Patching WFA2-lib to prevent C macros from bleeding into the environm…

…ent. See

- WFA2-lib smarco#35

and

- vcflib vcflib/vcflib#359

External tools will no longer include files wavefront_align.h or wavefront_aligner.h.
That includes the bindings that come with WFA2-lib itself - they included wavefront_aligner.h wrapped as external "C".
These no longer include files, such as commons.h, that define macros and variables that are mostly internal to WFA.

To bind external tools I have introduced a new include file 'wfa.h' that contains the function definitions and minimal data structures. This file is now included by wavefront_aligner.h so WFA2-lib sees the same definitions. Also I included wfa.hpp that uses a C calling convention for C++, so there is no need to wrap the include file in extern "C".

To make it work I had to remove almost all includes to utils/commons.h from the headers. If this causes compile errors in other tools, have them include the file at the .c level (not .h). I also merged wavefront_align.h into wfa.h, so as not to have to include wavefront_alignment.h fully every time.

To validate that it works I called some macros by vcflib - after this patch they no longer are seen.

Finally I included a guix.scm file that allows us to create an isolated build environment using different versions of C compilers etc. See the header of guix.scm for instructions.

Notes: it is better to use newer `#pragma once' for header files that the ifndef construct

.gitignore

@@ -35,3 +35,9 @@ build/
 *.out
 *.app
 
+# Test output files
+tests/wfa.utest.log.correct
+tests/wfa.utest.log.mem
+tests/wfa.utest.log.time
+
+

alignment/affine2p_penalties.h

@@ -32,8 +32,6 @@
 #ifndef AFFINE2P_PENALTIES_H_
 #define AFFINE2P_PENALTIES_H_
 
-#include "utils/commons.h"
-
 /*
  * Affine 2-piece penalties
  */

alignment/affine_penalties.h

@@ -32,8 +32,6 @@
 #ifndef AFFINE_PENALTIES_H_
 #define AFFINE_PENALTIES_H_
 
-#include "utils/commons.h"
-
 /*
  * Affine penalties
  */

alignment/cigar.c

@@ -29,6 +29,7 @@
  * DESCRIPTION: Edit cigar data-structure (match/mismatch/insertion/deletion)
  */
 
+#include "utils/commons.h"
 #include "cigar.h"
 
 /*
@@ -506,5 +507,3 @@ void cigar_print_pretty(
   mm_allocator_free(mm_allocator,ops_alg);
   mm_allocator_free(mm_allocator,text_alg);
 }
-
-

alignment/cigar.h

@@ -32,7 +32,6 @@
 #ifndef CIGAR_H_
 #define CIGAR_H_
 
-#include "utils/commons.h"
 #include "system/mm_allocator.h"
 #include "alignment/linear_penalties.h"
 #include "alignment/affine_penalties.h"

alignment/score_matrix.c

@@ -29,8 +29,10 @@
  * DESCRIPTION: Score matrix for alignment using dynamic programming
  */
 
+#include "utils/commons.h"
 #include "score_matrix.h"
 
+
 /*
  * Setup
  */
@@ -114,7 +116,3 @@ void score_matrix_print(
   }
   fprintf(stream,"\n");
 }
-
-
-
-

alignment/score_matrix.h

@@ -32,7 +32,6 @@
 #ifndef SCORE_MATRIX_H_
 #define SCORE_MATRIX_H_
 
-#include "utils/commons.h"
 #include "system/mm_allocator.h"
 #include "alignment/cigar.h"
 

bindings/cpp/WFAligner.hpp

@@ -34,9 +34,7 @@
 
 #include <string>
 
-extern "C" {
-  #include "../../wavefront/wavefront_aligner.h"
-}
+#include "../../wavefront/wfa.hpp"
 
 /*
  * Namespace

examples/wfa_adapt.c

@@ -29,6 +29,7 @@
  * DESCRIPTION: WFA Sample-Code
  */
 
+#include "utils/commons.h"
 #include "wavefront/wavefront_align.h"
 
 int main(int argc,char* argv[]) {

examples/wfa_basic.c

@@ -29,6 +29,7 @@
  * DESCRIPTION: WFA Sample-Code
  */
 
+#include "utils/commons.h"
 #include "wavefront/wavefront_align.h"
 
 int main(int argc,char* argv[]) {

examples/wfa_repeated.c

@@ -29,6 +29,7 @@
  * DESCRIPTION: WFA Sample-Code
  */
 
+#include "utils/commons.h"
 #include "wavefront/wavefront_align.h"
 
 int main(int argc,char* argv[]) {

guix.scm

@@ -0,0 +1,66 @@
+;; Set up build environment using GNU Guix packages
+;;
+;; CC0 license, Pjotr Prins (c) 2022
+;;
+;; To use this file to build HEAD:
+;;
+;;   guix build -f guix.scm
+;;
+;; To get a development container (emacs shell will work)
+;;
+;;   guix shell -C -D -f guix.scm
+;;
+;; For the tests you need /bin/bash. In a container create it with
+;;
+;;   mkdir -p /bin ; ln -s $GUIX_ENVIRONMENT/bin/bash /bin/bash
+;;
+;; and run the tests with
+;;
+;;   env CC=gcc make
+;;   ./tests/wfa.utest.sh
+
+(use-modules
+ ((guix licenses) #:prefix license:)
+  (guix gexp)
+  (guix packages)
+  (guix git-download)
+  (guix build-system cmake)
+  (gnu packages algebra)
+  (gnu packages autotools)
+  (gnu packages base)
+  (gnu packages bash)
+  (gnu packages compression)
+  (gnu packages build-tools)
+  (gnu packages check)
+  (gnu packages curl)
+  (gnu packages gcc)
+  (gnu packages gdb)
+  (gnu packages llvm)
+  (gnu packages parallel)
+  (gnu packages pkg-config)
+  (srfi srfi-1)
+  (ice-9 popen)
+  (ice-9 rdelim))
+
+(define %source-dir (dirname (current-filename)))
+
+(define %git-commit
+    (read-string (open-pipe "git show HEAD | head -1 | cut -d ' ' -f 2" OPEN_READ)))
+
+(define-public wfa2-lib-git
+  (package
+    (name "wfa2-lib-git")
+    (version (git-version "1.3" "HEAD" %git-commit))
+    (source (local-file %source-dir #:recursive? #f))
+    (build-system cmake-build-system)
+    (inputs
+     `(("bash" ,bash)
+       ("gdb" ,gdb)))
+    (native-inputs
+     `(("pkg-config" ,pkg-config)))
+    (home-page "https://github.com/smarco/WFA2-lib/")
+    (synopsis "Library for wavefront aligner")
+    (description "The wavefront alignment (WFA) algorithm is an **exact** gap-affine algorithm that takes advantage of homologous regions between the sequences to accelerate the alignment process. Unlike to traditional dynamic programming algorithms that run in quadratic time, the WFA runs in time `O(ns+s^2)`, proportional to the sequence length `n` and the alignment score `s`, using `O(s^2)` memory (or `O(s)` using the ultralow/BiWFA mode).")
+    (license license:expat))) ;; MIT license
+
+wfa2-lib-git

system/mm_allocator.c

@@ -32,6 +32,7 @@
  *   and dispatching memory segments in order.
  */
 
+#include "utils/commons.h"
 #include "mm_allocator.h"
 
 /*
@@ -611,6 +612,3 @@ void mm_allocator_print(
     mm_allocator_print_allocator_requests(stream,mm_allocator,false);
   }
 }
-
-
-

system/mm_stack.c

@@ -32,6 +32,7 @@
  *   requested at once.
  */
 
+#include "utils/commons.h"
 #include "mm_stack.h"
 
 /*
@@ -265,5 +266,3 @@ void mm_stack_print(
   fprintf(stream,"  => Segments.size      %" PRIu64 " MB\n",segment_size/(1024*1024));
   fprintf(stream,"  => Memory.available   %" PRIu64 " MB\n",num_segments*(segment_size/(1024*1024)));
 }
-
-

system/profiler_counter.c

@@ -30,6 +30,7 @@
  * DESCRIPTION: Simple profile counter
  */
 
+#include "utils/commons.h"
 #include "profiler_counter.h"
 
 /*

system/profiler_counter.h

@@ -33,8 +33,6 @@
 #ifndef PROFILER_COUNTER_H_
 #define PROFILER_COUNTER_H_
 
-#include "utils/commons.h"
-
 /*
  * Counters
  */

system/profiler_timer.c

@@ -30,6 +30,7 @@
  * DESCRIPTION: Simple time profiler
  */
 
+#include "utils/commons.h"
 #include "profiler_timer.h"
 
 #ifdef __MACH__

system/profiler_timer.h

@@ -33,7 +33,6 @@
 #ifndef PROFILER_TIMER_H
 #define PROFILER_TIMER_H
 
-#include "utils/commons.h"
 #include "profiler_counter.h"
 
 /*

tools/align_benchmark/edit/edit_bpm.c

@@ -22,6 +22,8 @@
  * DESCRIPTION: Edit-Distance based BPM alignment algorithm
  */
 
+#include "utils/commons.h"
+#include "system/mm_allocator.h"
 #include "edit/edit_bpm.h"
 #include "utils/dna_text.h"
 

tools/align_benchmark/edit/edit_bpm.h

@@ -25,7 +25,6 @@
 #ifndef EDIT_BPM_H_
 #define EDIT_BPM_H_
 
-#include "utils/commons.h"
 #include "alignment/cigar.h"
 #include "system/mm_allocator.h"
 

tools/align_benchmark/edit/edit_dp.c

@@ -29,6 +29,8 @@
  * DESCRIPTION: Dynamic-programming algorithm to compute Levenshtein alignment (edit)
  */
 
+#include "utils/commons.h"
+#include "system/mm_allocator.h"
 #include "edit_dp.h"
 
 /*

tools/align_benchmark/edit/edit_dp.h

@@ -32,7 +32,6 @@
 #ifndef EDIT_DP_H_
 #define EDIT_DP_H_
 
-#include "utils/commons.h"
 #include "alignment/score_matrix.h"
 
 /*

tools/align_benchmark/gap_affine/affine_matrix.c

@@ -29,6 +29,8 @@
  * DESCRIPTION: Gap-affine Matrix
  */
 
+#include "utils/commons.h"
+#include "system/mm_allocator.h"
 #include "gap_affine/affine_matrix.h"
 
 /*

tools/align_benchmark/gap_affine/affine_matrix.h

@@ -32,7 +32,6 @@
 #ifndef AFFINE_MATRIX_H_
 #define AFFINE_MATRIX_H_
 
-#include "utils/commons.h"
 #include "alignment/cigar.h"
 
 /*

tools/align_benchmark/gap_affine/swg.c

@@ -30,6 +30,8 @@
  *   pairwise alignment (Smith-Waterman-Gotoh - SWG)
  */
 
+#include "utils/commons.h"
+#include "system/mm_allocator.h"
 #include "gap_affine/swg.h"
 
 /*

tools/align_benchmark/gap_affine/swg.h

@@ -33,7 +33,6 @@
 #ifndef SWG_H_
 #define SWG_H_
 
-#include "utils/commons.h"
 #include "gap_affine/affine_matrix.h"
 
 /*

utils/bitmap.c

@@ -29,6 +29,7 @@
  * DESCRIPTION: Basic bitmap datastructure (static)
  */
 
+#include "utils/commons.h"
 #include "utils/bitmap.h"
 #include "system/mm_allocator.h"
 
@@ -122,6 +123,3 @@ uint64_t bitmap_erank(
   const uint64_t bitmap_count = POPCOUNT_64(bitmap_masked);
   return bitmap_block->counter + bitmap_count;
 }
-
-
-

utils/bitmap.h

@@ -35,7 +35,6 @@
 /*
  * Includes
  */
-#include "utils/commons.h"
 #include "system/mm_allocator.h"
 
 #define BITMAP_BLOCK_ELEMENTS 64

utils/commons.h

@@ -29,8 +29,7 @@
  * DESCRIPTION: Common functions/utilities and headers for C development
  */
 
-#ifndef COMMONS_H_
-#define COMMONS_H_
+#pragma once
 
 #include <stdio.h>
 #include <stdlib.h>
@@ -278,5 +277,3 @@ uint64_t nominal_prop_u64(const uint64_t base,const double factor);
   int i; \
   for (i=0;i<times;++i) fprintf(stream,"%c",character); \
 }
-
-#endif /* COMMONS_H_ */

utils/dna_text.c

@@ -39,6 +39,7 @@
 /*
  * Include
  */
+#include "utils/commons.h"
 #include "utils/dna_text.h"
 
 /*
@@ -58,4 +59,3 @@ const char dna_decode_table[DNA_EXTENDED_RANGE] =
   [ENC_DNA_CHAR_T] = DNA_CHAR_T,
   [ENC_DNA_CHAR_N] = DNA_CHAR_N,
 };
-

utils/dna_text.h

@@ -33,8 +33,6 @@
 #ifndef DNA_TEXT_H_
 #define DNA_TEXT_H_
 
-#include "utils/commons.h"
-
 /*
  * Range of DNA Nucleotides
  */

utils/heatmap.c

@@ -28,6 +28,7 @@
  * AUTHOR(S): Santiago Marco-Sola <[email protected]>
  */
 
+#include "utils/commons.h"
 #include "heatmap.h"
 
 /*
@@ -168,6 +169,3 @@ void heatmap_print(
     fprintf(stream,"\n");
   }
 }
-
-
-

utils/heatmap.h

@@ -31,8 +31,6 @@
 #ifndef HEATMAP_H_
 #define HEATMAP_H_
 
-#include "utils/commons.h"
-
 /*
  * Heatmap
  */