def shorten_title()

in readability/htmls.py [0:0]

• 40 lines of code
• 20 McCabe index (conditional complexity)

def shorten_title(doc):
    title = doc.find(".//title")
    if title is None or title.text is None or len(title.text) == 0:
        return ""

    title = orig = norm_title(title.text)

    candidates = set()

    for item in [".//h1", ".//h2", ".//h3"]:
        for e in list(doc.iterfind(item)):
            if e.text:
                add_match(candidates, e.text, orig)
            if e.text_content():
                add_match(candidates, e.text_content(), orig)

    for item in TITLE_CSS_HEURISTICS:
        for e in doc.cssselect(item):
            if e.text:
                add_match(candidates, e.text, orig)
            if e.text_content():
                add_match(candidates, e.text_content(), orig)

    if candidates:
        title = sorted(candidates, key=len)[-1]
    else:
        for delimiter in [" | ", " - ", " :: ", " / "]:
            if delimiter in title:
                parts = orig.split(delimiter)
                if len(parts[0].split()) >= 4:
                    title = parts[0]
                    break
                elif len(parts[-1].split()) >= 4:
                    title = parts[-1]
                    break
        else:
            if ": " in title:
                parts = orig.split(": ")
                if len(parts[-1].split()) >= 4:
                    title = parts[-1]
                else:
                    title = orig.split(": ", 1)[1]

    if not 15 < len(title) < 150:
        return orig

    return title