add progress option (closes #30) (#197)

Thanks @entrity and @t-oster for the work and the patience. This PR is
built on top of your work in #42
and #190. I made some
simplifications by using a std::function and caching the constant
output, instead of a function pointer.

closes #30

Author: pauldreik

Rdutil.cc

@@ -543,16 +543,22 @@ int
 Rdutil::fillwithbytes(enum Fileinfo::readtobuffermode type,
                       enum Fileinfo::readtobuffermode lasttype,
                       const long nsecsleep,
-                      const std::size_t buffersize)
+                      const std::size_t buffersize,
+                      std::function<void(std::size_t)> progress_cb)
 {
   // first sort on inode (to read efficiently from the hard drive)
   sortOnDeviceAndInode();
 
   const auto duration = std::chrono::nanoseconds{ nsecsleep };
 
   std::vector<char> buffer(buffersize, '\0');
+  std::size_t progress_count = 0;
 
   for (auto& elem : m_list) {
+    if (progress_cb) {
+      ++progress_count;
+      progress_cb(progress_count);
+    }
     elem.fillwithbytes(type, lasttype, buffer);
     if (nsecsleep > 0) {
       std::this_thread::sleep_for(duration);

Rdutil.hh

@@ -9,6 +9,7 @@
 #ifndef rdutil_hh
 #define rdutil_hh
 
+#include <functional>
 #include <vector>
 
 #include "Fileinfo.hh" //file container
@@ -90,7 +91,8 @@ public:
   int fillwithbytes(enum Fileinfo::readtobuffermode type,
                     enum Fileinfo::readtobuffermode lasttype,
                     long nsecsleep,
-                    std::size_t buffersize);
+                    std::size_t buffersize,
+                    std::function<void(std::size_t)> progress_cb);
 
   /// make symlinks of duplicates.
   std::size_t makesymlinks(bool dryrun) const;

rdfind.1

@@ -115,6 +115,9 @@ Delete (unlink) files. Default is false.
 .PP
 General options:
 .TP
+.BR \-progress " " \fItrue\fR|\fIfalse\fR
+Show progress during elimination. Defaults to false.
+.TP
 .BR \-sleep " " \fIX\fRms
 Sleeps X milliseconds between reading each file, to reduce
 load. Default is 0 (no sleep). Note that only a few values are

rdfind.cc

@@ -12,6 +12,7 @@ static_assert(__cplusplus >= 201703L,
 // std
 #include <iostream>
 #include <limits>
+#include <sstream>
 #include <string>
 #include <vector>
 
@@ -78,6 +79,7 @@ usage()
     << " -makeresultsfile  (true)| false  makes a results file\n"
     << " -outputname  name  sets the results file name to \"name\" "
        "(default results.txt)\n"
+    << " -progress          true |(false) output progress information"
     << " -deleteduplicates  true |(false) delete duplicate files\n"
     << " -sleep             Xms          sleep for X milliseconds between "
        "file reads.\n"
@@ -115,6 +117,7 @@ struct Options
   bool usesha512 = false;    // use sha512 checksum to check for similarity
   bool usexxh128 = false;    // use xxh128 checksum to check for similarity
   bool deterministic = true; // be independent of filesystem order
+  bool showprogress = false; // show progress while reading file contents
   std::size_t buffersize = 1 << 20; // chunksize to use when reading files
   long nsecsleep = 0; // number of nanoseconds to sleep between each file read.
   std::string resultsfile = "results.txt"; // results file name.
@@ -237,6 +240,8 @@ parseOptions(Parser& parser)
                   << nextarg << "\" is not among them.\n";
         std::exit(EXIT_FAILURE);
       }
+    } else if (parser.try_parse_bool("-progress")) {
+      o.showprogress = parser.get_parsed_bool();
     } else if (parser.current_arg_is("-help") || parser.current_arg_is("-h") ||
                parser.current_arg_is("--help")) {
       usage();
@@ -409,12 +414,28 @@ main(int narg, const char* argv[])
                        "xxh128 checksum");
   }
 
+  std::function<void(std::size_t)> progress_callback;
+  if (o.showprogress) {
+    progress_callback = []() {
+      // format the total count only once, not each iteration.
+      std::ostringstream oss;
+      oss << "/" << filelist.size() << ")"
+          << "\033[u"; // Restore the cursor to the saved position;
+      return [suffix = oss.str()](std::size_t completed) {
+        std::cout
+          << "\033[s\033[K" // Save the cursor position & clear following text
+          << "(" << completed << suffix << std::flush;
+      };
+    }();
+  }
+
   for (auto it = modes.begin() + 1; it != modes.end(); ++it) {
     std::cout << dryruntext << "Now eliminating candidates based on "
               << it->second << ": " << std::flush;
 
     // read bytes (destroys the sorting, for disk reading efficiency)
-    gswd.fillwithbytes(it[0].first, it[-1].first, o.nsecsleep, o.buffersize);
+    gswd.fillwithbytes(
+      it[0].first, it[-1].first, o.nsecsleep, o.buffersize, progress_callback);
 
     // remove non-duplicates
     std::cout << "removed " << gswd.removeUniqSizeAndBuffer()