Skip to content

Latest commit

 

History

History
100 lines (71 loc) · 4.11 KB

README_DEBUG_REPORT.md

File metadata and controls

100 lines (71 loc) · 4.11 KB

Sign-up Data

This document explains how we keep track of what data is being collected by the Orb, during a sign-up. The struct that is responsible for this data is DebugReport.

Files Tracking Sign-up Data Updates

In order to describe which fields have been updated, we associate each commit with 2 files: debug_report_schema.json and debug_report_schema.csv. With every commit pushed into master or prod branch, these files are uploaded:

  1. to wld-signup-data-schema AWS S3 Bucket.
  2. as GitHub Actions artifacts.

Schema Files Description

  • debug_report_schema.json: Contains the JSON schema of DebugReport struct. A JSON Schema file can be visualized as a schema-tree here.

  • debug_report_schema.csv: Contains a list of paths from root to each node of the (above) schema tree. This representation is useful when someone wants to annotate additional information for each data field. The columns of this CSV file are path and instance_type.

Downloading Schema Files

Steps:

  1. Go to wld-signup-data-schema AWS S3 Bucket.
    • If you do not have access, request access in #infrastructure channel on Slack.
  2. From the list, select the .zip file you are interested in. Each filename contains the date and the corresponding commit hash that produced the schema files.
  3. Download the .zip file.
  4. The .zip contains both debug_report_schema.json and debug_report_schema.csv files.

Track Changes Between Commits

Let debug_report_schema_old.csv and debug_report_schema_new.csv be the schema files of two different commits.

The command:

diff -w --color debug_report_schema_old.csv debug_report_schema_new.csv

will print which fields were added/removed between the two commits.

  • Lines that start with > are new fields.
  • Lines that start with < are deleted fields

Caution: Make sure that in the above command the first file is the older version.

Example output:

211a212
> metadata/backend_config/IrisModelConfig,"[String, Null]"
219,221d219
< metadata/backend_config/MaxOffgazeThreshold,"[Number, Null]"
< metadata/backend_config/MaxPupilToIrisRatioThreshold,"[Number, Null]"
< metadata/backend_config/MinOcclusion30Threshold,"[Number, Null]"
223,224d220
< metadata/backend_config/MinOcclusion90Threshold,"[Number, Null]"
< metadata/backend_config/MinPupilToIrisRatioThreshold,"[Number, Null]"
232d227
< metadata/backend_config/UseTokyoPipeline,"[Boolean, Null]"
750,755c745
< metadata/mega_agent_one_config/iris/max_offgaze_threshold,"[Number, Null]"
< metadata/mega_agent_one_config/iris/max_pupil_to_iris_ratio_threshold,"[Number, Null]"
< metadata/mega_agent_one_config/iris/min_occlusion_30_threshold,"[Number, Null]"
< metadata/mega_agent_one_config/iris/min_occlusion_90_threshold,"[Number, Null]"
< metadata/mega_agent_one_config/iris/min_pupil_to_iris_ratio_threshold,"[Number, Null]"
< metadata/mega_agent_one_config/iris/use_tokyo_pipeline,Boolean
---
> metadata/mega_agent_one_config/iris/config,"[String, Null]"

Maintaining Your Own Spreadsheet

If you maintain your own spreadsheet of sign-up data and you want to update it, then you cannot apply the above command directly.

Step 1: With the spreadsheet software of your preference, export a CSV file containing only the path and instance_type columns.

Step 2: After downloading the newest version of debug_report_schema.csv, execute the above command.

Step 3: Remove the rows indicated by the command from your spreadsheet.

Step 4: Add the rows indicated by the command to your spreadsheet.

Step 5: Make sure that your spreadsheet's rows are sorted based on path.

Extra Step: To be sure that you applied the changes correctly, execute Step 1 & Step 2 again. The command should now produce no output.