Merge branch 'master' of github.com:smithlabcode/smithlab_cpp

Author: andrewdavidsmith

README.md

@@ -1,20 +1,22 @@
 # The smithlab_cpp library
 
-This library contains code that has been used in the Smith lab for
-several years, and that we now depend on for several of our data
-analysis tools. Many of those tools use older versions of this source
-code in subdirectories of other repos.
+This library contains code that has been used in the Smith lab for too
+many years, and that we depend on for several of our data analysis
+tools. Many of those tools use older versions of this source code in
+subdirectories of other repos.
 
 ## Requirements
 
-- A C++ compiler that knows C++11. The GNU `g++` compiler works well
-  for this after version 5.3.
-- The GNU Scientific Library, GSL, which is likely already on your
-  system, or easily installed through a package manager.
+- A C++ compiler that knows C++17. The GNU `g++` compiler works well
+  for this after version GCC9, and the default since GCC11. You can
+  get this from apt, conda and brew, but beware on macos where it's
+  not really installed by default even though it pretends to be.
 - The [zlib library](https://zlib.net), which we use for I/O of files
-  in gzip format. You likely have this on your system.
+  in gzip format. You likely have this on your system. You can get
+  this from apt, conda and brew, and it's likely installed already.
 - Optional: The [HTSLib library](http://htslib.org), which we use for
-  I/O of SAM and BAM format files.
+  I/O of SAM and BAM format files. You can get this from apt, conda
+  and brew.
 
 ## Building and installing the smithlab_cpp library
 
@@ -55,7 +57,7 @@ use it the way it has been used from 2010-2019, then you can use the
 `Makefile` in this repo without running the `./configure` script:
 ```
 $ make OptionParser.o
-g++ -Wall -std=c++11 -c -o OptionParser.o OptionParser.cpp
+g++ -Wall -std=c++17 -c -o OptionParser.o OptionParser.cpp
 ```
 Note: if you run the `./configure` script it will overwrite the
 `Makefile` indicated above. If that happens, just get a new one.  The
@@ -72,7 +74,7 @@ result should be less total code overall in smithlab_cpp.
 - `QualityScore.*pp` likely should be removed, as we only use
   sequencing quality scores in specific places, and in those places
   have chosen to re-implement anything that would be here.
-- `smithlab_os.* Any use of character arrays should be replaced with
+- `smithlab_os.*` Any use of character arrays should be replaced with
   strings for filenames. Implementation of many functions in the cpp
   file is sloppy.
 - `smithlab_utils.*pp`: lots to replace here. Many functions seem

chromosome_utils.cpp

@@ -74,7 +74,7 @@ adjust_start_pos(const size_t orig_start, const string &chrom_name) {
 
 static size_t
 adjust_region_size(const size_t orig_start,
-                   const string &chrom_name,
+                   const string &chrom_name, // ADS: remove this soon
                    const size_t orig_size) {
   static const double LINE_WIDTH = 50.0;
   const size_t preceding_newlines_start =

sam_record.cpp

@@ -112,13 +112,13 @@ sam_rec::sam_rec(const string &line) {
     istringstream iss; // ADS: change to set the buffer from "line"
     iss.rdbuf()->pubsetbuf(const_cast<char*>(line.c_str()), line.size());*/
   istringstream iss(line); // ADS: unfortunate macos stuff?
-  uint32_t will_become_mapq = 0; // to not read mapq as character
-                                 // since it's uint8_t
+  int32_t will_become_mapq = 0; // to not read mapq as character
+                                // since it's uint8_t
   if (!(iss >>
         qname >> flags >> rname >> pos >> will_become_mapq >>
         cigar >> rnext >> pnext >> tlen >> seq >> qual))
     throw runtime_error("incorrect SAM record:\n" + line);
-  if (mapq > 255)
+  if (will_become_mapq < 0 || will_become_mapq > 255)
     throw runtime_error("invalid mapq in SAM record: " + line);
   mapq = static_cast<uint8_t>(will_become_mapq);
 

smithlab_os.cpp

@@ -24,6 +24,7 @@
 #include <cmath>
 #include <unordered_map>
 #include <exception>
+#include <filesystem>
 
 #include "smithlab_os.hpp"
 #include "smithlab_utils.hpp"
@@ -535,9 +536,10 @@ read_dir(const string& dirname, vector<string> &filenames) {
 
 bool
 is_valid_output_file(const string &filename) {
-  const bool file_exists = (access(filename.c_str(), F_OK) == 0);
-  if (file_exists)
-    return (!isdir(filename.c_str()) &&
+  // ADS: seems like there is no way around "access" and apparently
+  // access is not a great solution anyway.
+  if (std::filesystem::exists(filename))
+    return (!std::filesystem::is_directory(filename) &&
             access(filename.c_str(), W_OK) == 0);
   else {
     // ADS: check if dir exists and is writeable

smithlab_utils.hpp

@@ -300,7 +300,6 @@ template <class T> std::string toa(T t) {
   return s.str();
 }
 
-
 ////////////////////////////////////////////////////////////////////////
 // Code for dealing with the DNA alphabet
 
@@ -338,8 +337,6 @@ bits2string_masked(size_t mask, size_t bits) {
   return s;
 }
 
-
-
 inline std::string
 bits2string_for_positions(size_t positions, size_t bits) {
   std::string s;
@@ -356,22 +353,14 @@ percent(const size_t a, const size_t b) {
   return static_cast<size_t>((100.0*a)/b);
 }
 
-
-
-
-
-////////////////////Code from Alphabet//////////////////////
-
-
-
-bool
-inline valid_base(char c) {
+inline bool
+valid_base(char c) {
   char i = std::toupper(c);
   return (i == 'A' || i == 'C' || i == 'G' || i == 'T');
 }
 
-size_t
-inline mer2index(const char *s, size_t n) {
+inline size_t
+mer2index(const char *s, size_t n) {
   size_t multiplier = 1, index = 0;
   do {
     --n;
@@ -381,9 +370,9 @@ inline mer2index(const char *s, size_t n) {
   return index;
 }
 
-size_t
-inline kmer_counts(const std::vector<std::string> &seqs,
-                   std::vector<size_t> &counts, size_t k) {
+inline size_t
+kmer_counts(const std::vector<std::string> &seqs,
+            std::vector<size_t> &counts, size_t k) {
   counts.clear();
   size_t nwords =
     static_cast<size_t>(pow(static_cast<float>(smithlab::alphabet_size),
@@ -404,7 +393,6 @@ inline kmer_counts(const std::vector<std::string> &seqs,
   return total;
 }
 
-
 /*
  * How to use the ProgressBar:
  *