The goal of dir2json is to provide a utility for
converting directories into JSON format and decoding JSON back into
directory structures. This is particularly useful for archiving,
sharing, or analyzing directory contents in a structured format. The
package can handle a variety of file types within the directory,
including text and binary files (e.g., images, PDFs), converting them to
and from JSON. It can be used
with a Shiny app to allow users to upload files, encode them into
JSON format, and process or download the resulting JSON file. JSON, in
this manner, can be used as a format for sharing multiple files with
LLMs.
Note that this JSON representation is it not an alternative to ZIP files, but rather a way to represent the contents of a directory in a structured format that can be easily parsed and manipulated. ZIP files are a superior solution for compressing and archiving files, while JSON is a text-based format that is more suitable for data interchange and manipulation.
Encode a directory: Convert a directory structure (including its files) into a JSON object.
Decode a JSON object: Convert a JSON object back into the original directory structure, restoring files and folders.
Supports mixed file types: Handles both text and binary files by encoding binary files in base64 and text files as plain text.
JSON schema: The structure of the JSON output follows a defined schema, allowing for efficient storage and easy parsing.
The JSON schema used by dir2json is very simple and
flat. Each file (or directory) is represented by an object with two main
fields:
name: The file or directory
name.
content: The content of the file.
For text files, this is the raw content as a string. For binary files,
it is a base64-encoded string.
You can install the released version of lzstring from CRAN with:
install.packages("dir2json")You can install the development version of dir2json from
GitHub with:
# install.packages("devtools")
devtools::install_github("parmsam/dir2json-r")This is a basic example which shows you how to encode a directory into JSON and decode it back:
library(dir2json)
# Create a temporary directory with a file
example_dir <- tempfile()
dir.create(example_dir)
file.create(file.path(example_dir, "example.txt"))
#> [1] TRUE
writeLines("Hello, dir2json!", file.path(example_dir, "example.txt"))
# Encode the directory to JSON
json_data <- json_encode_dir(example_dir)
cat(json_data)
#> [{"name":"example.txt","content":"Hello, dir2json!\n"}]
# Decode the JSON back to a new directory
new_dir <- tempfile()
json_decode_dir(json_data, new_dir)
# Verify the contents
list.files(new_dir, recursive = TRUE)
#> [1] "example.txt"
readLines(file.path(new_dir, "example.txt"))
#> [1] "Hello, dir2json!"Encoding a Directory: The
json_encode_dir function traverses the directory
recursively, reading each file and determining if it is a text or binary
file. Text files are directly stored as their contents in the JSON,
while binary files are base64-encoded for safe transport within the JSON
format.
Decoding a JSON Object: The
json_decode_dir function reads the JSON object, restores
the directory structure, and writes files back to the file system.
Binary files are decoded from base64 back to their original binary
format, and text files are written as plain text.