github/ghidra
1
1
Fork 0
mirror of https://github.com/NationalSecurityAgency/ghidra.git synced 2026-01-10 06:27:59 -05:00
Code Issues Packages Projects Releases Wiki Activity

C/CPP: Eliminate static state from generated parsers

Browse Source 
This fixes issue #1421 by tracking the parenthesis depth per CPP
tokenizer instead of globally across all instances. The former behavior
allowed one failed parser run to break all subsequent parser runs until
Ghidra was restarted. There is no reason this variable needs to be
static, anyway, since it's a per-file piece of state.

Also make the same fix to the C tokenizer and to the CPP verbosity
level, the only other static variables in either parser. Neither of
those pieces of state should persist across parser runs either.
This commit is contained in:
Tom Hebb
2021-08-19 17:51:58 -07:00
committed by Ryan Kurtz
parent 00157dfd59
commit 410155810f
2 changed files with 4 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
Ghidra/Features/Base/src/main/javacc/ghidra/app/util/cparser/C/C.jj
View File

@@ -718,7 +718,7 @@ PARSER_END(CParser)
TOKEN_MGR_DECLS :
{
static int parenNesting = 0;
int parenNesting = 0;
}
SKIP :

5
Ghidra/Features/Base/src/main/javacc/ghidra/app/util/cparser/CPP/CPP.jj
View File

@@ -437,7 +437,7 @@ public class PreProcessor {
private HashMap<String, Integer> alreadyDone;
// Toggle printing
private static int verboseLevel = 0;
private int verboseLevel = 0;
public int verboseLevel() {
int vl = verboseLevel;
@@ -1125,6 +1125,7 @@ public class PreProcessor {
this.pathList = parent.pathList;
this.shift = parent.shift;
this.alreadyDone = parent.alreadyDone;
this.verboseLevel = parent.verboseLevel;
}
public PreProcessor(String[] args) throws ParseException {
@@ -2174,7 +2175,7 @@ PPToken ValueExpression() :
// LEXICAL SCANNER SECTION
TOKEN_MGR_DECLS :
{
static int parenNesting = 0;
int parenNesting = 0;
}
<DEFAULT>