Add option to dump taint graph (#5080)

* Add option to dump taint graph

* Fix types

* Simplify types

Co-authored-by: Matthew Brown <github@muglug.com>

docs/security_analysis/index.md

@@ -81,3 +81,12 @@ To generate a SARIF report run Psalm with the `--report` flag and a `.sarif` ext
 ```bash
 psalm --report=results.sarif
 ```
+
+## Debugging the taint graph
+
+Psalm can output the taint graph using the DOT language. This is useful when expected taints are not detected. To generate a DOT graph run Psalm with the `--dump-taint-graph` flag. For example:
+
+```bash
+psalm --taint-analysis --dump-taint-graph=taints.dot
+dot -Tsvg -o taints.svg taints.dot
+```

src/Psalm/Internal/Codebase/DataFlowGraph.php

@@ -8,6 +8,9 @@
 use function strlen;
 use function array_reverse;
 use function array_sum;
+use function array_walk;
+use function array_merge;
+use function array_keys;
 
 abstract class DataFlowGraph
 {
@@ -134,4 +137,18 @@ public function getEdgeStats() : array
 
         return [$count, \count($origin_counts), \count($destination_counts), $mean];
     }
+
+    /**
+     * @psalm-return list<list<string>>
+     */
+    public function summarizeEdges(): array
+    {
+        $edges = [];
+
+        foreach ($this->forward_edges as $source => $destinations) {
+            $edges[] = array_merge([$source], array_keys($destinations));
+        }
+
+        return $edges;
+    }
 }

src/command_functions.php

@@ -345,6 +345,9 @@ function getPsalmHelpText(): string
     --taint-analysis
         Run Psalm in taint analysis mode – see https://psalm.dev/docs/security_analysis for more info
 
+    --dump-taint-graph=OUTPUT_PATH
+        Output the taint graph using the DOT language – requires --taint-analysis
+
 Issue baselines:
     --set-baseline=PATH
         Save all current error level issues to a file, to mark them as info in subsequent runs

src/psalm.php

@@ -126,6 +126,7 @@
     'track-tainted-input',
     'taint-analysis',
     'security-analysis',
+    'dump-taint-graph:',
     'find-unused-psalm-suppress',
     'error-level:',
 ];
@@ -282,6 +283,9 @@ function () use ($current_dir, $options, $vendor_dir): ?\Composer\Autoload\Class
     || isset($options['security-analysis'])
     || isset($options['taint-analysis']));
 
+/** @var string|null $dump_taint_graph */
+$dump_taint_graph = $options['dump-taint-graph'] ?? null;
+
 if (array_key_exists('v', $options)) {
     echo 'Psalm ' . PSALM_VERSION . PHP_EOL;
     exit;
@@ -683,6 +687,18 @@ function (string $arg): bool {
     $project_analyzer->findReferencesTo($find_references_to);
 }
 
+$flow_graph = $project_analyzer->getCodebase()->taint_flow_graph;
+if ($flow_graph !== null && $dump_taint_graph !== null) {
+    file_put_contents($dump_taint_graph, "digraph Taints {\n\t".
+        implode("\n\t", array_map(
+            function (array $edges) {
+                return '"'.implode('" -> "', $edges).'"';
+            },
+            $flow_graph->summarizeEdges()
+        )) .
+        "\n}\n");
+}
+
 if (isset($options['set-baseline']) && is_string($options['set-baseline'])) {
     fwrite(STDERR, 'Writing error baseline to file...' . PHP_EOL);
 

tests/EndToEnd/PsalmEndToEndTest.php

@@ -169,6 +169,25 @@ public function testTaintingWithoutInit(): void
         $this->assertSame(1, $result['CODE']);
     }
 
+    public function testTaintGraphDumping(): void
+    {
+        $this->runPsalmInit(1);
+        $result = $this->runPsalm(
+            [
+                '--taint-analysis',
+                '--dump-taint-graph='.self::$tmpDir.'/taints.dot',
+            ],
+            self::$tmpDir,
+            true
+        );
+
+        $this->assertSame(1, $result['CODE']);
+        $this->assertFileEquals(
+            __DIR__ . '/../fixtures/expected_taint_graph.dot',
+            self::$tmpDir.'/taints.dot'
+        );
+    }
+
     public function testLegacyConfigWithoutresolveFromConfigFile(): void
     {
         $this->runPsalmInit(1);

tests/fixtures/expected_taint_graph.dot

@@ -0,0 +1,15 @@
+digraph Taints {
+	"$_GET:src/FileWithErrors.php:345" -> "$_GET['abc']-src/FileWithErrors.php:345-349"
+	"$_GET['abc']-src/FileWithErrors.php:345-349" -> "coalesce-src/FileWithErrors.php:345-363"
+	"$s-src/FileWithErrors.php:109-110" -> "acme\sampleproject\bar"
+	"$s-src/FileWithErrors.php:162-163" -> "acme\sampleproject\baz"
+	"$s-src/FileWithErrors.php:215-216" -> "acme\sampleproject\bat"
+	"$s-src/FileWithErrors.php:269-270" -> "acme\sampleproject\bang"
+	"acme\sampleproject\bang#1" -> "$s-src/FileWithErrors.php:269-270"
+	"acme\sampleproject\bar#1" -> "$s-src/FileWithErrors.php:109-110"
+	"acme\sampleproject\bat#1" -> "$s-src/FileWithErrors.php:215-216"
+	"acme\sampleproject\baz#1" -> "$s-src/FileWithErrors.php:162-163"
+	"acme\sampleproject\foo#1" -> "$s-src/FileWithErrors.php:57-58"
+	"call to echo-src/FileWithErrors.php:335-364" -> "echo#1-src/filewitherrors.php:330"
+	"coalesce-src/FileWithErrors.php:345-363" -> "call to echo-src/FileWithErrors.php:335-364"
+}