美丽的lxml汤不会完整处理一个巨大的文件

Question

我正在解析一个包含750万行的巨大XML（700+ Mb）。当我的代码完成运行时，我只获得前160,000行左右的解析信息。程序以代码0结束。 我没有足够的记忆吗？或者它是图书馆的限制（lxml或美丽的汤）？我该如何解决这个问题？

这是我的代码：

from bs4 import BeautifulSoup
import html5lib
import lxml
import csv
import gc


def parse(soup, filename):
    scvfile = open(filename, 'w')
    writer = csv.writer(scvfile, quoting=csv.QUOTE_ALL)
    writer.writerow(['name', 'primary_id', 'drug_interaction_list', 'food_interaction_list', 'dosage_list'])

    i = 0
    soup.prettify()
    drugs = soup.drugbank.findAll(name="drug", recursive=False)
    print(len(soup.findAll(name="drugbank")))
    print(len(drugs))
    for drug in drugs:
        #gc.collect()
        name = drug.find(name='name', recursive=False).string
        primary_id = drug.find(name='drugbank-id', attrs={'primary': 'true'}).string

        drug_interactions = drug.find(name='drug-interactions', recursive=False)
        drug_interactions = drug_interactions.findAll()
        drug_interacts = []
        for drug_interaction in drug_interactions:
            if not drug_interaction.description is None:
                drug_interacts.append({"drugbank_id": drug_interaction.find(name='drugbank-id', recursive=False).string,
                                       "name": drug_interaction.find(name='name', recursive=False).string,
                                       "description": drug_interaction.find(name='description',
                                                                            recursive=False).string})

        food_interactions = drug.find(name='food-interactions', recursive=False)
        food_interacts = []
        for food_interaction in food_interactions.findAll(name='food-interaction', recursive=False):
            food_interacts.append(food_interaction.string)

        dosages = drug.dosages
        dosage_arr = []
        for dosage in dosages.findAll(name='dosage'):
            dosage_arr.append({"form": "\n\n" + dosage.form.string, "route": dosage.route.string,
                               "Strength": dosage.strength.string})

        writer.writerow([name, primary_id, drug_interacts, food_interacts, dosage_arr])
        i = i + 1
        print("Processed " + str(i) + "rows for csv")


soup = BeautifulSoup(open("/Users/kamilsaitov/Desktop/full database.xml"), "lxml")
parse(soup, "DrugDatabase.csv")