added .de region specific support #3
Conversation
|
It seems good to go - the requirement of test.py output is primarily to make sure that people have tested it, which seems to be the case here :) I do have one question before merging, though: is all WHOIS data for the entire I also notice that this doesn't yet parse registrant data for whois.denic.de - while this is of course not required for this pull request, I'll make sure to implement that later (unless you were planning on doing so and submitting a pull request, in which case I'll avoid duplication of effort). |
|
That's a good question actually... I couldn't find any standard for such unusual behaviour to be honest. But since Refererer services like
So yes, there should be check for the two domains pointing to the whois DeNIC service. Although there may be more domains pointing to this very IP, but I guess those are quite exotic ones, right? So a simple check should do it. |
|
...I also added an simple rfc3490 encoding feature (with two tests and documentation). Have a look - although quite uncommon, some domains of local businesses are using it. |
…are different from domain to domain
…ds returned from DeNIC
|
Whoops, sorry for not responding earlier! I've been very busy for the past two weeks... Could you add the test output after your .de parsing changes? I should also add that if you name the matching groups street1, street2 etc. they will be combined into a single newline-delimited address :) |
|
No problem about the delay. You can find an example for the .de parsing in Maybe you like to do an automated testing with travis-ci and use mocked data for simulating the whois services. By doing so every single pull request will be tested and the result is visible within seconds for everybody. Also no changed in whois-records are to break the tests (like it is the case currently). To be honest I find it kind of annoying to copy'n'paste the whole bunch of testing output here 😄 The street-solution is not ideal but resilient. Because I have no idea what the maximum amount of |
|
After some more delays, I've finally gotten around to testing your pull request. It caused an error in Python 3.3 (which is supported since a few days ago after #4, and now tested through Tox), but this just turned out to be spaces as whitespace in your code, where the rest of the code uses tabs - so I've gone ahead and fixed that. I've manually merged your pull request - it was targeted at |
…468-detect-requests-exceeded to develop * commit 'b0f201e9797179e61c31076009d9750d1ee4253e': ENH: Wording in a comment REF: Removed the fix for parsing empty responses and moved it to the whois application that uses this REF: Processed Wytse his comments REF: Renamed whois_response.py to raw_whois_response.py REF: Made timeout an argument, but with a default value REF: Increased default cool down length from 1 second to 2 REF: Extracted the starting of a cool down to a separate method ADD: Can now check whether a rate limit has been exceeded or not FIX: If there are now results at all, parse.py simply returns an empty dictionary instead of crashing ADD: A holder for WHOIS responses. It contains information about the success of the retrieval.
Merge develop into master
Ignore everything but the registrar from the whois data.
You seem to query whois.denic.de for the .de TLD. Unfortunately no data is received by your module as you're lacking the
-T dn,ace -C US-ASCIIstring before the actually domain to query.You can easily check the behaviour through
echo "-T dn,ace -C US-ASCII test.de" | nc whois.denic.de 43andecho "test.de" | nc whois.denic.de 43Added domain
test.deto test-cases as this very domain is unlikely to change over time due to the state-sponsored organisation behind.Test-cases are green and I guess there is no need to print the whole output there considered the case that this change is a quite trivial one. I do hope you agree 😄