#!/bin/tcsh -f
#(ie run the tshell on this but don't read the .cshrc or .tcshrc)

echo version = 1.30 of yvpg 2011 Mar 28
# 2011 Mar 28, 1.30: fail for 'Genome Res. 2004 14: 1975-1986' - all 3 are years!
# 2011 Mar 14, 1.29: duplicated "if" statement caused bug - fixed
# 2011 Feb 17, 1.28: had year, page, didn't do volume??
#                    35 (2007) 5995-6003.
# 2011 Feb 17, 1.28: Wu.Barker2002 PMID not found - too long?
# 2011 Feb 17, 1.27: pronounce bibtex better usin "byb tech"!
# 2010 Nov 25, 1.26: less delays when reporting missing parts
# 2010 Nov 25, 1.25: delay before closing to allow browser to catch up?
#                    no - just build mechanisms for finding other scripts
# 2010 Nov 15, 1.24: "^@" means bibtex entry - parse accordingly
# 2010 Oct 13, 1.23: non-years inside parenthesis are the issue - ignore
# 2010 Oct 12, 1.22: report failure - screen delay and verbally
# 2010 Oct 04, 1.21: no ',' in long token mode
# 2010 Oct 01, 1.20: finish test 6
# 2010 Oct 01, 1.19: J.C. Angulo and J.S. Dehesa, J. Chem. Phys. 97 (1992) 6485.
#                    = test6
# 2010 Sep 10, 1.18: another kind of dash
# 2010 Sep 10, 1.17: 22(22) should be a volume - drop the issue number
# 2010 Sep 10, 1.16: clean up and report more to output; use first biggest token
# 2010 Sep 10, 1.15: add the biggest token to the search!
# 2010 Sep 10, 1.14: another example
# 2010 Apr 28, 1.13: colon should become a space
# 2010 Apr 28, 1.12: Handle new dash on Sun computer.
# 2010 Apr 28, 1.11: another dash –
# 2010 Apr 23, 1.10: tmp files should be in /tmp!
# 2010 Apr 19, 1.09: page dash from Biochemistry 2009 (failed)
# 2010 Feb 26, 1.08: crash
# 2010 Feb 03, 1.07: another example dash
# 2010 Feb 03, 1.06: tokenize everything except dashes but for each word!
# 2010 Feb 03, 1.05: tokenize everything except dashes!
# 2010 Feb 02, 1.04: handle all non-ascii forms of dash
# 2010 Jan 30, 1.03: fix if statement crash
# 2010 Jan 28, 1.02: year overrides volume only if range ok?? ... unsure
# 2010 Jan 28, 1.01: activate
# 2010 Jan 28, 1.00: origin, test

set delaygood  =  0  # seconds to delay for good results
set delaybad   = 10  # seconds to delay for problems
# say $delaygood $delaybad
# buffer for testing yvp call:
# 2006, 45, 6570

set tmp1 = /tmp/`whoami`-1.yvpg
set tmp2 = /tmp/`whoami`-2.yvpg
set tmp3 = /tmp/`whoami`-3.yvpg
set dotoken = '0' # use biggest token

if ($#argv == 0) then
   echo 'usage: yvpg [string] [-]'
   echo 'Guess the reference of a paper from some text and call yvp.'
   echo 'Use: highlight a reference in a paper, put it in the cut/paste buffer'
   echo '(use command-v for example), then call this script.'
   echo 'Yvpg will guess what the year, volume and page (YVP) are'
   echo 'from the text.  It will add on the first longest alphabetic token'
   echo 'in the buffer to reduce alternative hits.'
   echo 'If the first or second argument is a dash ("-") then'
   echo 'the longest token is used (see test2).'
   echo
   echo 'If non string is given, the cut/paste buffer is used (Mac OSX only),'
   echo 'If the string is "testN" (N=1 to 4) then a test example is run.'
   echo
   if (`uname` == "Darwin") then
      set string = "`pbpaste`"
      echo 'using argument string'
   else
      echo 'This is not a Mac and no string is given so halt.'
      say  'This is not a Mac and no string is given so halt.'
      sleep $delaybad
      exit
   endif
else
   if ("$1" == '-') then
      set dotoken = '1'
   endif
   if ("$#argv" == '2') then
      if ("$1" == '-') then
         set dotoken = '1'
      endif
   endif

   if ("$1" == 'test1') then
      cat > $tmp1 << EOF
      PUBLICATION
      E.J. Enemark and L. Joshua-Tor, Mechanism of DNA translocation in
      a replicative hexameric helicase, Nature 442 270-275 (2006).
EOF
   endif

   if ("$1" == 'test2') then
      cat > $tmp1 << EOF
Beecher, H. K. (1955) "The powerful placebo" Journal of the American Medical Association 159 pp.1602-1606 [Original article, most cited one, claiming a widespread placebo effect]
EOF
   endif

   if ("$1" == 'test3') then
      cat > $tmp1 << EOF
sequences (RefSeq): a curated non-redundant sequence
database of genomes, transcripts and proteins. Nucleic Acids
Res 2007;35:D61–5.
EOF
   endif

   if ("$1" == 'test4') then
      cat > $tmp1 << EOF
Marr,A.G. (1991) Microbiol. Rev., 55, 316-333.
EOF
   endif

   if ("$1" == 'test5') then
      cat > $tmp1 << EOF
Thompson, J.D., Higgins, D.G. and Gibson, T.J. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice (1994) Nucleic Acids Res., 22(22), 4673-4680.
EOF
   endif

   if ("$1" == 'test6') then
      cat > $tmp1 << EOF
J.C. Angulo and J.S. Dehesa, J. Chem. Phys. 97 (1992)
6485.
EOF
   endif

   if ("$1" == '-') then
      # token control: just use buffer
      set string = "`pbpaste`"
   else
      if (`echo "$1" | grep test|wc -l` > 0) then
         #cat /tmp1/hold | pbcopy
         set string = "`cat $tmp1`"
      else
         # use all arguments as the string
         set string = "$argv"
      endif
   endif

endif

# avoid Illegal byte problem for tr on Macs:
setenv LC_ALL en_US.ISO8859-1

set baseyear   = "1900"
set todayyear = `date +%Y`
echo "The current year is $todayyear"

# Look for yvp program
if (1) then # debug locating yvp
   which yvp
   echo
   echo "env ---------------------"
         env
   echo
   echo "pwd ---------------------"
         pwd
   echo
   echo " path -------------------"
   echo "$path"
   echo
   echo "whoami ------------------"
         whoami
   echo
endif
if (`which yvp| grep '/'|wc -l`> 0) then
   echo 'found the yvp program'
   set yvp = yvp
   # say  'found the yvp program' &
else
   # Ok, it's being called as a 'button' on the desktop
   set me = `whoami`
   set foundyvp = 0
   if (-d ~/script) then
      if (-f ~/script/yvp) then
         set yvp = `ls ~/script/yvp`
         set foundyvp = 1
# say found yvp
#sleep $delaybad
      else
         set foundyvp = 0
      endif
   else
      set foundyvp = 0
   endif
   if !($foundyvp) then
      echo 'Could not find the yvp program\!'
      say  'Could not find the yvp program\!'
      echo 'yvp should be in your script directory at'
      say  'yvp should be in your script directory at' 
      echo "~/script/yvp"
      say  "~/script/yvp"
      sleep $delaybad
      exit
   endif
endif

# this goes into yvp:
# Look for wgetas program
if (1) then # debug locating wgetas
   which wgetas
   echo
   echo "env ---------------------"
         env
   echo
   echo "pwd ---------------------"
         pwd
   echo
   echo " path -------------------"
   echo "$path"
   echo
   echo "whoami ------------------"
         whoami
   echo
endif
if (`which wgetas| grep '/'|wc -l`> 0) then
   echo 'found the wgetas program'
   set wgetas = wgetas
   # say  'found the wgetas program' &
else
   # Ok, it's being called as a 'button' on the desktop
   set me = `whoami`
   set foundwgetas = 0
   if (-d ~/script) then
      if (-f ~/script/wgetas) then
         set wgetas = `ls ~/script/wgetas`
         set foundwgetas = 1
# say found wgetas
#sleep $delaybad
      else
         set foundwgetas = 0
      endif
   else
      set foundwgetas = 0
   endif
   if !($foundwgetas) then
      echo 'Could not find the wgetas program\!'
      say  'Could not find the wgetas program\!'
      echo 'wgetas should be in your script directory at'
      say  'wgetas should be in your script directory at' 
      echo "~/script/wgetas"
      say  "~/script/yvp"
      sleep $delaybad
      exit
   endif
endif

echo
echo "--- Captured string to examine: -----"
echo "$string"

set count = 0
set year   = ''
set volume = ''
set page   = ''

# if $y is set, no other variable can override it
set y = ''
set v = ''
set p = ''

# *************************************************************************
# Special case: if the file contains @ on th start of a line,
# assume it is a bibtex entry and parse out the yvp from that.
if (`pbpaste|grep '^@'|wc -l`> 0) then
   pbpaste > $tmp1
   # 2011 Feb 17: byb tech: finally a good pronunciation for 'bibtex'!
   say  byb tech entry &
   echo bibtex entry
   set year = `grep 'year = ' $tmp1|tr -dc "[:digit:]"`
   set volume = `grep 'volume = ' $tmp1|tr -dc "[:digit:]"`
   set page = `grep 'pages = ' $tmp1|tr '-' '\n'|head -1|tr -dc "[:digit:]"`

#echo -=----- $tmp1
#cat $tmp1
#echo -=-----
   #set authors = `cat $tmp1|tr -d '\n'|tr ',' '\n'|grep "author ="`
#echo $authors
if (1) then
   # find longest word
   set biggesttokensize = 0
   set count = 0
   foreach token (`cat "$tmp1"|tr -c '[:alnum:]()' ' '`)
      @ count = $count + 1
# zzz
      echo -n "token ${count} IS '${token}'"
      # track token size - of alphabetic strings
      set thistokensize = `echo "$token"|tr -dc '[:alpha:]'|wc -c`
      if ($thistokensize > $biggesttokensize) then
         set biggesttokensize = $thistokensize 
         set biggesttoken = $token
         echo ", the biggest so far ("${biggesttokensize}" characters)"
      else
         echo
      endif
   end
   set key = $biggesttoken
else
   # use the pmid to nail the entry:
   set key = `cat $tmp1|tr -d '\n'|tr ',' '\n'|grep "pmid ="|tr -dc '[:digit:]'`
   # that failed for Wu.Barker2002!
endif
   echo "$yvp $year $volume $page $key"
         $yvp $year $volume $page $key
   sleep $delaygood
   exit
endif
# *************************************************************************

echo
set zap = ";:."
echo "---- Remove some characters (${zap}) from the string: ---"
# echo "---- modified string ----------------"
# set string = `echo "$string"|tr ";:." "   "|tr -dc '[:print:]'`
# don't do that!  It zaps the dashes which are "non printable"!
# set string = `echo "$string"|tr "$zap" "   "`
# 2010 Oct 1: That fails - loses spaces
#set newstring = "`echo $string|tr ';:.' '   '> $tmp3`"
#echo "$newstring"
# Doing the above causes loss of the spaces!!  Put it in a file.
# Force parenthesis to be separate tokens!!!
echo "$string"|tr ';:.' '   '|sed 's/)/) /g'|sed 's/(/ (/g'> $tmp3
set string = "`cat $tmp3`"
echo "$string"
echo "-------------------------------------"

set biggesttokensize = 0 # Track size of tokens
set biggesttoken = '' # the biggest token found so far

# 2010 oct 04: : remove non al-numeric characters from the string:
# foreach token (`echo "$string"|tr -c '[:alnum:]' ' '`)
# 2010 oct 13: : remove non al-numeric characters from the string, keep ():
foreach ptoken (`echo "$string"|tr -c '[:alnum:]()' ' '`)
echo $ptoken
   @ count = $count + 1
   echo -n "token ${count} IS '${ptoken}'" # for string ${string}"
#echo "$ptoken" | egrep "\(|\)"
#exit

   if (`echo "$ptoken" | egrep '\(|\)' | wc -l` > 0) then
      set token = `echo "$ptoken"|tr -d '()'`
      set notparenthesis = 0
echo -n " PARENTHESIS"
   else
      set token = "$ptoken"
      set notparenthesis = 1
   endif

   # track token size - of alphabetic strings
   # set thistokensize = `wc -c "$token"` # picks up page numbers!
   set thistokensize = `echo "$token"|tr -dc '[:alpha:]'|wc -c`
   if ($thistokensize > $biggesttokensize) then
      set biggesttokensize = $thistokensize 
      set biggesttoken = $token
      echo ", the biggest so far ("${biggesttokensize}" characters)"
   else
      echo
   endif

# Dash of Street.Barrick2008.pdf reference: Greene Pace 1974
# direct cut/paste into vim gives:
#
# Greene, R. F., Jr., and Pace, C. N. (1974). Urea and guanidine
# hydrochloride denaturation of ribonuclease, lysozyme,
# alpha-chymotrypsin, and beta-lactoglobulin. J. Biol. Chem. 249,
# 5388–5393.
# This is not converted to a dash!
# pbpaste >/tmp/t
# then :r /tmp/t
# gives:
# Greene, R. F., Jr., and Pace, C. N. (1974). Urea and guanidine
# hydrochloride denaturation of ribonuclease,
lysozyme,
# alpha-chymotrypsin, and beta-lactoglobulin. J. Biol. Chem. 249,
# 5388?5393.

   # Convert all forms of dash one may find into pure ascii
   # See the u2a script.
   echo "$token" |\
   sed "s/–/-/g" |\
   sed 's/?/-/g' |\
   sed "s/–/-/g" |\
   sed "s/\\u2013/-/g" |\
   sed 's/Ð/-/g' |\
   tr "abcdefghijklmnopqrstuvwxyz" "                          " |\
   tr "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "                          " |\
   # remove leading space as in D615: \
   sed "s/^ //" |\
   cat > $tmp2
   set word = "`cat $tmp2`"

   # set year and page computations for just this word
   set ynow = ''
   set pnow = ''

   # Look for pure numbers, remove periods too.
   set digits = `echo $word | tr -dc "[:digit:]"|tr -d "."`
   if ("$digits" != '') then
      # echo "digits '$digits'"
      echo "  Token ${count}: $word -> $digits"

      if (("$digits" > "1900")&&("$digits" <= "$todayyear")) then
         set couldbeyear = 1 # it could be a year
         echo "  given the value ${digits}, it could be a year"
      else
         set couldbeyear = 0 # it probably is not a year
         echo "  given the value ${digits}, it probably is not a year"
      endif

      # if the string has parenthesis, it is almost certainly a year.
      if (`echo "$word" | egrep '\(|\)' | wc -l` > 0) then
         echo "  has parenthesis"
         # if (("$digits" > "1900")&&("$digits" <= "$todayyear")) then
         if ($couldbeyear) then
            echo "  is greater than $baseyear and less or equal to $todayyear"
            echo "  $digits is the year."
            set year = "$digits"
            set y = "$year"
            set ynow = "$year"
         endif
      endif

      # if the string has a dash, it is probably pages.
      if (`echo "$word" | egrep "-" | wc -l` > 0) then
         echo "  has dash"
         set fp = `echo "$word" | tr '-' '\n' | head -1`
         set sp = `echo "$word" | tr '-' '\n' | tail -1`
         # Parts could be 'pp.1602' - so purify them to be just digits
         set firstpart  = `echo "$fp" | tr -dc '[:digit:]'`
         set secondpart = `echo "$sp" | tr -dc '[:digit:]'`
         echo "  first  part is $firstpart"
         echo "  second part is $secondpart"
         echo -n "  first part = $firstpart"
         if ("$firstpart" < "$secondpart") then
            echo " is less than second part = $secondpart"
         else
            echo " is greater than second part = $secondpart"
            echo " ... use it anyway\!"
         endif
         echo "  $firstpart is the page."
         set page = "$firstpart"
         set p = "$page"
         set pnow = "$page"
      endif

      # If it's not a year or page, assign it to be the year or volume
      if (("$ynow" == '')&&("$pnow" == '')) then
         # Wait! Woah!  It might be it was a year
         # Don't assign if year already assigned!
         if (($couldbeyear)&&("$y" == '')) then
            echo "  could be a year, so assign to be year"
            set year = "$digits"
            echo "  $year is the year."
            set y = "$year"
         else
            if ("$volume" == '') then
               echo "  not year and volume is empty, so assign to be volume"
               # the form 22(22) implies that we should drop the issue
               # see test5
               if (`echo "$token"|grep '('|wc -c`>0) then
                  # token has a '(' in it - assume material before that
                  # is the actual volume and after is the issue
                  set volume = `echo "$token"|tr '(' '\n'|head -1`
                  if ("$volume" == '') then # safety test
                     set volume = "$digits"
                  endif
               else
                  set volume = "$digits"
               endif
               echo "  $volume is the volume."
if ("$volume" == '') then # safety test
 echo 'program error'
 say  'program error'
 sleep $delaybad
 exit
endif
               set v = "$volume"
            else
               if (("$page" == '')&&($notparenthesis)) then
# echo "====== $word"

                  # retokanize ANYTHING not a digit, to break away
                  # first part of token: set digits = "`echo "$word"|
                  # tr -c '[:digit:]' '\n'|head -1`" double quote used
                  # 4 times is no good!
                  set digits = `echo "$word"| tr -c '[:digit:]' '\n'|head -1`
                  echo "  year and volume assigned already, so assign to be page"
                  set page = "$digits"
                  echo "  $page is the page."
                  set p = "$page"
               endif
            endif
         endif
      endif

      echo ---

   endif

end

if (("$y" != '')&&("$v" != '')&& ("$p" != '')) then
   if !($dotoken) then
      set biggesttoken = '' # zap the hard-earned token
   endif
   echo "The call by yvpg to yvp is:"
   echo "$yvp $year $volume $page $biggesttoken"
         $yvp $year $volume $page $biggesttoken
   echo "---"
   echo "The call by yvpg to yvp was:"
   echo "$yvp $year $volume $page $biggesttoken"
else
   echo "no yvp because:"
   set gavemessage = 0
   if ("$y" == '') then
      echo "no year found"
      say  "no year found"
      set gavemessage = 1
   endif
   if ("$v" == '') then
      echo "no volume found"
      say  "no volume found"
      set gavemessage = 1
   endif
   if ("$p" == '') then
      echo "no page found"
      say  "no page found"
      set gavemessage = 1
   endif
   #if (`` == 'Darwin') then
   #   echo 'not found'
   #   say  'not found'
   #   set gavemessage = 1
   #endif
   if ($gavemessage) then
      echo "yvpg ok"
   else
      echo "no message, vypg FAILED"
      say  "no message, vypg FAILED"
   endif
   sleep $delaybad
endif

# say exiting
sleep $delaygood

echo done yvpg
exit

********************************************************************************
*** END OF PROGRM **************************************************************
********************************************************************************
********************************************************************************
********************************************************************************
********************************************************************************

Junk below this point

example string:

from:
http://www.nsls.bnl.gov/newsroom/science/2006/12-Joshua-Tor.htm

PUBLICATION
E.J. Enemark and L. Joshua-Tor, β€œMechanism of DNA translocation in a replicative hexameric helicase,” Nature 442 270-275 (2006).

********************************************************************************

# echo -n 'Give user name: '
# # read user string (only works with one line, no good for this code)
# set user = $<

Unicode dash example:

Ofria, C., Huang, W., Torng, E., 2008. On the gradual evolution of complexity and the sudden emergence of complex features. Artif. Life 14, 255–263.

********************************************************************************

2010 Feb 03:
tokenize everything except dashes because of:

sequences (RefSeq): a curated non-redundant sequence
database of genomes, transcripts and proteins. Nucleic Acids
Res 2007;35:D61–5.

Street T O, Courtemanche N and Barrick D 2008 Protein folding
and stability using denaturants Methods Cell. Biol.
84 295–325

–

some old code:

# foreach token (`echo $string|sed "s/\\u2013/-/g"`)
# foreach token (`echo "$string"|tr ";:." "   "`)
# foreach token (`echo "$string"|tr ";:." "   "|tr -dc '[:print:]'`)
#                                ^ more tokanization

#    # Convert all forms of dash one may find into pure ascii
#    # See the u2a script
#    echo "$token" |\
#    sed "s/–/-/g" |\
#    sed "s/–/-/g" |\
#    sed "s/\\u2013/-/g" |\
#    sed 's/Ð/-/g' |\
#    cat > $tmp2
#    set word = "`cat $tmp2`"
#    # echo $word is the word
#   set word = "$token"

   # page dash from Biochemistry pdf 2009: \
   # sed 's/β€šΓ„Γ/-/g' |\
   # sed 's/–/-/g' |\
   # pbpaste and direct transfer: \
   # sed 's/Π/-/g' |\
   # that failed too. \
   # Mol Microb 56 1481 2005 dash of first reference \
   # CANNOT WRITE THE FILE FROM VIM on Mac OSX!  Sun is ok. \
   # CONVERSION ERROR 327L \
   # turns out to be – which is already handled \
   # sed 's//-/g' |\
   #yyy \
   # Tokanize everything else other than dash. \
   # This works on Mac OSX but not on Sun :-( \
   # tr -c "[:digit:]-" " " |\
   #tr -c "[:digit:]-" " " |\
   #tr -c "[:digit:]-" "#" |\
   # do this for the Sun OS: \
   # tr ";:" " " |\

   # the above is zapping the dash!!!
   # set word = "$token"

#    # Convert all forms of dash one may find into pure ascii
#    # See the u2a script
#    echo "$string" |\
#    sed "s/–/-/g" |\
#    sed "s/–/-/g" |\
#    sed "s/\\u2013/-/g" |\
#    sed 's/Ð/-/g' |\
#    # Tokanize everything else other than dash. \
#    # This works on Mac OSX but not on Sun :-( \
#    tr -c "[:digit:]-" " " |\
#    # do this for the Sun OS: \
#    tr ";:" " " |\
#    tr "abcdefghijklmnopqrstuvwxyz" "                          " |\
#    tr "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "                          " |\
#    # tr -d "." |\
#    # \
#    cat > $tmp2
#    set string = "`cat $tmp2`"

   # tr -d "." |\
   # \

# foreach token (`echo "$string"|tr -d "'\n"`)

# 2010 oct 01: : remove non printable characters from the string:
# foreach token (`echo "$string"|tr -cd '[:print:]'`)
That failed on
D. Sellis and Y. Almirantis, Gene 447, 18 (2009).
it kept the comma in 'Almirantis,'!