Lucene Query Builder Unification

components/common/test/ome/util/search/LuceneQueryBuilderTest.java

@@ -40,6 +40,9 @@ public void test_buildLuceneQuery() throws InvalidQueryException, ParseException
         String raw, expected;
         List<String> fields = new ArrayList<String>();
 
+        // Invalid
+        raw = "%"; expected = "";
+        checkQuery(fields, raw, expected);
 
         // No fields are provided
 

components/server/src/ome/services/search/FullText.java

@@ -21,6 +21,7 @@
 import ome.conditions.ApiUsageException;
 import ome.model.IAnnotated;
 import ome.model.IObject;
+import ome.model.core.Image;
 import ome.system.ServiceFactory;
 import ome.util.search.InvalidQueryException;
 import ome.util.search.LuceneQueryBuilder;
@@ -132,11 +133,24 @@ public FullText(SearchValues values, String fields, String from,
             throw new ApiUsageException(
                     "Invalid date format, dates must be in format YYYYMMDD.");
         }
-
+
+        if (LuceneQueryBuilder.DATE_ACQUISITION.equals(dateType) &&
+                !values.onlyTypes.contains(Image.class)) {
+            // Ignore acquisition ranges for non-images
+            dFrom = null;
+            dTo = null;
+        }
+
         try {
             this.queryStr = LuceneQueryBuilder.buildLuceneQuery(fieldsArray, dFrom,
                     dTo, dateType, query);
-            log.info("Generated Lucene query: "+this.queryStr);
+            if (this.queryStr.isEmpty()) {
+                q = null;
+                log.info("Generated empty Lucene query");
+                return; // EARLY EXIT!
+            } else {
+                log.info("Generated Lucene query: "+this.queryStr);
+            }
         } catch (InvalidQueryException e1) {
             throw new ApiUsageException(
                     "Invalid query: "+e1.getMessage());
@@ -293,6 +307,10 @@ protected void initializeQuery(FullTextQuery ftQuery) {
     @Transactional(readOnly = true)
     public Object doWork(Session s, ServiceFactory sf) {
 
+        if (q == null) {
+            return null;
+        }
+
         final Class<?> cls = values.onlyTypes.get(0);
         FullTextSession session = Search.createFullTextSession(s);
         Criteria criteria = criteria(session);

components/tools/OmeroPy/src/omero/gateway/__init__.py

@@ -3493,7 +3493,7 @@ def buildSearchQuery(self, text, fields=()):
 
         fields = [str(f) for f in fields]
         # # for each phrase or token, we strip out all non alpha-numeric
-        # except when inside double quotes. 
+        # except when inside double quotes.
         # To preserve quoted phrases we split by "
         phrases = text.split('"')
         tokens = []
@@ -3537,14 +3537,16 @@ def buildSearchQuery(self, text, fields=()):
         return text, leadingWc
 
 
-    def searchObjects(self, obj_types, text, created=None, fields=(), batchSize=1000, page=0, searchGroup=None, ownedBy=None):
+    def searchObjects(self, obj_types, text, created=None, fields=(), batchSize=1000, page=0, searchGroup=None, ownedBy=None,
+                      useAcquisitionDate=False):
         """
         Search objects of type "Project", "Dataset", "Image", "Screen", "Plate"
         Returns a list of results
 
         @param obj_types:   E.g. ["Dataset", "Image"]
         @param text:        The text to search for
         @param created:     L{omero.rtime} list or tuple (start, stop)
+        @param useAcquisitionDate if True, then use Image.acquisitionDate rather than import date for queries.
         @return:            List of Object wrappers. E.g. L{ImageWrapper}
         """
         if not text:
@@ -3573,24 +3575,36 @@ def getWrapper(obj_type):
                 details.setOwner(omero.model.ExperimenterI(ownedBy, False))
                 search.onlyOwnedBy(details, ctx)
 
-        text, leadingWc = self.buildSearchQuery(text, fields)
-        if leadingWc:
-            search.setAllowLeadingWildcard(True, ctx)
+        # Matching OMEROGateway.search()
+        search.setAllowLeadingWildcard(True)
+        search.setCaseSentivice(False)
 
-        if len(text) == 0:
-            return []
+        def parse_time(c, i):
+            try:
+                t = c[i]
+                t = unwrap(t)
+                if t is not None:
+                    t = time.localtime(t / 1000)
+                    t = time.strftime("%Y%m%d", t)
+                    return t
+            except:
+                pass
+            return None
 
-        logger.debug("Searching for: '%s'" % text);
+        d_from = parse_time(created, 0)
+        d_to = parse_time(created, 1)
+        d_type = useAcquisitionDate and "acquisitionDate" or "details.creationEvent.time"
 
         try:
-            if created:
-                search.onlyCreatedBetween(created[0], created[1], ctx);
             rv = []
             for t in types:
                 def actualSearch ():
                     search.onlyType(t().OMERO_CLASS, ctx)
-                    # search.bySomeMustNone(some, [], [])
-                    search.byFullText(text, ctx)
+                    search.byLuceneQueryBuilder(
+                        ",".join(fields),
+                        d_from, d_to, d_type,
+                        text, ctx)
+
                 timeit(actualSearch)()
                 # get results
                 def searchProcessing ():

components/tools/OmeroPy/src/omero/plugins/hql.py

@@ -35,6 +35,10 @@ def _configure(self, parser):
         parser.add_argument(
             "--all", help="Perform query on all groups", default=False,
             action="store_true", dest="admin")
+        parser.add_argument(
+            "--ids-only",
+            action="store_true",
+            help="Show only the ids of returned objects")
         parser.add_limit_arguments()
         parser.add_style_argument()
         parser.add_login_arguments()
@@ -64,7 +68,8 @@ def hql(self, args, loop=False):
         p = ParametersI()
         p.page(args.offset, args.limit)
         rv = self.project(q, args.query, p, ice_map)
-        has_details = self.display(rv, style=args.style)
+        has_details = self.display(rv, style=args.style,
+                                   idsonly=args.ids_only)
         if args.quiet or not sys.stdout.isatty():
             return
 
@@ -93,9 +98,9 @@ def hql(self, args, loop=False):
                 self.ctx.dbg("\nCurrent page: offset=%s, limit=%s\n" %
                              (p.theFilter.offset.val, p.theFilter.limit.val))
                 rv = self.project(q, args.query, p, ice_map)
-                self.display(rv, style=args.style)
+                self.display(rv, style=args.style, idsonly=args.ids_only)
             elif id.startswith("r"):
-                self.display(rv, style=args.style)
+                self.display(rv, style=args.style, idsonly=args.ids_only)
             else:
                 try:
                     id = long(id)
@@ -122,7 +127,7 @@ def hql(self, args, loop=False):
                     self.ctx.out("%s = %s" % (key, value))
             continue
 
-    def display(self, rv, cols=None, style=None):
+    def display(self, rv, cols=None, style=None, idsonly=False):
         import omero.all
         import omero.rtypes
         from omero.util.text import TableBuilder
@@ -136,7 +141,7 @@ def display(self, rv, cols=None, style=None):
             id = ""
             values = {}
             # Handling for simple lookup
-            if len(object_list) == 1 and \
+            if not idsonly and len(object_list) == 1 and \
                     isinstance(object_list[0], omero.rtypes.RObjectI):
                 has_details.append(idx)
                 o = object_list[0].val

components/tools/OmeroPy/src/omero/plugins/search.py

@@ -12,9 +12,12 @@
 
 
 import sys
+import time
 
+from Ice import OperationNotExistException
 from omero.cli import CLI
 from omero.plugins.hql import HqlControl
+from omero.rtypes import robject
 
 
 HELP = """Search for object ids by string.
@@ -43,14 +46,46 @@ def _configure(self, parser):
         parser.add_argument(
             "--index", action="store_true", default=False,
             help="Index an object as a administrator")
+        parser.add_argument(
+            "--no-parse",
+            action="store_true",
+            help="Pass the search string directly to Lucene with no parsing")
+        parser.add_argument(
+            "--field", nargs="*",
+            default=(),
+            help=("Fields which should be searched "
+                  "(e.g. name, description, annotation)"))
+        parser.add_argument(
+            "--from",
+            dest="_from",
+            metavar="YYYY-MM-DD",
+            type=self.date,
+            help="Start date for limiting searches (YYYY-MM-DD)")
+        parser.add_argument(
+            "--to",
+            dest="_to",
+            metavar="YYYY-MM-DD",
+            type=self.date,
+            help="End date for limiting searches (YYYY-MM-DD")
+        parser.add_argument(
+            "--date-type",
+            default="acquisitionDate",
+            choices=("acquisitionDate", "import"),
+            help=("Which field to use for --from/--to "
+                  "(default: acquisitionDate)"))
         parser.add_argument(
             "type",
             help="Object type to search for, e.g. 'Image' or 'Well'")
-        parser.add_argument(
-            "search_string", nargs="?",
-            help="Lucene search string")
+        HqlControl._configure(self, parser)
         parser.set_defaults(func=self.search)
-        parser.add_login_arguments()
+
+    def date(self, user_string):
+        try:
+            t = time.strptime(user_string, "%Y-%m-%d")
+            return time.strftime("%Y%m%d", t)
+        except Exception, e:
+            self.ctx.dbg(str(e))
+            raise
 
     def search(self, args):
         c = self.ctx.conn(args)
@@ -75,17 +110,48 @@ def search(self, args):
             c.sf.getUpdateService().indexObject(obj)
 
         else:
+            group = None
+            if args.admin:
+                group = "-1"
+            ctx = c.getContext(group)
             search = c.sf.createSearchService()
             try:
                 try:
+                    # Matching OMEROGateway.search()
+                    search.setAllowLeadingWildcard(True)
+                    search.setCaseSentivice(False)
                     search.onlyType(args.type)
-                    search.byFullText(args.search_string)
-                    if not search.hasNext():
+
+                    if args.no_parse:
+                        if args._from or args._to or args.field:
+                            self.ctx.err("Ignoring from/to/fields")
+                        search.byFullText(args.query)
+                    else:
+                        try:
+                            if args.date_type == "import":
+                                args.date_type = "details.creationEvent.time"
+                            search.byLuceneQueryBuilder(
+                                ",".join(args.field),
+                                args._from, args._to, args.date_type,
+                                args.query, ctx)
+                        except OperationNotExistException:
+                            self.ctx.err(
+                                "Server does not support byLuceneQueryBuilder")
+                            search.byFullText(args.query)
+
+                    if not search.hasNext(ctx):
                         self.ctx.die(433, "No results found.")
-                    while search.hasNext():
-                        results = search.results()
+
+                    self.ctx.set("search.results", [])
+                    while search.hasNext(ctx):
+                        results = search.results(ctx)
+                        self.ctx.get("search.results").extend(results)
                         results = [[x] for x in results]
-                        self.display(results)
+                        if not args.ids_only:
+                            results = [[robject(x[0])] for x in results]
+                        self.display(results,
+                                     style=args.style,
+                                     idsonly=args.ids_only)
                 except omero.ApiUsageException, aue:
                     self.ctx.die(434, aue.message)