excerpt.py

import re, collections

p = re.compile(r"(<em>.+</em>|[^\s]+)")

EXAMPLE = "Experienced Software Developer Specialties: Software Engineering :Design,Development,Test Programming :Java/J2EE,C#,C++,Erlang,SQL Development: Servlet, JSP, JSF, RESTful web services,JSON, Spring , Hibernate, Jackrabbit, Java SE ( AWT, Swing, Thread, JDBC ,RMI), J2EE (EJB, JSP, Webservice), ASP.NET Test :TDD,Junit,Mokito,Jmeter Software development methodology : Agile(XP, Scrum) DBMS :SQL Server, MySQL, Oracle <em>Distributed Systems</em> Big Data Processing :Hadoop MapReduce"

MAX_LEN = 50

EM_INVISIBLE_CHARS = 9

def excerpt(x):
    band = collections.deque()
    queue_len = 0
    center = -1
    toks = re.findall(p, x)
    for s in toks:
        band.append(s)
        queue_len += len(s)
        while queue_len > MAX_LEN:
            removed = band.popleft()
            queue_len -= len(removed)
        if s.startswith("<em>"):
            center = queue_len - len(s) if center == -1 else (center + queue_len) / 2
            queue_len -= EM_INVISIBLE_CHARS
        elif center != -1 and queue_len - center > MAX_LEN / 2:
            yield list(band)
            band.clear()
            queue_len = 0
            center = -1

    if center != -1:
        yield list(band)



res = ([" ".join(x) for x in excerpt(EXAMPLE)])

for x in res:
    print(len(x.replace("<em>", "").replace("</em>", "")))
    print(x)
    print()

"... :SQL Server, MySQL, Oracle Distributed Systems Big Data..."