Ruby/Python re-introduce normalCharacterSequence

aibaars · aibaars · commit 4fd8b1a9d54c · 2022-02-25T12:57:59.000+01:00
diff --git a/python/ql/lib/semmle/python/RegexTreeView.qll b/python/ql/lib/semmle/python/RegexTreeView.qll
@@ -40,7 +40,7 @@ newtype TRegExpParent =
   TRegExpSpecialChar(Regex re, int start, int end) { re.specialCharacter(start, end, _) } or
   /** A normal character */
   TRegExpNormalChar(Regex re, int start, int end) {
-    re.simpleCharacterSequence(start, end)
+    re.normalCharacterSequence(start, end)
     or
     re.escapedCharacter(start, end) and
     not re.specialCharacter(start, end, _)
diff --git a/python/ql/lib/semmle/python/regex.qll b/python/ql/lib/semmle/python/regex.qll
@@ -427,7 +427,7 @@ abstract class RegexString extends Expr {
   }
 
   predicate normalCharacter(int start, int end) {
-    this.character(start, end) and
+    this.simpleCharacter(start, end) and
     not this.specialCharacter(start, end, _)
   }
 
@@ -447,16 +447,16 @@ abstract class RegexString extends Expr {
   }
 
   /**
-   * A sequence of 'simple' characters.
+   * A sequence of 'normal' characters.
    */
-  predicate simpleCharacterSequence(int start, int end) {
-    // a simple character inside a character set is interpreted on its own
-    this.simpleCharacter(start, end) and
+  predicate normalCharacterSequence(int start, int end) {
+    // a normal character inside a character set is interpreted on its own
+    this.normalCharacter(start, end) and
     this.inCharSet(start)
     or
-    // a maximal run of simple characters is considered as one constant
+    // a maximal run of normal characters is considered as one constant
     exists(int s, int e |
-      e = max(int i | simpleCharacterRun(s, i)) and
+      e = max(int i | normalCharacterRun(s, i)) and
       not this.inCharSet(s)
     |
       // 'abc' can be considered one constant, but
@@ -474,17 +474,17 @@ abstract class RegexString extends Expr {
     )
   }
 
-  private predicate simpleCharacterRun(int start, int end) {
+  private predicate normalCharacterRun(int start, int end) {
     (
-      simpleCharacterRun(start, end - 1)
+      normalCharacterRun(start, end - 1)
       or
       start = end - 1 and not normalCharacter(start - 1, start)
     ) and
-    this.simpleCharacter(end - 1, end)
+    this.normalCharacter(end - 1, end)
   }
 
   private predicate characterItem(int start, int end) {
-    this.simpleCharacterSequence(start, end) or
+    this.normalCharacterSequence(start, end) or
     this.escapedCharacter(start, end) or
     this.specialCharacter(start, end, _)
   }
diff --git a/ruby/ql/lib/codeql/ruby/security/performance/ParseRegExp.qll b/ruby/ql/lib/codeql/ruby/security/performance/ParseRegExp.qll
@@ -402,16 +402,16 @@ class RegExp extends AST::RegExpLiteral {
   }
 
   /**
-   * A sequence of 'simple' characters.
+   * A sequence of 'normal' characters.
    */
-  predicate simpleCharacterSequence(int start, int end) {
-    // a simple character inside a character set is interpreted on its own
-    this.simpleCharacter(start, end) and
+  predicate normalCharacterSequence(int start, int end) {
+    // a normal character inside a character set is interpreted on its own
+    this.normalCharacter(start, end) and
     this.inCharSet(start)
     or
-    // a maximal run of simple characters is considered as one constant
+    // a maximal run of normal characters is considered as one constant
     exists(int s, int e |
-      e = max(int i | simpleCharacterRun(s, i)) and
+      e = max(int i | normalCharacterRun(s, i)) and
       not this.inCharSet(s)
     |
       // 'abc' can be considered one constant, but
@@ -429,17 +429,17 @@ class RegExp extends AST::RegExpLiteral {
     )
   }
 
-  private predicate simpleCharacterRun(int start, int end) {
+  private predicate normalCharacterRun(int start, int end) {
     (
-      simpleCharacterRun(start, end - 1)
+      normalCharacterRun(start, end - 1)
       or
       start = end - 1 and not normalCharacter(start - 1, start)
     ) and
-    this.simpleCharacter(end - 1, end)
+    this.normalCharacter(end - 1, end)
   }
 
   private predicate characterItem(int start, int end) {
-    this.simpleCharacterSequence(start, end) or
+    this.normalCharacterSequence(start, end) or
     this.escapedCharacter(start, end) or
     this.specialCharacter(start, end, _)
   }
diff --git a/ruby/ql/lib/codeql/ruby/security/performance/RegExpTreeView.qll b/ruby/ql/lib/codeql/ruby/security/performance/RegExpTreeView.qll
@@ -229,7 +229,7 @@ newtype TRegExpParent =
   TRegExpGroup(RegExp re, int start, int end) { re.group(start, end) } or
   TRegExpSpecialChar(RegExp re, int start, int end) { re.specialCharacter(start, end, _) } or
   TRegExpNormalChar(RegExp re, int start, int end) {
-    re.simpleCharacterSequence(start, end)
+    re.normalCharacterSequence(start, end)
     or
     re.escapedCharacter(start, end) and
     not re.specialCharacter(start, end, _)

Original file line number	Diff line number	Diff line change
`@@ -427,7 +427,7 @@ abstract class RegexString extends Expr {`
`427`	`427`	`}`
`428`	`428`
`429`	`429`	`predicate normalCharacter(int start, int end) {`
`430`		`- this.character(start, end) and`
	`430`	`+ this.simpleCharacter(start, end) and`
`431`	`431`	`not this.specialCharacter(start, end, _)`
`432`	`432`	`}`
`433`	`433`
`@@ -447,16 +447,16 @@ abstract class RegexString extends Expr {`
`447`	`447`	`}`
`448`	`448`
`449`	`449`	`/**`
`450`		`- * A sequence of 'simple' characters.`
	`450`	`+ * A sequence of 'normal' characters.`
`451`	`451`	`*/`
`452`		`- predicate simpleCharacterSequence(int start, int end) {`
`453`		`- // a simple character inside a character set is interpreted on its own`
`454`		`- this.simpleCharacter(start, end) and`
	`452`	`+ predicate normalCharacterSequence(int start, int end) {`
	`453`	`+ // a normal character inside a character set is interpreted on its own`
	`454`	`+ this.normalCharacter(start, end) and`
`455`	`455`	`this.inCharSet(start)`
`456`	`456`	`or`
`457`		`- // a maximal run of simple characters is considered as one constant`
	`457`	`+ // a maximal run of normal characters is considered as one constant`
`458`	`458`	`exists(int s, int e \|`
`459`		`- e = max(int i \| simpleCharacterRun(s, i)) and`
	`459`	`+ e = max(int i \| normalCharacterRun(s, i)) and`
`460`	`460`	`not this.inCharSet(s)`
`461`	`461`	`\|`
`462`	`462`	`// 'abc' can be considered one constant, but`
`@@ -474,17 +474,17 @@ abstract class RegexString extends Expr {`
`474`	`474`	`)`
`475`	`475`	`}`
`476`	`476`
`477`		`- private predicate simpleCharacterRun(int start, int end) {`
	`477`	`+ private predicate normalCharacterRun(int start, int end) {`
`478`	`478`	`(`
`479`		`- simpleCharacterRun(start, end - 1)`
	`479`	`+ normalCharacterRun(start, end - 1)`
`480`	`480`	`or`
`481`	`481`	`start = end - 1 and not normalCharacter(start - 1, start)`
`482`	`482`	`) and`
`483`		`- this.simpleCharacter(end - 1, end)`
	`483`	`+ this.normalCharacter(end - 1, end)`
`484`	`484`	`}`
`485`	`485`
`486`	`486`	`private predicate characterItem(int start, int end) {`
`487`		`- this.simpleCharacterSequence(start, end) or`
	`487`	`+ this.normalCharacterSequence(start, end) or`
`488`	`488`	`this.escapedCharacter(start, end) or`
`489`	`489`	`this.specialCharacter(start, end, _)`
`490`	`490`	`}`