initial commit

README.md

@@ -0,0 +1,10 @@
+nullblocks
+---
+Scan a file for contiguous blocks of 'null' bytes.
+
+Usage:
+```bash
+nullblocks <file>
+```
+The default minimum block size to be reported is 8192 bytes. Maybe this will 
+be adjustable as a command line argument in the future...

build

@@ -0,0 +1,2 @@
+#!/bin/bash
+g++ -o nullblocks src/nullblocks.cc

src/nullblocks.cc

@@ -0,0 +1,66 @@
+#include <iostream>
+#include <iomanip>
+#include <fstream>
+#include <string>
+using namespace std;
+
+const long IO_BUFFSIZE=(long)1024*1024;  // read 1 KiB at a time
+const long MIN_NULL_BLOCK_SIZE=(long)8*1024; // minimum number of consecutive NULL bytes
+
+int main(int argc,char *argv[]){
+  int ifile, iblock;
+  char* buffer = new char[IO_BUFFSIZE];
+  long filesize, iosize, nullblock_len, nullblock_beg;
+  
+  cout << "Checking files for NULL blocks (threshold: " 
+       << MIN_NULL_BLOCK_SIZE << " bytes)" << endl;
+  string filename;
+
+  for(ifile=1;ifile<argc;ifile++){
+    filename = argv[ifile];
+    ifstream file(filename,ios::in|ios::binary|ios::ate);
+    if(!file)
+      cerr << "Cannot open file: " << filename << endl;
+    filesize = (long)file.tellg();
+    file.seekg(0,ios::beg);
+    cout << filename << " (" << filesize << " bytes)" <<  endl;
+
+    iblock=0;
+    nullblock_beg=0;
+    nullblock_len=0;
+    bool trunc = false;
+    while(!trunc){
+      trunc = (filesize-iblock*IO_BUFFSIZE)<IO_BUFFSIZE;
+      if(trunc)
+	iosize = filesize-iblock*IO_BUFFSIZE;
+      else
+	iosize = IO_BUFFSIZE;
+      file.read(buffer,iosize);
+      if(file.fail()){
+        cerr << "I/O Error: " << filename << " @ " << file.tellg() << endl;
+	break;
+      }
+      for(int ii=0;ii<iosize;ii++){
+        if((int)buffer[ii]==0){
+          if(nullblock_beg==0){
+            nullblock_beg = iblock*IO_BUFFSIZE+ii;
+	  }
+	  nullblock_len++;
+	}
+	else{
+	  if(nullblock_len>=MIN_NULL_BLOCK_SIZE){
+	    cout << "NULL block: position = " << setw(16) << nullblock_beg
+		 << ", size = " << setw(16) << nullblock_len << endl;
+	  }
+	  nullblock_len = 0;
+	  nullblock_beg = 0;
+	}
+      }
+      iblock++;
+    }
+    file.close();
+  }
+  delete[] buffer;
+	
+  return 0;
+}