#!/usr/bin/env python
# rewritten from O(n^2) in bash, to O(n) in python, Thu Mar 24, 2005
import shutil
import time
import os
import sys
import string
import popen2
sys.stderr.write('Running web-extract at %s\n' % time.ctime())
# to use this script:
# you'll need par (which is part of the prc.tgz package) and txt2html
# (which is a nice python script for an obvious purpose). And of
# course, we're assuming you're sync'ing your palm via jpilot from time
# to time at least.
# the following patch to txt2html should make it better suited for
# converting palm memos; after this patch, the first line of the memo
# will be used as the title, rather than the filename:
#--- txt2html.t 2004-10-28 15:58:34.000000000 -0700
#+++ txt2html 2004-10-28 15:59:12.000000000 -0700
#@@ -34,8 +34,9 @@
# return '
' % (img, img)
#
# def main(filename):
#- title = re.sub('(.*)[.].*$', '\\1', os.path.basename(filename))
#+ #title = re.sub('(.*)[.].*$', '\\1', os.path.basename(filename))
# file = open(filename)
#+ title = file.readline()
#
# prev_blank = 1
# img = None
# the script has a special case as a sort of safety check:
# It won't add a memo whose title is less than 5 characters long.
# This should hopefully prevent the mishap where you accidentally
# add a blank line to the titles list, and suddenly find every last one
# of your memos on the internet. :)
# the script is O(n^2). This isn't a problem until I have a huge number
# of memos though. :)
def fn_from_title(title):
# sed \
# -e 's#[^ /A-Za-z0-9\.=\+-]*##g' \
# -e 's#[ /]#-#g' \
# -e 's/$/.html/')"
fn = ''
for ch in title:
if ch in 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789.=+-':
fn += ch
elif ch in ' /':
fn += '-'
fn += '.html'
#sys.stderr.write('Converted title %s to filename %s\n' % (title, fn))
return fn
def write_memo(title,srcfn,dstfn):
# txt2html "$file" | tr -d '\0' | sed 's/^------*$/
/' | \
# while read line
# do
# # Reformat long lines. This probably should be
# # a command line option, or better, an option specified
# # in the titles file.
# if [ "$(echo $line | wc -c)" -gt 80 ]
# then
# echo "$line" | fmt
# else
# echo "$line"
# fi
# done
# echo "
Back to Dan's palm memos"
sys.stderr.write('Converting %s to %s\n' % (title, dstfn))
file = open(os.path.expanduser('~/public_html/'+dstfn),'w')
pipe = os.popen("txt2html %s | tr -d '\\0' | sed 's/^-----*$/
/'" % srcfn)
while 1:
line = pipe.readline()
if not line:
break
if len(line) > 80:
# reformat the line to fit the screen a bit better
(stdouterr,stdin) = popen2.popen4('fmt')
stdin.write(line)
stdin.close()
while 1:
fmtline = stdouterr.readline()
if not fmtline:
break
file.write(fmtline)
stdin.close()
stdouterr.close()
else:
# this line isn't that long, so just write it without opening a
# pipe to fmt first - which can slow things down a lot
file.write(line)
#echo "
Back to Dan's palm memos"
file.write('
'+"Back to Dan's palm memos")
file.close()
pipe.close()
# make the directory... It probably already exists though
os.system('mkdir -p ~/.jpilot/web-extract/files')
os.chdir(os.path.expanduser('~/.jpilot/web-extract/files'))
# remove previous extracted palm memos
for file in os.listdir('.'):
os.unlink(file)
# extract current palm memos
os.system(os.path.expanduser('~/trees/`/dcs/bin/systype`/bin/par') + ' x ../../Memo32DB.pdb')
# copy the top of the top-level document
shutil.copyfile(os.path.expanduser('~/public_html/theme.html'),os.path.expanduser('~/public_html/palm-memos.html.new'))
file=open(os.path.expanduser('~/.jpilot/web-extract/titles'),'r')
do_titles=file.readlines()
file.close()
for title_no in range(len(do_titles)):
do_titles[title_no] = string.strip(do_titles[title_no])
do_titles.sort()
class title_class:
def __init__(self,title,filename):
self.title = title
self.filename = filename
def title(self):
return self.title
def filename(self):
return self.filename
def __cmp__(self,other):
if self.title < other.title:
return -1
elif self.title > other.title:
return 1
else:
return 0
have_titles=[]
for filename in os.listdir('.'):
file=open(os.path.expanduser(filename))
# the title is always the first line
have_titles.append(title_class(string.strip(file.readline()),filename))
file.close()
have_titles.sort()
memoshtml = open(os.path.expanduser('~/public_html/palm-memos.html.new'),'a')
len_do = len(do_titles)
len_have = len(have_titles)
do_no = 0
have_no = 0
while do_no < len_do and have_no < len_have:
#sys.stderr.write('Comparing names %s and %s\n' % (do_titles[do_no], have_titles[have_no].title))
if len(do_titles[do_no]) < 5:
# if "do title" has an extremely short length, then ignore it.
# Otherwise, we might typo in the list of titles, and
# (practically) everything would match it.
sys.stderr.write('Skipping very short title: %s\n' % do_titles[do_no])
do_no += 1
elif do_titles[do_no] == have_titles[have_no].title:
# titles match - output a memo as HTML!
#sys.stderr.write('Found %s, writing...\n' % do_titles[do_no])
fn = fn_from_title(do_titles[do_no])
memoshtml.write('
%s\n' % (fn, do_titles[do_no]))
write_memo(do_titles[do_no],have_titles[have_no].filename,fn)
do_no += 1
have_no += 1
elif do_titles[do_no] < have_titles[have_no].title:
sys.stderr.write('Warning: title %s not found!!!!!!!!!!!!!!!!!!!!!!!\n' % string.strip(do_titles[do_no]))
do_no += 1
else: # do > have
have_no += 1
while do_no < len_do:
sys.stderr.write('Warning: title %s not found!!!!!!!!!!!!!!!!!!!!!!!\n' % string.strip(do_titles[do_no]))
do_no += 1
# we don't really need to cover this case - if we have titles that
# aren't in do_titles, we can just ignore them
#while have_no < len_have:
# sys.stderr.write('Work on finishing up have_no\n')
memoshtml.write('\nPages last created at %s\n
\n' % time.ctime())
memoshtml.write('\n')
memoshtml.close()
os.rename(os.path.expanduser('~/public_html/palm-memos.html.new'),os.path.expanduser('~/public_html/palm-memos.html'))
sys.stderr.write('Done with web-extract at %s\n' % time.ctime())