find your publications in a bibtex file

June 30th, 2011

starting from a bib file with many entries, among them publications where you are co-author, you would like to extract only these and list them in a tex file and on an html webpage.

you could simply use bibtex2html which does a very nice job but doesn’t allow you to fine-tune the ordering of the results (i think). so the approach i’m presenting here combines this tool with pybtex which i guess would be able to do the whole job by itself.

this script produces tex and html file

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
#!/usr/bin/env python
# Does two tasks:
#
# 1) Generates pubs.html as refereed pubs then non-refereed pubs,
#    sorted reverse-chronologically
# 2) Generates LaTeX-formatted publications, same format
 
from operator import itemgetter
from pybtex.database.input import bibtex
 
bibfile = "/home/buschi/cv/db.bib"
texout = "sbusch_publications.tex"
htmout = "sbusch_publications.html"
 
parser = bibtex.Parser()
bib_data = parser.parse_file(bibfile)
 
# take everything with my name
sbusch_all = {}
for key in bib_data.entries.keys():
    try:
        authors = bib_data.entries[key].persons['author']
        for author in authors:
            try:
                if ((author.first()[0] == u'Sebastian') and (author.last()[0] == u'Busch')):
                    sbusch_all[key] = bib_data.entries[key]
            except IndexError: # no first / last name
                pass
    except KeyError: # no author (e.g. a collection)
        pass
 
# categorise
sbusch_nonref = {}
sbusch_nonref_sort = []
sbusch_proc = {}
sbusch_proc_sort = []
sbusch_ref = {}
sbusch_ref_sort = []
for key in sbusch_all.keys():
    publ = sbusch_all[key]
    year = -int(publ.fields['year'].strip('-')) # strip for 2009--; - to get the ones with largest years first
    for i in range(len(publ.persons['author'])):
        if ((publ.persons['author'][i].first()[0] == 'Sebastian') and (publ.persons['author'][i].last()[0] == 'Busch')):
            mypos = i
            break
    if ((publ.type == "techreport") or (publ.type == "mastersthesis") or (publ.type == "phdthesis") or (publ.type == "misc") or ("nonrefereed" in publ.fields.keys())):
        sbusch_nonref[key] = sbusch_all[key]
        sbusch_nonref_sort.append((key, year, mypos))
    elif (publ.type == "inproceedings"):
        sbusch_proc[key] = sbusch_all[key]
        sbusch_proc_sort.append((key, year, mypos))
    elif (publ.fields['journal'] != "in preparation"):
        try:
            vol = publ.fields['volume']
        except KeyError:
            vol = None
        if (vol != "submitted"):
            sbusch_ref[key] = sbusch_all[key]
            sbusch_ref_sort.append((key, year, mypos))
 
# sort
# the newest publications first
#     the ones where i'm first author first
sbusch_nonref_sorted = [i[0] for i in sorted(sbusch_nonref_sort, key=itemgetter(1,2))]
sbusch_proc_sorted = [i[0] for i in sorted(sbusch_proc_sort, key=itemgetter(1,2))]
sbusch_ref_sorted = [i[0] for i in sorted(sbusch_ref_sort, key=itemgetter(1,2))]
for i in [(sbusch_nonref_sorted, 'nonref.txt'), (sbusch_proc_sorted, 'proc.txt'), (sbusch_ref_sorted, 'ref.txt')]:
    f = open(i[1], 'w')
    for line in i[0]:
        f.write(str(line)+"\n")
    f.close()
 
from os import system, remove
from re import compile, DOTALL
 
pubs_html = ''
 
# iterate over refereed and non-refereed
for o in [ ['Refereed Publications', '--no-footer', 'ref.txt'], ['Proceedings', '--no-footer', 'proc.txt'], ['Non-Refereed Publications', '', 'nonref.txt'] ]:
    # output is pubs.html
    pubs_html += '<h1>%s</h1>' % o[0]
 
    # add non-/refereed pubs to pubs.html
    # sort by reverse-date; don't generate keys; use sbusch_web.bbl
    # writes into sbusch.html
    system("bibtex2html -q -d -r -dl -nobibsource -nokeys -m macros.tex -citefile %s -s sbusch_web -nodoc %s -o auxfile %s" % (o[2], o[1], bibfile))
    try:
        sbusch_html = open('auxfile.html', 'r')
        pubs_html += sbusch_html.read()
        sbusch_html.close()
    except IOError:
        pass
 
# change \"[ bib ]\" into \"[&nbsp;bib&nbsp;]\"
biblinkRE = compile(r'\[ (<a href="[^"]+">bib</a>) ]')
pubs_html = biblinkRE.sub(r'[&nbsp;\1&nbsp;]', pubs_html)
# remove explicit line breaks
deletebrRE = compile(r'<br />')
pubs_html = deletebrRE.sub('', pubs_html)
 
# write into output file pubs.html
pubs_html_file = open(htmout, 'w')
pubs_html_file.write(pubs_html)
pubs_html_file.close()
 
#########
 
# now we're going to generate a LaTeX version of my pubs, also sorted
 
# RE selects bibitems from bbl (key stored in group(1), entry in group(2))
bibitemRE = compile(r'\\bibitem\[\]\{([^\}]+)\}(.*)', DOTALL)
# refs.tex will contain the LaTeX version of my pubs
refs_tex = open(texout, 'w')
 
refs_tex.write("\section{Publications}\n")
 
# separately loop through refereed and non-refereed pubs
for o in [ ['ref.txt', 'Articles in Refereed Scientific Journals'], ['proc.txt', 'Articles in Conference Proceedings'], ['nonref.txt', 'Other'] ]:
  auxfile = open(o[0], 'r')
  sorted_keys = auxfile.read().split('\n')
  auxfile.close()
 
  refs_tex.write('%s\n\\renewcommand\\refname{%s}\n\\begin{bibunit}[unsrt]\n' % ("%", o[1]))
  for key in sorted_keys[:-1]:
      refs_tex.write('\\nocite{'+str(key)+'}\n')
  refs_tex.write('\\putbib[%s]\n\\end{bibunit}\n' % bibfile[:-4])
 
refs_tex.close()
 
# clean up temp files
try:
    remove("auxfile.html")
except OSError:
    pass
try:
    remove("nonref.txt")
except OSError:
    pass
try:
    remove("proc.txt")
except OSError:
    pass
try:
    remove("ref.txt")
except OSError:
    pass

the tex file can then be included in another document and processed with bibtex.

find non-ascii characters

June 10th, 2011

some non-ascii characters in the bibtex file, for example the long hyphen −, result in an error message like


! Package inputenc Error: Unicode char \u8:− not set up for use with LaTeX.

in order to find non-ascii characters in the bibtex file, the following command lists all these characters which can then be searched with a text editor:


tr -d "\000-\011\013-\177" < file.bib | sed '/^$/d'

found on http://www.unix.com/302107579-post5.html

backref with natbib’s compress option

June 9th, 2011

in a latex document where the citations are handled by natbib and where the compress option is active, citations can end up as [1-4]. if the backref option is active, the items in the bibliography will show on which page they are cited. this does not work in the present case though: only references 1 and 4 will show which page they were cited on, 2 and 3 won’t. this can be fixed by changing natbib.sty:

--- natbib.sty.old 2009-07-23 10:44:10.000000000 -0400
+++ natbib.sty  2009-11-01 17:07:53.309765500 -0500
@@ -408,6 +408,7 @@
         \@ifnum{\NAT@nm=\@tempcnta}{%
          \@ifnum{\NAT@merge>\@ne}{}{\NAT@last@yr@mbox}%
         }{%
+           \Hy@backout{\@citeb\@extra@b@citeb}%
           \advance\@tempcnta by\@ne
           \@ifnum{\NAT@nm=\@tempcnta}{%
             \ifx\NAT@last@yr\relax

found on http://tex.stackexchange.com/questions/13653/hyperref-with-the-backref-page-option

find latex symbols easily

June 9th, 2011

more easily than in http://mirrors.ctan.org/info/symbols/comprehensive/symbols-a4.pdf, latex symbols can be found at http://detexify.kirelabs.org/classify.html

bash auto completion

May 31st, 2011

to get the final slash for symbolic links to directories, either press tab twice or add

set mark-symlinked-directories on

to the ~/.inputrc

found on http://superuser.com/questions/271626/bash-autocomplete-on-symlink-to-directory-complete-to-whole-directory-including

silent thunderbird

May 23rd, 2011

thunderbird displays a “sending message” notice which is notoriously in the way. it can be hidden in the config editor by setting the key mailnews.show_send_progress to false.

found on http://www.geoffblog.com/2006/06/hiding-thunderbird-sending-messages.html .

silent ubuntu

May 16th, 2011

ubuntu natty 11.04 announces the login screen always with the annoying drums. they can be disabled with

sudo -u gdm gconftool-2 --type=bool --set /desktop/gnome/sound/event_sounds false

found on http://ubuntuforums.org/showthread.php?t=1756504

bottom panel in unity

May 13th, 2011

in ubuntu natty 11.04 with unity, there is no bottom panel any more which shows the open windows etc. this can be changed following http://askubuntu.com/questions/39069/how-do-i-use-the-gnome-panel-in-unity:

  • Run gnome-panel in the terminal (Don’t close the terminal until finished with the instructions)
  • A Gnome Classic desktop will appear over unity
  • Remove the top panel for we can use the Unity panel
  • Add gnome-panel to Start with the System.

special characters in postscript

May 2nd, 2011

in the postscript files produced by FRIDA, there is a macro to insert special characters:

1
2
3
4
5
6
7
8
/grec {
   showif
   /Symbol findfont fontsize scalefont setfont
   } def
/endgr {
   showif
   regularfont setfont
   } def

so for example a small omega is produced by {() grec (w) endgr ()}.

it can be used to print special characters, using this map
a table of symbols accessable in postscript

take value of column times ten (that’s what the x means) and add the number of the column. example: degree sign: 26*10+0=260.

then use: {() grec (\260) endgr ()}

blog…

April 30th, 2011

i’ve decided to delete the main page and replace it with the blog — it tended to be incomplete, and i found myself not using it but the sidebar and the search, anyway.

i’ll use the opportunity to advertise the free software foundation newsletter. you might have guessed already: it’s free! 😉