Using an edgeNGram index analyzer and a simpler search analyzer for partial filename purposes.

analysis.yaml

# Switch to the Analysis-tab to see how these samples get analyzed.
text:
    - My_second_file_created_at_2012.01.13.pdf
    - Another file.txt

analyzer: 

    filename_index:
        type: custom
        tokenizer: filename
        filter: 
            - lowercase
            - edge_ngram

    filename_search:
        type: custom
        tokenizer: filename
        filter: 
            - lowercase

tokenizer: 
    filename:
        type: pattern
        pattern: "[^\\p{L}\\d]+"


filter: 
    edge_ngram:
        type: edge_ngram
        side: front
        min_gram: 1
        max_gram: 20
      

documents.yaml

_type: file
filename: My_first_file_created_at_2012.01.13.doc
---
_type: file
filename: My_second_file_created_at_2012.02.01.pdf
---
_type: file
filename: Another file.txt
---
_type: file
filename: And_again_another_file.docx
---
_type: file
filename: foo.bar.txt

mappings.yaml

file:
    properties: 
        filename: 
            type: string
            index_analyzer: filename_index
            search_analyzer: filename_search

play.sh

#!/bin/bash

export ELASTICSEARCH_ENDPOINT="http://localhost:9200"

# Create indexes

curl -XPUT "$ELASTICSEARCH_ENDPOINT/play" -d '{
    "settings": {
        "analysis": {
            "text": [
                "My_second_file_created_at_2012.01.13.pdf",
                "Another file.txt"
            ],
            "analyzer": {
                "filename_index": {
                    "type": "custom",
                    "tokenizer": "filename",
                    "filter": [
                        "lowercase",
                        "edge_ngram"
                    ]
                },
                "filename_search": {
                    "type": "custom",
                    "tokenizer": "filename",
                    "filter": [
                        "lowercase"
                    ]
                }
            },
            "tokenizer": {
                "filename": {
                    "type": "pattern",
                    "pattern": "[^\\p{L}\\d]+"
                }
            },
            "filter": {
                "edge_ngram": {
                    "type": "edge_ngram",
                    "side": "front",
                    "min_gram": 1,
                    "max_gram": 20
                }
            }
        }
    },
    "mappings": {
        "file": {
            "properties": {
                "filename": {
                    "type": "string",
                    "index_analyzer": "filename_index",
                    "search_analyzer": "filename_search"
                }
            }
        }
    }
}'


# Index documents
curl -XPOST "$ELASTICSEARCH_ENDPOINT/_bulk?refresh=true" -d '
{"index":{"_index":"play","_type":"file"}}
{"filename":"My_first_file_created_at_2012.01.13.doc"}
{"index":{"_index":"play","_type":"file"}}
{"filename":"My_second_file_created_at_2012.02.01.pdf"}
{"index":{"_index":"play","_type":"file"}}
{"filename":"Another file.txt"}
{"index":{"_index":"play","_type":"file"}}
{"filename":"And_again_another_file.docx"}
{"index":{"_index":"play","_type":"file"}}
{"filename":"foo.bar.txt"}
'

# Do searches

curl -XPOST "$ELASTICSEARCH_ENDPOINT/_search?pretty" -d '
{
    "query": {
        "match": {
            "filename": {
                "query": "fi"
            }
        }
    }
}
'

curl -XPOST "$ELASTICSEARCH_ENDPOINT/_search?pretty" -d '
{
    "query": {
        "bool": {
            "should": [
                {
                    "match": {
                        "filename": {
                            "query": 2012.01
                        }
                    }
                },
                {
                    "match_phrase": {
                        "filename": {
                            "query": 2012.01
                        }
                    }
                }
            ]
        }
    }
}
'

play.yaml

# Auto generated by Found's Play-tool at 2013-11-13T13:20:26+01:00
version: 0
title: Partial filename search
description: Using an edgeNGram index analyzer and a simpler search analyzer for partial filename purposes.

searches.yaml

# Example to match partial filenames, without being fuzzy.

# Press Ctrl+Enter to execute when you have made changes.
# Click "Help" in the upper right corner for more information.

# This example is due to Clinton Gormley:
# http://stackoverflow.com/questions/9421358/filename-search-with-elasticsearch/9432450#9432450
query: 
    match: 
        filename:
            query: fi
---
query: 
    bool: 
        should:
            # This will match both "2012.01.13" and "2012.02.01", which is okay.
            - match:
                filename: 
                    query: 2012.01
            # ... but "2012.01.13" is a better match:
            - match_phrase:
                filename:
                    query: 2012.01