Trouble using date_range in Elastic Search; aggregations not working - datetime

Question

I have a form that allows for several fields to search for records in a database. It's able to search by first name, last name, location, etc, but I'm having trouble adding date ranges to the query. Each record has a start year and an end year, and my form has fields Date To and From. My code checks if each form in the field is filled out, and if so appends to a query string; once the query string is complete, it is passed into a Connection.Post search. Everything works except for To and From, but when I add these fields I get a 404 error.
I have a QueryBuilder method that checks the fields and builds the query, here is how it starts:
string query = "{\"query\": { \"bool\": { \"must\": [ ";
bool emptyQ = true;
Here is the first_name part for example:
if (!String.IsNullOrWhiteSpace(q.Name.First))
{
if (exact == true)
{
query += "{ \"match\": { \"first_name\": \"" + q.Name.First + "\" }}";
}
else
{
// no fuzziness allowed in the search term; how can i fix this? (i.e. Sam -> Samuel, but Isac -\> Isaac)
query += "{ \"match_phrase_prefix\": { \"first_name\": \"" + q.Name.First + "\"}}";
}
emptyQ = false;
}
Here is the part that isn't working right:
bool aggs = false;
if (q.FromYear != 0 || q.ToYear != DateTime.Now.Year)
{
if (emptyQ == false)
{
query += "] } }, ";
}
if (emptyQ == true)
{
query += "{";
}
DateTime fromyear = Convert.ToDateTime("01/01/" + q.FromYear.ToString());
if (q.FromYear == 0)
{
fromyear = Convert.ToDateTime("00/00/0000" + q.FromYear.ToString());
}
DateTime toyear = Convert.ToDateTime("12/31/" + q.ToYear.ToString());
query += "\"aggregations\": { \"range\": { \"date_range\": { \"field\": \"start_date\", \"ranges\": [{\"from\": \"" + fromyear + "\"}]}}}";
query += "{\"range\": { \"date_range\": { \"field\": \"end_date\", \"ranges\": [{\"to\": \"" + toyear +
"\"}] }}}}";
aggs = true;
}
if (aggs == false)
{
query += " ] } } }";
}
i.e. when one or both of the date fields is filled out, the query is invalid. When every field is filled out, this is what the query builder returns:
{"query":
{ "bool":
{ "must": [
{ "match_phrase_prefix": { "first_name": "sam"}},
{ "fuzzy": { "last_name": "durand" }},
{ "match": { "record_type": { "query": "Baptism", "type": "phrase"} }},
{ "fuzzy_like_this":
{ "fields": [ "city", "region", "county", "country" ],
"like_text": "Massachusetts"
}
}
]}
},
"aggregations":
{ "range":
{ "date_range":
{ "field": "start_date",
"ranges": [{"from": "1/1/1600 12:00:00 AM"}]
}
}
},
{"range":
{ "date_range":
{ "field": "end_date",
"ranges": [{"to": "12/31/2014 12:00:00 AM"}]
}
}
}
}
Which gives me a 400 bad request error. Again, when the year fields are not filled out, it works just fine. I'm sure it's a problem with how I'm formatting my query, but I'm not sure what I'm doing wrong. Also, is there an easier way I can search for a date range without using aggregations? Is there a search method that I can put in the "query { " itself, rather than after it?

Answer 1

If you're trying to filter the results by a date range you don't want to use the date_range aggregation, you want to use the range filter. That would give you something more like:
{"query":
{ "bool":
{ "must": [
{
"range" : {
"start_date" : {
"gte": "1/1/1600 12:00:00 AM",
}
}
},
{
"range" : {
"end_date" : {
"lte": 12/31/2014 12:00:00 AM",
}
}
},
]}
}

Answer 2

Turns out I didn't have to do anything fancy at all, I just used a simple range:
query += "{ \"range\": { \"start_date.year\": { \"gte\": " + fyear + ", \"lte\": " + tyear + "} }}, " +
"{ \"range\": { \"end_date.year\": { \"gte\": " + fyear + ", \"lte\": " + tyear + "} }}] } } }, ";
My problem was in assuming that datetime objects would not be read correctly in a range query, but you can in fact use datetime objects in a range.