Index | Thread | Search

From:
Klemens Nanni <kn@openbsd.org>
Subject:
Re: security/acme.sh: new port (version 3.0.9)
To:
Landry Breuil <landry@openbsd.org>, OpenBSD ports <ports@openbsd.org>
Date:
Sun, 17 Nov 2024 19:43:27 +0000

Download raw body.

Thread
17.11.2024 22:35, Kirill A. Korinsky пишет:
> On Sun, 17 Nov 2024 20:26:40 +0100,
> Klemens Nanni <kn@openbsd.org> wrote:
>>
>> I was aiming for an explanation of the change, something like this.
>> The other patch I haven't found the for yet.
>>
>>
>> Feedback? OK?
>>
>> Index: patches/patch-dnsapi_dns_netcup_sh
>> ===================================================================
>> RCS file: /cvs/ports/security/acme.sh/patches/patch-dnsapi_dns_netcup_sh,v
>> diff -u -p -r1.1.1.1 patch-dnsapi_dns_netcup_sh
>> --- patches/patch-dnsapi_dns_netcup_sh	17 Nov 2024 14:10:23 -0000	1.1.1.1
>> +++ patches/patch-dnsapi_dns_netcup_sh	17 Nov 2024 19:25:00 -0000
>> @@ -1,7 +1,10 @@
>> -Fix syntax error by our sh:
>> +ksh(1) defaults to alias login='exec login' which expands before definition
>> +of the equally named function, i.e. 'exec login() ...' causing
>>  
>>    dnsapi/dns_netcup.sh[128]: syntax error: `(' unexpected
>>  
>> +Rename it to _login() for sh(1) -n to pass.
>> +
>>  Index: dnsapi/dns_netcup.sh
>>  --- dnsapi/dns_netcup.sh.orig
>>  +++ dnsapi/dns_netcup.sh
>> @@ -32,12 +35,3 @@ Index: dnsapi/dns_netcup.sh
>>     tmp=$(_post "{\"action\": \"login\", \"param\": {\"apikey\": \"$NC_Apikey\", \"apipassword\": \"$NC_Apipw\", \"customernumber\": \"$NC_CID\"}}" "$end" "" "POST")
>>     sid=$(echo "$tmp" | tr '{}' '\n' | grep apisessionid | cut -d '"' -f 4)
>>     _debug "$tmp"
>> -@@ -134,7 +134,7 @@ login() {
>> -     return 1
>> -   fi
>> - }
>> --logout() {
>> -+_logout() {
>> -   tmp=$(_post "{\"action\": \"logout\", \"param\": {\"apikey\": \"$NC_Apikey\", \"apisessionid\": \"$sid\", \"customernumber\": \"$NC_CID\"}}" "$end" "" "POST")
>> -   _debug "$tmp"
>> -   if [ "$(_getfield "$tmp" "4" | sed s/\"status\":\"//g | sed s/\"//g)" != "success" ]; then
>>
> 
> I think that this chunk should stay in patch. Am I wrong?

test passes without it, I don't see an existing alias, function or program
this script's logout() function could clash with:

	$ type logout
	logout not found

Plus, you only renamed the function definition, but not the call to it.
So this should actually unbreak the script.