Question

我有两个字段，一个是Symbol，另一个是Source。它们具有ManyToMany关系，因为每个符号可以出现在多个Source中，并且每个Source可以具有多个Symbol。符号也是唯一的。到目前为止一切顺利。

现在，我要允许用户批量添加Symbol-> Source映射的CSV。但是，符号可能会在一个或多个CSV中出现两次，我认为这会违反UNIQUE约束。

我当然可以在遍历CSV的每一行时检查该字符是否已在数据库中。但这不会造成大量数据库调用吗？有没有更有效的方法？

class Source(models.Model):
    title = models.CharField(max_length=200, unique=True)
    author = models.CharField(max_length=200)
    offset = models.IntegerField(default=0)

    def __str__(self):
        return self.title + " by " + self.author

    class Meta:
        ordering = ('title',)


class Symbol(models.Model):
    name = models.CharField(max_length=1, unique=True)
    location = models.ManyToManyField(Source, through='CharInSource')

    def __str__(self):
        return self.name

    class Meta:
        ordering = ('name',)


class SymbolInSource(models.Model):
    source = models.ForeignKey(Source, on_delete=models.CASCADE)
    symbol = models.ForeignKey(Symbol, on_delete=models.CASCADE)
    page = models.IntegerField()

    def __str__(self):
        return "p" + str(self.page)

以下是批量上传逻辑：

class BulkUpload(forms.Form):
    file = forms.FileField()
    source = forms.ModelChoiceField(queryset=Source.objects.all())

    def save(self):
        file = self.cleaned_data['file']
        source = self.cleaned_data['source']
        file.seek(0)
        to_be_added = []
        for line in file.readlines():
            data = line.decode('utf-8').split(',')
            name = data[0]
            if len(name) > 1:
                continue  # any faulty characters
            page = data[1].split('\n')[0]

            symbol = Symbol(name=name)
            symbol.save()
            location = SymbolInSource(source=source, page=page, symbol=symbol)
            location.save()

            to_be_added.append({
                'source': location.source,
                'location': location.page,
                'symbol': symbol
            })

        return to_be_added

Django：在添加具有不同ManyToMany字段的对象时，保持拥有方唯一

0 个答案: