Wikipedia Scraper with networkX graphing, faster setup than neo4j version

wikipedia_scraper_netx.py

import requests, bs4 as bs, time as t,re,networkx as nx,matplotlib.pyplot as plot
graph = nx.Graph()
def getBSoupFromLink(link):
    return bs.BeautifulSoup(requests.get(link).text,"html.parser")
for x in range(10):
    curLink = "/wiki/Special:Random"
    startingTitle = "Random"
    for x in range(5):
        startingLink = curLink
        bsoup = getBSoupFromLink("https://en.wikipedia.org"+curLink)
        heading = bsoup.find("h1",attrs={"id":"firstHeading"}).text
        tables = bsoup.find_all("table")
        if len(tables) > 0:
            infotable = tables
            for table in infotable:
                table.decompose()
        thumbs = bsoup.find_all("div",class_=True)
        if len(thumbs) > 0:
            thumb = thumbs[0]
            thumb.decompose()
        mwpo = bsoup.find("div",attrs={"class":"mw-parser-output"})
        if mwpo:
            links = mwpo.find_all("a",href=re.compile("^/w[^:#]*$"))#,class_=["mw-redirect",False])
        else:
            break
        for link in links:
            if not ("anguage" in link["href"] or bool(re.search("[#?]",link["href"]))):
                if not link.parent.has_attr("role") and not link.parent.has_attr("class") and not link.parent.name == u"i":
                    curLink = link["href"]
                    break

        print(startingTitle,"->",heading, "\n", curLink, "\n", "-" * 50)
        if not graph.has_node(heading):
            graph.add_node(heading)
        else:
            if not graph.has_edge(startingTitle,heading):
                graph.add_edge(startingTitle,heading)
            print("node " + heading + " already exists")
            break
        if not graph.has_node(startingTitle):
            graph.add_node(startingTitle)
        if not graph.has_edge(startingTitle,heading):
            graph.add_edge(startingTitle,heading)
        else:
            break
        startingTitle = heading
nx.draw_networkx(graph,font_size = 5,nodesize=50)
plot.show()
nx.node_link_data()