#! /usr/bin/python
# COPYRIGHT 2001 KEN SUGINO
# NEEDS
# 1) Python (www.python.org) (tested version 1.5.2)
# 2) wxPython (www.wxpython.org) (tested version:2.2.5)
# USAGE
# 0) change first line of this file appropriately, i.e. /usr/bin/python->actual path to python
# 1) give executable property to this file: chmod +x pmBib
# 2) ./pmBIb &
# * optionally set BROWSER_CMD below to appropriate one url will be inserted to %s
#-------- PREFERENCES ----------------
MAXRETRIEVE = 200
BROWSER_CMD = "netscape -remote 'openURL(%s)'"
WIN_WIDTH = 850
WIN_HEIGHT = 750
COL_NUM_W = 35
COL_YEAR_W = 45
COL_TITLE_W = 370
COL_AUTHOR_W = 200
COL_JOURNAL_W = 200
#-------- HELP TEXT ------------------
pmHelpText ="""
Pub Med Search Help
See
NCBI PubMed help page for the details.
Field Specification
example: Einstein A[au] : specifies author fields [au]
Boolean Operator
AND, OR : use parenthesis to create logical block
Date Range
example: 1999:2001[edat] : entrez date from 1999 to 2001
"""
#-------- PROGRAM CODE ---------------
from wxPython.wx import *
from wxPython.html import *
from wxPython.lib.anchors import LayoutAnchors
import urllib,string,os,re
TEMPLATE_NEW = """% bibtex file made by pmBib\n"""
app = None
class pmApp(wxApp):
def OnInit(self):
global app
app = self
self.frame = pmFrame(NULL,-1,"PubMedQueryToBibTex",app=self)
self.frame.Show(true)
self.SetTopWindow(self.frame)
return true
ID_ABOUT = 101
ID_EXIT = 102
ID_OPEN = 103
ID_CLOSEFILE = 104
ID_NEW = 105
ID_TERM = 501
ID_FIND = 502
ID_SPLITTER = 1000
ID_LIST = 1001
ID_NOTE = 200
ID_SEARCH_PANE = 201
ID_HELP_PANE = 202
ID_OFFSET = 10000
class pmFrame(wxFrame):
def __init__(self,parent,ID,title,app):
wxFrame.__init__(self,parent,ID,title,
wxDefaultPosition,wxSize(WIN_WIDTH,WIN_HEIGHT))
self.CreateStatusBar()
self.dirname = os.path.abspath('.')
self.filename = ''
self.files = []
self.panes = []
self.app = app
self.SetAutoLayout(true)
# query term input text and button
self.term = wxTextCtrl(self, ID_TERM, "", wxPoint(0,0), wxSize(WIN_WIDTH-60, 20))
self.term.SetConstraints(LayoutAnchors(self.term,true,true,true,false))
EVT_CHAR(self.term, self.OnEvtChar)
self.findbtn = wxButton(self, ID_FIND,"find",wxPoint(WIN_WIDTH-55,0),wxSize(50,20))
self.findbtn.SetConstraints(LayoutAnchors(self.findbtn,false,true,true,false))
EVT_BUTTON(self, ID_FIND, self.OnFind)
# tabs
self.nb = wxNotebook(self, ID_NOTE,pos=wxPoint(0,22),size=wxSize(WIN_WIDTH,WIN_HEIGHT-50))
self.nb.SetConstraints(LayoutAnchors(self.nb,true,true,true,true))
# panes
# HELP
self.helppage = pmHtmlWindow(self.nb,wxNewId())
self.helppage.SetConstraints(LayoutAnchors(self.helppage,true,true,true,true))
helptext = """\n""" + pmHelpText
helptext = helptext + 'Fields
\n'
keys = med2bib.keys()
keys.sort()
for i in range(len(keys)):
key = keys[i]
if i % 2 == 0:
rows = ""
rowe = "
"
else:
rows = ""
rowe = ""
helptext = helptext+"""%s%s | %s | %s\n
""" % (rows,key,med2bib[key],rowe)
helptext = helptext + '
'
self.helppage.SetPage(helptext)
self.nb.AddPage(self.helppage,"Help")
# SEARCH RESULT
self.searchresult = pmPanel(self.nb,ID_SEARCH_PANE,"SearchResult",
pos= wxPoint(0,0),size=self.nb.GetSize())
self.searchresult.SetConstraints(LayoutAnchors(self.searchresult,true,true,true,true))
self.nb.AddPage(self.searchresult,"SearchResult")
# menu
menu = wxMenu()
menu.Append(ID_ABOUT,"&About","More info about this program")
menu.AppendSeparator()
menu.Append(ID_NEW, "&New\tAlt-N","Make New .bib file")
menu.Append(ID_OPEN,"&Open\tAlt-O","Open .bib file")
menu.Append(ID_CLOSEFILE, "&Close\tAlt-W", "Close file")
menu.Append(ID_EXIT,"E&xit\tAlt-X","Terminate this program")
menubar = wxMenuBar()
menubar.Append(menu,"&File")
self.SetMenuBar(menubar)
EVT_MENU(self, ID_ABOUT, self.OnAbout)
EVT_MENU(self, ID_EXIT, self.OnExit)
EVT_MENU(self, ID_NEW, self.OnNew)
EVT_MENU(self, ID_OPEN, self.OnOpen)
EVT_MENU(self, ID_CLOSEFILE, self.OnCloseFile)
def OnEvtChar(self, event):
if event.GetKeyCode() == WXK_RETURN:
self.OnFind(event)
else:
event.Skip()
def OnAbout(self, event):
msg = """This program retrieves reference records
from PubMed and export them as bibtex records"""
dlg = wxMessageDialog(self, msg, "About this program",
wxOK|wxICON_INFORMATION)
dlg.ShowModal()
dlg.Destroy()
def OnExit(self, event):
self.Close(true)
def OnNew(self, event):
dlg = wxFileDialog(self, "Save as", ".", "", "*.bib", wxSAVE)
if dlg.ShowModal() == wxID_OK:
fname = dlg.GetFilename()
fpath = os.path.join(dlg.GetDirectory(),fname)
open(fpath,'w').write(TEMPLATE_NEW)
self.open_bibtex(fpath,fname)
dlg.Destroy()
def OnOpen(self, event):
dlg = wxFileDialog(self,"Choose a file", self.dirname,"","*.bib",wxOPEN)
if dlg.ShowModal() == wxID_OK:
fname = dlg.GetFilename()
fpath = os.path.join(dlg.GetDirectory(),fname)
self.open_bibtex(fpath,fname)
dlg.Destroy()
def open_bibtex(self, fpath, fname):
bibs = parse_bibtex(open(fpath,'r').readlines())
try:
idx = self.files.index(fpath)
self.panes[idx].set_bibrec(bibs)
self.nb.SetSelection(idx+2)
except:
pane = pmPanel(self.nb,wxNewId(),fname,wxPoint(0,0),self.nb.GetSize())
pane.SetConstraints(LayoutAnchors(pane,true,true,true,true))
pane.set_bibrec(bibs)
self.files.append(fpath)
self.panes.append(pane)
self.nb.AddPage(pane,fname)
self.nb.SetSelection(self.nb.GetPageCount()-1)
def OnCloseFile(self, event):
idx = self.nb.GetSelection()
if idx < 2:
return 0
pane = self.nb.GetPage(idx)
self.nb.RemovePage(idx)
pane.Destroy()
del self.files[idx-2]
del self.panes[idx-2]
def OnFind(self, event):
term = self.term.GetValue()
if term == '':
return 0
bibs = get_bibrec(term,MAXRETRIEVE)
self.searchresult.set_bibrec(bibs)
self.nb.SetSelection(1)
self.SetStatusText("%d hit(s)" % (len(bibs),))
# hit = get_num(term)
# msg = "%d hit(s): Retrieve or Cancel?" % hit
# dlg = wxMessageDialog(self, msg, "hits", wxOK|wxCANCEL|wxICON_QUESTION)
# if dlg.ShowModal() == wxID_OK:# retrieve
# bibs = get_bibrec(term, hit)
# self.searchresult.set_bibrec(bibs)
# self.nb.SetSelection(1)
# dlg.Destroy()
class pmHtmlWindow(wxHtmlWindow):
def __init__(self, parent, id):
wxHtmlWindow.__init__(self, parent, id)
def OnLinkClicked(self, linkinfo):
link = string.lower(linkinfo.GetHref())
#print linkinfo.GetHref(),link
cmd = BROWSER_CMD % link
os.system(cmd)
class pmPanel(wxSplitterWindow):
def __init__(self,parent,ID,title,pos,size):
wxSplitterWindow.__init__(self,parent,ID,pos,size,wxSP_3D,title)
self.bibs = None
#self.text = wxTextCtrl(self,1,style=wxTE_MULTILINE)
self.text = pmHtmlWindow(self,wxNewId())
listid = wxNewId()
self.list = wxListCtrl(self,listid,style=wxLC_REPORT|wxSUNKEN_BORDER)
self.list.InsertColumn(0,"#")
self.list.InsertColumn(1,"Year")
self.list.InsertColumn(2,"Title")
self.list.InsertColumn(3,"Author")
self.list.InsertColumn(4,"Journal")
self.list.SetColumnWidth(0,COL_NUM_W)
self.list.SetColumnWidth(1,COL_YEAR_W)
self.list.SetColumnWidth(2,COL_TITLE_W)
self.list.SetColumnWidth(3,COL_AUTHOR_W)
self.list.SetColumnWidth(4,COL_JOURNAL_W)
self.list.SetToolTip(wxToolTip(
"Shift/Ctrl+Click:multiple selection Right+Click:append to bib file"))
self.SetMinimumPaneSize(100)
self.SplitHorizontally(self.list,self.text)
self.SetSashPosition(200)
EVT_LIST_COL_CLICK(self,listid, self.OnColClick)
EVT_LIST_ITEM_SELECTED(self, listid, self.OnItemSelected)
EVT_RIGHT_DOWN(self.list, self.OnRightDown)
def OnColClick(self, event):
self.col = event.m_col
self.list.SortItems(self.ColumnSorter)
def ColumnSorter(self, i1, i2):
if self.col == 0:
return i1 - i2
key = ['year','title','author','journal'][self.col-1]
item1 = self.bibs[i1][key]
item2 = self.bibs[i2][key]
if item1 == item2: return 0
elif item1 < item2: return -1
else: return 1
def set_bibrec(self, bibs):
self.bibs = bibs
#print "# hits = ", len(bibs)
self.list.DeleteAllItems()
for i in range(len(bibs)):
self.list.InsertStringItem(i,"%d" % i)
if bibs[i].has_key('year'):
self.list.SetStringItem(i,1,bibs[i]['year'])
if bibs[i].has_key('title'):
self.list.SetStringItem(i,2,bibs[i]['title'])
if bibs[i].has_key('author'):
self.list.SetStringItem(i,3,bibs[i]['author'])
if bibs[i].has_key('journal'):
self.list.SetStringItem(i,4,bibs[i]['journal'])
self.list.SetItemData(i,i)
#print '#',i, bibs[i]['year'], bibs[i]['title'],bibs[i]['journal']
def append_bibrec(self,bibs):
num = self.list.GetItemCount()
self.bibs.extend(bibs)
for i in range(len(bibs)):
self.list.InsertStringItem(i+num,"%d" % (i+num))
if bibs[i].has_key('year'):
self.list.SetStringItem(i+num,1,bibs[i]['year'])
if bibs[i].has_key('title'):
self.list.SetStringItem(i+num,2,bibs[i]['title'])
if bibs[i].has_key('author'):
self.list.SetStringItem(i+num,3,bibs[i]['author'])
if bibs[i].has_key('journal'):
self.list.SetStringItem(i+num,4,bibs[i]['journal'])
self.list.SetItemData(i,i)
self.list.SetItemState(i+num, wxLIST_STATE_SELECTED, wxLIST_STATE_SELECTED)
self.list.EnsureVisible(num+len(bibs)-1)
def OnItemSelected(self, event):
self.currentItem = event.m_itemIndex
#print "selected:",self.currentItem
if self.bibs != None:
#bibrec = bibtex_record(self.bibs[self.currentItem])
bibrec = self.bibs[self.currentItem]
html = bibrec_to_html(bibrec)
#self.text.SetValue(bibrec)
self.text.SetPage(html)
def OnRightDown(self, event):
sel = self.get_selection()
if len(sel)==0:
return 0
menu = wxMenu()
global app
frame = app.frame
nb = frame.nb
me = nb.GetSelection() -2
if nb.GetPageCount() == 2: return 0
if (me > -1) and (nb.GetPageCount()==3): return 0
for i in range(nb.GetPageCount() - 2):
if me == i: continue
menu.Append(ID_OFFSET+i,"append to:"+frame.files[i])
EVT_MENU(self,ID_OFFSET+i,self.OnAppendTo)
self.PopupMenu(menu, wxPoint(event.GetX(),event.GetY()))
menu.Destroy()
#event.Skip()
def get_bibrec(self, i):
return self.bibs[i]
def OnAppendTo(self, event):
idx = event.GetId()-ID_OFFSET
global app
frame = app.frame
nb = frame.nb
target = frame.files[idx]
bibs = self.bibs
sel = map(self.get_bibrec, self.get_selection())
if len(sel)==0:return 0
# append to file
bibtex = string.join(map(bibrec_to_bibtex,sel),'\n')
open(target,'a').write(bibtex)
# append to display
pane = frame.panes[idx]
pane.append_bibrec(sel)
nb.SetSelection(idx+2)
def get_selection(self):
sel = []
item = -1
while 1:
item = self.list.GetNextItem(item,
wxLIST_NEXT_ALL,
wxLIST_STATE_SELECTED);
if ( item == -1 ):
break
sel.append(item)
return sel
def parse_bibtex(lines,head=re.compile(r'^@([A-Za-z ]+){(.+),(.*)$'),
fieldhead=re.compile(r'^([^=]+)=\s*(["{]*)(.*)$'),
endbrace=re.compile(r'^\s*}\s*$')):
bibs = []
curbib = None
curfld = None
curdata = ''
for line in lines:
# ignore empty line, brace, comment
if line == '':
continue
if endbrace.match(line):
continue
if line[0] in ['#', '%']:
continue
# new record?
m = head.match(line)
if m:
# save previous
if curbib != None:
if curfld != None:
curbib[curfld] = balance(curdata)
bibs.append(curbib)
# setup new
curbib={}
curbib['type'] = string.strip(m.group(1))
curbib['label'] = string.strip(m.group(2))
curdata = ''
curfld = None
continue
# new field?
m = fieldhead.match(line)
if m and curbib:
if (m.group(2) != None) or (len(string.split(m.group(3),' ')) ==1):
# save previous field
if curfld != None:
curbib[curfld] = balance(curdata)
# setup new
curfld = string.strip(m.group(1))
curdata = m.group(3)
continue
# body of field
if curfld != None:
curdata = curdata + ' '+ string.strip(line)
# save last entry
if curbib != None:
if curfld != None:
curbib[curfld] = balance(curdata)
bibs.append(curbib)
return bibs
def balance(text):
text = trim_trailing_camma(text)
text = balance_braces(text)
text = balance_quotes(text)
return text
def balance_braces(text):
if len(text)==0:
return text
left = string.count(text,"{")
right = string.count(text,"}")
str = text
if left > right:
str = text + "}"*(left-right)
if right > left:
str = "{"*(right-left)+text
if str[0] == '{' and str[-1]=='}':
str =string.strip(str[1:-1])
return str
def balance_quotes(text):
if text=='':
return ''
if text[0] == '"' and text[-1] != '"':
return text+'"'
if text[-1] == '"' and text[0] != '"':
return '"'+text
return text
def trim_trailing_camma(text):
text = string.strip(text)
if text == '':
return ''
if text[-1] == ',':
return text[:-1]
return text
#dbs = ['PubMed','Genome','Nucleotide','OMIM','PopSet','Protein','Structure']
#formats = ['Summary','Brief','Abstract','Citation','MEDLINE','ASN.1','XML','Text']
#cmds = ['Search','Retrieve','Link','Text']
med2bib = {
'AB' : 'abstract',
'AD' : 'address',
'AID': 'ariticle_id',
'AU' : 'author',
'CI' : 'copyright',
'CIN': 'comment_in',
'CN' : 'collective_name',
'CON': 'comment_on',
'CY' : 'country',
'DA' : 'date_created',
'DCOM': 'date_completed',
'DEP' : 'date_of_electronic_publication',
'DP' : 'date',
'EDAT': 'entrez_date',
'EIN': 'erratum_in',
'GS' : 'gene_symbol',
'ID' : 'id',
'IP' : 'number',
'IS' : 'ISSN',
'JC' : 'journal_code',
'JID': 'journal_id',
'LA' : 'language',
'LR' : 'last_revision_date',
'MH' : 'keywords',
'MHDA': 'MeSH_date',
'PG' : 'pages',
'PHST': 'publication_history',
'PMID': 'PMID',
'PS' : 'personal_name_as_subject',
'PST': 'publication_status',
'PT' : 'publication_type',
'RF' : 'number_of_ref',
'RIN': 'retraction_in',
'RN' : 'EC_RN_number',
'ROF': 'retraction_of',
'RPF': 'republished_from',
'RPI': 'republished_in',
'SB' : 'journal_subset',
'SI' : 'secondary_source_id',
'SO' : 'source',
'TA' : 'journal',
'TI' : 'title',
'TT' : 'transliterated_title',
'UI' : 'medlineref',
'UIN': 'updated_in',
'UOF': 'update_of',
'URL': 'url',
'URLF': 'url_fulltext',
'URLS': 'url_summary',
'VI' : 'volume',
}
def get_url(term,base = "http://www.ncbi.nlm.nih.gov/entrez/query.fcgi",
cmd='Search',db='PubMed',format='MEDLINE',maxnum=1):
term = urllib.quote_plus(string.upper(term))
url = """%s?cmd=%s&db=%s&term=%s&doptcmdl=%s&dispmax=%d
""" % (base,cmd,db,term,format,maxnum)
return url
def get_num(term):
numSt = 'Items 1-1 of'
numEd = ' | '
result = urllib.urlretrieve(get_url(term,format='Summary'))
try:
html = open(result[0]).read()
st = string.index(html,numSt)
st = st + len(numSt)
ed = string.index(html,numEd,st+1)
except:
return 0
urllib.urlcleanup()
return string.atoi(html[st:ed])
def get_medline(term,maxnum=100):
""" returns list of MEDLINE records"""
result = urllib.urlretrieve(get_url(term,maxnum=maxnum))
html = open(result[0]).read()
tmplist = string.split(html,"")[1:]
records = map(lambda x:string.split(x,"")[0], tmplist)
return map(parse_medline,records)
def get_bibtex(term,maxnum=100):
bibtex = map(bibrec_to_bibtex, get_bibrec(term,maxnum))
return string.join(bibtex,'\n')
def get_bibrec(term, maxnum=100):
return map(medline_to_bibrec,get_medline(term,maxnum))
def parse_medline(text,header = re.compile ('^(\w\w[\w ][\w ])- (.*)$'),
body = re.compile ('^ (.*)$')):
lines = string.split(text,'\n')
rec = {} # keeps paresed
key = None
data = ''
for line in lines:
m = header.match(line)
if m:
if key: # save previous
if rec.has_key(key):
rec[key].append(data)
else:
rec[key] = [data]
key = string.strip(m.group(1))
data = string.rstrip(m.group(2))
else:
m = body.match(line)
if m:
data = data + ' '+string.rstrip(m.group(1))
if key: # save last
if rec.has_key(key):
rec[key].append(data)
else:
rec[key] = [data]
return rec
def reverse_author(author):
tmp = string.split(author,' ')
tmp.reverse()
return string.join(tmp,' ')
def medline_to_bibrec(rec):
keys = rec.keys()
bib = {}
# type
bib['type'] = 'article'
# label
year = 'unknown'
if rec.has_key('DP'):
date = rec['DP'][0]
tmp = string.split(date,' ')
if len(tmp)==2:
year,month = tmp
else:
year = tmp[0]
month = ''
elif rec.has_key('EDAT'):
year = rec['EDAT'][0][:4]
bib['year'] = year
if rec.has_key('PMID'):
id = rec['PMID'][0]
else:
id = rec['UI'][0]
label = string.join(string.split(rec['AU'][0],' '),'')+year[-2:]+':'+id
bib['label'] = label
# author
au = rec['AU']
au = map(reverse_author,au)
bib['author'] = '{'+string.join(au,'} and {')+'}'
keys.remove('AU')
# title
bib['title'] = rec['TI'][0]
keys.remove('TI')
# journal
bib['journal'] = rec['TA'][0]
keys.remove('TA')
# year
bib['year'] = year
# the rest
for key in keys:
if med2bib.has_key(key):
bib[med2bib[key]] = rec[key][0]
else:
bib['medline_'+key] = rec[key][0]
return bib
def bibrec_to_html(bibrec):
html = "\n"
keys = bibrec.keys()
ths = "| "
the = " | "
first = ['label','year','author','title','abstract']
urls = ['url','url_fulltext','url_summary']
for key in first:
try:
html = html+"%s%s%s| %s |
\n" % (ths,key,the,bibrec[key])
keys.remove(key)
except:
pass
for key in urls:
try:
html=html+"""%s%s%s| %s |
\n
""" % (ths,key,the,bibrec[key],bibrec[key])
keys.remove(key)
except:
pass
# the rest
for key in keys:
html = html + """%s%s%s| %s |
\n""" % (ths,key,the, bibrec[key])
html = html + "
"
return html
def bibrec_to_bibtex(bib):
keys = bib.keys()
str = '@'+bib['type']+'{'+bib['label']+',\n'
keys.remove('type')
keys.remove('label')
for key in ['author','title','journal','year']:
str = str+key+' = {'+bib[key]+'},\n'
keys.remove(key)
for key in keys:
str = str+key+' = {'+bib[key]+'},\n'
str = str + '}\n\n'
return str
if __name__ == '__main__':
app = pmApp(0)
app.MainLoop()
# import sys
# if len(sys.argv) == 3:
# switch,term = sys.argv[1:3]
# num = 100
# elif len(sys.argv) == 4:
# switch,term,num = sys.argv[1:4]
# else:
# sys.stderr.write( "usage: pubmed.py -n|q term [maxnum]\n")
# sys.exit(0)
# if switch == '-n':
# num = get_num(term)
# sys.stdout.write("%d\n" % num)
# sys.exit(0)
# elif switch == '-q':
# bib = get_bibtex(term,num)
# sys.stdout.write(bib)
# sys.exit(0)
# else:
# sys.stderr.write("usage: pubmed.py -n|q term [maxnum]\n")
# sys.exit(0)