根据字符串比较展开并删除重复项

Question

我有一个像这样的mongodb（2.6版）文档：

{
    "Uniprot": {
        "GO": [
            "cytoplasm [GO:0005737]",
            "nucleolus [GO:0005730]",
            "calcium ion binding [GO:0005509]",
            "zinc ion binding [GO:0008270]"
        ],
        "GO cc": [
            "cytoplasm [GO:0005737]",
            "nucleolus [GO:0005730]"
        ],
        "GO bp": [
            ""
        ],
        "GO mf": [
            "calcium ion binding [GO:0005509]",
            "zinc ion binding [GO:0008270]"
        ],
        "GO_ID": [
            "GO:0005509",
            "GO:0005737",
            "GO:0005730",
            "GO:0008270"
        ]
    }
}

使用以下代码，其中test_3是前一个文档的集合：

project = {"$project":{"_id": False,
               "Uniprot": "$Uniprot.Uniprot",
               "GO Description": {
                    "$cond":[
                    {"$eq": ["$Uniprot.GO", "$Uniprot.GO cc"]},
                    "$Uniprot.GO cc",
                        {"$cond":[
                        {"$eq": ["$Uniprot.GO", "$Uniprot.GO mf"]},
                        "$Uniprot.GO mf",
                            {"$cond":[
                            {"$eq": ["$Uniprot.GO", "$Uniprot.GO bp"]},
                            "$Uniprot.GO bp", False]}]}]},
               "GO Type": {"$cond":[
                    {"$eq": ["$Uniprot.GO", "$Uniprot.GO cc"]},
                    "Cellular component",
                        {"$cond":[
                        {"$eq": ["$Uniprot.GO", "$Uniprot.GO mf"]},
                        "Molecular function",
                            {"$cond":[
                            {"$eq": ["$Uniprot.GO", "$Uniprot.GO bp"]},
                            "Biological Process", False]}]}]},
               "GO ID":"$Uniprot.GO_ID"}}
redact = {"$redact":{"$cond":[{"$or":[
                            {"$eq":["$Uniprot.GO", "$Uniprot.GO cc"]},
                            {"$eq":["$Uniprot.GO", "$Uniprot.GO mf"]},
                            {"$eq":["$Uniprot.GO", "$Uniprot.GO bp"]}]},
                                            "$$KEEP", "$$PRUNE"]}}
d = test_3.aggregate([{"$match":{"Uniprot.Uniprot": "P33764"}},
                      {"$unwind":"$Uniprot.GO"},
                      {"$unwind":"$Uniprot.GO cc"},
                      {"$unwind":"$Uniprot.GO bp"},
                      {"$unwind":"$Uniprot.GO mf"},
                      redact,
                      {"$unwind":"$Uniprot.GO_ID"},
                      project
                      ])

输出结果为：

{u'GO ID': u'GO:0005509', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm [GO:0005737]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus [GO:0005730]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding [GO:0005509]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005737', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding [GO:0008270]', u'Uniprot': u'P33764'}

现在我使用以下方法修改输出：

b = []
for i in d:
    go = i["GO Description"][-13:]
    if i["GO ID"] == go[2:-1]:
        entry = i.copy()
        entry["GO Description"] = i["GO Description"][:-13]
        if entry not in b:
            b.append(entry)
for i in b:
    print i

获得我预期的输出：

{u'GO ID': u'GO:0005737', u'GO Type': u'Cellular component', u'GO Description': u'cytoplasm', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005730', u'GO Type': u'Cellular component', u'GO Description': u'nucleolus', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0005509', u'GO Type': u'Molecular function', u'GO Description': u'calcium ion binding', u'Uniprot': u'P33764'}
{u'GO ID': u'GO:0008270', u'GO Type': u'Molecular function', u'GO Description': u'zinc ion binding', u'Uniprot': u'P33764'}

然而，这种方式很慢，我想在mongodb中进行，而无需在python中进一步处理。我该怎么做？ 我注意到的事情：由于GO和GO_ID的展开，聚合会产生重复的行（这就是为什么有if entry not in b），我需要检查GO ID上是否有GO Description，因为我无法通过$search或$text找到解决方法。