add tracking of strings to compile-sites for poly-redos, in the style of Ruby

This commit is contained in:
erik-krogh
2023-02-02 19:26:51 +01:00
parent 52959d7c0a
commit 6e712b293a
4 changed files with 23 additions and 4 deletions

View File

@@ -85,6 +85,17 @@ predicate used_as_regex(Expr s, string mode) {
)
}
private import semmle.python.Concepts
private import semmle.python.RegexTreeView
/** Gets a parsed regular expression term that is executed at `exec`. */
RegExpTerm getTermForExecution(RegexExecution exec) {
exists(RegexTracking t, DataFlow::Node source | t.hasFlow(source, exec.getRegex()) |
result.getRegex() = source.asExpr() and
result.isRootTerm()
)
}
/**
* Gets the canonical name for the API graph node corresponding to the `re` flag `flag`. For flags
* that have multiple names, we pick the long-form name as a canonical representative.

View File

@@ -13,6 +13,7 @@ private import semmle.python.dataflow.new.RemoteFlowSources
private import semmle.python.dataflow.new.BarrierGuards
private import semmle.python.RegexTreeView::RegexTreeView as TreeView
private import semmle.python.ApiGraphs
private import semmle.python.regex
/**
* Provides default sources, sinks and sanitizers for detecting
@@ -66,7 +67,7 @@ module PolynomialReDoS {
RegexExecutionAsSink() {
exists(RegexExecution re |
re.getRegex().asExpr() = t.getRegex() and
t = getTermForExecution(re) and
this = re.getString()
) and
t.isRootTerm()