README ---------------- googlevoice-to-sqlite 3.2.2 (2012-07-25) (c) Arithmomaniac (arithmomaniac@hotmail.com) under LGPL2 ------------- The gvoice-to-sqlite script takes the Google Voice records you can get from Google Takeout one step further - instead of thousands of tiny files, it produces a SQLite database that you can use to make lists (of past texts within a year, for example) and analyses (such as the people you talk to the most). It also can export the database to CSV. This script depends on Python 2.7 and both the dateutil and html5lib libraries. If this is too much for you, download the win32 package from the website. ------------- Basic file structure (all in root directory): - googlevoice_to_sqlite.py the main executable - initdb.sql The SQL that creates the database. ------------- To use this script: 1) Download and unzip your Voice data from Google Takeout. 2) Start googlevoice_to_sqlite.py 3) Provide the location of the "Voice\Calls" directory 4) Follow the directions 5) The results are stored in .\output. A database diagram exists on the project homepage. If you want to save the results, copy it out of the directory. Due to limitations in Google's contact persistence, the entire database is rewritten every time. If you need a SQLite client, try SQLite Studio (Windows) or SQLite Manager (Firefox). If you would like to import the data into Excel, Oracle, SQL Server, etc on Windows, use OLDEB, or export the tables to CSV and import manually. -------------- Changelog: 3.3 (2013-02-04) Made independent of user language Supports non-ASCII characters in texts 3.2.2 (2012-07-25) Bugfix for long contact names causing crash 3.2.1 (2012-07-23) Updated to match new Takeout filename formats 3.2 (2012-07-03) Updated to match new Takeout time formats 3.1 (2012-01-05) Fixed bug where 'Welcome to Google Voice' message causes crash Clearly labeled time fields as UTC 3.0 (2011-10-04) All file names written on same line EXE Packaging available HTML unescaping inside text messages 2.0 (2011-09-26) CSV exports and SQL views Better file-input name handling Performance enhancements - up to 80% faster 1.0 (2011-09-20) --------------- Performance Notes Since several thousand objects are created during execution, this script can take a while to run. here's an approximate performance breakdown for the main routine. 10% - File Reading 50% - XML tree creation 30% - Object parsing/creation 10% - Database insertion/processing