common.py 文件源码

def showStaticWords(firstPage, secondPage):
    """
    Prints words appearing in two different response pages
    """

    infoMsg = "finding static words in longest matching part of dynamic page content"
    logger.info(infoMsg)

    firstPage = getFilteredPageContent(firstPage)
    secondPage = getFilteredPageContent(secondPage)

    infoMsg = "static words: "

    if firstPage and secondPage:
        match = SequenceMatcher(None, firstPage, secondPage).find_longest_match(0, len(firstPage), 0, len(secondPage))
        commonText = firstPage[match[0]:match[0] + match[2]]
        commonWords = getPageWordSet(commonText)
    else:
        commonWords = None

    if commonWords:
        commonWords = list(commonWords)
        commonWords.sort(lambda a, b: cmp(a.lower(), b.lower()))

        for word in commonWords:
            if len(word) > 2:
                infoMsg += "'%s', " % word

        infoMsg = infoMsg.rstrip(", ")
    else:
        infoMsg += "None"

    logger.info(infoMsg)