Mercurial > repos > aaronquinlan > multi_intersect

comparison BEDTools-Version-2.14.3/src/nucBed/nucBedMain.cpp @ 0:dfcd8b6c1bda

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Uploaded
author aaronquinlan
date Thu, 03 Nov 2011 10:25:04 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:dfcd8b6c1bda
1 /*****************************************************************************
2 nucBedMain.cpp
3
4 (c) 2009 - Aaron Quinlan
5 Hall Laboratory
6 Department of Biochemistry and Molecular Genetics
7 University of Virginia
8 aaronquinlan@gmail.com
9
10 Licenced under the GNU General Public License 2.0 license.
11 ******************************************************************************/
12 #include "nucBed.h"
13 #include "version.h"
14
15 using namespace std;
16
17 // define our program name
18 #define PROGRAM_NAME "nucBed"
19
20
21 // define our parameter checking macro
22 #define PARAMETER_CHECK(param, paramLen, actualLen) (strncmp(argv[i], param, min(actualLen, paramLen))== 0) && (actualLen == paramLen)
23
24 // function declarations
25 void ShowHelp(void);
26
27 int main(int argc, char* argv[]) {
28
29 // our configuration variables
30 bool showHelp = false;
31
32 // input files
33 string fastaDbFile;
34 string bedFile;
35 string pattern;
36
37 // checks for existence of parameters
38 bool haveFastaDb = false;
39 bool haveBed = false;
40 bool printSeq = false;
41 bool hasPattern = false;
42 bool forceStrand = false;
43
44 // check to see if we should print out some help
45 if(argc <= 1) showHelp = true;
46
47 for(int i = 1; i < argc; i++) {
48 int parameterLength = (int)strlen(argv[i]);
49
50 if((PARAMETER_CHECK("-h", 2, parameterLength)) ||
51 (PARAMETER_CHECK("--help", 5, parameterLength))) {
52 showHelp = true;
53 }
54 }
55
56 if(showHelp) ShowHelp();
57
58 // do some parsing (all of these parameters require 2 strings)
59 for(int i = 1; i < argc; i++) {
60
61 int parameterLength = (int)strlen(argv[i]);
62
63 if(PARAMETER_CHECK("-fi", 3, parameterLength)) {
64 if ((i+1) < argc) {
65 haveFastaDb = true;
66 fastaDbFile = argv[i + 1];
67 i++;
68 }
69 }
70 else if(PARAMETER_CHECK("-bed", 4, parameterLength)) {
71 if ((i+1) < argc) {
72 haveBed = true;
73 bedFile = argv[i + 1];
74 i++;
75 }
76 }
77 else if(PARAMETER_CHECK("-seq", 4, parameterLength)) {
78 printSeq = true;
79 }
80 else if(PARAMETER_CHECK("-s", 2, parameterLength)) {
81 forceStrand = true;
82 }
83 else if(PARAMETER_CHECK("-pattern", 8, parameterLength)) {
84 if ((i+1) < argc) {
85 hasPattern = true;
86 pattern = argv[i + 1];
87 i++;
88 }
89 }
90 else {
91 cerr << "*****ERROR: Unrecognized parameter: " << argv[i] << " *****" << endl << endl;
92 showHelp = true;
93 }
94 }
95
96 if (!haveFastaDb || !haveBed) {
97 showHelp = true;
98 }
99
100 if (!showHelp) {
101
102 NucBed *nuc = new NucBed(fastaDbFile, bedFile, printSeq, hasPattern, pattern, forceStrand);
103 delete nuc;
104
105 return 0;
106 }
107 else {
108 ShowHelp();
109 }
110 }
111
112 void ShowHelp(void) {
113
114 cerr << endl << "Program: " << PROGRAM_NAME << " (v" << VERSION << ")" << endl;
115
116 cerr << "Author: Aaron Quinlan (aaronquinlan@gmail.com)" << endl;
117
118 cerr << "Summary: Profiles the nucleotide content of intervals in a fasta file." << endl << endl;
119
120 cerr << "Usage: " << PROGRAM_NAME << " [OPTIONS] -fi <fasta> -bed <bed/gff/vcf>" << endl << endl;
121
122 cerr << "Options: " << endl;
123 cerr << "\t-fi\tInput FASTA file" << endl << endl;
124 cerr << "\t-bed\tBED/GFF/VCF file of ranges to extract from -fi" << endl << endl;
125 cerr << "\t-s\tProfile the sequence according to strand." << endl << endl;
126 cerr << "\t-seq\tPrint the extracted sequence" << endl << endl;
127 cerr << "\t-pattern\tReport the number of times a user-defined sequence is observed (case-insensitive)." << endl << endl;
128
129
130 cerr << "Output format: " << endl;
131 cerr << "\tThe following information will be reported after each original BED entry:" << endl;
132 cerr << "\t 1) %AT content" << endl;
133 cerr << "\t 2) %GC content" << endl;
134 cerr << "\t 3) Number of As observed" << endl;
135 cerr << "\t 4) Number of Cs observed" << endl;
136 cerr << "\t 5) Number of Gs observed" << endl;
137 cerr << "\t 6) Number of Ts observed" << endl;
138 cerr << "\t 7) Number of Ns observed" << endl;
139 cerr << "\t 8) Number of other bases observed" << endl;
140 cerr << "\t 9) The length of the explored sequence/interval." << endl;
141 cerr << "\t 10) The sequence extracted from the FASTA file. (optional, if -seq is used)" << endl;
142 cerr << "\t 11) The number of times a user defined pattern was observed. (optional, if -pattern is used.)" << endl;
143
144 // end the program here
145 exit(1);
146
147 }