This is a public library of BigQuery persistent user-defined functions (UDFs). The goal of this project is to adapt useful algorithms and solutions from the broader Javascript open-source community that add significant value to BigQuery scenarios. Select code is packaged as a persistent UDF and made publically available for general use from any BigQuery query.
Below is a list of UDFs available in this library, all of which are directly callable from any BigQuery query. A link is given for each UDF citing the original source project upon which it is based.
| Function | Description | Source |
|---|---|---|
| levenshtein | Compute a Levenshtein distance between two strings. Use to determine the degree of similarity between words or sentences. | js-levenshtein |
| parse_useragent | Parses web-browser user-agent strings, breaking them down into components of browser, engine, OS, CPU, and device type/model. | ua-parser-js |
| text_diff | Creates a textual diff-analysis that can be used to analyze differences between strings, highlight before/after changes, or to create a textual patch between two similar documents. | diff-match-patch |
| text_patch | Applies a textual patch to transform one string into another. Useful for restoring copies of highly similar documents, for example different revisions of a document or code file. The patch can be created by calling text_diff. | diff-match-patch |
Is there an additional charge for using UDFs?
No, there are no additional charges for running UDFs on top of the regular data processing costs. So there is no additional cost overhead to using UDFs.
My query is taking a lot longer now, is the UDF slowing it down?
Probably. Queries that make use of UDFs may run substantially slower, depending on the amount of data to be processed and the algorithmic complexity of the UDF itself.
How does billing work when using a UDF from another project? Which project is billed?
Billing is charged to the project from which a query is run.
Are there any other limits to using UDFs?
The amount of data output by a UDF is limited to approximately 5 MB per row. See UDF limits for more info.
Can you add support for moment?
What about cheerio and jsdom for parsing HTML content?
Pull requests and suggestions for additional UDFs are always welcome!