Obj id bitmask endpoint

omeroweb/webgateway/urls.py

@@ -47,6 +47,25 @@
 Get omero table metadata
 """
 
+table_obj_id_bitmask = url(
+    r"^table/(?P<fileid>\d+)/obj_id_bitmask/$",
+    views.obj_id_bitmask,
+    name="webgateway_table_obj_id_bitmask",
+)
+"""
+Get object id bitmask
+The user specifies a fileid for an OMERO Table and a query, and
+optionally provides a "col_name" query parameter for the column
+name to get a bitmask for. By default, "object" is used.
+The server will return a bitmask with the nth bit flipped to 1
+if the query returns a row where the col_name has a value of n.
+The bits returned are 0-indexed.
+E.g. if your query returns col_name values of 1, 7, 11, and 12,
+you will get back 2 bytes and the bitmask will be 0100000100011000
+Note that the 1st, 7th, 11th, and 12th bits are flipped to 1 and
+the rest are 0.
+"""
+
 object_table_query = url(
     r"^table/(?P<objtype>[\w.]+)/(?P<objid>\d+)/query/$",
     views.object_table_query,
@@ -597,6 +616,7 @@
     annotations,
     table_query,
     table_metadata,
+    table_obj_id_bitmask,
     object_table_query,
     open_with_options,
 ]

omeroweb/webgateway/views.py

@@ -21,6 +21,7 @@
 import omero
 import omero.clients
 from past.builtins import unicode
+import numpy
 
 from django.http import (
     HttpResponse,
@@ -2892,38 +2893,16 @@ def _bulk_file_annotations(request, objtype, objid, conn=None, **kwargs):
 annotations = login_required()(jsonp(_bulk_file_annotations))
 
 
-def _table_query(request, fileid, conn=None, query=None, lazy=False, **kwargs):
-    """
-    Query a table specified by fileid
-    Returns a dictionary with query result if successful, error information
-    otherwise
-
-    @param request:     http request; querystring must contain key 'query'
-                        with query to be executed, or '*' to retrieve all rows.
-                        If query is in the format word-number, e.g. "Well-7",
-                        if will be run as (word==number), e.g. "(Well==7)".
-                        This is supported to allow more readable query strings.
-    @param fileid:      Numeric identifier of file containing the table
-    @param query:       The table query. If None, use request.GET.get('query')
-                        E.g. '*' to return all rows.
-                        If in the form 'colname-1', query will be (colname==1)
-    @param lazy:        If True, instead of returning a 'rows' list,
-                        'lazy_rows' will be a generator.
-                        Each gen.next() will return a list of row data
-                        AND 'table' returned MUST be closed.
-    @param conn:        L{omero.gateway.BlitzGateway}
-    @param **kwargs:    offset, limit
-    @return:            A dictionary with key 'error' with an error message
-                        or with key 'data' containing a dictionary with keys
-                        'columns' (an array of column names) and 'rows'
-                        (an array of rows, each an array of values)
-    """
-    if query is None:
-        query = request.GET.get("query")
-    if not query:
-        return dict(error="Must specify query parameter, use * to retrieve all")
-    col_names = request.GET.getlist("col_names")
-
+def perform_table_query(
+    conn,
+    fileid,
+    query,
+    col_names,
+    offset=0,
+    limit=None,
+    lazy=False,
+    check_max_rows=True,
+):
     ctx = conn.createServiceOptsDict()
     ctx.setOmeroGroup("-1")
 
@@ -2953,19 +2932,8 @@ def _table_query(request, fileid, conn=None, query=None, lazy=False, **kwargs):
         column_names = [col.name for col in cols]
         rows = t.getNumberOfRows()
 
-        offset = kwargs.get("offset", 0)
-        limit = kwargs.get("limit", None)
-        if not offset:
-            offset = int(request.GET.get("offset", 0))
-        if not limit:
-            limit = (
-                int(request.GET.get("limit"))
-                if request.GET.get("limit") is not None
-                else rows
-            )
-
         range_start = offset
-        range_size = limit
+        range_size = limit if limit is not None else rows
         range_end = min(rows, range_start + range_size)
 
         if query == "*":
@@ -2989,12 +2957,13 @@ def _table_query(request, fileid, conn=None, query=None, lazy=False, **kwargs):
             except Exception:
                 return dict(error="Error executing query: %s" % query)
 
-        if len(hits) > settings.MAX_TABLE_DOWNLOAD_ROWS:
-            error = (
-                "Trying to download %s rows exceeds configured"
-                " omero.web.max_table_download_rows of %s"
-            ) % (len(hits), settings.MAX_TABLE_DOWNLOAD_ROWS)
-            return {"error": error, "status": 404}
+        if check_max_rows:
+            if len(hits) > settings.MAX_TABLE_DOWNLOAD_ROWS:
+                error = (
+                    "Trying to download %s rows exceeds configured"
+                    " omero.web.max_table_download_rows of %s"
+                ) % (len(hits), settings.MAX_TABLE_DOWNLOAD_ROWS)
+                return {"error": error, "status": 404}
 
         def row_generator(table, h):
             # hits are all consecutive rows - can load them in batches
@@ -3041,6 +3010,53 @@ def row_generator(table, h):
             t.close()
 
 
+def _table_query(request, fileid, conn=None, query=None, lazy=False, **kwargs):
+    """
+    Query a table specified by fileid
+    Returns a dictionary with query result if successful, error information
+    otherwise
+
+    @param request:     http request; querystring must contain key 'query'
+                        with query to be executed, or '*' to retrieve all rows.
+                        If query is in the format word-number, e.g. "Well-7",
+                        if will be run as (word==number), e.g. "(Well==7)".
+                        This is supported to allow more readable query strings.
+    @param fileid:      Numeric identifier of file containing the table
+    @param query:       The table query. If None, use request.GET.get('query')
+                        E.g. '*' to return all rows.
+                        If in the form 'colname-1', query will be (colname==1)
+    @param lazy:        If True, instead of returning a 'rows' list,
+                        'lazy_rows' will be a generator.
+                        Each gen.next() will return a list of row data
+                        AND 'table' returned MUST be closed.
+    @param conn:        L{omero.gateway.BlitzGateway}
+    @param **kwargs:    offset, limit
+    @return:            A dictionary with key 'error' with an error message
+                        or with key 'data' containing a dictionary with keys
+                        'columns' (an array of column names) and 'rows'
+                        (an array of rows, each an array of values)
+    """
+    if query is None:
+        query = request.GET.get("query")
+    if not query:
+        return dict(error="Must specify query parameter, use * to retrieve all")
+    col_names = request.GET.getlist("col_names")
+
+    offset = kwargs.get("offset", 0)
+    limit = kwargs.get("limit", None)
+    if not offset:
+        offset = int(request.GET.get("offset", 0))
+    if not limit:
+        limit = (
+            int(request.GET.get("limit"))
+            if request.GET.get("limit") is not None
+            else None
+        )
+    return perform_table_query(
+        conn, fileid, query, col_names, offset=offset, limit=limit, lazy=lazy
+    )
+
+
 table_query = login_required()(jsonp(_table_query))
 
 
@@ -3089,6 +3105,58 @@ def _table_metadata(request, fileid, conn=None, query=None, lazy=False, **kwargs
 table_metadata = login_required()(jsonp(_table_metadata))
 
 
+@login_required()
+@jsonp
+def obj_id_bitmask(request, fileid, conn=None, query=None, lazy=False, **kwargs):
+    col_name = request.GET.get("col_name", "object")
+    if query is None:
+        query = request.GET.get("query")
+    if not query:
+        return dict(error="Must specify query parameter, use * to retrieve all")
+
+    offset = kwargs.get("offset", 0)
+    limit = kwargs.get("limit", None)
+    if not offset:
+        offset = int(request.GET.get("offset", 0))
+    if not limit:
+        limit = (
+            int(request.GET.get("limit"))
+            if request.GET.get("limit") is not None
+            else None
+        )
+
+    rsp_data = perform_table_query(
+        conn,
+        fileid,
+        query,
+        [col_name],
+        offset=offset,
+        limit=limit,
+        lazy=False,
+        check_max_rows=False,
+    )
+    if "error" in rsp_data:
+        return rsp_data
+    data = rowsToByteArray(rsp_data["data"]["rows"])
+    return HttpResponse(bytes(data), content_type="application/octet-stream")
+
+
+def rowsToByteArray(rows):
+    maxval = 0
+    for obj in rows:
+        obj_id = int(obj[0])
+        maxval = max(obj_id, maxval)
+    bitArray = numpy.zeros(maxval + 1, dtype='uint8')
+    for obj in rows:
+        obj_id = int(obj[0])
+        bitArray[obj_id] = 1
+    packed = numpy.packbits(bitArray, bitorder='big')
+    data = bytearray()
+    for val in packed:
+        data.append(val)
+    return data
+
+
 @login_required()
 @jsonp
 def object_table_query(request, objtype, objid, conn=None, **kwargs):

test/unit/test_webgateway.py

@@ -8,6 +8,7 @@
 
 from omeroweb.webgateway.webgateway_cache import FileCache, WebGatewayCache
 from omeroweb.webgateway.webgateway_cache import WebGatewayTempFile
+from omeroweb.webgateway import views
 import omero.gateway
 
 
@@ -384,3 +385,10 @@ def testJsonCache(self):
         assert self.wcache._json_cache._num_entries != 0
         self.wcache.clear()
         assert self.wcache._json_cache._num_entries == 0
+
+class TestViews(object):
+    def testRowstoByteArray(self):
+        rows = [[1],[2],[7],[11],[12]]
+        data = views.rowsToByteArray(rows)
+        assert data[0] == 97  #01100001 First, Second and 7th bits
+        assert data[1] == 24  #00011000 11th and 12th bits