hohn/codeql
1
0
Fork 0
mirror of https://github.com/github/codeql.git synced 2026-01-07 19:50:22 +01:00
Code Issues Packages Projects Releases Wiki Activity

Merge rc/1.19 into next.

Browse Source
This commit is contained in:
Aditya Sharad
2018-12-04 12:39:41 +00:00
parent 47e15a6f75 a23f7a6b0e
commit 3caf4e52a7
149 changed files with 13933 additions and 9317 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
javascript/extractor/src/com/semmle/js/extractor/FileExtractor.java
View File

@@ -5,6 +5,7 @@ import java.io.File;
import java.io.FileInputStream;
import java.io.FileReader;
import java.io.IOException;
import java.nio.charset.Charset;
import java.util.LinkedHashSet;
import java.util.Set;
import java.util.regex.Pattern;
@@ -38,6 +39,11 @@ public class FileExtractor {
*/
public static final Pattern JSON_OBJECT_START = Pattern.compile("^(?s)\\s*\\{\\s*\"([^\"]|\\\\.)*\"\\s*:.*");
/**
* The charset for decoding UTF-8 strings.
*/
private static final Charset UTF8_CHARSET = Charset.forName("UTF-8");
/**
* Information about supported file types.
*/
@@ -169,6 +175,11 @@ public class FileExtractor {
if (isXml(bytes, length))
return true;
// Avoid files with an unrecognized shebang header.
if (hasUnrecognizedShebang(bytes, length)) {
return true;
}
return false;
} catch (IOException e) {
Exceptions.ignore(e, "Let extractor handle this one.");
@@ -249,6 +260,38 @@ public class FileExtractor {
return false;
}
/**
* Returns true if the byte sequence starts with a shebang line that is not
* recognized as a JavaScript interpreter.
*/
private boolean hasUnrecognizedShebang(byte[] bytes, int length) {
// Shebangs preceded by a BOM aren't recognized in UNIX, but the BOM might only
// be present in the source file, to be stripped out in the build process.
int startIndex = skipBOM(bytes, length);
if (startIndex + 2 >= length) return false;
if (bytes[startIndex] != '#' || bytes[startIndex + 1] != '!') {
return false;
}
int endOfLine = -1;
for (int i = startIndex; i < length; ++i) {
if (bytes[i] == '\r' || bytes[i] == '\n') {
endOfLine = i;
break;
}
}
if (endOfLine == -1) {
// The shebang is either very long or there are no other lines in the file.
// Treat this as unrecognized.
return true;
}
// Extract the shebang text
int startOfText = startIndex + "#!".length();
int lengthOfText = endOfLine - startOfText;
String text = new String(bytes, startOfText, lengthOfText, UTF8_CHARSET);
// Check if the shebang is a recognized JavaScript intepreter.
return !NODE_INVOCATION.matcher(text).find();
}
@Override
public IExtractor mkExtractor(ExtractorConfig config, ExtractorState state) {
return new TypeScriptExtractor(config, state.getTypeScriptParser());