This is an automated email from the ASF dual-hosted git repository.
dkuppitz pushed a commit to branch TINKERPOP-2114
in repository https://gitbox.apache.org/repos/asf/tinkerpop.git
The following commit(s) were added to refs/heads/TINKERPOP-2114 by this push:
new 340c06f fixup! TINKERPOP-2114 Documented common Gremlin anti-patterns
(and solutions)
340c06f is described below
commit 340c06f711fca28a248757f38ab4443b6fce511b
Author: Daniel Kuppitz <[email protected]>
AuthorDate: Mon Dec 17 09:42:11 2018 -0700
fixup! TINKERPOP-2114 Documented common Gremlin anti-patterns (and
solutions)
---
docs/src/recipes/anti-patterns.asciidoc | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/docs/src/recipes/anti-patterns.asciidoc
b/docs/src/recipes/anti-patterns.asciidoc
index 88684db..3a44bf6 100644
--- a/docs/src/recipes/anti-patterns.asciidoc
+++ b/docs/src/recipes/anti-patterns.asciidoc
@@ -128,7 +128,7 @@ and under certain circumstance they will not attempt to
rewrite a traversal. For
enabled (e.g. by using certain steps, such as `path()`, `simplePath()`,
`otherV()`, etc.), then the assumption is that all steps are
required in order to produce the desired path.
-An often seen anti-pattern is the one that explicitely traversers to an edge
and then to a vertex without using any filters.
+An often seen anti-pattern is the one that explicitly traverses to an edge and
then to a vertex without using any filters.
[gremlin-groovy,modern]
----
@@ -138,7 +138,7 @@ g.V().hasLabel("software").inE("created").outV().count()
<2>
<1> The `created` edge is never really needed as the traversal only asks for
all things that were created by all persons in the graph.
These "things" are only represented by the adjacent vertices, not the
edges.
-<2> This traversals counts the persons in the graph who created software. The
interesting thing about this query is that it actually
+<2> This traversal counts the persons in the graph who created software. The
interesting thing about this query is that it actually
doesn't need to traverse all the way to the `person` vertices to count
them. In this case it's sufficient to count the edges
between the `software` and `person` vertices. The performance of this
query pretty much depends on the particular provider
implementation, but counting incident edges is usually much faster than
counting adjacent vertices.
