def getTable(self, id, format='odt'):
"""Returns the table into a new 'format' file."""
try:
template_path = getTemplatePath(format)
template = ZipFile(template_path)
content_xml = etree.fromstring(template.read('content.xml'))
template.close()
table_list = self.document.parsed_content.xpath(
'//table:table[@table:name="%s"]' % id,
namespaces=self.document.parsed_content.nsmap)
if len(table_list) == 0:
return None
table = table_list[0]
# Next line do this
# <office:content><office:body><office:text><table:table>
content_xml[-1][0].append(table)
# XXX: Next line replace the <office:automatic-styles> tag. This include
# a lot of unused style tags. Will be better detect the used styles and
# include only those.
content_xml.replace(content_xml[-2],
self.document.parsed_content[-2])
odf_document = self._odfWithoutContentXml(format)
odf_document.writestr('content.xml', etree.tostring(content_xml))
odf_document_as_string = odf_document.fp
odf_document.close()
odf_document_as_string.seek(0)
return odf_document_as_string.read()
except Exception, e:
logger.error(e)
return None
def convert(self, destination_format):
""" Convert the inputed file to output as format that were informed """
# XXX This implementation could use ffmpeg -i pipe:0, but
# XXX seems super unreliable currently and it generates currupted files in
# the end
logger.debug("FfmpegConvert: %s > %s" % (self.input.source_format, destination_format))
output_url = mktemp(suffix=".%s" % destination_format,
dir=self.input.directory_name)
command = ["ffmpeg",
"-i",
self.input.getUrl(),
"-y",
output_url]
# XXX ffmpeg has a bug that needs this options to work with webm format
if destination_format == "webm":
command.insert(3, "32k")
command.insert(3, "-ab")
try:
stdout, stderr = Popen(command,
stdout=PIPE,
stderr=PIPE,
close_fds=True,
env=self.environment).communicate()
self.input.reload(output_url)
if len(self.input.getContent()) == 0:
logger.error(stderr.split("\n")[-2])
return self.input.getContent()
finally:
self.input.trash()
def run_convert(self,
filename='',
data=None,
meta=None,
extension=None,
orig_format=None):
"""Method to support the old API. Wrapper getFileMetadataItemList but
returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
try:
response_dict = {}
metadata_dict = self.getFileMetadataItemList(data,
extension,
base_document=True)
response_dict['meta'] = metadata_dict
# XXX - Backward compatibility: Previous API expects 'mime' now we
# use 'MIMEType'
response_dict['meta']['mime'] = response_dict['meta']['MIMEType']
response_dict['data'] = response_dict['meta']['Data']
response_dict['mime'] = response_dict['meta']['MIMEType']
del response_dict['meta']['Data']
return (200, response_dict, "")
except Exception, e:
import traceback
traceback.print_exc()
logger.error(e)
return (402, {}, e.args[0])
def run_convert(self, filename='', data=None, meta=None, extension=None,
orig_format=None):
"""Method to support the old API. Wrapper getFileMetadataItemList but
returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
try:
response_dict = {}
metadata_dict = self.getFileMetadataItemList(data, extension,
base_document=True)
response_dict['meta'] = metadata_dict
# XXX - Backward compatibility: Previous API expects 'mime' now we
# use 'MIMEType'
response_dict['meta']['mime'] = response_dict['meta']['MIMEType']
response_dict['data'] = response_dict['meta']['Data']
response_dict['mime'] = response_dict['meta']['MIMEType']
del response_dict['meta']['Data']
return (200, response_dict, "")
except Exception, e:
import traceback; traceback.print_exc()
logger.error(e)
return (402, {}, e.args[0])
def getChapterItem(self, chapter_id):
"""Return the chapter in the form of (title, level)."""
chapter_list = self._getChapterList()
try:
chapter = chapter_list[chapter_id].encode('utf-8')
return [chapter, chapter_id]
except IndexError:
msg = "Unable to find chapter %s at chapter list." % chapter_id
logger.error(msg)
return None
def getAllowedTargetItemList(self, content_type):
"""Wrapper getAllowedExtensionList but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
response_dict = {}
try:
extension_list = self.getAllowedExtensionList({"mimetype": content_type})
response_dict['response_data'] = extension_list
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def getAllowedTargetItemList(self, content_type):
"""Wrapper getAllowedExtensionList but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
response_dict = {}
try:
extension_list = self.getAllowedExtensionList(
{"mimetype": content_type})
response_dict['response_data'] = extension_list
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def run_setmetadata(self, filename='', data=None, meta=None,
extension=None, orig_format=None):
"""Wrapper updateFileMetadata but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
response_dict = {}
try:
response_dict['data'] = self.updateFileMetadata(data, extension, meta)
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def getAllowedTargetItemList(self, content_type):
"""Wrapper getAllowedExtensionList but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
response_dict = {}
try:
mimetype_list = self.getAllowedConversionFormatList(content_type)
extension_list = []
for m, t in mimetype_list:
ext = BBB_guess_extension(m, t)
if ext:
extension_list.append((ext.lstrip('.'), t))
response_dict['response_data'] = extension_list
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def run_getmetadata(self, filename='', data=None, meta=None,
extension=None, orig_format=None):
"""Wrapper for getFileMetadataItemList.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
response_dict = {}
try:
response_dict['meta'] = self.getFileMetadataItemList(data, extension)
# XXX - Backward compatibility: Previous API expects 'title' now
# we use 'Title'"
response_dict['meta']['title'] = response_dict['meta']['Title']
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def getAllowedTargetItemList(self, content_type):
"""Wrapper getAllowedExtensionList but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
response_dict = {}
try:
mimetype_list = self.getAllowedConversionFormatList(content_type)
extension_list = []
for m, t in mimetype_list:
ext = BBB_guess_extension(m, t)
if ext:
extension_list.append((ext.lstrip('.'), t))
response_dict['response_data'] = extension_list
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def getParagraph(self, paragraph_id):
"""Returns the paragraph in the form of (text, class)."""
relevant_paragraph_list = self._getRelevantParagraphList()
try:
paragraph = relevant_paragraph_list[paragraph_id]
except IndexError:
msg = "Unable to find paragraph %s at paragraph list." % paragraph_id
logger.error(msg)
return None
text = ''.join(paragraph.itertext())
if TEXT_ATTRIB_STYLENAME not in paragraph.attrib.keys():
logger.error("Unable to find %s attribute at paragraph %s " % \
(TEXT_ATTRIB_STYLENAME, paragraph_id))
return None
p_class = paragraph.attrib[TEXT_ATTRIB_STYLENAME]
return (text, p_class)
def run_setmetadata(self,
filename='',
data=None,
meta=None,
extension=None,
orig_format=None):
"""Wrapper updateFileMetadata but returns a dict.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
response_dict = {}
try:
response_dict['data'] = self.updateFileMetadata(
data, extension, meta)
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def _callUnoConverter(self, *feature_list, **kw):
""" """
if not openoffice.status():
openoffice.start()
command_list = self._getCommand(*feature_list, **kw)
stdout, stderr = self._subprocess(command_list)
if not stdout and stderr:
first_error = stderr
logger.error(stderr)
self.document.restoreOriginal()
openoffice.restart()
kw['document_url'] = self.document.getUrl()
command = self._getCommand(*feature_list, **kw)
stdout, stderr = self._subprocess(command)
if not stdout and stderr:
second_error = "\nerror of the second run: " + stderr
logger.error(second_error)
raise Exception(first_error + second_error)
return stdout, stderr
def run_generate(self,
filename='',
data=None,
meta=None,
extension=None,
orig_format=''):
"""Wrapper convertFile but returns a dict which includes mimetype.
This is a Backwards compatibility provided for ERP5 Project, in order
to keep compatibility with OpenOffice.org Daemon.
"""
# calculate original extension required by convertFile from
# given content_type (orig_format)
original_extension = BBB_guess_extension(orig_format).strip('.')
# XXX - ugly way to remove "/" and "."
orig_format = orig_format.split('.')[-1]
orig_format = orig_format.split('/')[-1]
if "htm" in extension:
zip = True
else:
zip = False
try:
response_dict = {}
# XXX - use html format instead of xhtml
if orig_format in ("presentation", "graphics", "spreadsheet",
'text') and extension == "xhtml":
extension = 'html'
response_dict['data'] = self.convertFile(data, original_extension,
extension, zip)
# FIXME: Fast solution to obtain the html or pdf mimetypes
if zip:
response_dict['mime'] = "application/zip"
elif extension == 'xhtml':
response_dict['mime'] = "text/html"
else:
response_dict['mime'] = mimetypes.types_map.get(
'.%s' % extension,
mimetypes.types_map.get('.%s' % extension.split('.')[-1]))
return (200, response_dict, "")
except Exception, e:
logger.error(e)
return (402, response_dict, str(e))
def run_generate(self, filename='', data=None, meta=None, extension=None,
orig_format=''):
"""Wrapper convertFile but returns a dict which includes mimetype.
This is a Backwards compatibility provided for ERP5 Project, in order
to keep compatibility with OpenOffice.org Daemon.
"""
# calculate original extension required by convertFile from
# given content_type (orig_format)
original_extension = BBB_guess_extension(orig_format).strip('.')
# XXX - ugly way to remove "/" and "."
orig_format = orig_format.split('.')[-1]
orig_format = orig_format.split('/')[-1]
if "htm" in extension:
zip = True
else:
zip = False
try:
response_dict = {}
# XXX - use html format instead of xhtml
if orig_format in ("presentation",
"graphics",
"spreadsheet",
'text') and extension == "xhtml":
extension = 'html'
response_dict['data'] = self.convertFile(data,
original_extension,
extension,
zip)
# FIXME: Fast solution to obtain the html or pdf mimetypes
if zip:
response_dict['mime'] = "application/zip"
elif extension == 'xhtml':
response_dict['mime'] = "text/html"
else:
response_dict['mime'] = mimetypes.types_map.get('.%s' % extension,
mimetypes.types_map.get('.%s' % extension.split('.')[-1]))
return (200, response_dict, "")
except Exception, e:
import traceback;
logger.error(traceback.format_exc())
return (402, response_dict, str(e))
def run_getmetadata(self,
filename='',
data=None,
meta=None,
extension=None,
orig_format=None):
"""Wrapper for getFileMetadataItemList.
This is a Backwards compatibility provided for ERP5 Project, in order to
keep compatibility with OpenOffice.org Daemon.
"""
if not extension:
extension = filename.split('.')[-1]
response_dict = {}
try:
response_dict['meta'] = self.getFileMetadataItemList(
data, extension)
# XXX - Backward compatibility: Previous API expects 'title' now
# we use 'Title'"
response_dict['meta']['title'] = response_dict['meta']['Title']
return (200, response_dict, '')
except Exception, e:
logger.error(e)
return (402, {}, e.args[0])
def removeDirectory(path):
"""Remove directory"""
try:
rmtree(path)
except OSError, msg:
logger.error(msg)
class OpenOffice(Application):
"""Object to control one OOo Instance and all features instance."""
implements(ILockable)
name = "openoffice"
def __init__(self):
"""Creates the variable to save the pid, port and hostname of the object.
The lock is a simple lock python that is used when one requisition is
using the openoffice.
"""
self._bin_soffice = 'soffice.bin'
self._lock = Lock()
self._cleanRequest()
def _testOpenOffice(self, host, port):
"""Test if OpenOffice was started correctly"""
logger.debug("Test OpenOffice %s - Pid %s" % (self.getAddress()[-1],
self.pid()))
python = join(self.office_binary_path, "python")
args = [exists(python) and python or "python",
pkg_resources.resource_filename("cloudooo",
join('handler', 'ooo',
"helper", "openoffice_tester.py")),
"--hostname=%s" % host,
"--port=%s" % port,
"--uno_path=%s" % self.uno_path]
logger.debug("Testing Openoffice Instance %s" % port)
stdout, stderr = Popen(args, stdout=PIPE,
stderr=PIPE, close_fds=True).communicate()
stdout_bool = convertStringToBool(stdout.replace("\n", ""))
if stdout_bool and stderr != "":
logger.debug("%s\n%s" % (stderr, stdout))
return False
else:
logger.debug("Instance %s works" % port)
return True
def _cleanRequest(self):
"""Define request attribute as 0"""
self.request = 0
def loadSettings(self, hostname, port, path_run_dir,
office_binary_path, uno_path, default_language,
environment_dict=None, **kw):
"""Method to load the configuratio to control one OpenOffice Instance
Keyword arguments:
office_path -- Full Path of the OOo executable.
e.g office_binary_path='/opt/openoffice.org3/program'
uno_path -- Full path of the Uno Library
"""
if environment_dict is None:
environment_dict = {}
Application.loadSettings(self, hostname, port, path_run_dir)
self.office_binary_path = office_binary_path
self.uno_path = uno_path
self.default_language = default_language
self.environment_dict = environment_dict
def _startProcess(self, command, env):
"""Start OpenOffice.org process"""
for i in range(5):
self.stop()
waitStopDaemon(self, self.timeout)
self.process = Popen(command,
close_fds=True,
env=env)
if not waitStartDaemon(self, self.timeout):
continue
if self._testOpenOffice(self.hostname, self.port):
return
def _releaseOpenOfficePort(self):
for process in psutil.process_iter():
try:
if process.exe == join(self.office_binary_path, self._bin_soffice):
for connection in process.get_connections():
if connection.status == "LISTEN" and \
connection.local_address[1] == self.port:
process.terminate()
except psutil.error.AccessDenied, e:
pass
except TypeError, e:
# exception to prevent one psutil issue with zombie processes
logger.debug(e)
except NotImplementedError, e:
logger.error("lsof isn't installed on this machine: " + str(e))
def removeDirectory(path):
"""Remove directory"""
try:
rmtree(path)
except OSError, msg:
logger.error(msg)