#!/usr/local/bin/perl -w
###############################################################################
#              Mike Galsworthy's "Free Text to MeSH" converter                #
#                                                                             #
#                                                                             #
#          PROGRAM 1 :  Retrieving PubMed articles in Medline format          #
#                                                                             #
#     (credit to Oleg Khovayko for key Medline download parts of script)      #
#                                                                             #
#                                                                             #
###############################################################################


#### PART 1. SETTING COMMAND-LINE TERMS #######################################

use Getopt::Long;

my ($term, $keep, $retmax);

GetOptions ("query:s" => \$term, "records=i" => \$keep, "batch=i" => \$retmax);

$term =~ s/~q~q~/\'/g;

if (!($term)) {
    die "No query term given. Use the --query 'term here' option.\n";
    }


#### PART 2. E-SEARCH ON THE QUERY TERM (USING "HISTORY") #####################

use LWP::Simple;

$base = "http://www.ncbi.nlm.nih.gov/entrez/eutils/esearch.fcgi?".
        "db=Pubmed".     "&".
        "retmax=100".    "&".
        "usehistory=y".  "&".
        "term=";

$signature = '&tool=medsum&email=mike_galsworthy@yahoo.co.uk';

$esearch = get($base . $term . $signature);

$esearch =~
m|<Count>(\d+)</Count>.*<QueryKey>(\d+)</QueryKey>.*<WebEnv>(\S+)</WebEnv>|s;

my $Count    = $1;
my $QueryKey = $2;
my $WebEnv   = $3;

print "QRY - $term\n";    # For info at end of output
print "CNT - $Count\n";   # For info at end of output
print "KP  - $keep\n";    # For info at end of output

#### PART 3. E-SEARCH ON THE QUERY TERM (USING "HISTORY") #####################

if (!($keep)) {$keep = 1000;}         # default value for records to keep
if ($Count < $keep) {$keep = $Count;}   # in case keeping more than query count
if (!($retmax)) {$retmax = 500;}      # default batch value (recommended max)
if ($keep < $retmax) {$retmax = $keep;} # in case batch larger than records

####### LOOP to write data into file in batches (of 300 max) ##################
                                                                              #
my $retstart;                                                                 #
for($retstart = 0; $retstart < $keep; $retstart += $retmax) {                 #
my $efetch = "http://www.ncbi.nlm.nih.gov/entrez/eutils/efetch.fcgi?".        #
              "rettype=medline".      "&".                                    #
              "retmode=text".         "&".                                    #
              "retstart=$retstart".   "&".                                    #
              "retmax=$retmax".       "&".                                    #
              "db=Pubmed".            "&".                                    #
              "query_key=$QueryKey".  "&".                                    #
              "WebEnv=$WebEnv".       "$signature";                           #
                                                                              #
my $efetch_result = get($efetch);                                             #
                                                                              #
  print "$efetch_result\n";                                                   #
  if ($retstart > 0) {sleep 3;} #3 second time delay between batch calls  #
}                                                                             #
########################## END OF LOOP ########################################

print "KP  - $keep\n";    # For info at end of output


open FILE, ">>log.txt" or die $!;
print FILE "MI - $term  ($Count) \r\n";
close FILE;