Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon.

Pages: 1-

gelbooru scraper

Name: Anonymous 2012-10-13 12:21

I MADE YOU A PRESENT, /PROG/! I HOPE YOU LIKE IT

(actually I made it for me, then spent two minutes tacking on python2 support so I could publish)

Does what it says on the tin.

https://gist.github.com/3885201

Name: Anonymous 2012-10-13 12:28

PLEASE TELL ME THIS WORKS WITH DANBOORU FAGGOT

Name: Anonymous 2012-10-13 12:33

Gelbooru is ad-ridden low quality crap.

Make a Danbooru version, ``please''.

Name: Anonymous 2012-10-13 12:36

Screw that, where's the pixiv scraper?

Name: Anonymous 2012-10-13 12:40

>>4
pixiv is worse than Gelbooru in quality terms, you dirty weeaboo.

Still, OP, thank you for making this.

Name: OP 2012-10-13 12:52

ugh I will make danbooru version. Check back soon.

Gelbooru is ad-ridden...
lol this doesn't scrape the ads.

Name: Anonymous 2012-10-13 13:01

>>6
of course it doesn't

that doesn't make gelbooru any better though

Name: OP 2012-10-13 13:12

Updated to support danbooru. You need an account there, and their API is hilariously insecure for users, but that's not my problem.

https://gist.github.com/3885201

Name: OP 2012-10-13 13:14

shit uh
forgot to change the number of images per request back to 20 (instead of 2).

Use this one.
https://gist.github.com/3885201

Name: Anonymous 2012-10-13 13:20

this is unpythonic

Name: Anonymous 2012-10-13 13:28

>>8
how is it unsecure? i'm downloading pics of naked touhous, how can that get me hacked?

Name: Anonymous 2012-10-13 13:30

Downloading a bunch of shitty pictures automatically instead of selecting the good ones manually?

Where's your autism?

Name: Anonymous 2012-10-13 13:31

>>10
sanitize looks pythonic as fuck

Name: Anonymous 2012-10-13 13:55

>>10
fuck you, 'Now is better than never.' is pythonic.

>>11
It authenticates you using a hash of your password, which completely defeats the purpose of hashing passwords in the first place. Anyone evesdropping your connection can authenticate as you. Not like it really matters, of course.

As I recall, Windows (at least pre-Vista) did something similar for LAN filesharing authentication, which meant you didn't even need to decrypt the password hash to authenticate. Of course, these days reversing NTLM hashes is trivial to the point of being laughable, but it's still a stupid design decision.

>>12
Easier to sort them after they're downloaded. Also, your autism allows you not to archive everything regardless of quality? That must be nice for you.

Name: Anonymous 2012-10-13 14:56

Awesomeee! I'm going to go scrape the toddlercon guro tag right now!

Name: Anonymous 2012-10-13 15:16

>>14
I thought autists were picky as hell.

Name: Anonymous 2012-10-13 16:36

>>1
Someone is bad with wget, cURL, and fetch under regular expression and AWK.

Name: Anonymous 2012-10-13 16:43

>>1,8,9
# can we all move on from python2, please?
Fully agreed.

Name: Anonymous 2012-10-13 16:54

>>14
It authenticates you using a hash of your password, which completely defeats the purpose of hashing passwords in the first place.
It doesn't. The purpose is to protect users reusing the same password for different applications, over an insecure connection that may include passive eavesdroppers. If the eavesdropper intercepts the hash, he can use it to log in to Danbooru, but it's useless for anything else.
It's nowhere near as good as using TLS, but if that's not an option, it's better than nothing. If you have active men in the middle, of course, it's completely meaningless.

The reason everyone insists that passwords need to be hashed for storage isn't primarily because accounts of the application using the database could be compromised if the database becomes public information; if that happens, they almost certainly already are.

Name: Anonymous 2012-10-13 16:55

the wheel already exists: wget. but if this is just to practice then okay.

Name: Anonymous 2012-10-13 23:18

someone download all the jewkari pictures for me

Name: Anonymous 2012-10-13 23:39

dubs scraper

Name: Anonymous 2012-10-14 19:53

>>22 nice dubs

Don't change these.
Name: Email:
Entire Thread Thread List