• • 根据您的详细说明修改我的答案：**：

你能试试XMLPullParser()吗，应该很快，而且不会阻塞你的机器。如果你有非常大的文件，你可以决定读取的哪一部分应该馈送到解析器。在我的例子中，我加载了整个XML，但在实际工作中一定不是这样的。

import xml.etree.ElementTree as ET
import pandas as pd

with open('jfontana.xml', 'r') as input_file:
    xml = input_file.read()

parser = ET.XMLPullParser(['end'])
parser.feed(xml)

data = []
for event, elem in parser.read_events():
    #print(elem)
    if elem.tag == 'dtok':
        #print(elem.tag, elem.text, elem.attrib)
        data.append(elem.attrib)
        
df = pd.DataFrame.from_dict(data)
df.to_csv("jfontana.csv")
print(df)

输出：

form         id ord lemma     xpos
0  dar  d-99771-1  25   dar  VMN0000
1  los  d-99771-2  26   els    L3CP0

或者，如果您对所有tok和dtok感兴趣：

import xml.etree.ElementTree as ET
import pandas as pd

with open('jfontana.xml', 'r') as input_file:
    xml = input_file.read()

parser = ET.XMLPullParser(['end'])
parser.feed(xml)

data = []
for event, elem in parser.read_events():
    #print(elem)
    if event =='end' and 'tok' in elem.tag:
        #print(elem.tag, elem.text, elem.attrib)
        data.append(elem.attrib)

        
df = pd.DataFrame.from_dict(data)
df.to_csv("jfontana_all.csv")
print(df)

输出：

form         id  ord   lemma      xpos
0    IX      w-384    1      IX         Z
1   NaN      w-385    2   morir      TMMS
2   NaN      w-386    3      de     SPC00
3   NaN      w-387    4    sant   NCMS000
4   NaN      w-388    5  Vicent   NP00000
5   NaN      w-389    6  Ferrer    NPCS00
6   NaN    w-99769   23       i        CC
7   NaN    w-99770   24      jo  PP1CSN00
8   dar  d-99771-1   25     dar   VMN0000
9   los  d-99771-2   26     els     L3CP0
10  NaN    w-99771  NaN     NaN       NaN
11  NaN    w-99772   27   haver   V0IF3S0
12  NaN    w-99773   28   diner   NCMP000
13  NaN      w-390    7     any        CC

基于编译XPath表达式一次的建议，以及我的评论，即您实际上多次执行XPath求值前后的所有操作，而不是我尝试使用的一次

match_xpath = et.XPath('//tok[re:match(., "^[EeLl][LlOoAa][Ss]*$") and not(starts-with(@xpos, "D"))]',
        namespaces={"re": "http://exslt.org/regular-expressions"})

preceding_xpath = et.XPath('preceding::tok[position() >= 1 and not(position() > 6)]')

following_xpath = et.XPath('following::tok[position() >= 1 and not(position() > 6)]')

def xml_extract(root_element):

    for el in match_xpath(root_element):

        target = el.text
        # allRelevantElements = el.xpath('preceding::tok[position() >= 1 and not(position() > 6)]/following::tok[position() >= 1 and not(position() > 6)]')
        RelevantPrecedingElements = preceding_xpath(el)
        prec1 = RelevantPrecedingElements[-1]
        RelevantFollowingElements = following_xpath(el)
        foll1 = RelevantFollowingElements[0]
        context_list = []

        for elem in RelevantPrecedingElements:
            elem_text = "".join(elem.itertext())
            assert elem_text is not None
            context_list.append(elem_text)

        # adjective = '<' + str(el.text) + '>'
        target = f"<{el.text}>"
        print(target)
        context_list.append(target)

        following_context = []
        for elem in RelevantFollowingElements:
            elem_text = "".join(elem.itertext())
            assert elem_text is not None
            following_context.append(elem_text)

        lema_fol = foll1.get('lemma') if foll1 is not None else None
        lema_prec = prec1.get('lemma') if prec1 is not None else None
        xpos_fol = foll1.get('xpos') if foll1 is not None else None
        xpos_prec = prec1.get('xpos') if prec1 is not None else None
        form_fol = foll1.text if foll1 is not None else None
        form_prec = prec1.text if prec1 is not None else None

        context = " ".join(context_list)
        print(f"Context is: {context}")

        llista = [
            context,
            lema_prec,
            xpos_prec,
            form_prec,
            target,
            lema_fol,
            xpos_fol,
            form_fol,
        ]

        writer = csv.writer(csv_file, delimiter=";")
        writer.writerow(llista)

with open(myCSV_FILE, "a+", encoding="UTF8", newline="") as csv_file:

    for root, dirs, files in os.walk(XMLDIR):
        for file in files:
            if file.endswith(ext):
                file_path = os.path.join(XMLDIR, file)
                file_root = et.parse(file_path).getroot()
                doc = file
                xml_extract(file_root)

并检查这是否提高了性能。
作为另一种选择，我现在还尝试在Python列表上实现大多数比较，而不是在lxmlXPath中实现，我在XSLT 3中尝试过的东西就像Python3一样实现了（例如，首先用XPath选择所有tok元素，然后在XSLT中查找序列，或者在Python中查找所有toks的列表，以查找regexp匹配toks之一的索引，并查找前面的和/或在该序列/列表中跟随）：

import os
import csv
import re

from lxml import etree as et

XMLDIR = "original-samples"
myCSV_FILE = "lxml-single-xpath-py-list-original-samples.csv"

ext = ".xml"

tok_path = et.XPath('//tok')

def xml_extract(root_element):

    all_toks = tok_path(root_element)

    matching_toks = filter(lambda tok: re.match(r'^[EeLl][LlOoAa][Ss]*$', "".join(tok.itertext())) is not None and not(tok.get('xpos').startswith('D')), all_toks)

    for el in matching_toks: 

        target = "".join(el.itertext())
        pos = all_toks.index(el)
        
        RelevantPrecedingElements = all_toks[max(pos - 6, 0):pos]

        prec1 = RelevantPrecedingElements[-1]
        foll1 = all_toks[pos + 1]

        context_list = []

        for elem in RelevantPrecedingElements:
            elem_text = "".join(elem.itertext())
            assert elem_text is not None
            context_list.append(elem_text)

        # adjective = '<' + str(el.text) + '>'
        target = f"<{target}>"
        print(target)
        context_list.append(target)

        lema_fol = foll1.get('lemma') if foll1 is not None else None
        lema_prec = prec1.get('lemma') if prec1 is not None else None
        xpos_fol = foll1.get('xpos') if foll1 is not None else None
        xpos_prec = prec1.get('xpos') if prec1 is not None else None
        form_fol = foll1.text if foll1 is not None else None
        form_prec = prec1.text if prec1 is not None else None

        context = " ".join(context_list)
        print(f"Context is: {context}")

        llista = [
            context,
            lema_prec,
            xpos_prec,
            form_prec,
            target,
            lema_fol,
            xpos_fol,
            form_fol,
        ]

        writer = csv.writer(csv_file, delimiter=";")
        writer.writerow(llista)

with open(myCSV_FILE, "a+", encoding="UTF8", newline="") as csv_file:

    for root, dirs, files in os.walk(XMLDIR):
        for file in files:
            if file.endswith(ext):
                file_path = os.path.join(XMLDIR, file)
                file_root = et.parse(file_path).getroot()
                doc = file
                xml_extract(file_root)

作为一种替代方法，首先对于单个文件，SaxonC（https://saxonica.com/saxon-c/1199/）和XSLT 3如何使用下面这样的XSLT样式表（但是为了测试，我不得不注解掉所有示例中的<!DOCTYPE ..>节点，因为没有提供引用的DTD），这会很有趣：

<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:xs="http://www.w3.org/2001/XMLSchema"
    exclude-result-prefixes="#all"
    version="3.0">
  
  <xsl:output method="text"/>

  <xsl:template match="/">
    <xsl:variable name="toks" select="//tok"/>
    <xsl:variable name="toks-id" select="$toks/generate-id()"/>
    <xsl:for-each select="$toks[matches(., '^[EeLl][LlOoAa][Ss]*$') and not(starts-with(@xpos, 'D'))]">
      <xsl:variable name="pos" select="index-of($toks-id, generate-id())"/>
      <xsl:variable name="target" select="'&lt;' || . || '>'"/>
      <xsl:variable name="prec-tok" select="$toks[$pos - 1]"/>
      <xsl:variable name="foll-tok" select="$toks[$pos + 1]"/>
      <xsl:value-of 
        select="let $s := string-join(($toks[position() = ($pos - 6) to ($pos - 1)], $target), ' ') return if (contains($s, '&quot;')) then '&quot;' || replace($s, '&quot;', '&quot;&quot;') || '&quot;' else $s, 
                $prec-tok/@lemma => string(),
                $prec-tok/@xpos => string(),
                $prec-tok => string(),
                $target,
                $foll-tok/@lemma => string(),
                $foll-tok/@xpos => string(),
                $foll-tok => string()" 
                separator=";"/>
      <xsl:text>&#10;</xsl:text>
    </xsl:for-each>
  </xsl:template>
  
</xsl:stylesheet>

当然，XSLT 3 with Saxon也可以轻松地处理一个目录中的所有.xml文件，以生成单个输出文件（例如'csv'）：

<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
    xmlns:xs="http://www.w3.org/2001/XMLSchema"
    exclude-result-prefixes="#all"
    version="3.0">

  <xsl:param name="csv-path" select="'saxonc-original-samples-single-xslt-to-file.csv'"/>

  <xsl:param name="directory-uri" as="xs:string" select="'original-samples-no-dtd/'"/>

  <xsl:param name="ext" as="xs:string" select="'*.xml'"/>

  <xsl:template name="xsl:initial-template">
    <xsl:result-document href="{$csv-path}">
      <xsl:apply-templates select="uri-collection($directory-uri || '?select=' || $ext)!doc(.)"/>
    </xsl:result-document>
  </xsl:template>
  
  <xsl:output method="text"/>

  <xsl:template match="/">
    <xsl:message expand-text="yes">Processing {base-uri()}</xsl:message>
    <xsl:variable name="toks" select="//tok"/>
    <xsl:variable name="toks-id" select="$toks/generate-id()"/>
    <xsl:for-each select="$toks[matches(., '^[EeLl][LlOoAa][Ss]*$') and not(starts-with(@xpos, 'D'))]">
      <xsl:variable name="pos" select="index-of($toks-id, generate-id())"/>
      <xsl:variable name="target" select="'&lt;' || . || '>'"/>
      <xsl:variable name="prec-tok" select="$toks[$pos - 1]"/>
      <xsl:variable name="foll-tok" select="$toks[$pos + 1]"/>
      <xsl:value-of 
        select="let $s := string-join(($toks[position() = ($pos - 6) to ($pos - 1)], $target), ' ') return if (contains($s, '&quot;')) then '&quot;' || replace($s, '&quot;', '&quot;&quot;') || '&quot;' else $s, 
                $prec-tok/@lemma => string(),
                $prec-tok/@xpos => string(),
                $prec-tok => string(),
                $target,
                $foll-tok/@lemma => string(),
                $foll-tok/@xpos => string(),
                $foll-tok => string()" 
                separator=";"/>
      <xsl:text>&#10;</xsl:text>
    </xsl:for-each>
  </xsl:template>
  
</xsl:stylesheet>

最后一个示例可以使用SaxonC的Python API运行，例如：

from saxonc import *

with PySaxonProcessor(license=True) as proc:
    print(proc.version)

    proc.set_cwd('.')
    
    xslt30_processor = proc.new_xslt30_processor()

    xslt30_executable = xslt30_processor.compile_stylesheet(stylesheet_file = 'xslt3-original-samples-to-csv.xsl')

    if xslt30_processor.exception_occurred:
        print(xslt30_processor.error_message)
    else:

        xslt30_executable.call_template_returning_file(template_name = None, output_file = 'xslt3-result.xml')
        if xslt30_executable.exception_occurred:
            print(xslt30_executable.error_message)