file

#Specifying escape characters in the name like \n also will not work, 
#as Python
#will make it \\n, I'm not really sure how to fix this

#currently the script prints the names it is searching for
#to STDOUT, so you can check to see if they are what you intend

import re
import string
import sys


if len(sys.argv) < 3:
	print 'usage is ./DSAextract <dot_file_to_modify> \
			<output_file> [list of nodes to extract]'

#open the input file
input = open(sys.argv[1], 'r')

#construct a set of node names
node_name_set = set()
for name in sys.argv[3:]:
	node_name_set |= set([name])

#construct a list of compiled regular expressions from the 
#node_name_set
regexp_list = []
for name in node_name_set:
	regexp_list.append(re.compile(name))

#used to see what kind of line we are on
nodeexp = re.compile('Node')
#used to check to see if the current line is an edge line
arrowexp = re.compile('->')

node_set = set()

#read the file one line at a time
buffer = input.readline()
while buffer != '':
	#filter out the unnecessary checks on all the edge lines
	if not arrowexp.search(buffer):
		#check to see if this is a node we are looking for
		for regexp in regexp_list:
			#if this name is for the current node, add the dot variable name
			#for the node (it will be Node(hex number)) to our set of nodes
			if regexp.search(buffer):
				node_set |= set([re.split('\s+',buffer,2)[1]])
				break
	buffer = input.readline()


#test code
#print '\n'

print node_name_set

#print node_set
	

#open the output file
output = open(sys.argv[2], 'w')
#start the second pass over the file
input = open(sys.argv[1], 'r')

buffer = input.readline()
while buffer != '':
	#there are three types of lines we are looking for
	#1) node lines, 2) edge lines 3) support lines (like page size, etc)
	
	#is this an edge line?
	#note that this is no completely robust, if a none edge line
	#for some reason contains -> it will be missidentified
	#hand edit the file if this happens
	if arrowexp.search(buffer):
		#check to make sure that both nodes are in the node list
		#if they are print this to output
		nodes = arrowexp.split(buffer)
		nodes[0] = string.strip(nodes[0])
		nodes[1] = string.strip(nodes[1])
		if nodes[0][:13] in node_set and \
				nodes[1][:13] in node_set:
					output.write(buffer)
	elif nodeexp.search(buffer): #this is a node line
		node = re.split('\s+', buffer,2)[1]
		if node in node_set:
			output.write(buffer)
	else: #this is a support line
		output.write(buffer)
	buffer = input.readline()