Searching

indoorGIS mainly makes use of three python frameworks. Django, Networkx and Geopandas. Django is a web framework which is used to make web applications with python as the backend (obviously). Networkx is used to get shortest routes and nearest facilities and other fucntions related to routing. Geopandas is used to read shapefiles on-the-fly and make queries based on attributes or lcoation

views.py
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
def searchBox(request):

    # Build BlksDF dictionary
    blksDF = {ix: gpd.read_file(blk) for ix, blk in enumerate(blks)}
    rqst = request.POST

    # Query for the keyword
    kwrd, lvlCode = name(rqst["keyword"]), level(rqst["keyword"])
    res = blksDF[lvlCode].query("PlaceName==@kwrd | PersonName==@kwrd | PlaceNode==@kwrd")

    # Log to Stat model
    Stat.objects.create(functionality="Search",
                        keyword=request.POST["keyword"],
                        returned=True)

    fac = [nearest_facility(lvlCode, int(res.PlaceNode), "Facility,male"),
           nearest_facility(lvlCode, int(res.PlaceNode), "Facility,female"),
           nearest_facility(lvlCode, int(res.PlaceNode), "Steps"),
           nearest_facility(lvlCode, int(res.PlaceNode), "Lift")]

    return HttpResponse(str([int(res.OBJECTID), lvlCode, fac]))

searchBox function just takes request as the parameter which is the HTML request itself and it a immutable python dictionary which contains the keyword to be searched. It returns a response to browser with object ids of nearest facilities, its PlaceNode and its name itself

views.py
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
def nearest_facility(lvlCode, frmOID, f):

    blksDF = {ix: gpd.read_file(blk) for ix, blk in enumerate(blks)}
    blk = blksDF[lvlCode]
    places = placesDF[lvlCode]
    _ND = ND[lvlCode]

    cond1 = blk.PlaceName.str.startswith(f, na=False)
    cond2 = blk.PlaceType.str.startswith(f, na=False)

    result = blk[cond1 | cond2]
    res_nodes = result.PlaceNode.tolist()
    res_coords = [get_coor(places, node) for node in res_nodes]

    start = get_coor(places, frmOID)

    pt_distance = {}
    for pn, pt in zip(res_nodes, res_coords):
        pt_distance[pn] = nx.astar_path_length(_ND, start, pt, dist)

    shortest_pnode = sorted(pt_distance, key=lambda x: pt_distance[x])[0]
    return int(result.query("PlaceNode==@shortest_pnode").OBJECTID)

searchBox function is used when user enters a keyword and searches for it. This will also returns the nearest facilites for the given keyword. nearest_facility fucntion will look for all the facilities in all the categories and returns the one which can be accessed quickly.

Workflow

digraph Searching { label="Working of Searching tab"; "Input in the 'Search For' box" -> "Click results from the search results dropdown" -> "Click Find button" -> "Function call search() \n(home.html)" -> "Ajax call searchBox to searchBox() function \n(views.py)" -> "Extract Level and keyword" -> "Read SHP files" -> "Check if keyword is equal to PlaceName or PersonName" -> "Append its OBJECTID to RETURN_LIST" -> "Get Nearest Facilities and append to RETURN_LIST" -> "Return RETURN_LIST to js function" -> "Get all the information and display accordingly" }