// Copyright 2017 Google Inc.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

syntax = "proto3";

package google.privacy.dlp.v2beta1;

import "google/api/annotations.proto";
import "google/longrunning/operations.proto";
import "google/privacy/dlp/v2beta1/storage.proto";
import "google/protobuf/empty.proto";
import "google/protobuf/timestamp.proto";
import "google/type/date.proto";
import "google/type/timeofday.proto";

option csharp_namespace = "Google.Cloud.Dlp.V2Beta1";
option go_package = "google.golang.org/genproto/googleapis/privacy/dlp/v2beta1;dlp";
option java_multiple_files = true;
option java_outer_classname = "DlpProto";
option java_package = "com.google.privacy.dlp.v2beta1";


// The DLP API is a service that allows clients
// to detect the presence of Personally Identifiable Information (PII) and other
// privacy-sensitive data in user-supplied, unstructured data streams, like text
// blocks or images.
// The service also includes methods for sensitive data redaction and
// scheduling of data scans on Google Cloud Platform based data sets.
service DlpService {
  // Finds potentially sensitive info in a list of strings.
  // This method has limits on input size, processing time, and output size.
  rpc InspectContent(InspectContentRequest) returns (InspectContentResponse) {
    option (google.api.http) = { post: "/v2beta1/content:inspect" body: "*" };
  }

  // Redacts potentially sensitive info from a list of strings.
  // This method has limits on input size, processing time, and output size.
  rpc RedactContent(RedactContentRequest) returns (RedactContentResponse) {
    option (google.api.http) = { post: "/v2beta1/content:redact" body: "*" };
  }

  // De-identifies potentially sensitive info from a list of strings.
  // This method has limits on input size and output size.
  rpc DeidentifyContent(DeidentifyContentRequest) returns (DeidentifyContentResponse) {
    option (google.api.http) = { post: "/v2beta1/content:deidentify" body: "*" };
  }

  // Schedules a job scanning content in a Google Cloud Platform data
  // repository.
  rpc CreateInspectOperation(CreateInspectOperationRequest) returns (google.longrunning.Operation) {
    option (google.api.http) = { post: "/v2beta1/inspect/operations" body: "*" };
  }

  // Schedules a job to compute risk analysis metrics over content in a Google
  // Cloud Platform repository.
  rpc AnalyzeDataSourceRisk(AnalyzeDataSourceRiskRequest) returns (google.longrunning.Operation) {
    option (google.api.http) = { post: "/v2beta1/dataSource:analyze" body: "*" };
  }

  // Returns list of results for given inspect operation result set id.
  rpc ListInspectFindings(ListInspectFindingsRequest) returns (ListInspectFindingsResponse) {
    option (google.api.http) = { get: "/v2beta1/{name=inspect/results/*}/findings" };
  }

  // Returns sensitive information types for given category.
  rpc ListInfoTypes(ListInfoTypesRequest) returns (ListInfoTypesResponse) {
    option (google.api.http) = { get: "/v2beta1/rootCategories/{category=*}/infoTypes" };
  }

  // Returns the list of root categories of sensitive information.
  rpc ListRootCategories(ListRootCategoriesRequest) returns (ListRootCategoriesResponse) {
    option (google.api.http) = { get: "/v2beta1/rootCategories" };
  }
}

// Configuration description of the scanning process.
// When used with redactContent only info_types and min_likelihood are currently
// used.
message InspectConfig {
  // Max findings configuration per info type, per content item or long running
  // operation.
  message InfoTypeLimit {
    // Type of information the findings limit applies to. Only one limit per
    // info_type should be provided. If InfoTypeLimit does not have an
    // info_type, the DLP API applies the limit against all info_types that are
    // found but not specified in another InfoTypeLimit.
    InfoType info_type = 1;

    // Max findings limit for the given infoType.
    int32 max_findings = 2;
  }

  // Restricts what info_types to look for. The values must correspond to
  // InfoType values returned by ListInfoTypes or found in documentation.
  // Empty info_types runs all enabled detectors.
  repeated InfoType info_types = 1;

  // Only returns findings equal or above this threshold.
  Likelihood min_likelihood = 2;

  // Limits the number of findings per content item or long running operation.
  int32 max_findings = 3;

  // When true, a contextual quote from the data that triggered a finding is
  // included in the response; see Finding.quote.
  bool include_quote = 4;

  // When true, excludes type information of the findings.
  bool exclude_types = 6;

  // Configuration of findings limit given for specified info types.
  repeated InfoTypeLimit info_type_limits = 7;
}

// Additional configuration for inspect long running operations.
message OperationConfig {
  // Max number of findings per file, Datastore entity, or database row.
  int64 max_item_findings = 1;
}

// Container structure for the content to inspect.
message ContentItem {
  // Type of the content, as defined in Content-Type HTTP header.
  // Supported types are: all "text" types, octet streams, PNG images,
  // JPEG images.
  string type = 1;

  // Data of the item either in the byte array or UTF-8 string form.
  oneof data_item {
    // Content data to inspect or redact.
    bytes data = 2;

    // String data to inspect or redact.
    string value = 3;

    // Structured content for inspection.
    Table table = 4;
  }
}

// Structured content to inspect. Up to 50,000 `Value`s per request allowed.
message Table {
  message Row {
    repeated Value values = 1;
  }

  repeated FieldId headers = 1;

  repeated Row rows = 2;
}

// All the findings for a single scanned item.
message InspectResult {
  // List of findings for an item.
  repeated Finding findings = 1;

  // If true, then this item might have more findings than were returned,
  // and the findings returned are an arbitrary subset of all findings.
  // The findings list might be truncated because the input items were too
  // large, or because the server reached the maximum amount of resources
  // allowed for a single API call. For best results, divide the input into
  // smaller batches.
  bool findings_truncated = 2;
}

// Container structure describing a single finding within a string or image.
message Finding {
  // The specific string that may be potentially sensitive info.
  string quote = 1;

  // The specific type of info the string might be.
  InfoType info_type = 2;

  // Estimate of how likely it is that the info_type is correct.
  Likelihood likelihood = 3;

  // Location of the info found.
  Location location = 4;

  // Timestamp when finding was detected.
  google.protobuf.Timestamp create_time = 6;
}

// Specifies the location of a finding within its source item.
message Location {
  // Zero-based byte offsets within a content item.
  Range byte_range = 1;

  // Character offsets within a content item, included when content type
  // is a text. Default charset assumed to be UTF-8.
  Range codepoint_range = 2;

  // Location within an image's pixels.
  repeated ImageLocation image_boxes = 3;

  // Key of the finding.
  RecordKey record_key = 4;

  // Field id of the field containing the finding.
  FieldId field_id = 5;

  // Location within a `ContentItem.Table`.
  TableLocation table_location = 6;
}

// Location of a finding within a `ContentItem.Table`.
message TableLocation {
  // The zero-based index of the row where the finding is located.
  int64 row_index = 1;
}

// Generic half-open interval [start, end)
message Range {
  // Index of the first character of the range (inclusive).
  int64 start = 1;

  // Index of the last character of the range (exclusive).
  int64 end = 2;
}

// Bounding box encompassing detected text within an image.
message ImageLocation {
  // Top coordinate of the bounding box. (0,0) is upper left.
  int32 top = 1;

  // Left coordinate of the bounding box. (0,0) is upper left.
  int32 left = 2;

  // Width of the bounding box in pixels.
  int32 width = 3;

  // Height of the bounding box in pixels.
  int32 height = 4;
}

// Request to search for potentially sensitive info in a list of items
// and replace it with a default or provided content.
message RedactContentRequest {
  message ReplaceConfig {
    // Type of information to replace. Only one ReplaceConfig per info_type
    // should be provided. If ReplaceConfig does not have an info_type, the DLP
    // API matches it against all info_types that are found but not specified in
    // another ReplaceConfig.
    InfoType info_type = 1;

    // Content replacing sensitive information of given type. Max 256 chars.
    string replace_with = 2;
  }

  // Configuration for determing how redaction of images should occur.
  message ImageRedactionConfig {
    // Type of information to redact from images.
    oneof target {
      // Only one per info_type should be provided per request. If not
      // specified, and redact_all_text is false, the DLP API will redact all
      // text that it matches against all info_types that are found, but not
      // specified in another ImageRedactionConfig.
      InfoType info_type = 1;

      // If true, all text found in the image, regardless whether it matches an
      // info_type, is redacted.
      bool redact_all_text = 2;
    }

    // The color to use when redacting content from an image. If not specified,
    // the default is black.
    Color redaction_color = 3;
  }

  // Configuration for the inspector.
  InspectConfig inspect_config = 1;

  // The list of items to inspect. Up to 100 are allowed per request.
  repeated ContentItem items = 2;

  // The strings to replace findings text findings with. Must specify at least
  // one of these or one ImageRedactionConfig if redacting images.
  repeated ReplaceConfig replace_configs = 3;

  // The configuration for specifying what content to redact from images.
  repeated ImageRedactionConfig image_redaction_configs = 4;
}

// Represents a color in the RGB color space.
message Color {
  // The amount of red in the color as a value in the interval [0, 1].
  float red = 1;

  // The amount of green in the color as a value in the interval [0, 1].
  float green = 2;

  // The amount of blue in the color as a value in the interval [0, 1].
  float blue = 3;
}

// Results of redacting a list of items.
message RedactContentResponse {
  // The redacted content.
  repeated ContentItem items = 1;
}

// Request to de-identify a list of items.
message DeidentifyContentRequest {
  // Configuration for the de-identification of the list of content items.
  DeidentifyConfig deidentify_config = 1;

  // Configuration for the inspector.
  InspectConfig inspect_config = 2;

  // The list of items to inspect. Up to 100 are allowed per request.
  // All items will be treated as text/*.
  repeated ContentItem items = 3;
}

// Results of de-identifying a list of items.
message DeidentifyContentResponse {
  repeated ContentItem items = 1;

  // A review of the transformations that took place for each item.
  repeated DeidentificationSummary summaries = 2;
}

// Request to search for potentially sensitive info in a list of items.
message InspectContentRequest {
  // Configuration for the inspector.
  InspectConfig inspect_config = 1;

  // The list of items to inspect. Items in a single request are
  // considered "related" unless inspect_config.independent_inputs is true.
  // Up to 100 are allowed per request.
  repeated ContentItem items = 2;
}

// Results of inspecting a list of items.
message InspectContentResponse {
  // Each content_item from the request has a result in this list, in the
  // same order as the request.
  repeated InspectResult results = 1;
}

// Request for scheduling a scan of a data subset from a Google Platform data
// repository.
message CreateInspectOperationRequest {
  // Configuration for the inspector.
  InspectConfig inspect_config = 1;

  // Specification of the data set to process.
  StorageConfig storage_config = 2;

  // Optional location to store findings.
  OutputStorageConfig output_config = 3;

  // Additional configuration settings for long running operations.
  OperationConfig operation_config = 5;
}

// Cloud repository for storing output.
message OutputStorageConfig {
  oneof type {
    // Store findings in a new table in the dataset.
    BigQueryTable table = 1;

    // The path to a Google Cloud Storage location to store output.
    // The bucket must already exist and
    // the Google APIs service account for DLP must have write permission to
    // write to the given bucket.
    // Results are split over multiple csv files with each file name matching
    // the pattern "[operation_id]_[count].csv", for example
    // `3094877188788974909_1.csv`. The `operation_id` matches the
    // identifier for the Operation, and the `count` is a counter used for
    // tracking the number of files written.
    //
    // The CSV file(s) contain the following columns regardless of storage type
    // scanned:
    // - id
    // - info_type
    // - likelihood
    // - byte size of finding
    // - quote
    // - timestamp
    //
    // For Cloud Storage the next columns are:
    //
    // - file_path
    // - start_offset
    //
    // For Cloud Datastore the next columns are:
    //
    // - project_id
    // - namespace_id
    // - path
    // - column_name
    // - offset
    //
    // For BigQuery the next columns are:
    //
    // - row_number
    // - project_id
    // - dataset_id
    // - table_id
    CloudStoragePath storage_path = 2;
  }
}

// Statistics regarding a specific InfoType.
message InfoTypeStatistics {
  // The type of finding this stat is for.
  InfoType info_type = 1;

  // Number of findings for this info type.
  int64 count = 2;
}

// Metadata returned within GetOperation for an inspect request.
message InspectOperationMetadata {
  // Total size in bytes that were processed.
  int64 processed_bytes = 1;

  // Estimate of the number of bytes to process.
  int64 total_estimated_bytes = 4;

  repeated InfoTypeStatistics info_type_stats = 2;

  // The time which this request was started.
  google.protobuf.Timestamp create_time = 3;

  // The inspect config used to create the Operation.
  InspectConfig request_inspect_config = 5;

  // The storage config used to create the Operation.
  StorageConfig request_storage_config = 6;

  // Optional location to store findings.
  OutputStorageConfig request_output_config = 7;
}

// The operational data.
message InspectOperationResult {
  // The server-assigned name, which is only unique within the same service that
  // originally returns it. If you use the default HTTP mapping, the
  // `name` should have the format of `inspect/results/{id}`.
  string name = 1;
}

// Request for the list of results in a given inspect operation.
message ListInspectFindingsRequest {
  // Identifier of the results set returned as metadata of
  // the longrunning operation created by a call to InspectDataSource.
  // Should be in the format of `inspect/results/{id}`.
  string name = 1;

  // Maximum number of results to return.
  // If 0, the implementation selects a reasonable value.
  int32 page_size = 2;

  // The value returned by the last `ListInspectFindingsResponse`; indicates
  // that this is a continuation of a prior `ListInspectFindings` call, and that
  // the system should return the next page of data.
  string page_token = 3;

  // Restricts findings to items that match. Supports info_type and likelihood.
  //
  // Examples:
  //
  // - info_type=EMAIL_ADDRESS
  // - info_type=PHONE_NUMBER,EMAIL_ADDRESS
  // - likelihood=VERY_LIKELY
  // - likelihood=VERY_LIKELY,LIKELY
  // - info_type=EMAIL_ADDRESS,likelihood=VERY_LIKELY,LIKELY
  string filter = 4;
}

// Response to the ListInspectFindings request.
message ListInspectFindingsResponse {
  // The results.
  InspectResult result = 1;

  // If not empty, indicates that there may be more results that match the
  // request; this value should be passed in a new `ListInspectFindingsRequest`.
  string next_page_token = 2;
}

// Description of the information type (infoType).
message InfoTypeDescription {
  // Internal name of the infoType.
  string name = 1;

  // Human readable form of the infoType name.
  string display_name = 2;

  // List of categories this infoType belongs to.
  repeated CategoryDescription categories = 3;
}

// Request for the list of info types belonging to a given category,
// or all supported info types if no category is specified.
message ListInfoTypesRequest {
  // Category name as returned by ListRootCategories.
  string category = 1;

  // Optional BCP-47 language code for localized info type friendly
  // names. If omitted, or if localized strings are not available,
  // en-US strings will be returned.
  string language_code = 2;
}

// Response to the ListInfoTypes request.
message ListInfoTypesResponse {
  // Set of sensitive info types belonging to a category.
  repeated InfoTypeDescription info_types = 1;
}

// Info Type Category description.
message CategoryDescription {
  // Internal name of the category.
  string name = 1;

  // Human readable form of the category name.
  string display_name = 2;
}

// Request for root categories of Info Types supported by the API.
// Example values might include "FINANCE", "HEALTH", "FAST", "DEFAULT".
message ListRootCategoriesRequest {
  // Optional language code for localized friendly category names.
  // If omitted or if localized strings are not available,
  // en-US strings will be returned.
  string language_code = 1;
}

// Response for ListRootCategories request.
message ListRootCategoriesResponse {
  // List of all into type categories supported by the API.
  repeated CategoryDescription categories = 1;
}

// Request for creating a risk analysis operation.
message AnalyzeDataSourceRiskRequest {
  // Privacy metric to compute.
  PrivacyMetric privacy_metric = 1;

  // Input dataset to compute metrics over.
  BigQueryTable source_table = 3;
}

// Privacy metric to compute for reidentification risk analysis.
message PrivacyMetric {
  // Compute numerical stats over an individual column, including
  // min, max, and quantiles.
  message NumericalStatsConfig {
    // Field to compute numerical stats on. Supported types are
    // integer, float, date, datetime, timestamp, time.
    FieldId field = 1;
  }

  // Compute numerical stats over an individual column, including
  // number of distinct values and value count distribution.
  message CategoricalStatsConfig {
    // Field to compute categorical stats on. All column types are
    // supported except for arrays and structs. However, it may be more
    // informative to use NumericalStats when the field type is supported,
    // depending on the data.
    FieldId field = 1;
  }

  // k-anonymity metric, used for analysis of reidentification risk.
  message KAnonymityConfig {
    // Set of fields to compute k-anonymity over. When multiple fields are
    // specified, they are considered a single composite key. Structs and
    // repeated data types are not supported; however, nested fields are
    // supported so long as they are not structs themselves or nested within
    // a repeated field.
    repeated FieldId quasi_ids = 1;

    // Optional message indicating that each distinct `EntityId` should not
    // contribute to the k-anonymity count more than once per equivalence class.
    EntityId entity_id = 2;
  }

  // l-diversity metric, used for analysis of reidentification risk.
  message LDiversityConfig {
    // Set of quasi-identifiers indicating how equivalence classes are
    // defined for the l-diversity computation. When multiple fields are
    // specified, they are considered a single composite key.
    repeated FieldId quasi_ids = 1;

    // Sensitive field for computing the l-value.
    FieldId sensitive_attribute = 2;
  }

  oneof type {
    NumericalStatsConfig numerical_stats_config = 1;

    CategoricalStatsConfig categorical_stats_config = 2;

    KAnonymityConfig k_anonymity_config = 3;

    LDiversityConfig l_diversity_config = 4;
  }
}

// Metadata returned within the
// [`riskAnalysis.operations.get`](/dlp/docs/reference/rest/v2beta1/riskAnalysis.operations/get)
// for risk analysis.
message RiskAnalysisOperationMetadata {
  // The time which this request was started.
  google.protobuf.Timestamp create_time = 1;

  // Privacy metric to compute.
  PrivacyMetric requested_privacy_metric = 2;

  // Input dataset to compute metrics over.
  BigQueryTable requested_source_table = 3;
}

// Result of a risk analysis
// [`Operation`](/dlp/docs/reference/rest/v2beta1/inspect.operations)
// request.
message RiskAnalysisOperationResult {
  // Result of the numerical stats computation.
  message NumericalStatsResult {
    // Minimum value appearing in the column.
    Value min_value = 1;

    // Maximum value appearing in the column.
    Value max_value = 2;

    // List of 99 values that partition the set of field values into 100 equal
    // sized buckets.
    repeated Value quantile_values = 4;
  }

  // Result of the categorical stats computation.
  message CategoricalStatsResult {
    // Histogram bucket of value frequencies in the column.
    message CategoricalStatsHistogramBucket {
      // Lower bound on the value frequency of the values in this bucket.
      int64 value_frequency_lower_bound = 1;

      // Upper bound on the value frequency of the values in this bucket.
      int64 value_frequency_upper_bound = 2;

      // Total number of records in this bucket.
      int64 bucket_size = 3;

      // Sample of value frequencies in this bucket. The total number of
      // values returned per bucket is capped at 20.
      repeated ValueFrequency bucket_values = 4;
    }

    // Histogram of value frequencies in the column.
    repeated CategoricalStatsHistogramBucket value_frequency_histogram_buckets = 5;
  }

  // Result of the k-anonymity computation.
  message KAnonymityResult {
    // The set of columns' values that share the same k-anonymity value.
    message KAnonymityEquivalenceClass {
      // Set of values defining the equivalence class. One value per
      // quasi-identifier column in the original KAnonymity metric message.
      // The order is always the same as the original request.
      repeated Value quasi_ids_values = 1;

      // Size of the equivalence class, for example number of rows with the
      // above set of values.
      int64 equivalence_class_size = 2;
    }

    // Histogram bucket of equivalence class sizes in the table.
    message KAnonymityHistogramBucket {
      // Lower bound on the size of the equivalence classes in this bucket.
      int64 equivalence_class_size_lower_bound = 1;

      // Upper bound on the size of the equivalence classes in this bucket.
      int64 equivalence_class_size_upper_bound = 2;

      // Total number of records in this bucket.
      int64 bucket_size = 3;

      // Sample of equivalence classes in this bucket. The total number of
      // classes returned per bucket is capped at 20.
      repeated KAnonymityEquivalenceClass bucket_values = 4;
    }

    // Histogram of k-anonymity equivalence classes.
    repeated KAnonymityHistogramBucket equivalence_class_histogram_buckets = 5;
  }

  // Result of the l-diversity computation.
  message LDiversityResult {
    // The set of columns' values that share the same l-diversity value.
    message LDiversityEquivalenceClass {
      // Quasi-identifier values defining the k-anonymity equivalence
      // class. The order is always the same as the original request.
      repeated Value quasi_ids_values = 1;

      // Size of the k-anonymity equivalence class.
      int64 equivalence_class_size = 2;

      // Number of distinct sensitive values in this equivalence class.
      int64 num_distinct_sensitive_values = 3;

      // Estimated frequencies of top sensitive values.
      repeated ValueFrequency top_sensitive_values = 4;
    }

    // Histogram bucket of sensitive value frequencies in the table.
    message LDiversityHistogramBucket {
      // Lower bound on the sensitive value frequencies of the equivalence
      // classes in this bucket.
      int64 sensitive_value_frequency_lower_bound = 1;

      // Upper bound on the sensitive value frequencies of the equivalence
      // classes in this bucket.
      int64 sensitive_value_frequency_upper_bound = 2;

      // Total number of records in this bucket.
      int64 bucket_size = 3;

      // Sample of equivalence classes in this bucket. The total number of
      // classes returned per bucket is capped at 20.
      repeated LDiversityEquivalenceClass bucket_values = 4;
    }

    // Histogram of l-diversity equivalence class sensitive value frequencies.
    repeated LDiversityHistogramBucket sensitive_value_frequency_histogram_buckets = 5;
  }

  // Values associated with this metric.
  oneof result {
    NumericalStatsResult numerical_stats_result = 3;

    CategoricalStatsResult categorical_stats_result = 4;

    KAnonymityResult k_anonymity_result = 5;

    LDiversityResult l_diversity_result = 6;
  }
}

// A value of a field, including its frequency.
message ValueFrequency {
  // A value contained in the field in question.
  Value value = 1;

  // How many times the value is contained in the field.
  int64 count = 2;
}

// Set of primitive values supported by the system.
message Value {
  oneof type {
    int64 integer_value = 1;

    double float_value = 2;

    string string_value = 3;

    bool boolean_value = 4;

    google.protobuf.Timestamp timestamp_value = 5;

    google.type.TimeOfDay time_value = 6;

    google.type.Date date_value = 7;
  }
}

// The configuration that controls how the data will change.
message DeidentifyConfig {
  oneof transformation {
    // Treat the dataset as free-form text and apply the same free text
    // transformation everywhere.
    InfoTypeTransformations info_type_transformations = 1;

    // Treat the dataset as structured. Transformations can be applied to
    // specific locations within structured datasets, such as transforming
    // a column within a table.
    RecordTransformations record_transformations = 2;
  }
}

// A rule for transforming a value.
message PrimitiveTransformation {
  oneof transformation {
    ReplaceValueConfig replace_config = 1;

    RedactConfig redact_config = 2;

    CharacterMaskConfig character_mask_config = 3;

    CryptoReplaceFfxFpeConfig crypto_replace_ffx_fpe_config = 4;

    FixedSizeBucketingConfig fixed_size_bucketing_config = 5;

    BucketingConfig bucketing_config = 6;

    ReplaceWithInfoTypeConfig replace_with_info_type_config = 7;

    TimePartConfig time_part_config = 8;

    CryptoHashConfig crypto_hash_config = 9;
  }
}

// For use with `Date`, `Timestamp`, and `TimeOfDay`, extract or preserve a
// portion of the value.
message TimePartConfig {
  enum TimePart {
    TIME_PART_UNSPECIFIED = 0;

    // [000-9999]
    YEAR = 1;

    // [1-12]
    MONTH = 2;

    // [1-31]
    DAY_OF_MONTH = 3;

    // [1-7]
    DAY_OF_WEEK = 4;

    // [1-52]
    WEEK_OF_YEAR = 5;

    // [0-24]
    HOUR_OF_DAY = 6;
  }

  TimePart part_to_extract = 1;
}

// Pseudonymization method that generates surrogates via cryptographic hashing.
// Uses SHA-256.
// Outputs a 32 byte digest as an uppercase hex string
// (for example, 41D1567F7F99F1DC2A5FAB886DEE5BEE).
// Currently, only string and integer values can be hashed.
message CryptoHashConfig {
  // The key used by the hash function.
  CryptoKey crypto_key = 1;
}

// Replace each input value with a given `Value`.
message ReplaceValueConfig {
  // Value to replace it with.
  Value new_value = 1;
}

// Replace each matching finding with the name of the info_type.
message ReplaceWithInfoTypeConfig {

}

// Redact a given value. For example, if used with an `InfoTypeTransformation`
// transforming PHONE_NUMBER, and input 'My phone number is 206-555-0123', the
// output would be 'My phone number is '.
message RedactConfig {

}

// Characters to skip when doing deidentification of a value. These will be left
// alone and skipped.
message CharsToIgnore {
  enum CharacterGroup {
    CHARACTER_GROUP_UNSPECIFIED = 0;

    // 0-9
    NUMERIC = 1;

    // A-Z
    ALPHA_UPPER_CASE = 2;

    // a-z
    ALPHA_LOWER_CASE = 3;

    // US Punctuation, one of !"#$%&'()*+,-./:;<=>?@[\]^_`{|}~
    PUNCTUATION = 4;

    // Whitespace character, one of [ \t\n\x0B\f\r]
    WHITESPACE = 5;
  }

  oneof characters {
    string characters_to_skip = 1;

    CharacterGroup common_characters_to_ignore = 2;
  }
}

// Partially mask a string by replacing a given number of characters with a
// fixed character. Masking can start from the beginning or end of the string.
// This can be used on data of any type (numbers, longs, and so on) and when
// de-identifying structured data we'll attempt to preserve the original data's
// type. (This allows you to take a long like 123 and modify it to a string like
// **3.
message CharacterMaskConfig {
  // Character to mask the sensitive values&mdash;for example, "*" for an
  // alphabetic string such as name, or "0" for a numeric string such as ZIP
  // code or credit card number. String must have length 1. If not supplied, we
  // will default to "*" for strings, 0 for digits.
  string masking_character = 1;

  // Number of characters to mask. If not set, all matching chars will be
  // masked. Skipped characters do not count towards this tally.
  int32 number_to_mask = 2;

  // Mask characters in reverse order. For example, if `masking_character` is
  // '0', number_to_mask is 14, and `reverse_order` is false, then
  // 1234-5678-9012-3456 -> 00000000000000-3456
  // If `masking_character` is '*', `number_to_mask` is 3, and `reverse_order`
  // is true, then 12345 -> 12***
  bool reverse_order = 3;

  // When masking a string, items in this list will be skipped when replacing.
  // For example, if your string is 555-555-5555 and you ask us to skip `-` and
  // mask 5 chars with * we would produce ***-*55-5555.
  repeated CharsToIgnore characters_to_ignore = 4;
}

// Buckets values based on fixed size ranges. The
// Bucketing transformation can provide all of this functionality,
// but requires more configuration. This message is provided as a convenience to
// the user for simple bucketing strategies.
// The resulting value will be a hyphenated string of
// lower_bound-upper_bound.
// This can be used on data of type: double, long.
// If the bound Value type differs from the type of data
// being transformed, we will first attempt converting the type of the data to
// be transformed to match the type of the bound before comparing.
message FixedSizeBucketingConfig {
  // Lower bound value of buckets. All values less than `lower_bound` are
  // grouped together into a single bucket; for example if `lower_bound` = 10,
  // then all values less than 10 are replaced with the value “-10”. [Required].
  Value lower_bound = 1;

  // Upper bound value of buckets. All values greater than upper_bound are
  // grouped together into a single bucket; for example if `upper_bound` = 89,
  // then all values greater than 89 are replaced with the value “89+”.
  // [Required].
  Value upper_bound = 2;

  // Size of each bucket (except for minimum and maximum buckets). So if
  // `lower_bound` = 10, `upper_bound` = 89, and `bucket_size` = 10, then the
  // following buckets would be used: -10, 10-20, 20-30, 30-40, 40-50, 50-60,
  // 60-70, 70-80, 80-89, 89+. Precision up to 2 decimals works. [Required].
  double bucket_size = 3;
}

// Generalization function that buckets values based on ranges. The ranges and
// replacement values are dynamically provided by the user for custom behavior,
// such as 1-30 -> LOW 31-65 -> MEDIUM 66-100 -> HIGH
// This can be used on
// data of type: number, long, string, timestamp.
// If the bound `Value` type differs from the type of data being transformed, we
// will first attempt converting the type of the data to be transformed to match
// the type of the bound before comparing.
message BucketingConfig {
  // Buckets represented as ranges, along with replacement values. Ranges must
  // be non-overlapping.
  message Bucket {
    // Lower bound of the range, inclusive. Type should be the same as max if
    // used.
    Value min = 1;

    // Upper bound of the range, exclusive; type must match min.
    Value max = 2;

    // Replacement value for this bucket. If not provided
    // the default behavior will be to hyphenate the min-max range.
    Value replacement_value = 3;
  }

  repeated Bucket buckets = 1;
}

// Replaces an identifier with an surrogate using FPE with the FFX
// mode of operation.
// The identifier must be representable by the US-ASCII character set.
// For a given crypto key and context, the same identifier will be
// replaced with the same surrogate.
// Note that a given identifier must be either the empty string or be at
// least two characters long.
message CryptoReplaceFfxFpeConfig {
  // These are commonly used subsets of the alphabet that the FFX mode
  // natively supports. In the algorithm, the alphabet is selected using
  // the "radix". Therefore each corresponds to particular radix.
  enum FfxCommonNativeAlphabet {
    FFX_COMMON_NATIVE_ALPHABET_UNSPECIFIED = 0;

    // [0-9] (radix of 10)
    NUMERIC = 1;

    // [0-9A-F] (radix of 16)
    HEXADECIMAL = 2;

    // [0-9A-Z] (radix of 36)
    UPPER_CASE_ALPHA_NUMERIC = 3;

    // [0-9A-Za-z] (radix of 62)
    ALPHA_NUMERIC = 4;
  }

  // The key used by the encryption algorithm. [required]
  CryptoKey crypto_key = 1;

  // A context may be used for higher security since the same
  // identifier in two different contexts likely will be given a distinct
  // surrogate. The principle is that the likeliness is inversely related
  // to the ratio of the number of distinct identifiers per context over the
  // number of possible surrogates: As long as this ratio is small, the
  // likehood is large.
  //
  // If the context is not set, a default tweak will be used.
  // If the context is set but:
  //
  // 1. there is no record present when transforming a given value or
  // 1. the field is not present when transforming a given value,
  //
  // a default tweak will be used.
  //
  // Note that case (1) is expected when an `InfoTypeTransformation` is
  // applied to both structured and non-structured `ContentItem`s.
  // Currently, the referenced field may be of value type integer or string.
  //
  // The tweak is constructed as a sequence of bytes in big endian byte order
  // such that:
  //
  // - a 64 bit integer is encoded followed by a single byte of value 1
  // - a string is encoded in UTF-8 format followed by a single byte of value 2
  //
  // This is also known as the 'tweak', as in tweakable encryption.
  FieldId context = 2;

  oneof alphabet {
    FfxCommonNativeAlphabet common_alphabet = 4;

    // This is supported by mapping these to the alphanumeric characters
    // that the FFX mode natively supports. This happens before/after
    // encryption/decryption.
    // Each character listed must appear only once.
    // Number of characters must be in the range [2, 62].
    // This must be encoded as ASCII.
    // The order of characters does not matter.
    string custom_alphabet = 5;

    // The native way to select the alphabet. Must be in the range [2, 62].
    int32 radix = 6;
  }
}

// This is a data encryption key (DEK) (as opposed to
// a key encryption key (KEK) stored by KMS).
// When using KMS to wrap/unwrap DEKs, be sure to set an appropriate
// IAM policy on the KMS CryptoKey (KEK) to ensure an attacker cannot
// unwrap the data crypto key.
message CryptoKey {
  oneof source {
    TransientCryptoKey transient = 1;

    UnwrappedCryptoKey unwrapped = 2;

    KmsWrappedCryptoKey kms_wrapped = 3;
  }
}

// Use this to have a random data crypto key generated.
// It will be discarded after the operation/request finishes.
message TransientCryptoKey {
  // Name of the key. [required]
  // This is an arbitrary string used to differentiate different keys.
  // A unique key is generated per name: two separate `TransientCryptoKey`
  // protos share the same generated key if their names are the same.
  // When the data crypto key is generated, this name is not used in any way
  // (repeating the api call will result in a different key being generated).
  string name = 1;
}

// Using raw keys is prone to security risks due to accidentally
// leaking the key. Choose another type of key if possible.
message UnwrappedCryptoKey {
  // The AES 128/192/256 bit key. [required]
  bytes key = 1;
}

// Include to use an existing data crypto key wrapped by KMS.
// Authorization requires the following IAM permissions when sending a request
// to perform a crypto transformation using a kms-wrapped crypto key:
// dlp.kms.encrypt
message KmsWrappedCryptoKey {
  // The wrapped data crypto key. [required]
  bytes wrapped_key = 1;

  // The resource name of the KMS CryptoKey to use for unwrapping. [required]
  string crypto_key_name = 2;
}

// A type of transformation that will scan unstructured text and
// apply various `PrimitiveTransformation`s to each finding, where the
// transformation is applied to only values that were identified as a specific
// info_type.
message InfoTypeTransformations {
  // A transformation to apply to text that is identified as a specific
  // info_type.
  message InfoTypeTransformation {
    // Info types to apply the transformation to. Empty list will match all
    // available info types for this transformation.
    repeated InfoType info_types = 1;

    // Primitive transformation to apply to the info type. [required]
    PrimitiveTransformation primitive_transformation = 2;
  }

  // Transformation for each info type. Cannot specify more than one
  // for a given info type. [required]
  repeated InfoTypeTransformation transformations = 1;
}

// The transformation to apply to the field.
message FieldTransformation {
  // Input field(s) to apply the transformation to. [required]
  repeated FieldId fields = 1;

  // Only apply the transformation if the condition evaluates to true for the
  // given `RecordCondition`. The conditions are allowed to reference fields
  // that are not used in the actual transformation. [optional]
  //
  // Example Use Cases:
  //
  // - Apply a different bucket transformation to an age column if the zip code
  // column for the same record is within a specific range.
  // - Redact a field if the date of birth field is greater than 85.
  RecordCondition condition = 3;

  // Transformation to apply. [required]
  oneof transformation {
    // Apply the transformation to the entire field.
    PrimitiveTransformation primitive_transformation = 4;

    // Treat the contents of the field as free text, and selectively
    // transform content that matches an `InfoType`.
    InfoTypeTransformations info_type_transformations = 5;
  }
}

// A type of transformation that is applied over structured data such as a
// table.
message RecordTransformations {
  // Transform the record by applying various field transformations.
  repeated FieldTransformation field_transformations = 1;

  // Configuration defining which records get suppressed entirely. Records that
  // match any suppression rule are omitted from the output [optional].
  repeated RecordSuppression record_suppressions = 2;
}

// Configuration to suppress records whose suppression conditions evaluate to
// true.
message RecordSuppression {
  RecordCondition condition = 1;
}

// A condition for determing whether a transformation should be applied to
// a field.
message RecordCondition {
  // The field type of `value` and `field` do not need to match to be
  // considered equal, but not all comparisons are possible.
  //
  // A `value` of type:
  //
  // - `string` can be compared against all other types
  // - `boolean` can only be compared against other booleans
  // - `integer` can be compared against doubles or a string if the string value
  // can be parsed as an integer.
  // - `double` can be compared against integers or a string if the string can
  // be parsed as a double.
  // - `Timestamp` can be compared against strings in RFC 3339 date string
  // format.
  // - `TimeOfDay` can be compared against timestamps and strings in the format
  // of 'HH:mm:ss'.
  //
  // If we fail to compare do to type mismatch, a warning will be given and
  // the condition will evaluate to false.
  message Condition {
    // Field within the record this condition is evaluated against. [required]
    FieldId field = 1;

    // Operator used to compare the field or info type to the value. [required]
    RelationalOperator operator = 3;

    // Value to compare against. [Required, except for `EXISTS` tests.]
    Value value = 4;
  }

  message Conditions {
    repeated Condition conditions = 1;
  }

  // A collection of expressions
  message Expressions {
    enum LogicalOperator {
      LOGICAL_OPERATOR_UNSPECIFIED = 0;

      AND = 1;
    }

    // The operator to apply to the result of conditions. Default and currently
    // only supported value is `AND`.
    LogicalOperator logical_operator = 1;

    oneof type {
      Conditions conditions = 3;
    }
  }

  Expressions expressions = 3;
}

// High level summary of deidentification.
message DeidentificationSummary {
  // Total size in bytes that were transformed in some way.
  int64 transformed_bytes = 2;

  // Transformations applied to the dataset.
  repeated TransformationSummary transformation_summaries = 3;
}

// Summary of a single tranformation.
message TransformationSummary {
  // A collection that informs the user the number of times a particular
  // `TransformationResultCode` and error details occurred.
  message SummaryResult {
    int64 count = 1;

    TransformationResultCode code = 2;

    // A place for warnings or errors to show up if a transformation didn't
    // work as expected.
    string details = 3;
  }

  // Possible outcomes of transformations.
  enum TransformationResultCode {
    TRANSFORMATION_RESULT_CODE_UNSPECIFIED = 0;

    SUCCESS = 1;

    ERROR = 2;
  }

  // Set if the transformation was limited to a specific info_type.
  InfoType info_type = 1;

  // Set if the transformation was limited to a specific FieldId.
  FieldId field = 2;

  // The specific transformation these stats apply to.
  PrimitiveTransformation transformation = 3;

  // The field transformation that was applied. This list will contain
  // multiple only in the case of errors.
  repeated FieldTransformation field_transformations = 5;

  // The specific suppression option these stats apply to.
  RecordSuppression record_suppress = 6;

  repeated SummaryResult results = 4;
}

// Categorization of results based on how likely they are to represent a match,
// based on the number of elements they contain which imply a match.
enum Likelihood {
  // Default value; information with all likelihoods is included.
  LIKELIHOOD_UNSPECIFIED = 0;

  // Few matching elements.
  VERY_UNLIKELY = 1;

  UNLIKELY = 2;

  // Some matching elements.
  POSSIBLE = 3;

  LIKELY = 4;

  // Many matching elements.
  VERY_LIKELY = 5;
}

// Operators available for comparing the value of fields.
enum RelationalOperator {
  RELATIONAL_OPERATOR_UNSPECIFIED = 0;

  // Equal.
  EQUAL_TO = 1;

  // Not equal to.
  NOT_EQUAL_TO = 2;

  // Greater than.
  GREATER_THAN = 3;

  // Less than.
  LESS_THAN = 4;

  // Greater than or equals.
  GREATER_THAN_OR_EQUALS = 5;

  // Less than or equals.
  LESS_THAN_OR_EQUALS = 6;

  // Exists
  EXISTS = 7;
}