gistfile1.py

ADDRESSES_RE = re.compile(r"""(?x)
    (?<!-|/|:|,|\.|\$) # These various characters are not allowed before an address/intersection.
    \b

    # Ignore things that look like dates -- e.g., "21 May 2009".
    # This is a problem e.g. in cases where there's a May Street.
    (?!
        \d+\s+
        (?:January|February|March|April|May|June|July|August|September|October|November|December)
        ,?\s+
        \d\d\d\d
    )

    # Ignore intersections that are prefixed by "University of", like
    # "University of Texas at Austin". This is a common false positive.
    (?<!
        [Uu][Nn][Ii][Vv][Ee][Rr][Ss][Ii][Tt][Yy]\s[Oo][Ff]\s
    )

    (?:
        # SEGMENT ("FOO BETWEEN BAR AND BAZ")
        (?:
            %(STREET_NAME_CAPTURE)s (,?\ + between \ +) %(STREET_NAME_CAPTURE)s (,?\ + and \ +) %(STREET_NAME_CAPTURE)s
            |
            %(STREET_NAME_CAPTURE)s (,?\ + from \ +) %(STREET_NAME_CAPTURE)s (,?\ + to \ +) %(STREET_NAME_CAPTURE)s
        )

        |

        # BLOCK/ADDRESS
        (?:
            (
                (?:
                    (?:\d+|[Ff][Ii][Rr][Ss][Tt])[-\ ]
                        (?:(?:[Nn][Oo][Rr][Tt][Hh]|[Ss][Oo][Uu][Tt][Hh]|[Ee][Aa][Ss][Tt]|[Ww][Ee][Ss][Tt])\ )?
                    [Bb][Ll][Oo][Cc][Kk]\ [Oo][Ff]
                    |
                    \d+\ *-\ *\d+
                    |
                    \d+
                )
                \ +
            )
            %(STREET_NAME_CAPTURE)s

            # ignore the intersection in parenthesis so that it's not picked
            # up as a separate location. We do this by consuming the string
            # but *not* capturing it.
            (?:
                \ +
                \(?
                between
                \ +
                %(STREET_NAME_NOCAPTURE)s
                \ +
                and
                \ +
                %(STREET_NAME_NOCAPTURE)s
                \)?
            )?
        )

        |

        # INTERSECTION
        (?:
            # Common intersection prefixes. They're included here so that the
            # regex doesn't include them as part of the street name.
            (?:
                (?:
                    [Nn]ear |
                    [Aa]t |
                    [Oo]n |
                    [Tt]o |
                    [Aa]round |
                    [Ii]ntersection\ of |
                    [Cc]orner\ of |
                    [Aa]rea\ of |
                    [Aa]reas?\ surrounding |
                    vicinity\ of |
                    ran\ down |
                    running\ down |
                    crossed
                )
                \ +
            )?
            \b
            (?:%(STREET_NAME_CAPTURE)s)
            (\ +)
            (
                (?:
                    [Aa][Nn][Dd] |
                    [Aa][Tt] |
                    [Nn][Ee][Aa][Rr] |
                    & |
                    [Aa][Rr][Oo][Uu][Nn][Dd] |
                    [Tt][Oo][Ww][Aa][Rr][Dd][Ss]? |
                    [Oo][Ff][Ff] |
                    (?:[Jj][Uu][Ss][Tt]\ )?(?:[Nn][Oo][Rr][Tt][Hh]|[Ss][Oo][Uu][Tt][Hh]|[Ee][Aa][Ss][Tt]|[Ww][Ee][Ss][Tt])\ [Oo][Ff] |
                    (?:[Jj][Uu][Ss][Tt]\ )?[Pp][Aa][Ss][Tt]
                )
                \ +
            )
            (?:%(STREET_NAME_CAPTURE)s)
        )
    )

    # OPTIONAL CITY SUFFIX
    (?:
        (?:
            ,?\s+in |
            ,
        )
        \s+

        # CITY NAME
        (
            [A-Z][a-z][A-Za-z]*                   # One initial-capped word
            (?:
                ,?\ Jr\.?,?
                |
                \ [A-Z][a-z][A-Za-z]*
                |
                -[A-Za-z]+                        # Hyphenated words (e.g. "Croton-on-Hudson" in NY)
            ){0,4}  # Initial-capped words
        )
    )?