RECENT TOPICS » View all
- « Previous
- 1
- Next »
Let me just bump this one.
I too spend all my commuting time studying Japanese. I was considering writing all my stories up into a notebook so that I can read up on them while sitting on the train. But how good would it be if you could just export them and print them off ?
Amazing site anyway !! Fast becoming one of my favorite websites !! 
Oh genius ! Why didn't I think of that ?
Got it all into Excel with a looping web query in a couple of minutes. Thanks
Here is an export script. I've hard-coded a delay of 60 seconds between requests to prevent hammering the server too hard. Don't abuse this script or our beloved site will die under the stress 
To use it, save the script to a file e.g. export.py, edit the parameters username, password, startingKanji, and endingKanji, and run it with "python export.py". The output is in a comma-separated file "stories.txt" which can probably be imported into any spreadsheet. The file is written incrementally (closed and re-opened) after every kanji story is fetched, so even if the program crashes in the middle of downloading, you should still be able to read all of the stories downloaded so far.
Are there bugs? Probably. Will this script fail if the website changes formatting? Definitely.
Enjoy. Code is public domain.
#!/usr/bin/env python
################## start of user-configurable parameters ######################
username="YOURNAME"
password="YOURPASSWORD"
startingKanji = 205
endingKanji = 210
outputFilename = "stories.txt"
delay=60
################## end of user-configurable parameters #######################
from ClientForm import ParseResponse
from ClientCookie import urlopen
import re
import time
outfile = open(outputFilename,"w")
response = urlopen("http://kanji.koohii.com/login.php")
forms = ParseResponse(response, backwards_compat=False)
form = forms[0]
form["uname"] = username
form["passwd"] = password
urlopen(form.click()).read()
for ikanji in range(205,210):
kanjiurl="http://kanji.koohii.com/study/?mode=failed&framenum=%d" % (ikanji)
rawtext = urlopen(kanjiurl).read()
match_keyword = re.search('(<div class="keyword">)([^<]*)(</div>)',rawtext)
if(match_keyword):
g = match_keyword.groups()
keyword = g[1]
match_story = re.search('(id="frmStory">)(.*)(</textarea>)',rawtext)
if(match_story):
g = match_story.groups()
story = g[1]
outstring = '%d,"%s","%s"' % (ikanji, keyword, story)
outfile.write(outstring+"\n")
outfile.close()
outfile = open(outputFilename,"a")
print outstring
time.sleep(delay)
outfile.close()
*little bump* Hi there! I have tried the Greasemonkey script but it doesn't seem to be working 
I'm using Opera and User-scripts but it's the same principle ... has the coding changed to make it not work, or am I being a n00b? ^_^
Doog
shinkyu wrote:
I have loaded this script too but it doesnt seem to do anything when I click the "export button" - has something changed or am I missing something?
Still works fine here.
Are there any errors in Firefox error console when you click the button?
this_is_douglas wrote:
I have tried the Greasemonkey script but it doesn't seem to be working sad I'm using Opera and User-scripts but it's the same principle ... has the coding changed to make it not work, or am I being a n00b?
It's possible to create scripts that work both on Firefox and Opera, but I was a bit lazy at the time, and coded it specifically to Firefox. Sorry. I may fix it later.
ファブリス wrote:
There ought to be a way to export your stories. I don't want to write this now on the old codebase, so I've put it on the refactoring to-do list as this should be fairly simple to add. The refactored site should be online before the end of April *fingers crossed*.
That would be great.
No more workarounds.
Last edited by iSoron (2009 April 02, 12:08 pm)
- « Previous
- 1
- Next »