php (#107)

Signed-off-by: Prabhu Subramanian <prabhu@appthreat.com>

README.md

@@ -164,14 +164,15 @@ docker run --rm -v /tmp:/tmp -v $HOME:$HOME -v $(pwd):/app:rw -it ghcr.io/appthr
 
 ## Languages supported
 
-- C/C++ (Requires Java 17 or above)
+- C/C++
 - H (C/C++ Header files alone)
 - Java (Requires compilation)
 - Jar
 - Android APK (Requires Android SDK. Set the environment variable `ANDROID_HOME`)
 - JavaScript
 - TypeScript
 - Python
+- PHP
 
 ## Atom Specification
 
@@ -210,7 +211,8 @@ Apache-2.0
 
 ## Developing / Contributing
 
-Install Java 17 or 21 (Recommended)
+Install Java 21
+Node.js > 21
 
 ```shell
 sbt clean stage scalafmt test createDistribution

build.sbt

@@ -1,9 +1,9 @@
 name                     := "atom"
 ThisBuild / organization := "io.appthreat"
-ThisBuild / version      := "1.8.0"
+ThisBuild / version      := "1.8.1"
 ThisBuild / scalaVersion := "3.3.1"
 
-val chenVersion      = "1.1.1"
+val chenVersion      = "1.1.3"
 
 lazy val atom = Projects.atom
 
@@ -24,6 +24,7 @@ libraryDependencies ++= Seq(
   "io.appthreat"      %% "javasrc2cpg"       % Versions.chen,
   "io.appthreat"      %% "jssrc2cpg"         % Versions.chen,
   "io.appthreat"      %% "jimple2cpg"        % Versions.chen,
+  "io.appthreat"      %% "php2atom"        % Versions.chen,
   "io.appthreat"      %% "semanticcpg"       % Versions.chen % Test classifier "tests",
   "io.appthreat"      %% "x2cpg"             % Versions.chen % Test classifier "tests",
   "io.appthreat"      %% "pysrc2cpg"         % Versions.chen % Test classifier "tests",

codemeta.json

@@ -7,8 +7,8 @@
   "downloadUrl": "https://github.com/AppThreat/atom",
   "issueTracker": "https://github.com/AppThreat/atom/issues",
   "name": "atom",
-  "version": "1.8.0",
-  "description": ".",
+  "version": "1.8.1",
+  "description": "Atom is a novel intermediate representation for next-generation code analysis.",
   "applicationCategory": "code-analysis",
   "keywords": [
     "static-analysis",

contrib/Dockerfile-graalvm21

@@ -5,14 +5,14 @@ LABEL maintainer="appthreat" \
       org.opencontainers.image.authors="Team AppThreat <cloud@appthreat.com>" \
       org.opencontainers.image.source="https://github.com/appthreat/chen" \
       org.opencontainers.image.url="https://github.com/appthreat/atom" \
-      org.opencontainers.image.version="1.7.x" \
+      org.opencontainers.image.version="1.8.x" \
       org.opencontainers.image.vendor="appthreat" \
       org.opencontainers.image.licenses="Apache-2.0" \
       org.opencontainers.image.title="atom" \
       org.opencontainers.image.description="Container image for testing AppThreat atom with Oracle GraalVM" \
       org.opencontainers.docker.cmd="docker run --rm -v /tmp:/tmp -v $HOME:$HOME -v $(pwd):/app:rw -it ghcr.io/appthreat/atom"
 
-ARG MAVEN_VERSION=3.9.5
+ARG MAVEN_VERSION=3.9.6
 
 ENV MAVEN_VERSION=$MAVEN_VERSION \
     MAVEN_HOME="/opt/maven/${MAVEN_VERSION}" \
@@ -36,7 +36,7 @@ RUN set -e; \
             ;; \
         *) echo >&2 "error: unsupported architecture: '$ARCH_NAME'"; exit 1 ;; \
     esac; \
-    echo -e "[nodejs]\nname=nodejs\nstream=21\nprofiles=\nstate=enabled\n" > /etc/dnf/modules.d/nodejs.module \
+    echo -e "[nodejs]\nname=nodejs\nstream=20\nprofiles=\nstate=enabled\n" > /etc/dnf/modules.d/nodejs.module \
     && microdnf install -y gcc git-core wget bash glibc-common glibc-all-langpacks \
         pcre2 findutils which tar gzip zip unzip sudo nodejs \
     && curl -s "https://get.sdkman.io" | bash \

src/main/scala/io/appthreat/atom/Atom.scala

@@ -15,6 +15,8 @@ import io.appthreat.jssrc2cpg.passes.{
     JavaScriptInheritanceNamePass
 }
 import io.appthreat.jssrc2cpg.{JsSrc2Cpg, Config as JSConfig}
+import io.appthreat.php2atom.passes.PhpSetKnownTypesPass
+import io.appthreat.php2atom.{Php2Atom, Config as PhpConfig}
 import io.appthreat.pysrc2cpg.{
     DynamicTypeHintFullNamePass,
     Py2CpgOnFileSystem,
@@ -514,6 +516,18 @@ object Atom:
                               new AstLinkerPass(ag).createAndApply()
                               ag
                           }
+                  case Languages.PHP =>
+                      new Php2Atom().createCpgWithOverlays(
+                        PhpConfig()
+                            .withDisableDummyTypes(true)
+                            .withInputPath(config.inputPath.pathAsString)
+                            .withOutputPath(outputAtomFile)
+                            .withDefaultIgnoredFilesRegex(List("\\..*".r))
+                            .withIgnoredFilesRegex(".*(samples|examples|docs|tests).*")
+                      ).map { ag =>
+                          new PhpSetKnownTypesPass(ag).createAndApply()
+                          ag
+                      }
                   case _ => Failure(
                         new RuntimeException(
                           s"No language frontend supported for language '$language'"

src/main/scala/io/appthreat/atom/slicing/ReachableSlicing.scala

@@ -69,6 +69,14 @@ object ReachableSlicing:
                 .map(toSlice)
                 .toList
         end if
+        if language == Languages.PHP
+        then
+            flowsList ++= atom.ret.where(_.tag.name(config.sinkTag)).reachableByFlows(
+              atom.tag.name(config.sourceTag).parameter
+            ).map(toSlice).toList
+            flowsList ++= atom.tag.name(FRAMEWORK_TAG).parameter.reachableByFlows(
+              atom.tag.name(config.sourceTag).parameter
+            ).map(toSlice).toList
         if language == Languages.NEWC || language == Languages.C
         then
             flowsList ++= atom.tag.name(LIBRARY_CALL_TAG).call.reachableByFlows(atom.tag.name(

src/main/scala/io/appthreat/atom/slicing/UsageSlicing.scala

@@ -45,14 +45,18 @@ object UsageSlicing:
         val slices       = usageSlices(atom, () => getDeclarations, typeMap)
         val language     = atom.metaData.language.headOption
         val userDefTypes = userDefinedTypes(atom)
-        if language.get == Languages.NEWC || language.get == Languages.C then
+        if language.get == Languages.NEWC || language.get == Languages.C
+        then
             ProgramUsageSlice(slices ++ importsAsSlices(atom), userDefTypes ++ routesAsUDT(atom))
         else if language.get == Languages.PYTHON || language.get == Languages.PYTHONSRC
         then
             ProgramUsageSlice(
               slices ++ externalCalleesAsSlices(atom, typeMap),
               userDefTypes ++ routesAsUDT(atom)
             )
+        else if language.get == Languages.PHP
+        then
+            ProgramUsageSlice(slices, userDefTypes ++ routesAsUDT(atom))
         else
             ProgramUsageSlice(slices ++ unusedTypeDeclAsSlices(atom), userDefTypes)
     end calculateUsageSlice

wrapper/nodejs/astgen.js

@@ -98,6 +98,7 @@ const vueTemplateRegex = /(<template.*>)([\s\S]*)(<\/template>)/gi;
 const vueCommentRegex = /<!--[\s\S]*?-->/gi;
 const vueBindRegex = /(:\[)([\s\S]*?)(\])/gi;
 const vuePropRegex = /\s([.:@])([a-zA-Z]*?=)/gi;
+const vueOpenImgTag = /(<img)((?!>)[\s\S]+?)( [^/]>)/gi;
 
 /**
  * Convert a single vue file to AST
@@ -115,7 +116,10 @@ const toVueAst = (file) => {
       return grA.replaceAll(/\S/g, " ") + grB + grC.replaceAll(/\S/g, " ");
     })
     .replace(vuePropRegex, function (match, grA, grB) {
-      return " " + grA.replace(/[.:@]/g, " ") + grB;
+      return " " + grA.replace(/[.:@]/g, " ") + grB.replaceAll(".", "-");
+    })
+    .replace(vueOpenImgTag, function (match, grA, grB, grC) {
+      return grA + grB + grC.replace(" >", "/>");
     })
     .replace(vueTemplateRegex, function (match, grA, grB, grC) {
       return grA + grB.replaceAll("{{", "{ ").replaceAll("}}", " }") + grC;
@@ -256,7 +260,7 @@ const createJSAst = async (options) => {
  * Generate AST for .vue files
  */
 const createVueAst = async (options) => {
-  const srcFiles = getAllFiles(options.src, ".vue");
+  const srcFiles = await getAllFiles(options.src, ".vue");
   for (const file of srcFiles) {
     try {
       const ast = toVueAst(file);

wrapper/nodejs/build.sh

@@ -7,4 +7,13 @@ if [ -e "../../target/atom.zip" ]; then
 else
     echo "Build the atom project using 'sbt createDistribution' before running this script"
 fi
+php -r "copy('https://getcomposer.org/installer', 'composer-setup.php');"
+php -r "if (hash_file('sha384', 'composer-setup.php') === 'e21205b207c3ff031906575712edab6f13eb0b361f2085f1f1237b7126d785e826a450292b6cfd1d64d92e6563bbde02') { echo 'Installer verified'; } else { echo 'Installer corrupt'; unlink('composer-setup.php'); } echo PHP_EOL;"
+php composer-setup.php
+php -r "unlink('composer-setup.php');"
+export COMPOSER_VENDOR_DIR=plugins
+php composer.phar require nikic/php-parser --ignore-platform-reqs --optimize-autoloader
+
 npm install
+
+rm composer.phar composer.json composer.lock

wrapper/nodejs/index.js

@@ -19,6 +19,7 @@ const _version = selfPJson.version;
 export const LOG4J_CONFIG = join(dirName, "plugins", "log4j2.xml");
 export const ATOM_HOME = join(dirName, "plugins");
 export const APP_LIB_DIR = join(ATOM_HOME, "lib");
+export const PHP_PARSER_BIN = join(ATOM_HOME, "bin", "php-parse");
 const freeMemoryGB = Math.max(Math.floor(freemem() / 1024 / 1024 / 1024), 4);
 export const JVM_ARGS = "-XX:+UseG1GC -XX:+UseStringDeduplication";
 export const JAVA_OPTS = `${
@@ -66,7 +67,8 @@ export const executeAtom = (atomArgs) => {
     ]);
   const env = {
     ...process.env,
-    ATOM_HOME
+    ATOM_HOME,
+    PHP_PARSER_BIN
   };
   const cwd = process.env.ATOM_CWD || process.cwd();
   spawnSync(JAVACMD, args, {

wrapper/nodejs/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@appthreat/atom",
-  "version": "1.8.0",
+  "version": "1.8.1",
   "description": "Create atom (⚛) representation for your application, packages and libraries",
   "exports": "./index.js",
   "type": "module",
@@ -18,7 +18,8 @@
   },
   "bin": {
     "atom": "./index.js",
-    "astgen": "./astgen.js"
+    "astgen": "./astgen.js",
+    "phpastgen": "./phpastgen.js"
   },
   "engines": {
     "node": ">=16.0.0"

wrapper/nodejs/phpastgen.js

@@ -0,0 +1,31 @@
+#!/usr/bin/env node
+
+import { dirname, join } from "node:path";
+import { spawnSync } from "node:child_process";
+import { fileURLToPath } from "node:url";
+import { detectPhp } from "./utils.mjs";
+
+let url = import.meta.url;
+if (!url.startsWith("file://")) {
+  url = new URL(`file://${import.meta.url}`).toString();
+}
+const dirName = import.meta ? dirname(fileURLToPath(url)) : __dirname;
+export const PLUGINS_HOME = join(dirName, "plugins");
+const PHP_PARSER_BIN =
+  process.env.PHP_PARSER_BIN || join(PLUGINS_HOME, "bin", "php-parse");
+
+function main(argvs) {
+  if (!detectPhp()) {
+    console.warn("PHP is not installed!");
+    return false;
+  }
+  const cwd = process.env.ATOM_CWD || process.cwd();
+  spawnSync(PHP_PARSER_BIN, argvs, {
+    encoding: "utf-8",
+    cwd,
+    stdio: "inherit",
+    stderr: "inherit",
+    timeout: process.env.ATOM_TIMEOUT || process.env.ASTGEN_TIMEOUT
+  });
+}
+main(process.argv.slice(2));

wrapper/nodejs/utils.mjs

@@ -86,3 +86,13 @@ export const detectJava = () => {
   }
   return true;
 };
+
+export const detectPhp = () => {
+  let result = spawnSync(process.env.JAVA_CMD || "php", ["--version"], {
+    encoding: "utf-8"
+  });
+  if (result.status !== 0 || result.error) {
+    return false;
+  }
+  return true;
+};