libreoffice python 操作word及excel文档

1、开始、关闭libreoffice服务；

开始之前同步字体文件时间，是因为创建soffice服务时，服务会检查所需加载的文件的时间，如果其认为时间不符，则其可能会重新加载，耗时较长，因此需事先统一时间。

使用时如果需要多次调用，最后每次调用均开启后关闭，否则libreoffice会创建一个缓存文档并越用越大，处理时间会增加。

class OfficeProcess(object):
    def __init__(self):
        self.p = 0
        subprocess.Popen('find /usr/share/fonts | xargs touch -m -t 201801010000.00', shell=True)

    def start_office(self):
        self.p = subprocess.Popen('soffice --pidfile=sof.pid --invisible --accept="socket,host=localhost,port=2002;urp;"', shell=True)
        while True:
            try:
                local_context = uno.getComponentContext()
                resolver = local_context.getServiceManager().createInstanceWithContext('com.sun.star.bridge.UnoUrlResolver', local_context)
                resolver.resolve('uno:socket,host=localhost,port=2002;urp;StarOffice.ComponentContext')
                return
            except:
                print(ts(), "wait for connecting soffice...")
                time.sleep(1)
                continue

    def stop_office(self):
        with open("sof.pid", "rb") as f:
            try:
                os.kill(int(f.read()), signal.SIGTERM)
                self.p.wait()
            except:
                pass

2、init service manager

local_context = uno.getComponentContext()
        service_manager = local_context.getServiceManager()
        resolver = service_manager.createInstanceWithContext('com.sun.star.bridge.UnoUrlResolver', local_context)
        self.ctx = resolver.resolve('uno:socket,host=localhost,port=2002;urp;StarOffice.ComponentContext')
        self.smgr = self.ctx.ServiceManager
        self.desktop = self.smgr.createInstanceWithContext('com.sun.star.frame.Desktop', self.ctx)

3、从二进制数据中读取doc文档

def ImportFromMemory(self, data):
        istream = self.smgr.createInstanceWithContext('com.sun.star.io.SequenceInputStream', self.ctx)
        istream.initialize((uno.ByteSequence(data), ))
        pv = PropertyValue()
        pv.Name = 'InputStream'
        pv.Value = istream
        self.doc = {'doc': []}
        try:
            self.document = self.desktop.loadComponentFromURL('private:stream/swriter', '_blank', 0, (pv, ))
            self.text = self.document.getText()
        except:
            self.text = None

4、读取doc文档中的数据

def ExportToJson(self):
        try:
            l = self.__ParseText(self.text, self.__Callback(self.doc['doc']))
            self.doc['length'] = l
        except:
            self.doc = {'doc': [], 'length': 0}
        return json.dumps(self.doc)

@staticmethod
    def __Callback(alist):
        def Append(sth):
            alist.append(sth)
        return Append

def __ParseText(self, text, func):
        l = 0
        text_it = text.createEnumeration()
        while text_it.hasMoreElements():
            element = text_it.nextElement()
            if element.supportsService('com.sun.star.text.Paragraph'):
                l += self.__ParseParagraph(element, func)
            elif element.supportsService('com.sun.star.text.TextTable'):
                l += self.__ParseTable(element, func)
            else:
                pass
        return l

def __ParseParagraph(self, paragraph, func):
        p = {'paragraph': []}
        l = 0
        paragraph_it = paragraph.createEnumeration()
        while paragraph_it.hasMoreElements():
            portion = paragraph_it.nextElement()
            if portion.TextPortionType == 'Text':
                l += self.__ParsePortionText(portion, self.__Callback(p['paragraph']))
            elif portion.TextPortionType == 'SoftPageBreak':
                pass
            elif portion.TextPortionType == 'TextField':
                l += self.__ParsePortionText(portion, self.__Callback(p['paragraph']))
            else:
                l += self.__ParseTextContent(portion, self.__Callback(p['paragraph']))
        if hasattr(paragraph, 'createContentEnumeration'):
            l += self.__ParseTextContent(paragraph, self.__Callback(p['paragraph']))
        p['length'] = l
        func(p)
        return l

    def __ParseTextContent(self, textcontent, func):
        l = 0
        content_it = textcontent.createContentEnumeration('com.sun.star.text.TextContent')
        while content_it.hasMoreElements():
            element = content_it.nextElement()
            if element.supportsService('com.sun.star.text.TextGraphicObject'):
                l += self.__ParsePortionGraphic(element, func)
            elif element.supportsService('com.sun.star.text.TextEmbeddedObject'):
                pass
            elif element.supportsService('com.sun.star.text.TextFrame'):
                l += self.__ParseFrame(element, func)
            elif element.supportsService('com.sun.star.drawing.GroupShape'):
                l += self.__ParseGroup(element, func)
            else:
                pass
        return l

    def __ParseFrame(self, frame, func):
        f = {'frame': []}
        l = self.__ParseText(frame.getText(), self.__Callback(f['frame']))
        f['length'] = l
        func(f)
        return l

    def __ParseGroup(self, group, func):
        l = 0
        for i in range(group.getCount()):
            it = group.getByIndex(i)
            if it.supportsService('com.sun.star.drawing.Text'):
                l += self.__ParseFrame(it, func)
            else:
                pass
        return l

    def __ParsePortionText(self, portion_text, func):
        func({'portion': portion_text.String, 'length': len(portion_text.String)})
        return len(portion_text.String)

    def __ParsePortionGraphic(self, portion_graphic, func):
        gp = self.smgr.createInstanceWithContext('com.sun.star.graphic.GraphicProvider', self.ctx)
        stream = self.smgr.createInstanceWithContext('com.sun.star.io.TempFile', self.ctx)
        pv1 = PropertyValue()
        pv1.Name = 'OutputStream'
        pv1.Value = stream
        pv2 = PropertyValue()
        pv2.Name = 'MimeType'
        pv2.Value = 'image/png'
        gp.storeGraphic(portion_graphic.Graphic, (pv1, pv2))
        stream.getOutputStream().flush()
        stream.seek(0)
        l = stream.getInputStream().available()
        b = uno.ByteSequence(b'')
        stream.seek(0)
        l, b = stream.getInputStream().readBytes(b, l)
        img = {'image': base64.b64encode(b.value).decode('ascii')}
        img['height'] = portion_graphic.Height
        img['width'] = portion_graphic.Width
        img['actualheight'] = portion_graphic.ActualSize.Height
        img['actualwidth'] = portion_graphic.ActualSize.Width
        img['croptop'] = portion_graphic.GraphicCrop.Top
        img['cropbottom'] = portion_graphic.GraphicCrop.Bottom
        img['cropleft'] = portion_graphic.GraphicCrop.Left
        img['cropright'] = portion_graphic.GraphicCrop.Right
        img['length'] = 0
        func(img)
        return 0

    def __ParseTable(self, table, func):
        l = 0
        try:
            matrix = self.__GetTableMatrix(table)
            seps = self.__GetTableSeparators(table)
            t = {}
            count = 0
            for ri in matrix.keys():
                t[ri] = {}
                for ci in matrix[ri].keys():
                    t[ri][ci] = dict(matrix[ri][ci])
                    del t[ri][ci]['cell']
                    t[ri][ci]['content'] = []
                    l += self.__ParseText(matrix[ri][ci]['cell'], self.__Callback(t[ri][ci]['content']))
                    count += t[ri][ci]['rowspan'] * t[ri][ci]['colspan']
            if count != len(t) * len(seps):
                raise ValueError('count of cells error')
            func({'table': t, 'row': len(t), 'column': len(seps), 'length': l, 'tableid': self.table_id})
            self.table_id += 1
        except:
            l = 0
            print('discard wrong table')
        return l

    @staticmethod
    def __GetTableSeparators(table):
        result = [table.TableColumnRelativeSum]
        for ri in range(table.getRows().getCount()):
            result += [s.Position for s in table.getRows().getByIndex(ri).TableColumnSeparators]
        result = sorted(set(result))
        for i in range(len(result) - 1):
            result[i] += 1 if result[i] + 1 == result[i + 1] else 0
        return sorted(set(result))

    @staticmethod
    def __NameToRC(name):
        r = int(re.sub('[A-Za-z]', '', name)) - 1
        cstr = re.sub('[0-9]', '', name)
        c = 0
        for i in range(len(cstr)):
            if cstr[i] >= 'A' and cstr[i] <= 'Z':
                c = c * 52 + ord(cstr[i]) - ord('A')
            else:
                c = c * 52 + 26 + ord(cstr[i]) - ord('a')
        return r, c

    @staticmethod
    def __GetTableMatrix(table):
        result = {}
        for name in table.getCellNames():
            ri, ci = WordToJson.__NameToRC(name)
            cell = table.getCellByName(name)
            if ri not in result:
                result[ri] = {}
            result[ri][ci] = {'cell': cell, 'rowspan': cell.RowSpan, 'name': name}

        seps = WordToJson.__GetTableSeparators(table)
        for ri in result.keys():
            sep = [s.Position for s in table.getRows().getByIndex(ri).TableColumnSeparators] + [table.TableColumnRelativeSum]
            sep = sorted(set(sep))
            for ci in result[ri].keys():
                right = seps.index(sep[ci]) if sep[ci] in seps else seps.index(sep[ci] + 1)
                left = -1 if ci == 0 else seps.index(sep[ci - 1]) if sep[ci - 1] in seps else seps.index(sep[ci - 1] + 1)
                result[ri][ci]['colspan'] = right - left
        return result

5、写doc文档

self.doco = self.desktop.loadComponentFromURL('private:factory/swriter', '_blank', 0, ())
        self.texto = self.doco.getText()
        self.cursoro = self.texto.createTextCursor()
        self.cursoro.ParaBottomMargin = 500

def __WriteText(self, text, texto, cursoro):
        for it in text:
            if 'paragraph' in it:
                self.__WriteParagraph(it, texto, cursoro)
            elif 'image' in it:
                self.__WritePortionGraphic(it, texto, cursoro)
            elif 'table' in it:
                self.__WriteTable(it, texto, cursoro)

    def __WriteParagraph(self, paragraph, texto, cursoro):
        if paragraph['length'] > 0:
            if 'result' in paragraph:
                for it in paragraph['result']:
                    texto.insertString(cursoro, it['trans_sen'], False)
            else:
                texto.insertString(cursoro, paragraph['paragraph'], False)
            texto.insertControlCharacter(cursoro, ControlCharacter.PARAGRAPH_BREAK, False)

    def __WritePortionGraphic(self, portion_graphic, texto, cursoro):
        png_base64 = portion_graphic['image']
        png = base64.b64decode(png_base64)
        gp = self.smgr.createInstanceWithContext('com.sun.star.graphic.GraphicProvider', self.ctx)
        istream = self.smgr.createInstanceWithContext('com.sun.star.io.SequenceInputStream', self.ctx)
        istream.initialize((uno.ByteSequence(png), ))
        pv = PropertyValue()
        pv.Name = 'InputStream'
        pv.Value = istream

        actualsize = uno.createUnoStruct('com.sun.star.awt.Size')
        actualsize.Height = portion_graphic['actualheight'] if 'actualheight' in portion_graphic else portion_graphic['height']
        actualsize.Width = portion_graphic['actualwidth'] if 'actualwidth' in portion_graphic else portion_graphic['width']
        graphiccrop = uno.createUnoStruct('com.sun.star.text.GraphicCrop')
        graphiccrop.Top = portion_graphic['croptop'] if 'croptop' in portion_graphic else 0
        graphiccrop.Bottom = portion_graphic['cropbottom'] if 'cropbottom' in portion_graphic else 0
        graphiccrop.Left = portion_graphic['cropleft'] if 'cropleft' in portion_graphic else 0
        graphiccrop.Right = portion_graphic['cropright'] if 'cropright' in portion_graphic else 0

        image = self.doco.createInstance('com.sun.star.text.TextGraphicObject')
        image.Surround = NONE
        image.Graphic = gp.queryGraphic((pv, ))
        image.Height = portion_graphic['height']
        image.Width = portion_graphic['width']
        image.setPropertyValue('ActualSize', actualsize)
        image.setPropertyValue('GraphicCrop', graphiccrop)
        texto.insertTextContent(cursoro, image, False)
        texto.insertControlCharacter(cursoro, ControlCharacter.PARAGRAPH_BREAK, False)

    def __WriteTable(self, table, texto, cursoro):
        tableo = self.doco.createInstance('com.sun.star.text.TextTable')
        tableo.initialize(table['row'], table['column'])
        texto.insertTextContent(cursoro, tableo, False)
#        texto.insertControlCharacter(cursoro, ControlCharacter.PARAGRAPH_BREAK, False)
        tcursoro = tableo.createCursorByCellName("A1")

        hitbug = False
        if table['row'] > 1:
            tcursoro.goDown(1, True)
            hitbug = tcursoro.getRangeName() == 'A1'

        for ri in sorted([int(r) for r in table['table'].keys()]):
            rs = table['table'][str(ri)]
            for ci in sorted([int(c) for c in rs.keys()]):
                cell = rs[str(ci)]
                if hitbug == False and (cell['rowspan'] > 1 or cell['colspan'] > 1):
                    tcursoro.gotoCellByName(cell['name'], False)
                    if cell['rowspan'] > 1:
                        tcursoro.goDown(cell['rowspan'] - 1, True)
                    if cell['colspan'] > 1:
                        tcursoro.goRight(cell['colspan'] - 1, True)
                    tcursoro.mergeRange()
                ctexto = tableo.getCellByName(cell['name'])
                if ctexto == None:
                    continue
                ccursoro = ctexto.createTextCursor()
                ccursoro.CharWeight = FontWeight.NORMAL
                ccursoro.CharWeightAsian = FontWeight.NORMAL
                ccursoro.ParaAdjust = LEFT
                self.__WriteText(cell['content'], ctexto, ccursoro)

6、生成二进制的doc文档数据

        streamo = self.smgr.createInstanceWithContext('com.sun.star.io.Pipe', self.ctx)
        self.doco.storeToURL('private:stream', (PropertyValue('FilterName', 0, 'MS Word 2007 XML', 0), PropertyValue('OutputStream', 0, streamo, 0)))
        streamo.flush()
        _, datao = streamo.readBytes(None, streamo.available())

7、从doc文档数据生成pdf的二进制数据

        streamo = self.smgr.createInstanceWithContext('com.sun.star.io.Pipe', self.ctx)
        self.doco.storeToURL('private:stream', (PropertyValue('FilterName', 0, 'writer_pdf_Export', 0), PropertyValue('OutputStream', 0, streamo, 0)))
        streamo.flush()
        _, datap = streamo.readBytes(None, streamo.available())

8、读取excel二进制数据

　　def ImportFromMemory(self, data):
        istream = self.smgr.createInstanceWithContext('com.sun.star.io.SequenceInputStream', self.ctx)
        istream.initialize((uno.ByteSequence(data), ))
        pv = PropertyValue()
        pv.Name = 'InputStream'
        pv.Value = istream
        self.doc = {'doc': []}
        try:
            print("before loadComponentFromURL")
            self.document = self.desktop.loadComponentFromURL('private:stream/scalc', '_blank', 0, (pv, ))
            self.sheets = self.document.getSheets()
            print("ImportFromMemory done")
        except:
            print("ImportFromMemory failed")
            self.sheets = None

9、读取excel的文本数据

    def ExportToJson(self):
        try:
            l = self.__ParseText(self.sheets, self.__Callback(self.doc['doc']))
            self.doc['length'] = l
        except:
            self.doc = {'doc': [], 'length': 0}
        return json.dumps(self.doc)

    def __ParseText(self, sheets, func):
        l = 0
        sheets_it = sheets.createEnumeration()
        while sheets_it.hasMoreElements():
            element = sheets_it.nextElement()
            if element.supportsService('com.sun.star.sheet.Spreadsheet'):
                l += self.__ParseSpreadsheet(element, func)
        return l

    def __ParseSpreadsheet(self, spreadsheet, func):
        l = 0
        p = {'spreadsheet': []}
        visible_cells_it = spreadsheet.queryVisibleCells().getCells().createEnumeration()
        while visible_cells_it.hasMoreElements():
            cell = visible_cells_it.nextElement()
            type = cell.getType()
            if type == self.EMPTY:
                print("cell.type==empty")
            elif type == self.VALUE:
                print("cell.type==VALUE", "value=", cell.getValue(), cell.getCellAddress ())
            elif type == self.TEXT:
                print("cell.type==TEXT","content=", cell.getString().encode("UTF-8"), cell.getCellAddress ())
                l += self.__ParseCellText(spreadsheet, cell, self.__Callback(p['spreadsheet']))
                print("__ParseCellText=", p)
            elif type == self.FORMULA:
                print("cell.type==FORMULA", "formula=", cell.getValue())
        p['length'] = l
        func(p)
        return l

    def __ParseCellText(self, sheet, cell, func):
        try:
            x = cell.getCellAddress().Column
            y = cell.getCellAddress().Row
            sheetname = sheet.getName()
        except:
            x = -1
            y = -1
            sheetname = None
        func({'celltext': cell.getString(), 'x': x, 'y': y, 'sheetname': sheetname, 'length': len(cell.getString())})
        return len(cell.getString())

　　　　　self.EMPTY = uno.Enum("com.sun.star.table.CellContentType", "EMPTY")
        self.TEXT = uno.Enum("com.sun.star.table.CellContentType", "TEXT")
        self.FORMULA = uno.Enum("com.sun.star.table.CellContentType", "FORMULA")
        self.VALUE = uno.Enum("com.sun.star.table.CellContentType", "VALUE")

10、替换excel的文本信息

    def ImportFromJson(self, data):
        doc = json.loads(data)
        try:
            self.__WriteText(doc['doc'])
        except:
            pass

    def __WriteText(self, text):
        print("__WriteText begin:", text)
        sheet = None
        for it in text:
            if 'paragraph' in it and 'sheetname' in it:
                if sheet == None or sheet.getName() != it['sheetname']:
                    try:
                        sheet = self.sheets.getByName(it['sheetname'])
                        print("getsheet:", it['sheetname'], "=", sheet.getName())
                    except:
                        sheet = None
                        continue
                self.__WriteParagraph(it, sheet)

    def __WriteParagraph(self, paragraph, sheet):
        print("__WriteParagraph")
        if paragraph['length'] > 0:
            try:
                x = paragraph['x']
                y = paragraph['y']
                print("getcell:", x, y)
                cell = sheet.getCellByPosition(x, y)
                print("getcell done")
            except:
                return
            if 'result' in paragraph:
                for it in paragraph['result']:
                    print("cell=", cell.getString())
                    cell.setString(it['trans_sen'])
                    print("cell,", cell.getString(), ",done")

11、生成excel文档二进制数据

　　　　  streamo = self.smgr.createInstanceWithContext('com.sun.star.io.Pipe', self.ctx)
        self.document.storeToURL('private:stream', (PropertyValue('FilterName', 0, 'Calc MS Excel 2007 XML', 0), PropertyValue('OutputStream', 0, streamo, 0)))
        streamo.flush()
        _, datao = streamo.readBytes(None, streamo.available())

12、生成excel的pdf文档

        streamo = self.smgr.createInstanceWithContext('com.sun.star.io.Pipe', self.ctx)
        self.document.storeToURL('private:stream', (PropertyValue('FilterName', 0, 'calc_pdf_Export', 0), PropertyValue('OutputStream', 0, streamo, 0)))
        streamo.flush()
        _, datap = streamo.readBytes(None, streamo.available())