countfa v1.0, changed output of shuffler -f

ChangeLog

@@ -1,3 +1,8 @@
+2019-04-26  Benjamin Jean-Marie Tremblay  <[email protected]>
+
+* Added countfa v1.0
+* shuffler -f now inserts newlines every 80 characters
+
 2019-04-25  Benjamin Jean-Marie Tremblay  <[email protected]>
 
 * Added seqgen v1.0

Makefile

@@ -1,9 +1,10 @@
 SRC := countlets.cpp klets.cpp shuffler.cpp shuffle_euler.cpp \
-  shuffle_linear.cpp shuffle_markov.cpp seqgen.cpp
+  shuffle_linear.cpp shuffle_markov.cpp seqgen.cpp countfa.cpp
 OBJ_COUNTLETS := countlets.o klets.o
 OBJ_SHUFFLER := shuffler.o klets.o shuffle_euler.o shuffle_linear.o \
   shuffle_markov.o
 OBJ_SEQGEN := seqgen.o
+OBJ_COUNTFA := countfa.o
 CC := g++
 
 all: build install
@@ -12,6 +13,10 @@ build:
 	cd src;\
 	$(CC) --std=c++11 -O2 -Wall -c $(SRC)
 
+countfa:
+	cd src;\
+	  $(CC) $(OBJ_COUNTFA) -o ../bin/countfa
+
 countlets:
 	cd src;\
 	$(CC) $(OBJ_COUNTLETS) -o ../bin/countlets
@@ -27,7 +32,7 @@ seqgen:
 makebin:
 	mkdir -p bin
 
-install: makebin countlets shuffler seqgen
+install: makebin countfa countlets shuffler seqgen
 
 clean:
 	cd src;\

README

@@ -14,13 +14,30 @@ Installation
 
 The following binaries are created:
 
+    bin/countfa
     bin/countlets
     bin/seqgen
     bin/shuffler
 
 Run these without any arguments or with the -h flag to see usage.
 
 
+countfa
+-------
+
+Counts the number of characters per sequence in a fasta file. For each sequence,
+the name followed by the character count are returned to stdout.
+
+Example usage:
+
+    echo ">1\nACAAG\n>2\nGCCCGGTTAT" | bin/countfa
+
+        >1
+        5
+        >2
+        10
+
+
 countlets
 ---------
 

src/countfa.cpp

@@ -0,0 +1,119 @@
+/*
+ * Copyright (C) 2019 Benjamin Jean-Marie Tremblay
+ *
+ * This file is part of sequenceshuffler.
+ *
+ * sequenceshuffler is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 3 of the License, or
+ * (at your option) any later version.
+ *
+ * sequenceshuffler is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with sequenceshuffler.  If not, see <https://www.gnu.org/licenses/>.
+ *
+ */
+
+#include <iostream>
+#include <fstream>
+#include <unistd.h>
+#include <string>
+using namespace std;
+
+void usage() {
+  printf(
+    "countfa v1.0  Copyright (C) 2019  Benjamin Jean-Marie Tremblay                  \n"
+    "                                                                                \n"
+    " Usage:  coutfa -i [filename]                                                   \n"
+    "         cat [filename] | coutfa                                                \n"
+    "                                                                                \n"
+    " -i <str>    Input filename. File must be fasta-formatted. Alternatively, takes \n"
+    "             input from a pipe.                                                 \n"
+    " -h          Print usage and exit.                                              \n"
+  );
+}
+
+void do_countfa(istream &input) {
+
+  string name, line;
+  int counter{0};
+
+  while (getline(input, line).good()) {
+
+    if (line.empty() || line[0] == '>') {
+
+      if (!name.empty()) {
+        cout << name << endl;
+        name.clear();
+      }
+      if (!line.empty()) {
+        name = line;
+      }
+      if (counter > 0) {
+        cout << counter << endl;
+      }
+      counter = 0;
+
+    } else if (!name.empty()) {
+
+      if (line.find(' ') != string::npos) {
+        line.erase(remove(line.begin(), line.end(), ' '), line.end());
+      }
+
+      if (line.length() == 0) {
+        name.clear();
+        counter = 0;
+      } else {
+        counter += line.length();
+      }
+
+    }
+
+  }
+
+  if (!name.empty()) {
+    cout << name << endl;
+    cout << counter << endl;
+  }
+
+}
+
+int main(int argc, char **argv) {
+
+  int opt;
+  bool has_file{false};
+  ifstream seqfile;
+
+  while ((opt = getopt(argc, argv, "i:h")) != -1) {
+    switch (opt) {
+      case 'i': if (optarg) {
+                  seqfile.open(optarg);
+                  if (seqfile.bad()) {
+                    cerr << "Error: file not found" << endl;
+                    exit(EXIT_FAILURE);
+                  }
+                  has_file = true;
+                }
+                break;
+      case 'h': usage();
+                return 0;
+    }
+  }
+
+  if (!has_file) {
+    if (isatty(STDIN_FILENO)) {
+      usage();
+      exit(EXIT_FAILURE);
+    }
+    do_countfa(cin);
+  } else {
+    do_countfa(seqfile);
+  }
+
+  return 0;
+
+}

src/shuffler.cpp

@@ -41,7 +41,8 @@ void usage() {
     "                                                                                \n"
     " -i <str>   Input filename. All white space will be removed. Alternatively, can \n"
     "            take string input from a pipe.                                      \n"
-    " -o <str>   Output filename. Alternatively, prints to stdout.                   \n"
+    " -o <str>   Output filename. Alternatively, prints to stdout. For fasta input, a\n"
+    "            newline is inserted every 80 characters.                            \n"
     " -k <int>   K-let size. Defaults to 1.                                          \n"
     " -s <int>   RNG seed number. Defaults to time in seconds.                       \n"
     " -m         Use the markov shuffling method. Defaults to euler.                 \n"
@@ -291,6 +292,7 @@ int main(int argc, char **argv) {
     }
 
     for (int i = 0; i < fa_names.size(); ++i) {
+
       vector<char> letters2(fa_seqs[i].begin(), fa_seqs[i].end());
       if (k >= letters2.size()) {
         cerr << "Error: sequence length must be greater than k" << endl;
@@ -300,9 +302,23 @@ int main(int argc, char **argv) {
       outletters = do_shuffle(letters2, k, gen, false, method_i);
 
       if (has_out) {
-        outfile << fa_names[i] << "\n" << outletters << "\n";
+        outfile << fa_names[i] << endl;
+        for (int j = 0; j < outletters.length(); ++j) {
+          if (j % 80 == 0 && j != 0) {
+            outfile << endl;
+          }
+          outfile << outletters[j];
+        }
+        outfile << endl;
       } else {
-        cout << fa_names[i] << "\n" << outletters << "\n";
+        cout << fa_names[i] << endl;
+        for (int j = 0; j < outletters.length(); ++j) {
+          if (j % 80 == 0 && j != 0) {
+            cout << endl;
+          }
+          cout << outletters[j];
+        }
+        cout << endl;
       }
 
     }