def doc_title_with_good_suffix(doc):
"""Return title (perhaps unchanged) which reflects the MIME type.
For example, a text/plain 'Readme' will be mapped to 'Readme.txt'."""
orig_filename, orig_ext = os.path.splitext(originals_filename_for_doc(doc))
wpc = _check_for_webpage_complete(doc.originals_path())
if wpc:
orig_filespec, orig_ext = os.path.splitext(wpc)
filename = doc.get_metadata('title') or doc.original_name() or doc.id
filename = os.path.basename(filename).strip(' \t.,-;')
filename = re.sub('[^A-Za-z0-9 \._-]+', '_', filename) # sanitize charset
assert '\r' not in filename
assert '\n' not in filename
orig_ext = orig_ext.lower()
# e.g. for 'ReadMe', orig_ext == ''
if orig_ext and \
filename.lower().endswith(orig_ext) and \
is_good_extension(orig_ext):
return filename
# Popular entries from mimetypes.common_types not found in CONTENT_TYPES:
# image/pict .{pct,pic,pict}, application/rtf .rtf (no uplib parser)
# Popular entries from mimetypes.types_map not found in CONTENT_TYPES:
# audio .aiff + many more, video .avi + many more, text .bat .h .c .css,
# html .htm .xml, msword .dot, image .bmp .jpeg .tiff, postscript .eps
# Originals lacked an extension, or had a bad extension. Synthesize one.
fallback_ext = get_extension_for_type(doc.get_metadata('apparent-mime-type') or \
doc.get_metadata('content-type') or \
'text/plain')
ext = ((orig_ext and is_good_extension(orig_ext) and orig_ext) or fallback_ext)
assert is_good_extension(ext)
if not ext.startswith("."):
ext = "." + ext
return filename + ext
def is_good_extension(ext):
"""Boolean, true if uploading with this file extension could succeed.
That is, after mapping the extension to a content-type, we will want
extensions/UploadDocument.py's _add_internal to find it in CONTENT_TYPES.
"""
if ext.startswith('.'): # e.g. might be ''
ext = ext[1:]
# get_content_type() punts on '.JPG' &c., letting mimetypes.guess_type() deal with it.
ext = ext.lower()
ct = get_content_type('foo.' + ext)
if ct == 'application/octet-stream':
return False # This binary type is absolutely unacceptable to _add_internal.
return get_extension_for_type(ct) == ext and ext in CONTENT_TYPES.values()
def _separate_images (html):
images = {}
counter = 0
m = _DATA_URI_PATTERN.search(html)
while m:
maintype = m.group("maintype")
subtype = m.group("subtype")
encoding = m.group("encoding")
params = m.group("params")
data = m.group("data")
content_type = "%s/%s" % (maintype, subtype)
if encoding == "base64":
data = base64.decodestring(data)
image_name = "image-%s.%s" % (counter, get_extension_for_type(content_type))
counter += 1
images[image_name] = (content_type, data)
html = html[:m.start()] + ('src="images/%s"' % image_name) + html[m.end():]
m = _DATA_URI_PATTERN.search(html)
return html, images
