- src/web_exploit_detector.js: processRulesOnFile(): previously if a
   path match succeeded but the file requires a content check and cannot
   be read, the result of the test was the result of the path check.
   This has now been changed to avoid false positives; if a file passes
   a path check then the result will depend on the content check and
   file access. If the file can be accessed then it depends on the
   content check result as usual. If it cannot be read then the result
   be false unless the rule requires no content check, in which case the
   result will be set to the result of the path check.