Custom document loaders in LangChain - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ component_behavior

intermediate

What is the output of this custom loader's load method?

Consider this custom document loader class using LangChain. What will loader.load() return?

LangChain

from langchain.document_loaders import BaseLoader

class MyLoader(BaseLoader):
    def __init__(self, data):
        self.data = data

    def load(self):
        return [{'page_content': d, 'metadata': {'source': 'custom'}} for d in self.data]

loader = MyLoader(['doc1', 'doc2'])
result = loader.load()

A None

B [{'page_content': 'doc1', 'metadata': {'source': 'custom'}}, {'page_content': 'doc2', 'metadata': {'source': 'custom'}}]

C [{'content': 'doc1'}, {'content': 'doc2'}]

D ['doc1', 'doc2']

💡 Hint

Look at what the load method returns: a list of dictionaries with specific keys.

The load method returns a list of dicts, each with keys page_content and metadata. So option B matches exactly.

📝 Syntax

intermediate

Which option correctly defines a custom loader inheriting from BaseLoader?

Choose the correct syntax for a custom loader class that inherits from BaseLoader and implements load method.

A

class CustomLoader(BaseLoader):
    def load(self):
        return []

B

class CustomLoader:
    def load(self):
        return []

C

class CustomLoader(BaseLoader):
    def load():
        return []

D

class CustomLoader(BaseLoader):
    def load(self, data):
        return data

💡 Hint

Remember that methods inside classes need self as the first parameter.

Option A correctly inherits from BaseLoader and defines load(self). Option A misses inheritance. Option A misses self parameter. Option A adds an unexpected parameter.

🔧 Debug

advanced

What error does this custom loader code raise?

Given this custom loader code, what error will it raise when calling loader.load()?

LangChain

from langchain.document_loaders import BaseLoader

class BrokenLoader(BaseLoader):
    def load(self):
        return [{'page_content': d} for d in self.data]

loader = BrokenLoader()
result = loader.load()

A Returns empty list []

B TypeError: load() missing 1 required positional argument: 'self'

C SyntaxError: invalid syntax

D AttributeError: 'BrokenLoader' object has no attribute 'data'

💡 Hint

Check if self.data is defined before use.

The class does not define self.data anywhere, so accessing it raises an AttributeError.

❓ state_output

advanced

What is the value of loader.data after this code runs?

After running this code, what is the value of loader.data?

LangChain

from langchain.document_loaders import BaseLoader

class MyLoader(BaseLoader):
    def __init__(self, data):
        self.data = data

    def load(self):
        return [{'page_content': d} for d in self.data]

loader = MyLoader(['a', 'b'])
loader.data.append('c')

A ['a', 'b', 'c']

B ['a', 'b']

C ['c']

D Raises AttributeError

💡 Hint

Remember that loader.data is a list passed in and can be modified.

The list assigned to self.data is mutable. Appending 'c' adds it to the list.

🧠 Conceptual

expert

Why implement a custom document loader in LangChain?

Which reason best explains why you would create a custom document loader in LangChain?

A To replace the LangChain core with a different framework

B To speed up the LangChain library's internal processing

C To load documents from a unique or unsupported data source not covered by built-in loaders

D To automatically generate documents without user input

💡 Hint

Think about when built-in loaders are not enough.

Custom loaders let you handle special data sources or formats that LangChain's built-in loaders don't support.