-
hook54321
-
arkiverouch...
-
kiskaTime to contemplate a project :D
-
OrIdow6So I intend to finish the Niconino (removing some data March 1) script within the next hour or so
-
OrIdow6But this will have to wait for ark iver for dictionary setup, because it's mostly text data
-
OrIdow6I have not done multi-item, because it's a puzzle to write those, and instead am going with ranges of IDs
-
purplebotFotolog.com edited by Hook54321 (-13, shut down a couple of years ago) just now -- archiveteam.org/?diff=46345&oldid=29404
-
OrIdow6arkiver: Do you know if there's a way for wget-lua to keep the headers you give it after the first request?
-
OrIdow6My goal here is Content-Type: text/plain, as the main Niconino comments API endpoint doesn't take application/x-www-form-urlencoded
-
OrIdow6And optionally (though it looks like this can also be controlled by a cookie on this specific site) accept-language
-
OrIdow6If you don't know of a way, I can use the trick from Mercurial
-
OrIdow6Anyhow, I've done this for long enough today, I keep trying to write Python instead of Lua
-
JAAOrIdow6: I'm pretty sure the --header options are supposed to be applied to all requests. That's why the Mercurial project needed that awkward workaround.
-
OrIdow6JAA: Something in the Lua script seems to be stripping headers after the first request
-
OrIdow6It looks like it happens whenever wget.callbacks.get_urls is defined
-
OrIdow6And the issue (at least as I understood it) with Mercurial was that it needed to change headers in the middle of the script's execution
-
OrIdow6Which is different from what needs to happen here
-
JAAOrIdow6: Huh, if that's true, then that'd be a bug in the Lua hook implementation I think. It certainly shouldn't remove headers set by wget internally (like Host) or provided to the CLI unless you override them from Lua.
-
JAAThe Mercurial project needed to set headers for particular requests, which wasn't supported by wget-at yet at the time. So that's why it used that hack.
-
purplebot800notes edited by Hook54321 (-116) just now -- archiveteam.org/?diff=46346&oldid=24620
-
fuzzy8021is there an updated dnsmasq available? the one on warcforceone evidently has a vulnerability
-
fuzzy8021dnsmasq docker container (if that wasnt obvious) :)