Skip to main content

A library/command line utility for parsing Facebook chat history

Project description

Facebook Chat Archive Parser Build Status

A small tool and library for parsing chat history from a Facebook data archive into more usable formats.

What is a “Facebook Chat Archive”?

Facebook Messenger records all your conversation history since the beginning of time. If you want it back for some reason, you have two options:

  1. Create a scraper that constantly “scrolls up” in the conversation window you’re interested in (or simulates that with API calls), progressively getting more and more of your chat history.

  2. Ask Facebook for an archive of all your data here , and wait a couple of days for them to give it to you as a zip archive.

Number 2 is the only practical option if you want everything in a timely manner.

What does Facebook give me in this zip archive?

Facebook gives you literally everything you’ve ever posted to Facebook, which includes your pictures, videos, posts. etc in addition to your chat messages.

Your chat history comes in a single HTML page titled messages.htm. Unfortunately, the data is mostly unordered and impossible to load into your web browser (it can be several hundred megabytes in size). You’re essentially forced to parse it if you want to analyze the content.

Why would I ever want my Facebook chat history?

There are a number of reasons you may want to parse your Facebook chat history:

  1. To make a simulation of your friends using Markov chains.

  2. You’re deleting your Facebook account, but would like a record of your conversations.

  3. You need to analyze a copy of your conversations for legal reasons.

Here comes the Facebook Chat Archive Parser!

The Facebook Chat Archive Parser is a command line tool (and library for advanced users) for easily transforming your messages.htm file into something actually useful.

How do I get it?

Install the Facebook Chat Archive Parser via pip under at least Python 2.7:

pip install fbchat-archive-parser

How does it work?

Simply run the command fbcap in your terminal with your messages.htm file as the argument.

fbcap ./messages.htm

And watch as the parser sifts through your data!

Processing gif

When it’s done, it will dump all your conversation history is dumped to stdout. Obviously, this can be very long, so here is an example:

Results

What if I want JSON?

Simply supply the -f json option to the command line:

fbcap ./messages.htm -f json

The output format is as follows:

Messages are ordered from oldest to newest.

{
    "threads": [
        {
            "participants": ["participant_0", "...", "participant_n"],
            "messages": [
                {
                    "date": "ISO 8601 formatted date",
                    "sender": "sender name",
                    "message": "message text"
                },
                "..."
            ]
        },
        "..."
    ]
}

How about CSV?

Of course!

fbcap ./messages.htm -f csv
thread,sender,date,message
Third User,Third User,2013-10-04T15:05Z,1
Third User,Third User,2013-10-04T15:05Z,2
Third User,Third User,2013-10-04T15:05Z,3
Third User,First User,2013-10-04T15:05Z,4
Third User,Third User,2013-10-04T15:06Z,5
Third User,First User,2013-10-04T15:07Z,6
Third User,First User,2013-10-04T15:07Z,7
Second User,Second User,2013-10-04T15:04Z,X Y Z
Second User,Second User,2013-10-04T15:05Z,X? Y Z!
Second User,Second User,2013-10-04T15:05Z,This is a test
Second User,Second User,2013-10-04T15:05Z,"Yes, it is"
Second User,Second User,2013-10-04T15:05Z,The last message!
"Second User, Third User",Third User,2013-10-04T15:05Z,1
"Second User, Third User",Third User,2013-10-04T15:05Z,2
...

What about that YAML thing the kids these days are talking about?

For sure!

fbcap ./messages.htm -f yaml
user: First User
threads:
- participants:
  - Second User
  - Third User
  messages:
  - date: 2013-10-04T22:05-07:00
    message: '1'
    sender: Third User
  - date: 2013-10-04T22:05-07:00
    message: '2'
    sender: Third User
  - date: 2013-10-04T22:05-07:00
    message: '3'
    sender: Third User
...

What if I want to see some statistics?

See who you talk to the most among your friends and how much each of you contribute to the conversation.

fbcap ./messages.htm -f stats
stats image

How do I get any of the above into a file?

Just use standard file redirects.

fbcap ./messages.htm > my_file.txt

Can I get each conversation into a separate file?

Use the -d directive to send the output to a directory instead.

fbcap ./messages.htm -d some/random/directory

This will create a file per conversation titled thread_#.ext where # is the conversation number and ext is the extension of the format (e.g. json). A manifest.txt file is also created, which lists the participants in each thread number for navigational/search purposes.

What if I only want to parse out a specific conversation?

You can use the -t option to specify a particular conversation/thread you want to output. Just provide a comma-separated set of names. If you don’t remember a last name (or conversely, only remember the last name), the system will try to compensate.

fbcap ./messages.htm -t second
filter second
fbcap ./messages.htm -t second,third
filter second and third

What else can I do?

Take a look at the help options to find out more!

$ fbcap --help
fbcap: A program for converting Facebook chat history to a number of more usable formats

Usage: fbcap {{arguments}} {{options}}

Arguments:
  path [text]  Path of the messages.htm file to parse

Options:
  -h, --help              Show this help message and exit
  -f, --format [str]      Format to convert to (csv, json, pretty-json, text, yaml, stats) (default: text)
  -t, --thread [text]     Only include threads involving exactly the following comma-separated participants in output (-t 'Billy,Steve Jensson')
  -z, --timezones [text]  Timezone disambiguators (TZ=OFFSET,[TZ=OFFSET[...]])
  -d, --directory [text]  Write all output as a file per thread into a directory (subdirectory will be created)
  -u, --utc               Use UTC timestamps in the output
  -n, --nocolor           Do not colorize output
  -p, --noprogress        Do not show progress output
  -r, --resolve           [BETA] Resolve profile IDs to names by connecting to Facebook

Troubleshooting

Why are do some names appear as <some number>@facebook.com?

For some reason, Facebook seems to randomly swap names for IDs. In recent times, it has gotten worse. You can have the parser resolve the names via Facebook itself with the --resolve flag. Keep in mind, this is a beta feature and may not work perfectly.

$ fbcap ./messages.htm -t second --resolve
Facebook username/email: facebook_username
Facebook password:

This requires your Facebook credentials to get accurate results. This does not relay your credentials through any servers and is a direct connection from your computer to Facebook. Please look at the code if you are feeling paranoid or skeptical :)

Why are some of my chat threads missing?

This is a mysterious issue on Facebook’s end. From anecdotal evidence, it seems that what gets returned in your chat archive is generally conversations with people who you have most recently talked to. Fortunately, it always seems to be the complete history for each conversation and nothing gets truncated.

Unfortunately, this cannot be remedied unless Facebook fixes the problem on their end.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

fbchat_archive_parser-0.9.post9.tar.gz (34.1 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page