Match sets of keywords in freeform text

I could guarantee that the Record features arrive before the text features.

If the datasets are not too big, you could use an unconditional FeatureMerger (join on 1=1) to create a Cartesian join.

Then follow by a Tester to test on:

Text contains ID And Text contains Name ......

That would solve it using 2 transformers.

What format is your record dataset in? If it is stored as a csv/text or similar I'd be tempted to read the csv directly to a list of tuples and iterate over them for matches, e.g.

import fme
import fmeobjects
import csv

class FeatureProcessor(object):
    
    
    def __init__(self):
        self.inputfilename = FME_MacroValues['SourceDataset_CSV2']
        with open(self.inputfilename) as f:
            self.data=[tuple(line) for line in csv.reader(f)]

        
    def input(self,feature):
        text = feature.getAttribute('text')
        for y in self.data:
            value = 0
            for x in y:
                if x in text:
                    value +=1
            if value ==4:
                feature.setAttribute('value',value)
                feature.setAttribute('record',','.join(y))
                feature.setAttribute('ID',y[0])
                feature.setAttribute('Name',y[1])
                feature.setAttribute('Date',y[2])
                feature.setAttribute('Location',y[3])
                self.pyoutput(feature)
                break
                    

    def close(self):
        pass

I would say that it averages to about 2000 records, and there are 8 attributes to check.

With 2000 records, I would give this a try.

Assuming you use FME 2019, which is a lot faster and better with memory.

And in FME 2019 using the FeatureJoiner instead of FeatureMerger

A Feature Joiner with 2000 features on an unconditional merge would output 4 million features, all of which would have to be tested. Also how do you differentiate a record that has no matching text, from a record that just doesn't match that particular text?

import fme
import fmeobjects
import csv

class FeatureProcessor(object):
    
    
    def __init__(self):
        self.inputfilename = FME_MacroValues['SourceDataset_CSV2']
        with open(self.inputfilename) as f:
            self.data=[tuple(line) for line in csv.reader(f)]

        
    def input(self,feature):
        text = feature.getAttribute('text')
        for y in self.data:
            value = 0
            for x in y:
                if x in text:
                    value +=1
            if value ==4:
                feature.setAttribute('value',value)
                feature.setAttribute('record',','.join(y))
                feature.setAttribute('ID',y[0])
                feature.setAttribute('Name',y[1])
                feature.setAttribute('Date',y[2])
                feature.setAttribute('Location',y[3])
                self.pyoutput(feature)
                break
                    

    def close(self):
        pass

I like you idea of just tracking how many values passed.