Ress's questions

So at the moment I have a virtual host configured on my Oracle HTTP Server instance with a ProxyPass as such:

ProxyPass ^/test/home/ https://example.com/
ProxyPassMatch ^/test/home/(.*)$ https://example.com/$1
ProxyPassReverse ^/test/home/(.*)$ https://example.com/$1

When I attempt to access https://mywebsite.com/test/home/<url_from_other_server> the request seems to be working as expected. However, when I attempt to access https://mywebsite.com/test/home/ it is not proxying me to https://example.com/ but rather returning a 404.

The ProxyPassMatch wildcard seems to be working for all suburls I attempt to access, but the regular ProxyPass keyword is not.

I have also attempted removing the ProxyPass entirely and I get the same 404 error when attempting to access /test/home/

Does anyone have any idea what could be causing this strange behaviour?

Thank you.

I've been researching this for a while and it seems that installing an Oracle JDK rather than OpenJDK through Ansible is much more troublesome due to licensing from Oracle.
I've found some community solutions on Ansible Galaxy, namely these two:

• azavea.ansible-java which isn't working due to a deprecated PPA repository the Ansible roles are using

• ansiblebit.oracle-java which seems promising, however this does require an oracle_java_apt_repository to be set, which we currently do not have.

It needs to be Oracle JDK specifically unfortunately, so there is no work around to changing to OpenJDK.
Is there any other way to install Oracle Java using Ansible, or at least be able to acquire a PPA to install the JDK from?

Thanks.

I'm using grep -ril to be able to see all files that contain a certain string:

[user@machine]$ grep -ril "test string"
file1.log
file2.log
file3.log
file4.log
...

I then would like to zip all of the files that have been returned by that command. Is this possible?

I've tried things such as grep -ril myPattern | zip files.zip and grep -ril myPattern | zip > files.zip but nothing has worked so far.

Is there a workaround to this?

Thank you.

I have this task in my playbook:

- name: Update instance tags
    command: oci compute instance update -c {{ compartment }} --freeform-tags {{ tag_var_json }}

According to the oracle documentation for this command, the parameter --freeform-tags accepts a json that represents the key-value pair for the tag. I need to have this json be created dynamically in the playbook itself, so prior to running that task, I have this one for testing purposes:

  - name: Create a json object to use as tag
    set_fact:
      tag_var: '{ "test": "thisisatest" }'
    set_fact:
      tag_var_json: "{{ tag_var | to_json }}"

But I must be doing something wrong because I keep getting this error:
fatal: [localhost]: FAILED! => {"msg": "The task includes an option with an undefined variable. The error was: 'tag_var' is undefined

Is there an easier way of creating a json directly in the playbook and passing it as an argument to that parameter?

Thank you.

I have the following:

- set_fact:
   test_string: "{{ htmlres.content | regex_search('test-([0-9]+)', '\\1') | first}}"

This retrieves 2 which is part of a string contained in the htmlres.content content which is test-2.

So now I'm trying to compare the result of that output and fail the execution if it is not 2, so I tried this:

- name: Fail if test_string is not 2
  fail: msg="Incorrect string. Expected 2, but instead got {{ test_string }}"
  when: test_string != 2

However I've outputted the contents of test_string and I know for a fact that it is 2. Why is it failing?

I've tried adding | string and | int to the end of test_string since my first though is that it would be an issue with type comparison, but that didn't work either.

Thank you.

I have the following

- set_fact:
   test_result: " {{ htmlres.content | regex_search('http://website([0-9]+)', '\\1') }}"

Using debug, this returns the following " '[01]'"

Wanting only the number, I did some experimenting using the replace() function and was able to strip the [ ] by adding the following:

- set_fact:
   test_result: " {{ htmlres.content | regex_search('http://website([0-9]+)', '\\1')
| replace('[','') | replace(']','') }}"

My problem now is that the output is now " '01' and I can't seem to remove the ' or the whitespace.

Adding | trim to the end for some reason doesn't remove the whitespaces, and adding regex_search('\'','') also doesn't seem to escape the character and work.

Am I missing something?

Here's the output from the debug after the first removal:

"msg": [
    " '01'",
    ...

Thanks

So I have an html file that contains the following somewhere in the middle:

<span dir="ltr">http:(...).com</span>

I'm attempting to extract the url, but am having some issues doing so. Because that "ltr" is the only one that exists in the html, I came up with this regex:

(?<=ltr">)(.*)(?=<\/span>)

Using regex101 I confirmed that the regex expression works. However, because of how ansible deals with quotes and double quotes, I think it may be causing some issues.

I'm trying it like this:

    - set_fact:
       regex_test: " {{ htmlres.content | regex_search('(?<=ltr">)(.*)(?=<\/span>)') }}"  

Where htmlres.content is the html content received from an http get request done previously in the same playbook. However, running it:

    - set_fact:
       regex_pubdest: " {{ htmlres.content | regex_search('(?<=ltr">)(.*)(?=<\/span>)' }}"
                                                                    ^ here

Is there any way to circumvent this issue with quotes in regex in ansible? I've managed to achieve the desired output by doing something slightly different, which is this:

 shell:  grep -oP 'ltr">\K.*?(?=</span>)' /dir/htmlcontent.txt

The issue is the previous only works when reading from a file, and I'm trying to avoid saving the html.content to a file before passing a regex through it. I've tried replacing the path to the folder in the grep with "{{html.content}}", but unfortunately that causes ansible to not run correctly due to the quotes.

Any ideas?

Thank you!