/**
* Tagging support.
* Tags are a system of providing identifiers for regions of the input image.
* Currently tags are just used for fragment type.
*/
void feature_recorder::write_tag(const pos0_t &pos0,size_t len,const string &tagName)
{
if(flags & FLAG_DISABLED) return; // disabled
stringstream ss;
string desc = pos0.alphaPart();
// This allows you to set a breakpoint at a specific position
// it could be a configurable variable, I guess...
#ifdef DEBUG_OFFSET
if(len==DEBUG_OFFSET){
std::cerr << "write_tag debug point pos0=" << pos0 << " len=" << len <<" name=" << tagName << "\n";
}
#endif
/* offset is either the sbuf offset or the path offset */
uint64_t offset = pos0.offset>0 ? pos0.offset : stoi64(pos0.path);
/** Create what will got to the feature file */
ss << offset << ":" << len << "\t";
if(desc.size()>0) ss << desc << '/';
ss << tagName;
this->write(ss.str());
}
void Database_LevelDB::listAllLoadableBlocks(std::list<v3s16> &dst)
{
leveldb::Iterator* it = m_database->NewIterator(leveldb::ReadOptions());
for (it->SeekToFirst(); it->Valid(); it->Next()) {
dst.push_back(getIntegerAsBlock(stoi64(it->key().ToString())));
}
ENSURE_STATUS_OK(it->status()); // Check for any errors found during the scan
delete it;
}
void Database_Redis::listAllLoadableBlocks(std::list<v3s16> &dst)
{
redisReply *reply;
reply = (redisReply*) redisCommand(ctx, "HKEYS %s", hash.c_str());
if(!reply)
throw FileNotGoodException(std::string("redis command 'HKEYS %s' failed: ") + ctx->errstr);
if(reply->type != REDIS_REPLY_ARRAY)
throw FileNotGoodException("Failed to get keys from database");
for(size_t i = 0; i < reply->elements; i++)
{
assert(reply->element[i]->type == REDIS_REPLY_STRING);
dst.push_back(getIntegerAsBlock(stoi64(reply->element[i]->str)));
}
freeReplyObject(reply);
}
void Database_Redis::listAllLoadableBlocks(std::vector<v3s16> &dst)
{
redisReply *reply = static_cast<redisReply *>(redisCommand(ctx, "HKEYS %s", hash.c_str()));
if (!reply) {
throw FileNotGoodException(std::string(
"Redis command 'HKEYS %s' failed: ") + ctx->errstr);
}
switch (reply->type) {
case REDIS_REPLY_ARRAY:
for (size_t i = 0; i < reply->elements; i++) {
assert(reply->element[i]->type == REDIS_REPLY_STRING);
dst.push_back(getIntegerAsBlock(stoi64(reply->element[i]->str)));
}
case REDIS_REPLY_ERROR:
throw FileNotGoodException(std::string(
"Failed to get keys from database: ") + reply->str);
}
freeReplyObject(reply);
}
DBRedis::DBRedis(const std::string &mapdir) :
m_blocksReadCount(0),
m_blocksQueriedCount(0)
{
Settings world_mt(mapdir + "/world.mt");
std::string address;
if (!world_mt.check("redis_address", address) || !world_mt.check("redis_hash", hash)) {
throw std::runtime_error("Set redis_address and redis_hash in world.mt to use the redis backend");
}
int port = stoi64(world_mt.get("redis_port", "6379"));
ctx = redisConnect(address.c_str(), port);
if(!ctx)
throw std::runtime_error("Cannot allocate redis context");
else if(ctx->err) {
std::string err = std::string("Connection error: ") + ctx->errstr;
redisFree(ctx);
throw std::runtime_error(err);
}
}
static void process_open_path(const image_process &p,std::string path,scanner_params::PrintOptions &po,
const size_t process_path_bufsize)
{
/* Check for "/r" in path which means print raw */
if(path.size()>2 && path.substr(path.size()-2,2)=="/r"){
path = path.substr(0,path.size()-2);
}
std::string prefix = get_and_remove_token(path);
int64_t offset = stoi64(prefix);
/* Get the offset into the buffer process */
u_char *buf = (u_char *)calloc(process_path_bufsize,1);
if(!buf){
std::cerr << "Cannot allocate " << process_path_bufsize << " buffer\n";
return;
}
int count = p.pread(buf,process_path_bufsize,offset);
if(count<0){
std::cerr << p.image_fname() << ": " << strerror(errno) << " (Read Error)\n";
return;
}
/* make up a bogus feature recorder set and with a disabled feature recorder.
* Then we call the path printer, which throws an exception after the printing
* to prevent further printing.
*
* The printer is called when a PRINT token is found in the
* forensic path, so that has to be added.
*/
feature_recorder_set fs(feature_recorder_set::SET_DISABLED,feature_recorder_set::null_hasher,
feature_recorder_set::NO_INPUT,feature_recorder_set::NO_OUTDIR);
pos0_t pos0(path+"-PRINT"); // insert the PRINT token
sbuf_t sbuf(pos0,buf,count,count,true); // sbuf system will free
scanner_params sp(scanner_params::PHASE_SCAN,sbuf,fs,po);
try {
process_path_printer(sp);
}
catch (path_printer_finished &e) {
}
}
static void bulk_process_feature_file(const std::string &fn)
{
if(ends_with(fn,".txt")==false) return; // don't process binary files
if(ends_with(fn,"_histogram.txt")==true) return; // ignore histogram files
const string features(fn.substr(fn.rfind('/')+1,fn.size()-fn.rfind('/')-5));
const string name(features+": ");
const string SPACE(" ");
const string UNKNOWN("UNKNOWN");
bool tagfile = ends_with(fn,"_tags.txt");
ifstream f(fn.c_str());
if(!f.is_open()){
cerr << "Cannot open tag input file: " << fn << "\n";
return;
}
try {
string line;
while(getline(f,line)){
if(line.size()==0 || line[0]=='#' || line.substr(0,4)=="\357\273\277#") continue;
vector<string> fields = split(line,'\t'); // fields of the feature file
if(fields.size()<2) continue; // improper formatting
std::string &taglocation = fields[0];
std::string &tagtype = fields[1];
uint64_t offset = stoi64(taglocation);
uint64_t sector = offset / opt_bulk_block_size;
/* If the array hasn't been expanded to the point of this element, expand it with blanks */
while(sector > sector_typetags.size()){
sector_typetags.push_back(sector_typetag()); // expand to fill gap
}
if(tagfile){ // first pass
/* Process a tag */
vector<string> vals = split(taglocation,':');
if(vals.size()!=2){
std::cerr << "Invalid tag file line: " << line << " (size=" << vals.size() << ")\n";
exit(1);
}
uint32_t len = stoi(vals[1]);
// If no data for this sector, simply append this type
// and then continue
if(sector_typetags.size()==sector){
sector_typetags.push_back(sector_typetag(len,tagtype,string("")));
continue;
}
// We have new data for the same element. Which is better?
if(sector_typetags[sector].specificity() < sector_typetag::specificity(tagtype)){
// New is more specific than the old.
// Preserve the old one
sector_typetags[sector].scomment = sector_typetags[sector].stype + string("; ") + sector_typetags[sector].scomment;
sector_typetags[sector].stype = tagtype; // specify new tag type
} else {
// New is less specific than the old, so just make new type a comment.
sector_typetags[sector].scomment = tagtype + string("; ") + sector_typetags[sector].scomment;
}
continue;
}
/* Process a feature, which will add specificity to the tag */
if(sector_typetags.size()==sector){
/* Hm... No tag (and all tags got processed first), so this is unknown */
sector_typetags.push_back(sector_typetag(opt_bulk_block_size,UNKNOWN,SPACE));
}
/* append what we've learned regarding this feature */
// If we got an MD5 as field1 and there is a second field, go with that
sector_typetag &s = sector_typetags[sector];
int field = 1;
if(fields.size()>2 && fields[1].size()==32 && fields[2].size()>0 && fields[2][0]=='<'){
field = 2; // go with the second field
}
s.scomment += " " + name + fields[field];
// append any XML if it is present
if(field==1 && fields.size()>2 && fields[2].size()>0 && fields[2][0]=='<'){
s.scomment += " " + name + " " + fields[2];
}
// Scan through the feature indicator table and if we find a match note the type
for(int i=0;feature_indicators[i].feature_file_name;i++){
if(features!=feature_indicators[i].feature_file_name) continue;
if(feature_indicators[i].feature_content==0
|| fields[1].find(feature_indicators[i].feature_content)!=string::npos
|| fields[2].find(feature_indicators[i].feature_content)!=string::npos){
s.stype = pos0_t(fields[0]).alphaPart();
if(s.stype.size()>1){
char lastchar = s.stype.at(s.stype.size()-1);
if(lastchar!='-' && lastchar!='/') s.stype += string("-");
}
s.stype += feature_indicators[i].dfrws_type;
}
}
}
}
catch (const std::exception &e) {
cerr << "ERROR: " << e.what() << " processing tagfile " << fn << "\n";
}
}
int64 VString::toi64() const { return stoi64((char*)data, sz); }
int main(int argc,char **argv)
{
#ifdef HAVE_MCHECK
mtrace();
#endif
/* setup */
feature_recorder::set_main_threadid();
const char *progname = argv[0];
word_and_context_list alert_list; /* shold be flagged */
word_and_context_list stop_list; /* should be ignored */
scanner_info::scanner_config s_config; // the bulk extractor phase 1 config created from the command line
BulkExtractor_Phase1::Config cfg;
cfg.num_threads = threadpool::numCPU();
/* Options */
const char *opt_path = 0;
int opt_recurse = 0;
int opt_zap = 0;
int opt_h = 0;
int opt_H = 0;
std::string opt_sampling_params;
std::string opt_outdir;
bool opt_write_feature_files = true;
bool opt_write_sqlite3 = false;
bool opt_enable_histograms=true;
/* Startup */
setvbuf(stdout,0,_IONBF,0); // don't buffer stdout
std::string command_line = dfxml_writer::make_command_line(argc,argv);
std::vector<std::string> scanner_dirs; // where to look for scanners
/* Add the default plugin_path */
add_if_present(scanner_dirs,"/usr/local/lib/bulk_extractor");
add_if_present(scanner_dirs,"/usr/lib/bulk_extractor");
add_if_present(scanner_dirs,".");
if (getenv("BE_PATH")) {
std::vector<std::string> dirs = split(getenv("BE_PATH"),':');
for(std::vector<std::string>::const_iterator it = dirs.begin(); it!=dirs.end(); it++){
add_if_present(scanner_dirs,*it);
}
}
#ifdef WIN32
setmode(1,O_BINARY); // make stdout binary
threadpool::win32_init();
#endif
/* look for usage first */
if(argc==1) opt_h=1;
/* Process options */
int ch;
while ((ch = getopt(argc, argv, "A:B:b:C:d:E:e:F:f:G:g:Hhij:M:m:o:P:p:q:Rr:S:s:VW:w:x:Y:z:Z")) != -1) {
switch (ch) {
case 'A': feature_recorder::offset_add = stoi64(optarg);break;
case 'b': feature_recorder::banner_file = optarg; break;
case 'C': feature_recorder::context_window_default = atoi(optarg);break;
case 'd':
{
if(strcmp(optarg,"h")==0) debug_help();
int d = atoi(optarg);
switch(d){
case DEBUG_ALLOCATE_512MiB:
if(calloc(1024*1024*512,1)){
std::cerr << "-d1002 -- Allocating 512MB of RAM; may be repeated\n";
} else {
std::cerr << "-d1002 -- CANNOT ALLOCATE MORE RAM\n";
}
break;
default:
cfg.debug = d;
break;
}
be13::plugin::set_scanner_debug(cfg.debug);
}
break;
case 'E':
be13::plugin::scanners_disable_all();
be13::plugin::scanners_enable(optarg);
break;
case 'e':
be13::plugin::scanners_enable(optarg);
break;
case 'F': FindOpts::get().Files.push_back(optarg); break;
case 'f': FindOpts::get().Patterns.push_back(optarg); break;
case 'G': cfg.opt_pagesize = scaled_stoi64(optarg); break;
case 'g': cfg.opt_marginsize = scaled_stoi64(optarg); break;
case 'i':
std::cout << "info mode:\n";
cfg.opt_info = true;
break;
case 'j': cfg.num_threads = atoi(optarg); break;
case 'M': scanner_def::max_depth = atoi(optarg); break;
case 'm': cfg.max_bad_alloc_errors = atoi(optarg); break;
case 'o': opt_outdir = optarg;break;
case 'P': scanner_dirs.push_back(optarg);break;
case 'p': opt_path = optarg; break;
case 'q':
if(atoi(optarg)==-1) cfg.opt_quiet = 1;// -q -1 turns off notifications
else cfg.opt_notify_rate = atoi(optarg);
break;
case 'r':
if(alert_list.readfile(optarg)){
err(1,"Cannot read alert list %s",optarg);
}
break;
case 'R': opt_recurse = 1; break;
case 'S':
{
std::vector<std::string> params = split(optarg,'=');
if(params.size()!=2){
std::cerr << "Invalid paramter: " << optarg << "\n";
exit(1);
}
s_config.namevals[params[0]] = params[1];
continue;
}
case 's':
#if defined(HAVE_SRANDOM) && !defined(HAVE_SRANDOMDEV)
srandom(time(0));
#endif
#if defined(HAVE_SRANDOMDEV)
srandomdev(); // if we are sampling initialize
#endif
opt_sampling_params = optarg;
break;
case 'V': std::cout << "bulk_extractor " << PACKAGE_VERSION << "\n"; exit (1);
case 'W':
fprintf(stderr,"-W has been deprecated. Specify with -S word_min=NN and -S word_max=NN\n");
exit(1);
break;
case 'w': if(stop_list.readfile(optarg)){
err(1,"Cannot read stop list %s",optarg);
}
break;
case 'x':
be13::plugin::scanners_disable(optarg);
break;
case 'Y': {
std::string optargs = optarg;
size_t dash = optargs.find('-');
if(dash==std::string::npos){
cfg.opt_offset_start = stoi64(optargs);
} else {
cfg.opt_offset_start = scaled_stoi64(optargs.substr(0,dash));
cfg.opt_offset_end = scaled_stoi64(optargs.substr(dash+1));
}
break;
}
case 'z': cfg.opt_page_start = stoi64(optarg);break;
case 'Z': opt_zap=true;break;
case 'H': opt_H++;continue;
case 'h': opt_h++;continue;
}
}
cfg.validate();
argc -= optind;
argv += optind;
if(cfg.debug & DEBUG_PRINT_STEPS) std::cerr << "DEBUG: DEBUG_PRINT_STEPS\n";
if(cfg.debug & DEBUG_PEDANTIC) validateOrEscapeUTF8_validate = true;
/* Create a configuration that will be used to initialize the scanners */
scanner_info si;
s_config.debug = cfg.debug;
si.config = &s_config;
/* Make individual configuration options appear on the command line interface. */
si.get_config("work_start_work_end",&worker::opt_work_start_work_end,
"Record work start and end of each scanner in report.xml file");
si.get_config("enable_histograms",&opt_enable_histograms,
"Disable generation of histograms");
si.get_config("debug_histogram_malloc_fail_frequency",&HistogramMaker::debug_histogram_malloc_fail_frequency,
"Set >0 to make histogram maker fail with memory allocations");
si.get_config("hash_alg",&be_hash_name,"Specifies hash algorithm to be used for all hash calculations");
si.get_config("dup_data_alerts",&be13::plugin::dup_data_alerts,"Notify when duplicate data is not processed");
si.get_config("write_feature_files",&opt_write_feature_files,"Write features to flat files");
si.get_config("write_feature_sqlite3",&opt_write_sqlite3,"Write feature files to report.sqlite3");
/* Make sure that the user selected a valid hash */
{
uint8_t buf[1];
be_hash_func(buf,0);
}
/* Load all the scanners and enable the ones we care about */
be13::plugin::load_scanner_directories(scanner_dirs,s_config);
be13::plugin::load_scanners(scanners_builtin,s_config);
be13::plugin::scanners_process_enable_disable_commands();
/* Print usage if necessary */
if(opt_H){ be13::plugin::info_scanners(true,true,scanners_builtin,'e','x'); exit(0);}
if(opt_h){ usage(progname);be13::plugin::info_scanners(false,true,scanners_builtin,'e','x'); exit(0);}
/* Give an error if a find list was specified
* but no scanner that uses the find list is enabled.
*/
if(!FindOpts::get().empty()) {
/* Look through the enabled scanners and make sure that
* at least one of them is a FIND scanner
*/
if(!be13::plugin::find_scanner_enabled()){
errx(1,"find words are specified with -F but no find scanner is enabled.\n");
}
}
if(opt_path){
if(argc!=1) errx(1,"-p requires a single argument.");
process_path(argv[0],opt_path,cfg.opt_pagesize,cfg.opt_marginsize);
exit(0);
}
if(opt_outdir.size()==0) errx(1,"error: -o outdir must be specified");
/* The zap option wipes the contents of a directory, useful for debugging */
if(opt_zap){
DIR *dirp = opendir(opt_outdir.c_str());
if(dirp){
struct dirent *dp;
while ((dp = readdir(dirp)) != NULL){
std::string name = dp->d_name;
if(name=="." || name=="..") continue;
std::string fname = opt_outdir + std::string("/") + name;
unlink(fname.c_str());
std::cout << "erasing " << fname << "\n";
}
}
if(rmdir(opt_outdir.c_str())){
std::cout << "rmdir " << opt_outdir << "\n";
}
}
/* Start the clock */
aftimer timer;
timer.start();
/* If output directory does not exist, we are not restarting! */
std::string reportfilename = opt_outdir + "/report.xml";
BulkExtractor_Phase1::seen_page_ids_t seen_page_ids; // pages that do not need re-processing
image_process *p = 0; // the image process iterator
/* Get image or directory */
if (*argv == NULL) {
if (opt_recurse) {
fprintf(stderr,"filedir not provided\n");
} else {
fprintf(stderr,"imagefile not provided\n");
}
exit(1);
}
std::string image_fname = *argv;
if(opt_outdir.size()==0){
fprintf(stderr,"output directory not provided\n");
exit(1);
}
if(directory_missing(opt_outdir) || directory_empty(opt_outdir)){
/* First time running */
/* Validate the args */
if ( argc !=1 ) errx(1,"Disk image option not provided. Run with -h for help.");
validate_fn(image_fname);
if (directory_missing(opt_outdir)) be_mkdir(opt_outdir);
} else {
/* Restarting */
std::cout << "Restarting from " << opt_outdir << "\n";
bulk_extractor_restarter r(opt_outdir,reportfilename,image_fname,seen_page_ids);
/* Rename the old report and create a new one */
std::string old_reportfilename = reportfilename + "." + itos(time(0));
if(rename(reportfilename.c_str(),old_reportfilename.c_str())){
std::cerr << "Could not rename " << reportfilename << " to " << old_reportfilename << ": " << strerror(errno) << "\n";
exit(1);
}
}
/* Open the image file (or the device) now */
p = image_process::open(image_fname,opt_recurse,cfg.opt_pagesize,cfg.opt_marginsize);
if(!p) err(1,"Cannot open %s: ",image_fname.c_str());
/***
*** Create the feature recording set.
*** Initialize the scanners.
****/
/* Determine the feature files that will be used */
feature_file_names_t feature_file_names;
be13::plugin::get_scanner_feature_file_names(feature_file_names);
uint32_t flags = 0;
if (stop_list.size()>0) flags |= feature_recorder_set::CREATE_STOP_LIST_RECORDERS;
if (opt_write_sqlite3) flags |= feature_recorder_set::ENABLE_SQLITE3_RECORDERS;
if (!opt_write_feature_files) flags |= feature_recorder_set::DISABLE_FILE_RECORDERS;
{
feature_recorder_set fs(flags,be_hash,image_fname,opt_outdir);
fs.init(feature_file_names);
if(opt_enable_histograms) be13::plugin::add_enabled_scanner_histograms_to_feature_recorder_set(fs);
be13::plugin::scanners_init(fs);
fs.set_stop_list(&stop_list);
fs.set_alert_list(&alert_list);
/* Look for commands that impact per-recorders */
for(scanner_info::config_t::const_iterator it=s_config.namevals.begin();it!=s_config.namevals.end();it++){
/* see if there is a <recorder>: */
std::vector<std::string> params = split(it->first,':');
if(params.size()>=3 && params.at(0)=="fr"){
feature_recorder *fr = fs.get_name(params.at(1));
const std::string &cmd = params.at(2);
if(fr){
if(cmd=="window") fr->set_context_window(stoi64(it->second));
if(cmd=="window_before") fr->set_context_window_before(stoi64(it->second));
if(cmd=="window_after") fr->set_context_window_after(stoi64(it->second));
}
}
/* See if there is a scanner? */
}
/* Store the configuration in the XML file */
dfxml_writer *xreport = new dfxml_writer(reportfilename,false);
dfxml_create(*xreport,command_line,cfg);
xreport->xmlout("provided_filename",image_fname); // save this information
/* provide documentation to the user; the DFXML information comes from elsewhere */
if(!cfg.opt_quiet){
std::cout << "bulk_extractor version: " << PACKAGE_VERSION << "\n";
#ifdef HAVE_GETHOSTNAME
char hostname[1024];
gethostname(hostname,sizeof(hostname));
std::cout << "Hostname: " << hostname << "\n";
#endif
std::cout << "Input file: " << image_fname << "\n";
std::cout << "Output directory: " << opt_outdir << "\n";
std::cout << "Disk Size: " << p->image_size() << "\n";
std::cout << "Threads: " << cfg.num_threads << "\n";
}
/****************************************************************
*** THIS IS IT! PHASE 1!
****************************************************************/
if ( fs.flag_set(feature_recorder_set::ENABLE_SQLITE3_RECORDERS )) {
fs.db_transaction_begin();
}
BulkExtractor_Phase1 phase1(*xreport,timer,cfg);
if(cfg.debug & DEBUG_PRINT_STEPS) std::cerr << "DEBUG: STARTING PHASE 1\n";
if(opt_sampling_params.size()>0) BulkExtractor_Phase1::set_sampling_parameters(cfg,opt_sampling_params);
xreport->add_timestamp("phase1 start");
phase1.run(*p,fs,seen_page_ids);
if(cfg.debug & DEBUG_PRINT_STEPS) std::cerr << "DEBUG: WAITING FOR WORKERS\n";
std::string md5_string;
phase1.wait_for_workers(*p,&md5_string);
delete p; // not strictly needed, but why not?
p = 0;
if ( fs.flag_set(feature_recorder_set::ENABLE_SQLITE3_RECORDERS )) {
fs.db_transaction_commit();
}
xreport->add_timestamp("phase1 end");
if(md5_string.size()>0){
std::cout << "MD5 of Disk Image: " << md5_string << "\n";
}
/*** PHASE 2 --- Shutdown ***/
if(cfg.opt_quiet==0) std::cout << "Phase 2. Shutting down scanners\n";
xreport->add_timestamp("phase2 start");
be13::plugin::phase_shutdown(fs);
xreport->add_timestamp("phase2 end");
/*** PHASE 3 --- Create Histograms ***/
if(cfg.opt_quiet==0) std::cout << "Phase 3. Creating Histograms\n";
xreport->add_timestamp("phase3 start");
if(opt_enable_histograms) fs.dump_histograms(0,histogram_dump_callback,0); // TK - add an xml error notifier!
xreport->add_timestamp("phase3 end");
/*** PHASE 4 --- report and then print final usage information ***/
xreport->push("report");
xreport->xmlout("total_bytes",phase1.total_bytes);
xreport->xmlout("elapsed_seconds",timer.elapsed_seconds());
xreport->xmlout("max_depth_seen",be13::plugin::get_max_depth_seen());
xreport->xmlout("dup_data_encountered",be13::plugin::dup_data_encountered);
xreport->pop(); // report
xreport->flush();
xreport->push("scanner_times");
fs.get_stats(xreport,stat_callback);
xreport->pop();
xreport->add_rusage();
xreport->pop(); // bulk_extractor
xreport->close();
if(cfg.opt_quiet==0){
float mb_per_sec = (phase1.total_bytes / 1000000.0) / timer.elapsed_seconds();
std::cout.precision(4);
printf("Elapsed time: %g sec.\n",timer.elapsed_seconds());
printf("Total MB processed: %d\n",int(phase1.total_bytes / 100000));
printf("Overall performance: %g MBytes/sec (%g MBytes/sec/thread)\n",
mb_per_sec,mb_per_sec/cfg.num_threads);
if (fs.has_name("email")) {
feature_recorder *fr = fs.get_name("email");
if(fr){
std::cout << "Total " << fr->name << " features found: " << fr->count() << "\n";
}
}
}
}
#ifdef HAVE_MCHECK
muntrace();
#endif
exit(0);
}
void process_path_printer(const scanner_params &sp)
{
/* 1. Get next token
* 2. if prefix part is a number, skip forward that much in sbuf and repeat.
* if the prefix is PRINT, print the buffer
* if next part is a string, strip it and run that decoder.
* if next part is a |, print
* 3. If we are print, throw an exception to prevent continued analysis of buffer.
*/
std::string new_path = sp.sbuf.pos0.path;
std::string prefix = get_and_remove_token(new_path);
/* Time to print ?*/
if(prefix.size()==0 || prefix=="PRINT"){
uint64_t print_start = 0;
uint64_t print_len = 4096;
/* Check for options */
scanner_params::PrintOptions::iterator it;
it = sp.print_options.find("Content-Length");
if(it!=sp.print_options.end()){
print_len = stoi64(it->second);
}
it = sp.print_options.find("Range");
if(it!=sp.print_options.end()){
if(it->second[5]=='='){
size_t dash = it->second.find('-');
std::string v1 = it->second.substr(6,dash-6);
std::string v2 = it->second.substr(dash+1);
print_start = stoi64(v1);
print_len = stoi64(v2)-print_start+1;
}
}
if(print_start>sp.sbuf.bufsize){
print_len = 0; // can't print anything
}
if(print_len>0 && print_start+print_len>sp.sbuf.bufsize){
print_len = sp.sbuf.bufsize-print_start;
}
switch(scanner_params::getPrintMode(sp.print_options)){
case scanner_params::MODE_HTTP:
std::cout << "Content-Length: " << print_len << HTTP_EOL;
std::cout << "Content-Range: bytes " << print_start << "-" << print_start+print_len-1 << HTTP_EOL;
std::cout << "X-Range-Available: bytes " << 0 << "-" << sp.sbuf.bufsize-1 << HTTP_EOL;
std::cout << HTTP_EOL;
sp.sbuf.raw_dump(std::cout,print_start,print_len); // send to stdout as binary
break;
case scanner_params::MODE_RAW:
std::cout << print_len << HTTP_EOL;
std::cout.flush();
sp.sbuf.raw_dump(std::cout,print_start,print_len); // send to stdout as binary
break;
case scanner_params::MODE_HEX:
sp.sbuf.hex_dump(std::cout,print_start,print_len);
break;
case scanner_params::MODE_NONE:
break;
}
throw printing_done;
//return; // our job is done
}
/* If we are in an offset block, process recursively with the offset */
if(isdigit(prefix[0])){
uint64_t offset = stoi64(prefix);
if(offset>sp.sbuf.bufsize){
printf("Error: %s only has %u bytes; can't offset to %u\n",
new_path.c_str(),(unsigned int)sp.sbuf.bufsize,(unsigned int)offset);
return;
}
process_path_printer(scanner_params(scanner_params::PHASE_SCAN,
sbuf_t(new_path,sp.sbuf+offset),
sp.fs,sp.print_options));
return;
}
/* Find the scanner and use it */
scanner_t *s = be13::plugin::find_scanner(lowerstr(prefix));
if(s){
(*s)(scanner_params(scanner_params::PHASE_SCAN,
sbuf_t(new_path,sp.sbuf),
sp.fs,sp.print_options),
recursion_control_block(process_path_printer,prefix));
return;
}
std::cerr << "Unknown name in path: " << prefix << "\n";
}