Directory loader for bulk documents in LangChain - Step-by-Step Execution

Concept Flow - Directory loader for bulk documents

Start

↓

Specify directory path

↓

Initialize DirectoryLoader

↓

Load all files in directory

↓

Read each document

↓

Store documents in list

↓

Return list of documents

↓

End

The loader starts by specifying a folder path, then reads all files inside, loads each document, and returns them as a list.

Execution Sample

LangChain

from langchain.document_loaders import DirectoryLoader
loader = DirectoryLoader('docs')
docs = loader.load()
print(len(docs))

This code loads all documents from the 'docs' folder and prints how many documents were loaded.

Execution Table

Step	Action	Files Found	Documents Loaded	Output
1	Initialize DirectoryLoader with 'docs'	N/A	0	Loader ready
2	Scan 'docs' directory	3 files found	0	Files listed
3	Read file 1	3 files found	1	Document 1 loaded
4	Read file 2	3 files found	2	Document 2 loaded
5	Read file 3	3 files found	3	Document 3 loaded
6	Return documents list	3 files found	3	List of 3 documents
7	Print length of docs	3 files found	3	Output: 3

💡 All files in directory processed and documents loaded

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	After Step 5	Final
loader	None	DirectoryLoader instance	DirectoryLoader instance	DirectoryLoader instance	DirectoryLoader instance	DirectoryLoader instance
files_found	None	3 files	3 files	3 files	3 files	3 files
docs	[]	[]	[doc1]	[doc1, doc2]	[doc1, doc2, doc3]	[doc1, doc2, doc3]

Key Moments - 2 Insights

Why does the loader return a list of documents instead of a single document?

Because the directory can contain multiple files, the loader reads each file and returns all documents as a list, as shown in execution_table steps 3 to 6.

What happens if the directory is empty?

The loader finds zero files, so the documents list remains empty. The process ends quickly with an empty list, similar to step 2 but with zero files.

Visual Quiz - 3 Questions

Look at the execution table, how many documents are loaded after step 4?

A 2

B 1

C 3

D 0

💡 Hint

Check the 'Documents Loaded' column at step 4 in the execution_table.

At which step does the loader finish reading all files?

A Step 3

B Step 5

C Step 6

D Step 7

💡 Hint

Look for the last 'Read file' action in the execution_table.

If the directory had 5 files instead of 3, what would change in the variable_tracker?

A loader variable would change to 5

B files_found would remain 3

C docs list would have 5 documents after final step

D docs list would be empty

💡 Hint

See how docs list grows with files_found in variable_tracker.

Concept Snapshot

DirectoryLoader loads all files from a folder.
Initialize with folder path.
Call load() to read all documents.
Returns a list of documents.
Useful for bulk document processing.

Full Transcript

This visual execution shows how DirectoryLoader from langchain works. First, you create a loader with the folder path. Then it scans the folder and finds all files. It reads each file one by one, loading documents into a list. Finally, it returns the list of all documents. The example code loads documents from a folder named 'docs' and prints how many documents were loaded. The execution table traces each step, showing files found and documents loaded. The variable tracker shows how the docs list grows as files are read. Key moments clarify why the loader returns a list and what happens if the folder is empty. The quiz tests understanding of document counts and process steps. This helps beginners see how bulk document loading works step-by-step.