#!/usr/bin/python # Copyright (c) 2009 Olivier Berger and Institut TELECOM # This tool is published under the terms of the GNU General Public # License version 3 or above. # This tool is used in order to convert a multi-framed HTML # documentation of an ontology generated with Protege's OWLDoc plugin # into a single HTML file, using htmldoc. # There are 2 steps : # - option 'book' helps generate a .book file for batch processing by # htmldoc : python ../owldoc2htmldoc.py book ../helios_bt/ontologies/index.html helios_bt >book.book # - option 'convert' helps get rid of external links to files not # considered in the book : python ../owldoc2htmldoc.py convert ../helios_bt/ontologies/index.html helios_bt book.html >index.html # This tool was developped in the frame of the Helios project to help # document the ontologies designed in the project. from BeautifulSoup import BeautifulSoup import sys import os import re arguments = sys.argv def print_usage(): print "USAGE:" print "grep.py action ontologies/index.html ontology" print " with action in : book, convert" ############################################### # PROGRAM STARTS HERE if len(arguments) < 2: print_usage() sys.exit(1) action = arguments[1] #print action if not action in ["book", "convert"] : print_usage() sys.exit(1) if action == "book" and len(arguments) != 4: print_usage() sys.exit(1) if action == "convert" and len(arguments) != 5: print_usage() sys.exit(1) filename = arguments[2] #print filename ontology = arguments[3] #print ontology if action == "convert" : bookHtmlFileName = arguments[4] ontologiesDir = os.path.dirname(filename) owldocDir = os.path.dirname(ontologiesDir) classesDir = os.path.normpath(owldocDir + "/classes") #print "classesDir :", classesDir #print "owldocDir :",owldocDir objectPropertiesDir = os.path.normpath(owldocDir + "/objectproperties") dataPropertiesDir = os.path.normpath(owldocDir + "/dataproperties") file = open(filename) html = file.read() soup = BeautifulSoup(html) #print soup.prettify() a = soup.find("a", { "class" : "active-ontology-uri"}) mainOnto = a.string if ontology != mainOnto : print "Error : found active ontology", mainOnto, "doesn't match wanted", ontology sys.exit(1) ontoHtmls=[] classesHtmls=[] objectPropertiesHtmls=[] dataPropertiesHtmls=[] ontoMainHtml = a['href'] #print ontoMainHtml ontoMainFilename = ontologiesDir + "/" + ontoMainHtml #print ontoMainFilename ontoHtmls.append(ontoMainFilename) ontoMainFile = open(ontoMainFilename) html = ontoMainFile.read() soup = BeautifulSoup(html) #print soup.prettify() for a in soup.findAll("a") : if a.string == "Classes" : indexClassesOntoFilename = a["href"] if a.string == "Object Properties": # print a indexObjectPropertiesFilename = a["href"] if a.string == "Data Properties": indexDataPropertiesFilename = a["href"] indexClassesOntoFilename = os.path.normpath(classesDir + "/" + indexClassesOntoFilename) classesHtmls.append(indexClassesOntoFilename) indexClassesOntoFile = open(indexClassesOntoFilename) html = indexClassesOntoFile.read() soup = BeautifulSoup(html) #print soup.prettify() codebox = soup.find("div", { "class" : "codebox"}) #print codebox for a in codebox.findAll("a") : classesHtmls.append(os.path.normpath(classesDir + "/" + a["href"])) indexObjectPropertiesFilename = os.path.normpath(objectPropertiesDir + "/" + indexObjectPropertiesFilename) objectPropertiesHtmls.append(indexObjectPropertiesFilename) indexObjectPropertiesFile = open(indexObjectPropertiesFilename) html = indexObjectPropertiesFile.read() soup = BeautifulSoup(html) #print soup.prettify() codebox = soup.find("div", { "class" : "codebox"}) #print codebox for a in codebox.findAll("a") : objectPropertiesHtmls.append(os.path.normpath(objectPropertiesDir + "/" + a["href"])) indexDataPropertiesFilename = os.path.normpath(dataPropertiesDir + "/" + indexDataPropertiesFilename) dataPropertiesHtmls.append(indexDataPropertiesFilename) indexDataPropertiesFile = open(indexDataPropertiesFilename) html = indexDataPropertiesFile.read() soup = BeautifulSoup(html) #print soup.prettify() codebox = soup.find("div", { "class" : "codebox"}) #print codebox for a in codebox.findAll("a") : dataPropertiesHtmls.append(os.path.normpath(dataPropertiesDir + "/" + a["href"])) ############################################### # Action : book if action == "book" : print '#HTMLDOC 1.8.27' print '-t html -f "book.html" --book --toclevels 3 --no-numbered --toctitle "Table of Contents" --title --linkstyle plain --fontsize 11.0 --fontspacing 1.2 --headingfont Helvetica --bodyfont Times --headfootsize 11.0 --headfootfont Helvetica --charset iso-8859-1 --browserwidth 680 --no-strict --no-overflow' print "about.html" for i in ontoHtmls : print i for i in classesHtmls : print i for i in objectPropertiesHtmls : print i for i in dataPropertiesHtmls : print i ############################################### # Action : convert #################################### # utility functions for action : convert def add_anchor(dict,prefix,file,tag) : # print tag a = tag.parent # print a anchor = None # print a.attrs for name, value in a.attrs : if name == 'name' : anchor = value dict[prefix+file] = anchor dict[file] = anchor def process_html_page(dict, htmlPage, title, path) : file = os.path.basename(htmlPage) classname = os.path.splitext(file)[0] #print file if classname == "index-"+ontology : for t in soup.findAll(text=re.compile(ontology+": "+path+" .*")) : #print t add_anchor(dict, "../"+path+"/", file, t) else : for t in soup.findAll(text=title+": "+classname) : add_anchor(dict, "../"+path+"/", file, t) #################################### # now the convert code if action == "convert" : bookHtmlFile = open(bookHtmlFileName) html = bookHtmlFile.read() soup = BeautifulSoup(html) #print soup.prettify() # remove footers of OWLDoc for p in soup.findAll('p', { "class": "footer"}) : p.extract() # Change style ? # construct a dict of pages and their anchors dict = {} for html in ontoHtmls : file = os.path.basename(html) onto = os.path.splitext(file)[0] for t in soup.findAll(text="Ontology: "+onto) : a = t.parent # print a.attrs anchor = None for name, value in a.attrs : if name == 'name' : anchor = value #print anchor dict["../ontologies/"+file] = anchor dict[file] = anchor for html in classesHtmls : process_html_page(dict,html,"Class","classes") for html in objectPropertiesHtmls : process_html_page(dict,html,"Object Property","objectproperties") for html in dataPropertiesHtmls : process_html_page(dict,html,"Data Property","dataproperties") #print dict # substitute links to external pages by internal links to the anchors for path in dict.keys() : #print "looking for path", path for a in soup.findAll('a', {'href' : path}) : a['href']= "#"+dict[path] # remove href of all links to other classes (other ontologies) for a in soup.findAll('a',href=re.compile('^\.\./.*')) : del(a['href']) # remove href of all links to other pages for a in soup.findAll('a',href=re.compile('^[^#]')) : del(a['href']) # remove the opening of links in other frame for a in soup.findAll('a',target="content") : del(a['target']) # remove blocks for expansion of long lists (already expanded by htmldoc) for a in soup.findAll('a', { "class" : "subsexpand" }, id="showSubs") : a.extract() print soup.prettify()