PDF files

[vasi26ro]

Hello,
Very nice tool.
I am trying to read restaurant menus from their website, but in most of the cases I encounter PDF files.
How would you tackle this situation?
Thank you

[VinciGit00]

Make in the main the http rquest for getting the pdf and then use the pdfscraper provided by our library.
Let us know!

[vasi26ro]

I have tried the following example with both local file and endpoint. Also I have tried with groq.

from scrapegraphai.graphs.pdf_scraper_graph import PDFScraperGraph

graph_config = {
    "llm": {
        "model": "ollama/llama3",
        "temperature": 0,
        "format": "json",  # Ollama needs the format to be specified explicitly
        "base_url": "http://localhost:11434",  # set Ollama URL
    },
    "embeddings": {
        "model": "ollama/nomic-embed-text",
        "base_url": "http://localhost:11434",  # set Ollama URL
    }
}

pdf_scraper = PDFScraperGraph(
    prompt="List or chapters",
    source="https://ww1.microchip.com/downloads/en/devicedoc/39632c.pdf",
    config=graph_config
)

result = pdf_scraper.run()
print(result)

In all for cases the following error was raised:

Traceback (most recent call last):
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/nodes/base_node.py", line 86, in get_input_keys
    input_keys = self._parse_input_keys(state, self.input)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/nodes/base_node.py", line 196, in _parse_input_keys
    raise ValueError("No state keys matched the expression.")
ValueError: No state keys matched the expression.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "ddd/scraperai-test.py", line 32, in <module>
    result = pdf_scraper.run()
             ^^^^^^^^^^^^^^^^^
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/graphs/pdf_scraper_graph.py", line 116, in run
    self.final_state, self.execution_info = self.graph.execute(inputs)
                                            ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/graphs/base_graph.py", line 106, in execute
    result = current_node.execute(state)
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/nodes/fetch_node.py", line 58, in execute
    input_keys = self.get_input_keys(state)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "ddd/.venv/lib/python3.11/site-packages/scrapegraphai/nodes/base_node.py", line 90, in get_input_keys
    raise ValueError(
ValueError: Error parsing input keys for Fetch: No state keys matched the expression.

What I am doing wrong?

[VinciGit00]

Please update it