mneudert/instream

Instream is an Elixir driver for InfluxDB that supports both v1.x and v2.x APIs through a unified connection and query interface. It enables users to execute Flux and InfluxQL queries, write data via line protocol or UDP, and decode responses into strongly typed series structs with compile-time validation. The system handles version-specific routing, authentication, and structured logging while abstracting HTTP communication and JSON serialization.

14 changes.

This period focused on refactoring the Instream library to support InfluxDB v2.x while maintaining v1.x compatibility, introducing version-specific connection runners and authentication methods. Significant work was done on data handling, including new series hydration, validation logic, and a dedicated line protocol encoder for efficient data ingestion. The development environment was also standardized with comprehensive test infrastructure, linting tools, and coverage settings.

11 changes.

This period focused on modernizing the library's infrastructure by introducing a Hackney-based HTTP client and implementing comprehensive structured logging with password redaction. Significant effort was dedicated to supporting InfluxDB v2.x through new CSV decoding capabilities and extensive test coverage for both v1 and v2 connection and writer modules.

4 changes.

This period focused on expanding test coverage for the InfluxDB v2 integration by adding comprehensive test suites for the series hydrator, data deleter, and Line protocol writers. The work included creating helper modules for connection and series definitions to support these new tests, ensuring correct data serialization, hydration, and deletion logic.

Users can now decode InfluxDB v2.x CSV responses that include datatype annotations (such as boolean, long, double, duration, base64Binary, and RFC3339 date/time formats) into properly typed Elixir maps. The new \Instream.Decoder.CSV\ module handles multiple CSV tables, applies default values for empty fields, and converts specific values like "true"/"false" booleans and "infinity" doubles to their corresponding Elixir atoms or types. A supporting \Instream.Decoder.RFC3339\ module ensures accurate nanosecond-precision timestamp conversion.

lib/instream/decoder · high confidence

The new \Instream\ module provides the main entry point for the InfluxDB driver, featuring detailed documentation on establishing connections for both InfluxDB v1.x and v2.x, executing Flux and InfluxQL queries with parameter binding and timeout configuration, writing data points via Series modules or plain maps, and deleting data in v2.x environments.

lib · high confidence

The repository now includes configuration files for the Credo linter, Dialyzer, and Elixir formatter, along with a CHANGELOG, README, license, and test coverage settings. These files establish the project's coding standards, documentation, and development environment setup.

(repo-wide) · high confidence

This change introduces a new default logging system for Instream that captures structured data for ping, query, status, write, and delete operations. For users, this means request details are now logged at the debug level, including the number of points written, the query result, and the delete predicate. Crucially, passwords included in query strings are automatically redacted before being logged to prevent credential leakage. The system also attaches common metadata, such as query time and response status, to each log entry.

lib/instream/log · high confidence

The library now includes a new HTTP client implementation using the \:hackney\ library (\Instream.HTTPClient.Hackney\), which wraps \:hackney.request/5\ to handle HTTP requests and extract response bodies. Additionally, a new utility module \Instream.HTTPClient.Headers\ provides case-insensitive header lookup by normalizing header names to lowercase during comparison, ensuring consistent header retrieval regardless of casing in the response.

_lib/instream/http\client · high confidence

A new dedicated module for encoding data points into the InfluxDB line protocol has been introduced. This module handles the serialization of measurements, tags, and fields, ensuring that keys are sorted for consistent output and special characters are properly escaped. It supports integer and RFC3339 binary timestamps, converting them to nanosecond precision integers, and allows timestamps to be omitted or set to nil.

lib/instream/encoder · high confidence

Users can now write data using the InfluxDB line protocol via new \Instream.Writer.Line\ (dispatching to v1 or v2 specific writers) and \Instream.Writer.UDP\ modules. The line protocol writers handle HTTP POST requests with appropriate headers and URL construction for their respective API versions, while the UDP writer enables asynchronous or synchronous point transmission via a process pool, allowing for high-throughput, low-latency data ingestion.

lib/instream/writer · high confidence

The application now includes a dedicated configuration file that sets the console logger format specifically for the test environment. This change ensures that test logs include query times and response status metadata, providing more detailed visibility into test execution without affecting other environments.

config · high confidence

The library now includes a dedicated hydrator and validator for InfluxDB series structs. The hydrator converts raw query result maps (including pivoted CSV results and InfluxQL GROUP BY outputs) into strongly typed series structs, handling timestamp conversion from both integers and RFC3339 strings. The validator enforces stricter series definitions by requiring explicit field and measurement definitions, forbidding the use of reserved keys like '\_field', '\_measurement', and 'time' as user-defined fields or tags, and raising errors if a name is used for both a field and a tag.

lib/instream/series · high confidence

The connection module has been restructured to support InfluxDB v1 and v2 through separate \QueryRunner\ and \ResponseParser\ implementations, ensuring version-specific behavior for queries, writes, and response parsing. A new \Config\ module centralizes connection configuration, supporting application, runtime, and inline setup methods, while a dedicated \JSON\ module abstracts encoding and decoding to allow flexible JSON library integration (defaulting to Jason). The \Supervisor\ now handles connection initialization and worker management, providing a cleaner supervision tree for connection processes.

lib/instream/connection · high confidence

The library introduces a new \Instream.Connection\ module that manages server connections and routes operations to version-specific runners (\QueryRunnerV1\/\QueryRunnerV2\), enabling support for InfluxDB v2.x features like buckets and organizations while maintaining v1.x compatibility. A new \Instream.Series\ module allows users to define structured data points with compile-time validation, default values, and hydration from query results. The \Instream.Writer\ behaviour is simplified to focus on point writing, and the \Instream.HTTPClient\ provides a switchable interface for HTTP requests.

lib/instream · high confidence

The library now supports InfluxDB v2 connections by routing queries to the \/api/v2/query\ endpoint and writes to \/api/v2/write\, while maintaining v1 compatibility via the legacy \/query\ and \/write\ paths. Users can now specify individual time precision (epoch) for read and write operations, choose response encoding formats such as CSV or JSON, and select the query language (Flux or InfluxQL). Authentication is handled via headers for Bearer tokens, API tokens, or Basic Auth, replacing the previous method of passing credentials as URL parameters for v2 connections.

lib/instream/query · high confidence

Added Line Protocol encoder benchmark; Added connection integration tests; Added connection tests for InfluxDB v2; Added test coverage for InfluxDB v1 connection behaviors; Added test coverage for InfluxDB v2 connection authorization, bucket/org priority, and response parsing; Added test helper modules for connection and series definitions; Added tests for InfluxDB v1 line and UDP writers; Added tests for InfluxDB v2 Line protocol writers; Added tests for InfluxDB v2 data deletion logic; Added tests for series definition defaults and metadata; Added tests for series hydration and validation; Added tests for the CSV decoder; Added tests for the InfluxDB v2 series hydrator; Added tests for the Line protocol encoder; Added tests for the default logger's request logging behavior; Adds test infrastructure for InfluxDB query headers and version-specific test configuration.

This entry introduces the initial project configuration for Instream, an InfluxDB driver for Elixir, setting the development version to 2.3.0-dev. It establishes a minimum Elixir requirement of 1.9 and defines the core runtime dependencies, including hackney for HTTP communication, influxql for query formatting, jason for JSON handling, nimble\_csv for CSV parsing, and poolboy for connection pooling. The configuration also includes development and test dependencies such as credo, dialyxir, ex\_doc, excoveralls, and mox, alongside a benchmarking tool (benchee). The project is configured to generate HTML documentation, run dialyzer checks with specific flags, and publish to Hex under the Apache-2.0 license.

(dependencies) · high confidence

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

• First survey — no prior run to compare against. CAI 65.

• Code Health 99
• Architecture 93
• Maturity 47
• Readiness 83
• Security 76

• 300 commits — 287 feature/other, 13 fixes

• (root) — 81 commits
• lib/instream — 79 commits
• .github/workflows — 65 commits
• test/influxdb_v1 — 23 commits
• test/influxdb_v2 — 18 commits
• test/instream — 18 commits
• lib/instream.ex — 9 commits
• config/config.exs — 3 commits
• test/test_helper.exs — 2 commits
• (repo) — 1 commit
• test/helpers — 1 commit

• fix: Fix (default) query language header for InfluxDB v2.x (#77)
• fix: Fix Elixir 1.15.0 deprecations
• fix: Fix InfluxDB v1.8 testing environment
• fix: Fix UDP writer test configuration
• fix: Fix multiple line breaks in response (#73)
• fix: Fix parsing double type without decimal point (#68)
• fix: Fix precision usage example
• fix: Fix setup of InfluxDB v2.x dbrp mapping
• fix: Fix socket test initializer
• fix: Fix spec of write/3 of Instream.Connection.{QueryRunnerV1, QueryRunnerV2} (#74)
• fix: Fix test values passed for bucket/org override testing
• fix: Fix the pool_timeout param usage (#81)
• fix: Fix typespec for Config.get/4
• change: Add "location: :keep" to relevant macros
• change: Add "test" branch to run-on-push action config
• change: Add @impl attributes to series functions
• change: Add Elixir 1.11.x to InfluxDB v2.x matrix
• change: Add Elixir 1.13.x to testing matrix
• change: Add Elixir 1.14.x to testing matrix
• change: Add Elixir 1.15.x to testing matrix
• …and 280 more

Written by watchdog.canine.dev from the codebase's own history, inside the signed delivery this page is composed from.

mneudert/instream was measured the same way every project in this corpus was: the same rubric, at a pinned commit, with the result published in full. Point a surveyor at a repository you know and see whether you agree with it.