Skip to content

Frequently Asked Questions

What is Phinder?

Phinder is a command-line application that identifies Personally Identifiable Information (PII) in various file formats, such as PDF, Word, Excel, and more. It uses the Phileas library for PII detection.

Which file formats are supported?

Phinder supports a wide range of formats including: - Plain Text (.txt, .log) - Adobe PDF (.pdf) - Microsoft Word (.doc, .docx) - Microsoft Excel (.xls, .xlsx) - Microsoft PowerPoint (.ppt, .pptx) - Comma-Separated Values (.csv) - Rich Text Format (.rtf) - Email Messages (.eml, .msg) - Images (.png, .jpg, .tiff) via OCR

How do I define what PII to look for?

You can provide a Phileas policy file in JSON format using the -p or --policy option. This policy defines the types of PII (e.g., names, SSNs, credit card numbers) to detect.

Can I scan entire directories?

Yes, Phinder can scan a single file or an entire directory. Use the -R or --recursive flag to scan subdirectories as well.

How does Phinder handle large files?

Phinder is designed for high performance. Large log files are processed line-by-line to minimize memory usage, and other formats are handled efficiently using specialized processors.

What are the Risk, Magnitude, and Density scores?

  • Risk Score: How sensitive the PII in a document is. It combines the severity of each entity type, how many occurrences were found, and how confident the detector was, and is reported with a risk level from NONE to CRITICAL. Use it to decide what to remediate first. See Risk Scoring.
  • Magnitude Score: A weighted score representing the total amount of PII found in a document. You can customize the weights for different PII types.
  • Density Score: The ratio of PII findings to the total amount of text in the document.

Why does a file with less PII sometimes have a higher Risk Score?

Because the Risk Score measures sensitivity rather than volume. Two Social Security numbers outrank fifty city names. The Magnitude Score is the one that answers "how much PII is in here"; the two are calculated and configured independently, so tuning one does not move the other.

Does the Risk Score tell me whether I am compliant?

No. It is a prioritization aid built from the entity types a scan found. It does not know how a file is stored, who can read it, or whether the detections are correct: detection is probabilistic, so a scan may both miss PII and flag text that is not PII. Review the findings against your own data before acting on the ranking.

How can I speed up subsequent scans?

You can use the --log and --skip-unchanged options. Phinder will maintain a log (in a MongoDB database) of the files it has already scanned and their hashes. On subsequent runs, it will skip files that haven't changed. You must provide the MongoDB connection string via the PHINDER_MONGODB_URI environment variable.

In what formats are the reports generated?

Phinder always generates both HTML (.html) and JSON (.json) reports.

How is OCR handled for images?

Phinder uses Tesseract for OCR. You may need to set the TESSDATA_PREFIX environment variable to the directory containing your Tesseract language data files (e.g., eng.traineddata).

Does Phinder require an internet connection?

No, Phinder performs all PII detection locally using the Phileas library and does not send your data to any external services.

Does Phinder redact the PII?

No, Phinder is designed for identification and reporting of PII. If you need to redact or anonymize PII, consider using Philter.