Python UNO (libreoffice): How to enable auto filters for a sheet

Question

I have a program, that creates a CSV file.

Now I'd like to use a Python UNO script, which shall do several things:

1.) open the csv file in a spread sheet

2.) enable auto filter for all columns

3.) create a macro and add it to the document

4.) save the file as ODS file

This question concerns only 2.)

1.) is working

for 3.) I will probably create another question 1.) Is working (with pyoo and with unotools)

My steps so far:

I start manually:

libreoffice --accept='socket,host=localhost,port=2202;urp;' --norestore --nologo --nodefault

My python script:

With pyoo

import pyoo
# Step 0) connect to UNO bridge
desktop = pyoo.Desktop('localhost', 2002)

# Step 1) open the doc and get the sheet
# This works only if the field separator is a comma.
# I don't know how for example to specify tab as separator instead
doc = desktop.open_spreadsheet('tst.csv')
# I see the spreadsheet opening
sheet = doc.sheets[0] # I get an object of type Sheet

# Step2) set autofilter for active sheet
# no idea how to do

# Step3) create a macro and add it to the document
# no idea how to do but will create another question as 
# soon as step2 is solved

# Step 4) save the sheet
doc.save("tst_pyoo.ods")

Or with unotools

import unotools
from unotools.component.calc import Calc
from unotools.unohelper import convert_path_to_url

# Step 0) connect to UNO bridge
context = unotools.connect(unotools.Socket('localhost', 2002))

# Step 1) open document
doc = Calc(ctx, convert_path_to_url('tst.csv')
# I see the spreadsheet opening
sheet = doc.get_sheet_by_index(0)
# I get an object of type unotools.component.calc.Spreadsheet

# Step2) set autofilter for active sheet
# no idea how to do

# Step3) create a macro and add it to the document
# no idea how to do but will create another question as 
# soon as step2 is solved

# Step 4)
doc.store_to_url(convert_path_to_url("tst_unotools.ods"))

Thanks in advance for any feedback

Answer 1

Here is code that uses straight PyUNO rather than a wrapper library. It is adapted from http://www.imaccanici.org/en.libreofficeforum.org/node/5413.html .

import os
import uno

class UnoObjs:
    "Initializes and stores UNO objects to connect to a document."""
    def __init__(self, filepath=None):
        localContext = uno.getComponentContext()
        resolver = localContext.ServiceManager.createInstanceWithContext(
            "com.sun.star.bridge.UnoUrlResolver", localContext )
        self.ctx = resolver.resolve(
            "uno:socket,host=localhost,port=2002;urp;"
            "StarOffice.ComponentContext")
        self.smgr = self.ctx.ServiceManager
        desktop = self.smgr.createInstanceWithContext(
            "com.sun.star.frame.Desktop", self.ctx)
        if filepath:
            fileUrl = uno.systemPathToFileUrl(os.path.realpath(filepath))
            self.document = desktop.loadComponentFromURL(
                fileUrl, "_default", 0, ())
        else:
            self.document = desktop.getCurrentComponent()

def add_autofilter(unoObjs):
    """This adds an autofilter by selecting only the filled spreadsheet area. 
    NOTE: If any cell in the header row of the selection is empty this will
    trigger a popup for interactive user action (must click Yes for the
    Autofilter column header message box)
    """
    dispatcher = unoObjs.smgr.createInstanceWithContext(
        "com.sun.star.frame.DispatchHelper", unoObjs.ctx)
    controller = unoObjs.document.getCurrentController()
    sheet = unoObjs.document.getSheets().getByIndex(0)
    # select a sufficiently big "guess" area, hopefully
    guessRange = sheet.getCellRangeByPosition(0, 0, 150, 10000)
    # look up the actual used area within the guess area
    cursor = sheet.createCursorByRange(guessRange)
    cursor.gotoEndOfUsedArea(False)
    lastcol = cursor.RangeAddress.EndColumn
    lastrow = cursor.RangeAddress.EndRow
    # select the filled part of the spreadsheet
    actualRange = sheet.getCellRangeByPosition(0, 0, lastcol, lastrow)
    controller.select(actualRange)
    # add autofilter
    dispatcher.executeDispatch(
        unoObjs.document.getCurrentController(), ".uno:DataFilterAutoFilter",
        "", 0, ())

add_autofilter(UnoObjs("tst.csv"))

Dispatcher calls such as .uno:DataFilterAutoFilter are difficult to figure out the parameters. In most cases it's better to use UNO API calls such as XTextCursor instead. However there are a few options to figure out dispatcher calls:

• Use the macro recorder.
• Look through a list like this one .
• Find the calls in the LibreOffice source code. This is the most reliable, but it can still sometimes be difficult to determine the arguments.

Regarding dispatcher calls, see https://forum.openoffice.org/en/forum/viewtopic.php?f=20&t=61127 .

Answer 2

Is it possible to modify the content.xml directly contained in the .ods file with python?

My aim is also to apply an autofilter in LibreOffice on Windows 10. I can't find any documentation whatsoever.

But I found that if you unzip an .ods file, in the content.xml file there are the following lines of xml

<table:database-ranges>
                <table:database-range table:name="__Anonymous_Sheet_DB__0" table:target-range-address="Sheet1.A1:Sheet1.B3" table:display-filter-buttons="true"/>
</table:database-ranges>

Is there a way to read and modify the content.xml using the lxml module of python and in that way apply the autofilter in LibreOffice?