A configurable tool for concurrent processing of U.S. Patent and Trademark Office (USPTO) bulk data zip files.

At this time, the tool supports the following USPTO bulk data products:

• Patent Grant Full Text Data (No Images) (2004 - Present)
• Patent Application Full Text Data (No Images) (2004 - Present)

Given a directory of USPTO zip files, the application will produce one of the following outputs:

• Complete XML files of individual documents split out from the zip
• JSON files of individual documents
  • Selective (non-exhaustive) parsing of main document fields
  • Structured patent claims representing referential relationships, as in the original PatentPublicData tool
  • HTML formatting of Abstract and Description fields
• Apache Parquet files corresponding to bulk zip files

Clone this repository. Edit the config.toml as needed - the most important config values are the first three:

[required]
inputdirectory = "data/in"
outputdirectory = "data/out"
outputmode = "json"

For the most basic setup, create data/in directories within the project root, and populate the /in directory with zip files to process.

Then, from the root of project directory:

make run

For more advanced usage running the application from somewhere other than the root of the project directory, the executable accepts a single optional argument specifying the path to a config.toml file.

MIT