From 83e63aeb078bf0ee4f1a9f9d0eab32bdffc6c7ff Mon Sep 17 00:00:00 2001
From: epriestley <git@epriestley.com>
Date: Fri, 4 Sep 2020 17:24:25 -0700
Subject: [PATCH] Allow AAST to extract string literal values from HEREDOCs

Summary:
Ref T13577. I'd like to `arc lint --everything` to find other bad calls to `pht()` and similar functions, but `n_HEREDOC` nodes currently can not generate a response to "getStringLiteralValue()".

Support literal extraction from heredocs.

Test Plan: Added a test, made it pass. Will lint everything.

Maniphest Tasks: T13577

Differential Revision: https://secure.phabricator.com/D21455
---
 .../formatted-string.lint-test                | 11 ++++
 src/parser/xhpast/api/XHPASTNode.php          | 54 +++++++++++++------
 2 files changed, 49 insertions(+), 16 deletions(-)

diff --git a/src/lint/linter/xhpast/rules/__tests__/formatted-string/formatted-string.lint-test b/src/lint/linter/xhpast/rules/__tests__/formatted-string/formatted-string.lint-test
index 67ec7746..f9e5a11b 100644
--- a/src/lint/linter/xhpast/rules/__tests__/formatted-string/formatted-string.lint-test
+++ b/src/lint/linter/xhpast/rules/__tests__/formatted-string/formatted-string.lint-test
@@ -14,6 +14,16 @@ foobar(null, null, '%s');
 
 pht('x %s y');
 pht('x %s y'.'z');
+
+pht(<<<HEREDOC
+a b c
+HEREDOC
+  );
+
+pht(<<<HEREDOC
+a %s c
+HEREDOC
+  );
 ~~~~~~~~~~
 error:3:1:XHP54:Formatted String
 error:7:1:XHP54:Formatted String
@@ -22,6 +32,7 @@ error:11:1:XHP54:Formatted String
 error:13:1:XHP54:Formatted String
 error:15:1:XHP54:Formatted String
 error:16:1:XHP54:Formatted String
+error:23:1:XHP54:Formatted String
 ~~~~~~~~~~
 ~~~~~~~~~~
 {
diff --git a/src/parser/xhpast/api/XHPASTNode.php b/src/parser/xhpast/api/XHPASTNode.php
index 4decaea1..0f685c5e 100644
--- a/src/parser/xhpast/api/XHPASTNode.php
+++ b/src/parser/xhpast/api/XHPASTNode.php
@@ -33,7 +33,9 @@ final class XHPASTNode extends AASTNode {
         return $this->getChildByIndex(0)->evalStatic();
         break;
       case 'n_STRING_SCALAR':
-        return (string)$this->getStringLiteralValue();
+        return phutil_string_cast($this->getStringLiteralValue());
+      case 'n_HEREDOC':
+        return phutil_string_cast($this->getStringLiteralValue());
       case 'n_NUMERIC_SCALAR':
         $value = $this->getSemanticString();
         if (preg_match('/^0x/i', $value)) {
@@ -186,31 +188,51 @@ final class XHPASTNode extends AASTNode {
   }
 
   public function getStringLiteralValue() {
-    if ($this->getTypeName() != 'n_STRING_SCALAR') {
-      return null;
+    $type_name = $this->getTypeName();
+
+    if ($type_name === 'n_HEREDOC') {
+      $value = $this->getSemanticString();
+      $value = phutil_split_lines($value);
+      $value = array_slice($value, 1, -1);
+      $value = implode('', $value);
+
+      // Strip the final newline from value, this isn't part of the string
+      // literal.
+      $value = preg_replace('/(\r|\n|\r\n)\z/', '', $value);
+
+      return $this->newStringLiteralFromSemanticString($value);
     }
 
-    $value = $this->getSemanticString();
-    $type  = $value[0];
-    $value = preg_replace('/^b?[\'"]|[\'"]$/i', '', $value);
-    $esc   = false;
-    $len   = strlen($value);
-    $out   = '';
+    if ($type_name === 'n_STRING_SCALAR') {
+      $value = $this->getSemanticString();
+      $type  = $value[0];
+      $value = preg_replace('/^b?[\'"]|[\'"]$/i', '', $value);
 
-    if ($type == "'") {
-      // Single quoted strings treat everything as a literal except "\\" and
-      // "\'".
-      return str_replace(
-        array('\\\\', '\\\''),
-        array('\\',   "'"),
-        $value);
+      if ($type == "'") {
+        // Single quoted strings treat everything as a literal except "\\" and
+        // "\'".
+        return str_replace(
+          array('\\\\', '\\\''),
+          array('\\',   "'"),
+          $value);
+      }
+
+      return $this->newStringLiteralFromSemanticString($value);
     }
 
+    return null;
+  }
+
+  private function newStringLiteralFromSemanticString($value) {
     // Double quoted strings treat "\X" as a literal if X isn't specifically
     // a character which needs to be escaped -- e.g., "\q" and "\'" are
     // literally "\q" and "\'". stripcslashes() is too aggressive, so find
     // all these under-escaped backslashes and escape them.
 
+    $len = strlen($value);
+    $esc = false;
+    $out = '';
+
     for ($ii = 0; $ii < $len; $ii++) {
       $c = $value[$ii];
       if ($esc) {