The User Interface's “Search” box accepts predicates of the form “[-]operator:value[:value]”. These predicates may be separated by ‘and’ or ‘or’ keywords, or spaces which mean the same as ‘and’. Expressions like this may be grouped with parenthesis. Grouped expressions are evaluated first. Grouped expressions may be negated.
An ‘and’ besides an ‘or’ is evaluated first. This means for example that
tag:foo or is:pano tag:bar
will return all images having tag foo together with the panorama images having tag bar.
Negation of a predicate is achieved by prepending a minus sign: -is:landscape will match with pictures of not landscape ratio. For example:
-(after:"2010-01-01" before:"2010-03-02T12:33:44") or loc:"Amsterdam"
will return all images having “modtime” outside the specified period, joined with all images taken in Amsterdam.
The logical grouping of the map predicate is an exception, see its definition.
When you need to match a value containing a space, you need to use double quotes around the value only. For example: tag:“Three word tagname” and not “tag:Three word tagname”. If your value contains double quotes you can use backslash escaping. For example:
attr:bar:"He said: \"Hi\""
loc : uses the available metadata, such as EXIF GPS fields, or check-in locations, to match nodes having a location near the specified location. Locations are resolved using maps.googleapis.com. For example:
loc:"new york, new york" This requires that you get a Geocoding API key from Google. See the the page on how to configure geocoding.
locrect : uses the various location metadata fields (such as EXIF GPS) to match nodes having a location within the specified location area. The area is defined by its North-West corner, followed and comma-separated by its South-East corner. Each corner is defined by its latitude, followed and comma-separated by its longitude. For example:
map : is defined like locrect, and it has a similar meaning. However, it is not defined server-side, and it is specifically meant to represent the area of the world that is visible in the screen when using the map aspect, and in particular when zooming or panning. As such, it follows stricter rules than the other predicates, which are:
height : use
height:min-max to match images having a height of at least min and at most max. Use
height:min- to specify only an underbound and
height:-max to specify only an upperbound. Exact matches should use
width : use width:min-max to match images having a width of at least min and at most max. Use width:min- to specify only an underbound and width:-max to specify only an upperbound. Exact matches should use