Elasticsearch Custom Analyzer

Question

I'm currently working on an Search Engine implementation with Scrapy as Crawler and Elasticsearch as server. Scrapy and Elasticsearch work fine, but what I'm currently struggling to enable is case-insensitive search with a german analyzer. I have a general structure (match_all query) like this:

"hits": {
    "total": 14,
    "max_score": 0.40951526,
    "hits": [
        {
            "_index": "uni",
            "_type": "items",
            "_id": "AVcuHuT6qni1Wq78foIA",
            "_score": 0.40951526,
            "_source": {
                "description": "...",
                "tags": [
                    "...",
                    "..."
                ],
                "url":"...",
                "author": "...",
                "content": "...",
                "date": "18.09.2015",
                "title": "..."
            },
            "highlight": {
                "content": [
                    "...",
                    "...",
                    "..."
                ]
            }
        }
    ]
}

And tried to add these settings just by "curl -XPUT localhost:9200/uni {...}":

{
    "mappings":{
        "_source":{
            "type":"object",
            "properties":{
                "title":{
                    "type":"string",
                    "analyzer":"german_lowercase"
                },
                "content":{
                    "type":"string",
                    "analyzer":"german_lowercase"
                },
                "description":{
                    "type":"string",
                    "analyzer":"german_lowercase"
                },
                "tags":{
                    "type":"array",
                    "analyzer":"german_lowercase"
                }
            }
        }
    },
    "settings":{
        "uni":{
            "analysis":{
                "analyzer":{
                    "german_lowercase":{
                        "type":"custom",
                        "tokenizer":"keyword",
                        "filter":[
                            "lowercase",
                            "german_stop",
                            "german_keywords",
                            "german_normalization",
                            "german_stemmer"
                        ]
                    }
                },
                "filter":{
                    "german_stop": {
                        "type": "stop",
                        "stopwords": "_german_"
                    },
                    "german_keywords": {
                        "type": "keyword_marker",
                        "keywords": []
                    },
                    "german_stemmer": {
                        "type": "stemmer",
                        "language": "light_german"
                    }
                }
            }
        }
    }
}

I'm not sure whats going wrong, can anybody help out?

EDIT: Elasticsearch wont allow me to put those settings into the index (already exists) and if I try to put the mapping separately Ill get an "missing mapping type" exception. In case of the settings it failes ot update the non dynamic settings. So I'm asking for a more general info how I should update those settings/mappings in order to enable case-insensitive search (other posts have the same problem).