trying to obtain number of times a word appears in SharePoint

[PAppsAlways1]

Hi,

I have a flow, which is taking input from PowerApps, via a input text box, allowing a user to search for a word, this is then triggering a flow, which calls a REST Api to SharePoint. I am able to obtain the files inside a document library which contain the searched word in the documents.

I am trying to enhance the flow to be able to obtain how many times a word appears in each document and this is where I need help please, below is my flow, and its failing in different stages:

[Chriddle]

This certainly needs some enhancements, but it might be a start.

1. Remove special characters such as ".", "!", "?" that may immediately follow a word (this must be done with several other characters such as ",", ";", ")" etc.)
2. lowercase the text
3. Split each file content into a word array
4. Count the occurrences of the search term using xpath

Map file:

item()['file']

Map content:

item()['content']

Map words (not really needed, but here you may check if the words are splitted correctly):

split(
	toLower(
		replace(
			replace(
				replace(
					replace(
						item()['content'],
						'.',
						''
					),
					'?',
					''
				),
				':',
				''
			),
			'!',
			''
		)
	),
	' '
)

Map count:

xpath(
	xml(json(concat('{"Root":{"Item":',
		split(
			toLower(
				replace(
					replace(
						replace(
							replace(
								item()['content'],
								'.',
								''
							),
							'?',
							''
						),
						':',
						''
					),
					'!',
					''
				)
			),
			' '
		),
		'}}'))
	),
	concat('count(//Item[.="', toLower(outputs('SearchWord')),'"])')
)

 Result:

Spoiler

[
  {
    "file": "file1",
    "content": "The file content of file 1.",
    "words": [
      "the",
      "file",
      "content",
      "of",
      "file",
      "1"
    ],
    "count": 1
  },
  {
    "file": "file2",
    "content": "Another file content, I's of file 2?",
    "words": [
      "another",
      "file",
      "content,",
      "i's",
      "of",
      "file",
      "2"
    ],
    "count": 0
  },
  {
    "file": "file3",
    "content": "A third file content: File 3 also has content!",
    "words": [
      "a",
      "third",
      "file",
      "content",
      "file",
      "3",
      "also",
      "has",
      "content"
    ],
    "count": 2
  },
  {
    "file": "file4",
    "content": "Content of file 4",
    "words": [
      "content",
      "of",
      "file",
      "4"
    ],
    "count": 1
  }
]

[PAppsAlways1]

Taking it to another level there wow, I'll need steps to see what you have done and used for each expression.

For starters I've never used xpath....

[Chriddle]

If  the search term does not need to match the whole word, you can use this xpath query: 

xpath(
	xml(json(concat('{"Root":{"Item":',
		split(
			toLower(
				item()['content']
			),
			' '
		),
		'}}'))
	),
	concat('count(//Item[contains(.,"', toLower(outputs('SearchWord')),'")])')
)

This saves annoying replaces

[Chriddle]

And if you only expect a few occurrences you can use an expression like this:

if(
	less(
		nthIndexOf(toLower(item()['content']), toLower(outputs('Searchword')), 1),
		0
	),
	0,
	if(
		less(
			nthIndexOf(toLower(item()['content']), toLower(outputs('Searchword')), 2),
			0
		),
		1,
		if(
			less(
				nthIndexOf(toLower(item()['content']), toLower(outputs('Searchword')), 3),
				0
			),
			2,
			if(
				less(
					nthIndexOf(toLower(item()['content']), toLower(outputs('Searchword')), 4),
					0
				),
				3,
				'>3'
			)
		)
	)
)

[PAppsAlways1]

@Chriddle 
There is no way for me to know how many times the word will appear, so I wouldn't want to limit the search 

[Chriddle]

As you can see, this flow expects a Compose action with an array of objects with properties file and content.

So you have to prepare your data 😉

[PAppsAlways1]

tbh, that is where I have been struggling, I cannot get my flow to function after it has got the files back, i've created a 2nd HTTP Get request (Send a HTTP request to SharePoint 2) and after that step, the subsequent steps just fail for different reasons, I don't think I am close to using what you have suggested in my flow yet, as I don't think I'm that far ahead

[Chriddle]

But maybe the exact amount does not matter if there are more than e.g. 3 occurences 😉

[Chriddle]

What errors do you get, and why do you use an API request and not just the "Get files" action?

[PAppsAlways1]

Unable to process template language expressions in action 'Compose_6' inputs at line '0' and column '0': 'The template language function 'int' was invoked with a parameter that is not valid. The value cannot be converted to the target type.'.

I've pointed it o a single file, rather than to search through all files in a document library, but the full flow errors