A JavaScript application to analyze WhatsApp chat history locally in your browser.
βοΈ Caffeine fund βοΈ
If you found this tool useful or fun, please consider buying me a coffee
Usage
- Clone the repository
git clone https://github.com/mowolf/ChatAnalyzer.git - Open
index.htmlin the root folder. - Follow instructions on the web page.
Alternatively you can use the official hosted version of this code.
What happens to my chat data?
No chat data is transferred to any remote server, all analysis is performed locally on your device
The application works completly offline once the page has loaded. You can turn off your internet before loading your data file if you are particularly concerned. You can also take a look at the source code and give it an audit.
Reporting Issues
Please open a new Github Issue if you find any issues or have suggestions/improvements for the project.
Please also supply your data format (e.g. [07.09.17, 6:44:16 PM] NAME: message) and Language/Region setting of our phone if you are having a specific problem with data not loading correctly.
Have a great idea for a new graph?
Head over to planned features and add your idea if it's missing!
Wanna help making this better?
Cheers! You are very welcome! Just submit a pull request. The goal of this tool is to automize the parsing as much as possible.
Add your language identifier for audio/video/pictures
Please see the header of main.js. There you can add your identifiers.
Markov Chain
To generate text that sounds like you check here: https://github.com/mowolf/markov_chain_js/tree/master
How the data is parsed & how you can build on that
The chat data is parsed via regex and splitted into it's parts. There are two variables that hold the data.
structArray:
Formatted line data in the form of a struct with the keys: name date time message
name: "Name Surname"date: "YYYY-MM-DD"time: "HH:MM:SS"message: text
each key represents an array, the index represents the line number e.g. structArray.date[0] is the date of the first line
userStruct:
Array of structs with the keys:
name: "Name Surname"date: "YYYY-MM-DD"time: "HH:MM:SS"message: text
For every person there is one struct, same format as above, the index represents messageNumber of name.