view flys-backend/contrib/shpimporter/utils.py @ 4887:1f6e544f7a7f

Importer: Use cp1252 instead of latin-9 to guess filename encodings
author Andre Heinecke <aheinecke@intevation.de>
date Mon, 28 Jan 2013 12:45:41 +0100
parents b457532dae63
children c0a58558b817
line wrap: on
line source
import os
import sys
from shpimporter import DEBUG, INFO, ERROR

SHP='.shp'

def findShapefiles(path):
    shapes = []

    for root, dirs, files in os.walk(path):
        if len(files) == 0:
            continue

        DEBUG("Processing directory '%s' with %i files " % (root, len(files)))

        for f in files:
            idx = f.find(SHP)
            if (idx+len(SHP)) == len(f):
                shapes.append((f.replace(SHP, ''), root + "/" + f))

    return shapes

def getUTF8Path(path):
    """
    Tries to convert path to utf-8 by first checking the filesystemencoding
    and trying the default windows encoding afterwards.
    Returns a valid UTF-8 encoded unicode object or throws a UnicodeDecodeError
    """
    try:
        return unicode.encode(unicode(path, sys.getfilesystemencoding()), "UTF-8")
    except UnicodeDecodeError:
        # Probably European Windows names so lets try again
        return unicode.encode(unicode(path, "cp1252"), "UTF-8")

http://dive4elements.wald.intevation.org