[ https://issues.apache.org/jira/browse/TINKERPOP-1545?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15657806#comment-15657806 ]
ASF GitHub Bot commented on TINKERPOP-1545: ------------------------------------------- Github user dkuppitz commented on the issue: https://github.com/apache/tinkerpop/pull/486 I think we have 2 options: 1. reset requirements (a simple `this.requirements = null;`) at the end of mutating methods (`addStep` and `removeStep`) 2. don't cache requirements before `applyStrategies()` The problem with 1 is, that we would have to propagate the reset to all child- and parent-traversals. The problem with 2 is, that if we don't cache requirements while startegies are being applied, why do we cache them at all? Are requirements used at runtime? Thoughts? > IncidentToAdjacentStrategy is buggy > ----------------------------------- > > Key: TINKERPOP-1545 > URL: https://issues.apache.org/jira/browse/TINKERPOP-1545 > Project: TinkerPop > Issue Type: Bug > Affects Versions: 3.1.4, 3.2.3 > Reporter: Daniel Kuppitz > Assignee: Daniel Kuppitz > > {{IncidentToAdjacentStrategy}} turns {{outE().inV().simplePath()}} into > {{out().simplePath()}}, which will return a wrong result as soon as there is > more than 1 edge between a pair of vertices. > The set {{INVALIDATING_STEP_CLASSES}} should only contain > {{PathProcessor.class}} and {{LambdaHolder.class}} Furthermore > {{SimplePathStep}} should implement {{PathProcessor}}. -- This message was sent by Atlassian JIRA (v6.3.4#6332)