SOLVED: Using TTP to extract an unknow number of words from a line

[Calab]

I'm using Python v 3.10.11 and ttp v 0.9.5. I am trying to build a TTP template that can extract an unknown number of words from a line into a group.

The data could look like this:

Output:
link-aggregate 0         AdminState:Up      OperState:IS          
Member Ports:   6/1 6/5 6/6 6/7 6/8 6/9 6/10 6/11
                6/12 
Description:  << The port description >>

or, the data could like like this:

Output:
link-aggregate 0         AdminState:Up      OperState:IS          
Member Ports:   6/1
Description:  << The port description >>

or anything in between.

The closest template that I can come up with is the following, but it misses the Member Port 6/12 in the first example, and returns the rest as a single string instead of a list of ports:

Output:
link-aggregate {{ interfaceId }}          AdminState:{{ AdminState }}       OperState:{{ OperState }}          
Member Ports: {{ MemberPorts | PHRASE }}
Description:  {{ Description | ORPHRASE }}

How can I get my Member Ports into a list, and include any that may follow on the next line?

[deanhystad]

I think the problem is that TTP uses re, and re does not support patterns with multiple matches. You can write a pattern where a match has multiple groups, but it is still just 1 match.

What's wrong with getting the ports as a string? Use split() to get a list of ports.

[Calab]

After a bit of work I think I have a template that I can use...

Output:
link-aggregate {{ interfaceId }} AdminState:{{ AdminState }} OperState:{{ OperState }}
Member Ports: {{ MemberPorts | ORPHRASE | split() }}
               {{ MemberPorts | ORPHRASE | split() | joinmatches }}
Description: {{ Description | ORPHRASE }}

[deanhystad]

You don't need the joinmatches.

from ttp import ttp


data = """
link-aggregate 0 AdminState:Up OperState:IS
Member Ports: 6/1 6/2 6/3
Description: << The port description >>
"""

template = """
link-aggregate {{ interfaceId }} AdminState:{{ AdminState }} OperState:{{ OperState }}
Member Ports: {{ MemberPorts | PHRASE | split() }}
Description: {{ Description | ORPHRASE }}
"""

parser = ttp(data=data, template=template)
parser.parse()
print(parser.result(format="json")[0])

Output:
[
    {
        "AdminState": "Up",
        "Description": "<< The port description >>",
        "MemberPorts": [
            "6/1",
            "6/2",
            "6/3"
        ],
        "OperState": "IS",
        "interfaceId": "0"
    }
]

[Pedroski55]

If ttp uses re, why not just use re and have more control?

I assume the key words are fixed: keys = ['link-aggregate', 'AdminState', 'OperState', 'Member Ports', 'Description'] If not, find them with a pattern.

import re
import json

data = """
link-aggregate 0 AdminState:Up OperState:IS
Member Ports: 6/1 6/2 6/3 6/4 6/5
6/6 6/7 6/8
7/1 7/2 /7/3
Description: << The port description >>
"""
#keys = ['link-aggregate', 'AdminState', 'OperState', 'Member Ports', 'Description']

p = re.compile(r'(link-aggregate) (\d+)') # get link-aggregate 0
q = re.compile(r'(AdminState):([A-Za-z]+)') # get AdminState Up
r = re.compile(r'(OperState):([A-Za-z]+)') # get OperState IS
s = re.compile(r'(Member Ports):') # get Member Ports
t = re.compile('([0-9]+/[0-9]+)') # get Member Ports values
u = re.compile('(Description:) <<([\s\S]*?)>>') # get Description

def getbits(data):
    d = {}
    resp = p.search(data) # res.group(2) = 0
    d[resp.group(1)] = resp.group(2)
    resq = q.search(data) # res.group(2) = 'Up'
    d[resq.group(1)] = resq.group(2)          
    resr = r.search(data) # res.group(2) = 'IS'
    d[resr.group(1)] = resr.group(2)
    # different because t uses .findall() not .search
    ress = s.search(data) # res.group(1) = 'Member Ports'   
    rest = t.findall(data) # res = ['6/1', '6/2', '6/3', '6/4', '6/5', '6/6', '6/7', '6/8', '7/1', '7/2', '7/3']
    d[ress.group(1)] = rest
    resu = u.search(data) # res.group(2) = ' The port description '
    d[resu.group(1)] = resu.group(2)
    return d

# if you have many sets of data, loop through them
# here there is only 1 data
data_dict = getbits(data)
for item in data_dict.items():
    print(item)

# save as json
mypath = '/home/pedro/temp/myjson.json'    
with open(mypath, 'a') as json_file:
    json.dump(data_dict, json_file)
    print('data saved to', mypath)

data_dict

Output:
for item in data_dict.items():
    print(item)

    
('link-aggregate', '0')
('AdminState', 'Up')
('OperState', 'IS')
('Member Ports', ['6/1', '6/2', '6/3', '6/4', '6/5', '6/6', '6/7', '6/8', '7/1', '7/2', '7/3'])
('Description:', ' The port description ')

I never had much luck with re.VERBOSE and a complex, multiline re expression, so I just keep it simple!

[deanhystad]

Writing new code when you can use a popular, supported package is not pythonic. TTP works great for this problem.