oraerr.py

#!/usr/bin/python2
"""Display the error message for an Oracle error either from the internet
or from the cache db, based on the error code ("the query") you get when
confronted with the error."""

import sys, os, re, urllib2, urlparse, anydbm
from BeautifulSoup import BeautifulSoup

BASE_URL = "http://download.oracle.com/docs/cd/B19306_01/server.102/b14219/"
TOC_URL = BASE_URL + "toc.htm"
DEFAULT_PREFIX = "ORA"
CACHE_PATH = "oraerr.cache"

db = anydbm.open(CACHE_PATH, "c")

def get_error_message(query):
    query = tidy_query(query)
    return db.get(query) or fetch_error_message(query)

def tidy_query(query):
    # turn a query into the canonical form, or raise an error if not possible.

    if query.isdigit():
        errprefix, errnum = DEFAULT_PREFIX, query
    else:
        errprefix, _, errnum = query.partition("-")

    if not errprefix.isalpha():
        raise ValueError("error prefix must be alphabetical")
    if not errnum or not errnum.isdigit() or len(str(int(errnum))) > 5:
        raise ValueError("error number can't be more then five digits")

    return "%s-%.5d" % (errprefix.upper(), int(errnum))

def fetch_error_message(query):
    url = find_query_url(query)
    if not url:
        raise ValueError("the error you requested does not exist on the server")
    soup = BeautifulSoup(get_html(url))

    try:
        errmsg = strip_tags(soup.find("a", id=query).parent.parent.parent.renderContents())
    except AttributeError:
        if not soup.find("a", id=query):
            return None
        else:
            raise IOError("html page layout is irregular")

    db[query] = errmsg
    return errmsg

def find_query_url(query):
    # Go to the index page, find the section with the
    # query prefix and range, and return its url.

    toc = BeautifulSoup(get_html(TOC_URL))
    errprefix, errnum = query.split("-")
    rangere = re.compile(r"%s-(\d+) to %s-(\d+)" % (errprefix, errprefix))

    for h2 in toc("h2", "tocheader"):
        s = rangere.search(h2.a.renderContents())
        if s and s.group(1) <= errnum <= s.group(2):
            return urlparse.urljoin(BASE_URL, h2.a["href"])
    return None

def get_html(url):
    try:
        return urllib2.urlopen(url).read()
    except urllib2.URLError:
        raise IOError("could not fetch data from server")

def strip_tags(html):
    return re.sub("\n+", "\n", re.sub("<[^>]*>", "", html)).strip()

if __name__ == "__main__":
    APPNAME = os.path.basename(__file__)

    if len(sys.argv) < 2:
        print >> sys.stderr, "usage:", APPNAME, "<prefix>-<number>"
        print >> sys.stderr, "default prefix:", DEFAULT_PREFIX
        sys.exit(2)

    try:
        print >> sys.stderr, "fetching error message...\n"
        print get_error_message(sys.argv[1])
    except (ValueError, IOError), e:
        print >> sys.stderr, "%s: error: %s" % (APPNAME, e)
        sys.exit(1)