only forward paths

Author: morria

.devcontainer/Dockerfile

@@ -6,10 +6,11 @@ RUN apt-get install -y git curl vim bash sudo tmux man ack zip unzip
 
 # nikic/php-ast
 RUN cd /tmp; git clone https://github.com/nikic/php-ast.git && cd php-ast && phpize && ./configure && make && make install
-RUN echo "extension=ast" > /usr/local/etc/php/conf.d/docker-php-ast.ini
+RUN echo "extension=ast" > /usr/local/etc/php/conf.d/php-ast.ini
 
 # xdebug
 RUN cd /tmp; curl -O https://xdebug.org/files/xdebug-3.2.1.tgz && tar -xvzf xdebug-3.2.1.tgz && cd xdebug-3.2.1 && phpize && ./configure && make && make install
+RUN echo "zend_extension=xdebug" > /usr/local/etc/php/conf.d/xdebug.ini
 
 # Composer
 RUN cd /tmp; curl -O https://getcomposer.org/installer && php ./installer --install-dir=/usr/local/bin --filename=composer

src/Paths/FunctionPaths.php

@@ -54,11 +54,11 @@ public function isEmpty(): bool
     public function toString(?int $max_length = null): string
     {
         if ($this->isEmpty()) {
-            return $this->function_name;
+            return '';
         }
 
         $paths = $this->paths;
-        if ($max_length !== null) {
+        if ($max_length !== null && count($paths) > $max_length) {
             shuffle($paths);
             $paths = array_slice($paths, 0, $max_length);
         }

src/Paths/GraphNode/NonTerminal.php

@@ -42,14 +42,32 @@ public function allPathsToTerminals(PartialPath $prefix): \Generator
         $previous_node = $prefix->lastNode();
         $prefix = $prefix->withNonTerminal($this);
 
-        // TODO: Only allow forward paths?
+        // Only follow "forward" paths by only going to children with
+        // greater indices than the previous node on the path.
+        //
+        // n.b.: This assumes that the indices of `$children` are
+        // monotonically incrementing by 1. If this ever isn't the case
+        // we'll need to call `array_values` on it.
+        $starting_index = array_search($previous_node, $this->children, $strict = true);
+        if ($starting_index === false) {
+            $starting_index = -1;
+        }
+
+        if ($starting_index + 1 <= count($this->children) - 1) {
+            foreach (range($starting_index + 1, count($this->children) - 1) as $i) {
+                yield from $this->children[$i]->allPathsToTerminals($prefix);
+            }
+        }
+
+        /*
         foreach ($this->children as $child) {
             // Skip the node we just popped out of
             if ($child === $previous_node) {
                 continue;
             }
             yield from $child->allPathsToTerminals($prefix);
         }
+        */
     }
 
     /**

src/Paths/GraphNodeVisitor.php

@@ -21,13 +21,18 @@ public function __construct(GraphNode $parent = null, bool $use_node_ids = false
         $this->use_node_ids = $use_node_ids;
     }
 
+    public function newGraphNodeVisitorWithParent(GraphNode $parent = null)
+    {
+        return new GraphNodeVisitor($parent, $this->use_node_ids);
+    }
+
     /**
      * @param $node Node|null|string|int
      */
     public function graphNodeFromNodeOrValue($node, GraphNode $parent): GraphNode
     {
         if ($node instanceof Node) {
-            return (new GraphNodeVisitor($parent, $this->use_node_ids))($node);
+            return $this->newGraphNodeVisitorWithParent($parent)($node);
         }
         return self::terminalFromNodeOrValue($node, $parent);
     }
@@ -109,7 +114,7 @@ public function visitParamList(Node $node): GraphNode
 
         $gn = new NonTerminal($this->nodeName($node), $this->parent);
         foreach ($node->children ?? [] as $param) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($param));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($param));
         }
         return $gn;
     }
@@ -147,22 +152,22 @@ public function visitForeach(Node $node): GraphNode
 
         $expr = $node->children['expr'] ?? null;
         if ($expr instanceof Node) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($expr));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($expr));
         }
 
         $value = $node->children['value'] ?? null;
         if ($value instanceof Node) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($value));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($value));
         }
 
         $key = $node->children['key'] ?? null;
         if ($key instanceof Node) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($key));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($key));
         }
 
         $stmts = $node->children['stmts'] ?? null;
         if ($stmts instanceof Node) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($stmts));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($stmts));
         }
 
         return $gn;
@@ -173,7 +178,7 @@ public function visitIf(Node $node): GraphNode
         $gn = new NonTerminal($this->nodeName($node), $this->parent);
 
         foreach ($node->children as $child) {
-            $gn->appendChild((new GraphNodeVisitor($gn))($child));
+            $gn->appendChild($this->newGraphNodeVisitorWithParent($gn)($child));
         }
 
         return $gn;

src/Paths/Subtokens.php

@@ -12,11 +12,16 @@ final class Subtokens
      */
     public static function fromString(string $string): array
     {
+
+        // Replace special characters
         $string = str_replace([',', '|', '\\', '"', "'"], ['comma', 'pipe', 'slash', 'quote', "quote"], $string);
 
         // Collapse whitespace
         $string = trim(preg_replace('/\s+/', ' ', $string));
 
+        // Remove non-ascii
+        $string = preg_replace('/[[:^print:]]/', '', $string);
+
         // Replace all-caps `FOO` with `Foo`.
         $string = preg_replace_callback('/([A-Z])([A-Z]+)/', function ($matches) {
             return $matches[1] . strtolower($matches[2]);

src/paths.php

@@ -64,6 +64,7 @@
         if ($function_paths->isEmpty()) {
             continue;
         }
+        // print(memory_get_usage()/1024/1024 . " " . memory_get_peak_usage()/1024/1024 . "\n");
         print $function_paths->toString($max_length) . "\n";
     }
 });

tests/FunctionPathsTest.php

@@ -24,9 +24,9 @@ public static function caseProvider(): array
     #[DataProvider('caseProvider')]
     public function testCases($actual_file, $expected_file): void
     {
-        $actual_paths = implode("\n", array_map(function (FunctionPaths $path): string {
+        $actual_paths = implode("\n", array_filter(array_map(function (FunctionPaths $path): string {
             return $path->toString();
-        }, iterator_to_array(FunctionPaths::fromFileName($actual_file))));
+        }, iterator_to_array(FunctionPaths::fromFileName($actual_file)))));
         $expected_paths = trim(file_get_contents($expected_file));
 
         $this->assertEquals($actual_paths, $expected_paths);

tests/SubtokensTest.php

@@ -37,7 +37,8 @@ public function testKebabCase(): void
     {
         $this->assertEquals(
             Subtokens::fromString('foo-bar'),
-            ['foo', 'bar']
+            ['foo', 'bar'],
+            'this'
         );
     }
 
@@ -112,5 +113,10 @@ public function testSpecialCharacters(): void {
             Subtokens::fromString(", | \\ \" '"),
             ['comma', 'pipe', 'slash', 'quote', 'quote']
         );
+
+        $this->assertEquals(
+            Subtokens::fromString('aAÂ'),
+            ['a', 'a']
+        );
     }
 }

tests/cases/0003_function.c2s

@@ -1 +0,0 @@
-f

tests/cases/0004_function.c2s

@@ -1 +1 @@
-f long,parameter,x long,parameter|function|return|variable,x x,parameter,long x,parameter|function|return|variable,x x,variable|return|function|parameter,long x,variable|return|function|parameter,x
+f long,parameter,x long,parameter|function|return|variable,x x,parameter|function|return|variable,x