def link_file(self, link):
# cleanup the path
isabs = link.startswith('/')
isdir = link.endswith('/')
link = link.replace('\\', '/')
parts = [p for p in link.split('/') if p and not p == '.']
# TODO fold '..'
link = '/'.join(parts)
if isabs and link != '/': link = '/' + link
if isdir and link != '/': link += '/'
if link.startswith('/'):
# document root
return url_encode('/+docs/' + link.lstrip('/'))
# TODO use script location as root for cgi-bin
# TODO allow alternative document root for cgi-bin
else:
# attachment or external file
try:
file = self.notebook.resolve_file(link, self.path)
if file.ischild(self.notebook.dir):
# attachment
relpath = file.relpath(self.notebook.dir)
# TODO: need abstract interface for this
return url_encode('/+file/' + relpath)
else:
# external file -> file://
return file.uri
except:
# typical error is a non-local file:// uri
return link
def link_file(self, link):
# cleanup the path
isabs = link.startswith('/')
isdir = link.endswith('/')
link = link.replace('\\', '/')
parts = [p for p in link.split('/') if p and not p == '.']
# TODO fold '..'
link = '/'.join(parts)
if isabs and link != '/': link = '/' + link
if isdir and link != '/': link += '/'
if link.startswith('/'):
# document root
return url_encode('/+docs/' + link.lstrip('/'))
# TODO use script location as root for cgi-bin
# TODO allow alternative document root for cgi-bin
else:
# attachment or external file
try:
file = self.notebook.resolve_file(link, self.path)
if file.ischild(self.notebook.dir):
# attachment
relpath = file.relpath(self.notebook.dir)
# TODO: need abstract interface for this
return url_encode('/+file/' + relpath)
else:
# external file -> file://
return file.uri
except:
# typical error is a non-local file:// uri
return link
def _joinuri(names):
# first element must be either drive letter or UNC host
if not re.match(r'^(\w:|\\\\\w)', names[0]):
raise ValueError('Not an absolute path: %s' % '\\'.join(names))
elif re.match(r'^\w:$', names[0]): # Drive letter - e.g. file:///C:/foo
return 'file:///' + names[0] + '/' + url_encode('/'.join(names[1:]))
elif re.match(r'^\\\\\w+$', names[0]): # UNC path - e.g. file://host/share
return 'file://' + url_encode(names[0].strip('\\') + '/' + '/'.join(names[1:]))
def uri(self):
'''File uri property with win32 logic'''
# win32 paths do not start with '/', so add another one
# and avoid url encoding the second ":" in "file:///C:/..."
path = self.path.replace('\\', '/')
if re.match('[A-Za-z]:/', path):
return 'file:///' + path[:2] + url_encode(path[2:])
else:
return 'file:///' + url_encode(path)
def email_page(self, _callback=open_url):
'''Menu action to open an email containing the current page.
Encodes the current page as "mailto:" URI and calls L{open_url()}
to start the preferred email client.
'''
text = ''.join(self.page.dump(format='plain'))
url = 'mailto:?subject=%s&body=%s' % (
url_encode(self.page.name, mode=URL_ENCODE_DATA),
url_encode(text, mode=URL_ENCODE_DATA),
)
_callback(self.widget, url)
def testCreateThumbnail(self):
manager = ThumbnailManager(preferences={})
dir = Dir('./data/pixmaps')
# Test API and functions
#~ for file in dir.list_objects(): # TODO
for filename in dir.list():
file = dir.file(filename)
# Remove and assert thumbnail does not exist
manager.remove_thumbnails(file)
for size in (THUMB_SIZE_NORMAL, THUMB_SIZE_LARGE):
thumbfile = manager.get_thumbnail_file(file, size)
self.assertFalse(thumbfile.exists(), msg="File exists: %s" % thumbfile)
# Get thumbnails - twice, first they don't exist, than they do
for (size, pixels) in (
(THUMB_SIZE_NORMAL, 128),
(THUMB_SIZE_LARGE, 256),
(THUMB_SIZE_NORMAL, 128),
(THUMB_SIZE_LARGE, 256),
):
thumb = manager.get_thumbnail(file, size)
self.assertIsInstance(thumb, gtk.gdk.Pixbuf)
self.assertEqual(thumb.get_width(), pixels)
self.assertEqual(thumb.get_height(), pixels)
self.assertTrue(thumb.get_option('tEXt::Thumb::URI').startswith('file:///'))
# Specific requirement of spec to use file:/// and not file://localhost/
self.assertEqual(thumb.get_option('tEXt::Thumb::URI'), url_encode(file.uri, URL_ENCODE_READABLE))
self.assertEqual(thumb.get_option('tEXt::Thumb::MTime'), str( int( file.mtime() ) ))
thumbfile = manager.get_thumbnail_file(file, size)
self.assertTrue(thumbfile.exists(), msg="Missing file: %s" % thumbfile)
basename = hashlib.md5(file.uri).hexdigest() + '.png'
self.assertEqual(thumbfile.basename, basename)
def pack_urilist(links): text = '' for link in links: if is_url_re.match(link): link = url_encode(link, mode=URL_ENCODE_READABLE) # just to be sure text += '%s\r\n' % link return text.encode()
def link_page(self, link):
try:
page = self.notebook.resolve_path(link, source=self.path)
except PageNameError:
return ''
else:
if page == self.path:
return ''
parent = page.commonparent(self.path)
if parent == self.path:
path = './' + self.path.basename + '/'
downpath = page.relname(parent)
path += downpath
elif parent == page:
uppath = self.path.relname(parent)
path = '../' * (uppath.count(':') + 1)
path += page.basename
else:
uppath = self.path.relname(parent)
downpath = page.relname(parent)
path = '../' * uppath.count(':') or './'
path += downpath
path = encode_filename(path) + self._extension
#~ print '>>>', path
return url_encode(path.replace(' ', '_'))
def link_page(self, link):
try:
page = self.notebook.resolve_path(link, source=self.path)
except PageNameError:
return ''
else:
return url_encode('/' + encode_filename(page.name) + '.html')
def file_object(self, file):
'''Turn a L{File} object in a relative link or URI'''
if file.ischild(self.notebook.folder):
# attachment
relpath = file.relpath(self.notebook.folder)
return url_encode('/+file/' + relpath)
elif self.notebook.document_root \
and file.ischild(self.notebook.document_root):
# document root
relpath = file.relpath(self.notebook.document_root)
return url_encode('/+docs/' + relpath)
# TODO use script location as root for cgi-bin
# TODO allow alternative document root for cgi-bin
else:
# external file -> file://
return file.uri
def get_tmpdir():
'''Get a folder in the system temp dir for usage by zim.
This zim specific temp folder has permission set to be readable
only by the current users, and is touched if it didn't exist yet.
Used as base folder by L{TmpFile}.
@returns: a L{LocalFolder} object for the zim specific tmp folder
'''
# We encode the user name using urlencoding to remove any non-ascii
# characters. This is because sockets are not always unicode safe.
root = tempfile.gettempdir()
name = url_encode(os.environ['USER'], URL_ENCODE_READABLE)
dir = LocalFolder(tempfile.gettempdir()).folder('zim-%s' % name)
try:
dir.touch(mode=0o700) # Limit to single user
os.chmod(dir.path,
0o700) # Limit to single user when dir already existed
# Raises OSError if not allowed to chmod
os.listdir(dir.path)
# Raises OSError if we do not have access anymore
except OSError:
raise AssertionError(
'Either you are not the owner of "%s" or the permissions are un-safe.\n'
'If you can not resolve this, try setting $TMP to a different location.'
% dir.path)
else:
# All OK, so we must be owner of a safe folder now ...
return dir
def runTest(self):
manager = ThumbnailManager(preferences={})
dir = Dir('./data/pixmaps')
# Test API and functions
#~ for file in dir.list_objects(): # TODO
for filename in dir.list():
file = dir.file(filename)
# Remove and assert thumbnail does not exist
manager.remove_thumbnails(file)
for size in (THUMB_SIZE_NORMAL, THUMB_SIZE_LARGE):
thumbfile = manager.get_thumbnail_file(file, size)
self.assertFalse(thumbfile.exists(), msg="File exists: %s" % thumbfile)
# Get thumbnails - twice, first they don't exist, than they do
for (size, pixels) in (
(THUMB_SIZE_NORMAL, 128),
(THUMB_SIZE_LARGE, 256),
(THUMB_SIZE_NORMAL, 128),
(THUMB_SIZE_LARGE, 256),
):
thumb = manager.get_thumbnail(file, size)
self.assertIsInstance(thumb, gtk.gdk.Pixbuf)
self.assertEqual(thumb.get_width(), pixels)
self.assertEqual(thumb.get_height(), pixels)
self.assertTrue(thumb.get_option('tEXt::Thumb::URI').startswith('file:///'))
# Specific requirement of spec to use file:/// and not file://localhost/
self.assertEqual(thumb.get_option('tEXt::Thumb::URI'), url_encode(file.uri, URL_ENCODE_READABLE))
self.assertEqual(thumb.get_option('tEXt::Thumb::MTime'), str( int( file.mtime() ) ))
thumbfile = manager.get_thumbnail_file(file, size)
self.assertTrue(thumbfile.exists(), msg="Missing file: %s" % thumbfile)
basename = hashlib.md5(file.uri).hexdigest() + '.png'
self.assertEqual(thumbfile.basename, basename)
def interwiki_link(link):
'''Convert an interwiki link into an url'''
assert isinstance(link, basestring) and '?' in link
key, page = link.split('?', 1)
url = None
for line in config_file('urls.list'):
if line.startswith(key+' ') or line.startswith(key+'\t'):
url = line[len(key):].strip()
break
else:
list = get_notebook_list()
for name, path in list.get_names():
if name.lower() == key.lower():
url = path + '?{NAME}'
break
if url and is_url_re.match(url):
if not ('{NAME}' in url or '{URL}' in url):
url += '{URL}'
url = url.replace('{NAME}', page)
url = url.replace('{URL}', url_encode(page))
return url
else:
return None
def link_page(self, link):
try:
page = self.notebook.resolve_path(link, source=self.path)
except PageNameError:
return ''
else:
if page == self.path:
return ''
parent = page.commonparent(self.path)
if parent == self.path:
path = './' + self.path.basename + '/'
downpath = page.relname(parent)
path += downpath
elif parent == page:
uppath = self.path.relname(parent)
path = '../' * (uppath.count(':') + 1)
path += page.basename
else:
uppath = self.path.relname(parent)
downpath = page.relname(parent)
path = '../' * uppath.count(':') or './'
path += downpath
path = encode_filename(path) + self._extension
#~ print '>>>', path
return url_encode(path.replace(' ', '_'))
def create_thumbnail(self, file, thumbfile, size):
options = {
'tEXt::Thumb::URI': url_encode(file.uri, mode=URL_ENCODE_READABLE), # No UTF-8 here
'tEXt::Thumb::MTime': str( int( file.mtime() ) ),
}
pixbuf = gtk.gdk.pixbuf_new_from_file_at_size(file.path, size, size)
pixbuf.save(thumbfile.path, 'png', options)
def link_page(self, link):
try:
page = self.notebook.resolve_path(link, source=self.path)
except PageNameError:
return ''
else:
return url_encode('/' + encode_filename(page.name) + '.html')
def file_object(self, file):
'''Turn a L{File} object in a relative link or URI'''
if file.ischild(self.notebook.dir):
# attachment
relpath = file.relpath(self.notebook.dir)
return url_encode('/+file/' + relpath)
elif self.notebook.document_root \
and file.ischild(self.notebook.document_root):
# document root
relpath = file.relpath(self.notebook.document_root)
return url_encode('/+docs/' + relpath)
# TODO use script location as root for cgi-bin
# TODO allow alternative document root for cgi-bin
else:
# external file -> file://
return file.uri
def dump_link(self, tag, attrib, strings=None):
href = self.linker.link(attrib['href'])
href = url_encode(href, URL_ENCODE_READABLE)
if strings:
text = u''.join(strings)
else:
text = href
return ['\\href{%s}{%s}' % (href, text)]
def encode_urls(self):
"""Calls encode_url() on all links that contain urls"""
for link in self.getiterator("link"):
href = link.attrib["href"]
if is_url_re.match(href):
link.attrib["href"] = url_encode(href)
if link.text == href:
link.text = link.attrib["href"]
def pack_urilist(links):
text = ''
for link in links:
link = link.encode('utf-8')
if is_url_re.match(link):
link = url_encode(link, mode=URL_ENCODE_READABLE) # just to be sure
text += '%s\r\n' % link
return text
def create_thumbnail(self, file, thumbfile, size):
options = {
'tEXt::Thumb::URI':
url_encode(file.uri, mode=URL_ENCODE_READABLE), # No UTF-8 here
'tEXt::Thumb::MTime': str(int(file.mtime())),
}
pixbuf = gtk.gdk.pixbuf_new_from_file_at_size(file.path, size, size)
pixbuf.save(thumbfile.path, 'png', options)
def dump_link(self, tag, attrib, strings=None):
href = self.linker.link(attrib['href'])
href = url_encode(href, URL_ENCODE_READABLE)
if strings:
text = ''.join(strings)
else:
text = href
return ['\\href{%s}{%s}' % (href, text)]
def encode_urls(self, mode=URL_ENCODE_READABLE):
'''Calls encode_url() on all links that contain urls.
See zim.parsing for details. Modifies the parse tree.
'''
for link in self._etree.getiterator('link'):
href = link.attrib['href']
if href and is_url_re.match(href):
link.attrib['href'] = url_encode(href, mode=mode)
if link.text == href:
link.text = link.attrib['href']
def encode_urls(self, mode=URL_ENCODE_READABLE):
'''Calls encode_url() on all links that contain urls.
See zim.parsing for details. Modifies the parse tree.
'''
for link in self._etree.getiterator('link'):
href = link.attrib['href']
if is_url_re.match(href):
link.attrib['href'] = url_encode(href, mode=mode)
if link.text == href:
link.text = link.attrib['href']
def normalize_win32_share(path):
if os.name == 'nt':
if path.startswith('smb://'):
# smb://host/share/.. -> \\host\share\..
path = path[4:].replace('/', '\\')
path = url_decode(path)
else:
if path.startswith('\\\\'):
# \\host\share\.. -> smb://host/share/..
path = 'smb:' + path.replace('\\', '/')
path = url_encode(path)
return path
def _link_notebook(self, link):
if link.startswith('zim+'):
link = link[4:]
if '?' in link:
link, path = link.split('?')
# FIXME: code below is not robust because we don't know the
# storage mode of linked notebook...
path = url_decode(path) # was already encoded by interwiki_link()
path = encode_filename(path).replace(' ', '_')
return link + '/' + url_encode(path) + '.txt'
else:
return link
def dump_link(self, tag, attrib, strings=None):
# TODO: how do you do page links within an exported document
# use \label{} per page start and \ref to link it ??
href = self.linker.link(attrib['href'])
if href.startswith('#'):
return ['\\ref{%s}' % href.lstrip('#')]
else:
href = url_encode(href, URL_ENCODE_READABLE)
if strings:
text = ''.join(strings)
else:
text = href
return ['\\href{%s}{%s}' % (href, text)]
def link_notebook(self, url):
if url.startswith('zim+'):
url = url[4:]
if '?' in url:
uri, path = url.split('?')
# FIXME: code below is not robust because we don't know the
# storage mode of linked notebook...
path = url_decode(path) # was already encoded by interwiki_link()
path = encode_filename(path).replace(' ', '_')
return uri + '/' + url_encode(path) + '.txt'
else:
return url
def dump_img(self, tag, attrib, strings=None):
src = attrib['src'] or ''
alt = attrib.get('alt')
opts = []
items = sorted(attrib.items())
for k, v in items:
if k in ('src', 'alt') or k.startswith('_'):
continue
elif v: # skip None, "" and 0
data = url_encode(str(v), mode=URL_ENCODE_DATA)
opts.append('%s=%s' % (k, data))
if opts:
src += '?%s' % '&'.join(opts)
if alt:
return ('{{', src, '|', alt, '}}')
else:
return ('{{', src, '}}')
def dump_img(self, tag, attrib, strings=None):
src = attrib['src']
alt = attrib.get('alt')
opts = []
items = attrib.items()
items.sort() # unit tests don't like random output
for k, v in items:
if k in ('src', 'alt') or k.startswith('_'):
continue
elif v: # skip None, "" and 0
data = url_encode(unicode(v), mode=URL_ENCODE_DATA)
opts.append('%s=%s' % (k, data))
if opts:
src += '?%s' % '&'.join(opts)
if alt:
return ('{{', src, '|', alt, '}}')
else:
return('{{', src, '}}')
def interwiki_link(link):
'''Convert an interwiki link into an url'''
assert isinstance(link, str) and '?' in link
key, page = link.split('?', 1)
lkey = key.lower()
# First check known notebooks
list = get_notebook_list()
info = list.get_interwiki(key)
if info:
url = 'zim+' + info.uri + '?{NAME}'
# Then search all "urls.list" in config and data dirs
else:
url = None
files = XDGConfigFileIter(
'urls.list') # FIXME, shouldn't this be passed in ?
for file in files:
for line in file.readlines():
if line.startswith('#') or line.isspace():
continue
try:
mykey, myurl = line.split(None, 1)
except ValueError:
continue
if mykey.lower() == lkey:
url = myurl.strip()
break
if url is not None:
break
# Format URL
if url:
if not ('{NAME}' in url or '{URL}' in url):
url += '{URL}'
url = url.replace('{NAME}', page)
url = url.replace('{URL}', url_encode(page))
return url
else:
return None
def normalize_win32_share(path):
'''Translates paths for windows shares in the platform specific
form. So on windows it translates C{smb://} URLs to C{\\host\share}
form, and vice versa on all other platforms.
Just returns the original path if it was already in the right form,
or when it is not a path for a share drive.
@param path: a filesystem path or URL
@returns: the platform specific path or the original input path
'''
if os.name == 'nt':
if path.startswith('smb://'):
# smb://host/share/.. -> \\host\share\..
path = path[4:].replace('/', '\\')
path = url_decode(path)
else:
if path.startswith('\\\\'):
# \\host\share\.. -> smb://host/share/..
path = 'smb:' + url_encode(path.replace('\\', '/'))
return path
def page_object(self, path):
'''Turn a L{Path} object in a relative link or URI'''
return url_encode('/' + encode_filename(path.name) + '.html')
def resource(self, path):
return url_encode('/+resources/%s' % path)
def icon(self, name):
return url_encode('/+resources/%s.png' % name)
def resource(self, path):
return url_encode('/+resources/%s' % path)
def _filepath(self, file, ref):
relpath = file.relpath(ref, allowupward=True)
if relpath and not relpath.startswith('.'):
relpath = './' + relpath
return url_encode(relpath) or file.uri
def icon(self, name):
return url_encode('/+resources/%s.png' % name)
def _joinuri(names):
if names[0][0] == '/':
return 'file://' + url_encode('/'.join(names))
else:
return 'file:///' + url_encode('/'.join(names))
def _filepath(self, file, ref):
relpath = file.relpath(ref, allowupward=True)
if relpath and not relpath.startswith('.'):
relpath = './' + relpath
return url_encode(relpath) or file.uri
def set_in_main_process(in_main_process):
'''Set whether we are in the main process or not'''
global _IN_MAIN_PROCESS
_IN_MAIN_PROCESS = in_main_process
_m = hashlib.md5()
_m.update(zim.ZIM_EXECUTABLE.encode('UTF-8'))
key = zim.__version__ + '-' + _m.hexdigest()[:8]
# Make name specific for the install location
# But don't worry about collisons, first few bytes should do it
if sys.platform == 'win32':
# Windows named pipe
userstring = url_encode(os.environ['USER'], URL_ENCODE_READABLE)
SERVER_ADDRESS = '\\\\.\\pipe\\zim-%s-%s' % (userstring, key)
SERVER_ADDRESS_FAMILY = 'AF_PIPE'
from multiprocessing.connection import PipeListener
Listener = PipeListener
else:
# Unix domain socket
runtime_dir = os.environ.get('XDG_RUNTIME_DIR')
if runtime_dir is None:
# Use tmpdir as fallback
import tempfile
tmpdir = tempfile.gettempdir()
userstring = url_encode(os.environ['USER'], URL_ENCODE_READABLE)
runtime_dir = tmpdir + '/zim-' + userstring
logger.warn('XDG_RUNTIME_DIR is not set, using %s as a fallback' %
runtime_dir)
def page_object(self, path):
'''Turn a L{Path} object in a relative link or URI'''
return url_encode('/' + encode_filename(path.name) + '.html')
def _dump_children(self, list, output, istoplevel=False):
for element in list.getchildren():
text = html_encode(element.text)
if not element.tag == 'pre':
# text that goes into the element
# always encode excepts for <pre></pre>
text = encode_whitespace(text)
if element.tag == 'h':
tag = 'h' + str(element.attrib['level'])
if self.isrtl(element):
output += ['<', tag, ' dir=\'rtl\'>', text, '</', tag, '>']
else:
output += ['<', tag, '>', text, '</', tag, '>']
elif element.tag == 'p':
tag = 'p'
if self.isrtl(element):
tag += ' dir=\'rtl\''
if 'indent' in element.attrib:
level = int(element.attrib['indent'])
tag += ' style=\'padding-left: %ipt\'' % (30 * level)
output += ['<', tag, '>\n', text]
self._dump_children(element, output) # recurs
output.append('</p>\n')
elif element.tag == 'pre':
if self.isrtl(element):
output += ['<pre dir=\'rtl\'>\n', text, '</pre>\n']
else:
output += ['<pre>\n', text, '</pre>\n']
elif element.tag is 'ul':
output += ['<ul>\n', text]
self._dump_children(element, output) # recurs
output.append('</ul>\n')
elif element.tag == 'li':
if 'bullet' in element.attrib and element.attrib['bullet'] != '*':
icon = url_encode(self.linker.icon(element.attrib['bullet']))
output += ['<li style="list-style-image: url(%s)">' % icon, text]
else:
output += ['<li>', text]
self._dump_children(element, output) # recurs
output.append('</li>\n')
elif element.tag == 'img':
src = url_encode(self.linker.img(element.attrib['src']))
opt = ''
for o in ('width', 'height'):
if o in element.attrib and int(element.attrib[o]) > 0:
opt = ' %s="%s"' % (o, element.attrib[o])
output.append('<img src="%s" alt="%s"%s>' % (src, text, opt))
elif element.tag == 'link':
href = url_encode(self.linker.link(element.attrib['href']))
title = text.replace('"', '"')
output.append('<a href="%s" title="%s">%s</a>' % (href, title, text))
elif element.tag in ['emphasis', 'strong', 'mark', 'strike', 'code']:
if element.tag == 'mark': tag = 'u'
elif element.tag == 'emphasis': tag = 'em'
else: tag = element.tag
output += ['<', tag, '>', text, '</', tag, '>']
else:
assert False, 'Unknown node type: %s' % element
if not element.tail is None:
tail = html_encode(element.tail)
if not (istoplevel and tail.isspace()):
# text in between elements, skip encoding
# for whitespace between headings, paras etc.
tail = encode_whitespace(tail)
output.append(tail)
def uri(self):
'''File uri property'''
return 'file://' + url_encode(self.path)
