AaronPo/the-biergarten-app
1
0
Fork 0
mirror of https://github.com/aaronpo97/the-biergarten-app.git synced 2026-06-01 01:54:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add future plans, new design, and ethics document updates to Pipeline (#212)

Browse Source
This commit is contained in:
Aaron Po
2026-04-27 15:41:28 -04:00
committed by GitHub
parent d47e3ed7f0
commit e8c5b8a80c
13 changed files with 1418 additions and 159 deletions
Download Patch File Download Diff File Expand all files Collapse all files

324
pipeline/ETHICS-AND-KNOWN-ISSUES.md Normal file
View File

@@ -0,0 +1,324 @@
# Ethics, Bias, and Known Issues
This document covers the ethical context of the Biergarten Pipeline's output,
the model's biases, and known issues including hallucinated brewing science and
low-resource language failures.
> Note that all testing was used using `google_gemma-4-E4B-it-Q6_K.gguf`.
## Table of Contents
- [What This Dataset Is](#what-this-dataset-is)
- [What This Dataset Is Not](#what-this-dataset-is-not)
- [Model Bias and Language Quality](#model-bias-and-language-quality)
- [Western and Eurocentric Lens](#western-and-eurocentric-lens)
- [Wikipedia Enrichment](#wikipedia-enrichment)
- [The "Avoid AI Phrases" Prompt Instruction](#the-avoid-ai-phrases-prompt-instruction)
- [Known Issues](#known-issues)
- [Hallucinated Brewing Techniques](#hallucinated-brewing-techniques)
- [Low-Resource Language Hallucination](#low-resource-language-hallucination)
---
## What This Dataset Is
This is AI-generated fixture data for a proof-of-concept version of The
Biergarten App. Anyone who interacts with an application seeded from this
pipeline must be told upfront that the content is AI-generated.
---
## What This Dataset Is Not
The pipeline is not intended to produce accurate brewing science, faithful
cultural representation, or reliable local-language text. Hallucinations such as
invented fermentation techniques, or incoherent local-language prose, are
expected, observed, and partially documented in [Known Issues](#known-issues)
below.
Human control sits at the context layer (i.e. prompt design, Wikipedia
enrichment). Statistical output shapes in future pipeline stages (check-in
distributions, rating skews, activity profiles) will be handled the same way.
**Treat this data as an exercise in prompt engineering and model behaviour, not
as a source of truth for brewing techniques or cultural representation.**
**Natural language processing, although a powerful tool for data analysis and
generation is to be taken with scrutiny. Human language is not simply just data
points to be analyzed, but it also carries deep cultural and human meaning that
artificial intelligence is incapable of.**
---
## Model Bias and Language Quality
The underlying model's training biases surface within this pipeline.
Output quality tracks with how well a language is represented in the training
corpus: standard French (`fr-FR`) produces coherent text; regional variants like
`fr-CD` and `fr-CI` are noticeably weaker; low-resource languages like Welsh,
Māori, and Sicilian produce output that is syntactically plausible but often
semantically broken.
This is a property of the training distribution, not something that can be
mitigated through prompt design. This is a well-documented characteristic of
large language models trained predominantly on English-language
material.[^llm-bias]
Mitigations are documented in
[Known Issues: Low-Resource Language Hallucination](#low-resource-language-hallucination).
### Western and Eurocentric Lens
The model's training data skews heavily Western and North American. When
generating brewery descriptions for Kinshasa, Abidjan, or Osaka, for example, it
defaults to framing and cultural reference points drawn from that perspective
rather than from the lived context of those cities. Wikipedia enrichment grounds
some generation in city-specific material, but it does not eliminate the skew.
**Output should be read with an understanding of this bias.**
---
## Wikipedia Enrichment
City and beer context is fetched from the Wikipedia API. Wikipedia text is
co-licensed under the **Creative Commons Attribution-ShareAlike 4.0
International License (CC BY-SA 4.0)** and the **GNU Free Documentation License
(GFDL)**.[^wp-license]
Wikipedia's own accuracy limitations and editorial biases can propagate into
generated descriptions.
---
## The "Avoid AI Phrases" Prompt Instruction
The system prompt instructs the model to avoid common AI-generated phrasing
patterns. This is a prompt engineering experiment:
> How far can a model be pushed against its own stylistic defaults?
This is not an attempt to disguise the content as human-written. All downstream
consumers are informed of the AI-generated origin before engagement.
---
## Known Issues
### Hallucinated Brewing Techniques
When forced by the system prompt to generate a "highly specialized technical
brewing detail," the model frequently hallucinates fermentation science and
brewing chemistry. While the resulting sounds confident, it will be nonsensical
to reader with brewing and/or scientific expertise.
Small-parameter models such as Gemma 4 E4B cannot reliably satisfy prompts
calling for specialist brewing detail. This is consistent with the **CHOKE**
failure mode (Certain Hallucinations Overriding Known Evidence) [^llm-choke]
where a specific-sounding prompt causes a model to produce confident, coherent,
and factually wrong output.
#### Example — Osaka, produced using Gemma 4 E4B
```log
[2026-04-21 15:04:40.258] [info] 11. city="Osaka" country="Japan" state="Osaka" iso3166_2=JP-27 lat=34.6937 lon=135.5023
[2026-04-21 15:04:40.258] [info] brewery_name_en="Kani-no-Kuni Brewing"
[2026-04-21 15:04:40.258] [info] brewery_description_en="The humid scent of takoyaki and savory grilled squid always settles over the neighborhood early in the morning, mingling strangely with the metallic tang of spent grain outside our drying shed. We decided to build this place here because Osaka is a city that never pauses, a place built around the constant exchange of goods and tastes, and that is what we want our beer to reflect. Many people here are accustomed to the dependable, clean taste of the major pale lagers, but we are interested in exploring the deeper, more textured expressions of fermentation. Our base malts are specifically crafted by incorporating strains of local rice, like Yamada Nishiki, which we malt and use not for sake, but for its complex, buttery flavor profile during the mash. This combination, followed by a slow, anaerobic aging process, gives our beers a profound, satisfying depth that stands apart from the hurried mass-market style. The heavy, deep red brick of the original warehouse wall has absorbed decades of Kansai humidity and seasonal rains, creating a patina that tells the exact story of this district's tireless movement. We chose this structure not for its charm, but for its resilience and the sheer density of the local history held within its mortar. Our goal is simply to serve a drink worthy of this powerful trading city. If you are looking for a quiet spot away from the main thoroughfare, look for us just off the side street near Shinsekai."
[2026-04-21 15:04:40.258] [info] brewery_name_local="カニの国ブルワリー"
[2026-04-21 15:04:40.258] [info] brewery_description_local="早朝の、たこ焼きや香ばしいイカ焼きの湿った匂いは、いつも乾燥小屋の外にある使用済み麦芽の金属的な匂いと奇妙に混ざり合って近隣に漂います。私たちはこの場所に店を構えることを決めたのです。なぜなら、大阪は決して止まることのない都市であり、商品と味が絶え間なく交換されることで築かれた場所だからです。地元の多くの方々は、信頼できる大規模な淡麗ラガーの味が習慣になっていますが、私たちは発酵の、より深く、より複雑な表現を探求することに関心があります。私たちのベースモルトは、山田錦のような地元の米の品種を意図的に組み込んで作られています。この米を酒ではなく、麦芽として、仕込みの最中にその複雑でバターのような風味を引き出すために使用しています。この組み合わせを、ゆっくりとした嫌気的な熟成プロセスに続けることで、私たちのビールは、慌ただしい市場のスタイルとは一線を画す、深みのある、満足感のある複雑さを持っています。オリジナルの倉庫の重く深紅のレンガ壁は、関西特有の湿気と季節の雨を何十年も吸収し、この地区の絶え間ない動きの正確な物語を語るような古色を帯びています。私たちはこの構造物を、その魅力のためではなく、その回復力とモルタルに込められた地域の歴史の密度ゆえに選びました。私たちの目標は、ただこの力強い交易都市に値する飲み物を提供することだけです。もしメインの通りから離れた静かな場所をお探しなら、新世界近くの脇道にある私たちを探してください。"
```
A review of the following text for brewing techniques reveals several
inaccuracies, and no comments could be made on the local-language version due to
my own lack of proficiency in Japanese:
#### 1. "Buttery flavours" framed as a desirable malt-derived flavour
**Incorrect.**
Diacetyl is a fermentation byproduct of yeast metabolism, not a malt-derived
compound.[^diacetyl-source] Diacetyl produces a buttery or butterscotch
off-flavour and is carefully managed in many beer styles, in particular lighter
beers, through a process called a _diacetyl rest_. In this process, fermentation
temperature is briefly raised to allow yeast to reabsorb the compound before
packaging.[^diacetyl-rest]
The Oxford Companion to Beer claims that, while low levels are tolerable in some
ales and stouts, diacetyl is considered undesirable at any perceptible
concentration when it results from bacterial contamination or stressed
fermentation.[^oxford-beer]
#### 2. Yamada Nishiki sake rice described as a self-saccharifying base malt
**Incorrect.**
Yamada Nishiki (_山田錦_) is a short-grain Japanese rice bred specifically for
sake production.[^yn-wiki] Its value lies in its large starchy core
(_shinpaku_), low protein content, and amenability to _koji_ mold penetration
during saccharification.[^yn-sakestreet] Sake brewing does not use the grain's
own enzymatic activity for saccharification — it relies on _Aspergillus oryzae_
(koji mold) grown on a portion of the steamed rice to convert starches to
fermentable sugars.[^yn-sakeonline]
#### 3. "Anaerobic aging" presented as a differentiating technique
**Misleading**
Anaerobic conditions during packaging and aging are not differentiating
technique. Anaerobic conditions are the standard baseline for all commercial
beer production. Breweries exclude oxygen as a top priority for packaging and
shelf stability; published research in _Microbiology Spectrum_ confirms that
packaged beer constitutes an anaerobic environment by definition.[^anaerobic]
Professional packaging lines use CO_2 purges and closed transfers specifically
to maintain this state.[^packaging] Framing anaerobic aging as a distinctive
practice is misleading and suggests hallucinated output.
### Low-Resource Language Hallucination
The generation pipeline passes local language codes to the model to retrieve a
translated `description_local`. Output quality is reliable for high-resource
languages such as French, though it may struggle with regional variants and
idiomatic phrasing.
```json
[
{
"city": "Kinshasa",
"state_province": "Kinshasa",
"iso3166_2": "CD-KN",
"country": "Democratic Republic of the Congo",
"iso3166_1": "CD",
"latitude": -4.4419,
"longitude": 15.2663,
"local_languages": ["fr-CD", "ln"]
},
{
"city": "Paris",
"state_province": "Île-de-France",
"iso3166_2": "FR-IDF",
"country": "France",
"iso3166_1": "FR",
"latitude": 48.8566,
"longitude": 2.3522,
"local_languages": ["fr-FR"]
},
{
"city": "Abidjan",
"state_province": "Abidjan",
"iso3166_2": "CI-AB",
"country": "Ivory Coast",
"iso3166_1": "CI",
"latitude": 5.36,
"longitude": -4.0083,
"local_languages": ["fr-CI"]
},
{
"city": "Montreal",
"state_province": "Quebec",
"iso3166_2": "CA-QC",
"country": "Canada",
"iso3166_1": "CA",
"latitude": 45.5017,
"longitude": -73.5673,
"local_languages": ["fr-CA"]
},
{
"city": "Brussels",
"state_province": "Brussels-Capital Region",
"iso3166_2": "BE-BRU",
"country": "Belgium",
"iso3166_1": "BE",
"latitude": 50.8503,
"longitude": 4.3517,
"local_languages": ["fr-BE", "nl-BE"]
}
]
```
This dataset, when fed into the pipeline will often times reason that a local variant of French is needed, but will often times just default to a standardized dialect of French, devoid of any cultural or linguistic nuance.
For languages such as Welsh (Wales), Māori (Aotearoa/New Zealand), or Sicilian
(Sicily, Italy), the model can generate text that looks syntactically plausible
but is semantically incoherent. This comes from limited training-data coverage
rather than prompt engineering.
Output sample:
[./out-sample/french-cities.example](out-sample/french-cities.example)
#### Proposed Mitigations
- **Prevention via allowlist:** introduce a high-resource language allowlist. If
a location's code is unlisted, skip `description_local` generation and fall
back to English.
- **Upstream sanitization:** strip known low-resource language codes from the
`locations.json` payload before generation.
- **Downstream flagging:** add a `description_local_confidence` column to the
SQLite schema so downstream applications can filter or flag potentially
hallucinated text by language tier.
---
## Footnotes
[^llm-choke]: CHOKE (Certain Hallucinations Overriding Known Evidence) is a hallucination failure mode defined by Simhi et al. (2025), in which a model that can consistently answer a question correctly produces a confident, wrong response when the prompt is trivially perturbed. Source: Trust Me, I'm Wrong: LLMs Hallucinate with Certainty Despite Knowing the Answer — Adi Simhi, Itay Itzhak, Fazl Barez, Gabriel Stanovsky, Yonatan Belinkov.
[^llm-bias]:
e.g., Blasi et al. (2022), "Systematic Inequalities in Language Technology
Performance across the World's Languages," _ACL Anthology_. The pattern is
consistent with models trained predominantly on English-language web
corpora.
[^wp-license]:
Source:
[Wikipedia:FAQ/Copyright](https://en.wikipedia.org/wiki/Wikipedia:FAQ/Copyright).
[^cc-sa]:
Creative Commons CC BY-SA 4.0 deed: "If you remix, transform, or build upon
the material, you must distribute your contributions under the same license
as the original." Source:
[creativecommons.org/licenses/by-sa/4.0](https://creativecommons.org/licenses/by-sa/4.0/deed.en).
[^diacetyl-source]:
White Labs confirms that diacetyl is a yeast-derived fermentation byproduct: specifically, a compound produced during amino acid metabolism that leaks out of the yeast cell and oxidises into its characteristic buttery off-flavour. It is generally considered undesirable at any perceived level in most styles, though low levels are tolerated in some English ales and European lagers.
Source:
[whitelabs.com — Compound Spotlight: Diacetyl](https://www.whitelabs.com/news-update-detail?id=54).
[^diacetyl-rest]:
Brewing Science Institute: diacetyl "is produced during the fermentation
process, primarily as a byproduct of yeast metabolism… generally considered
a flaw in most beer styles." Source:
[brewingscience.com — Diacetyl: Understanding Its Role as an Off-Flavor in Beer](https://brewingscience.com/diacetyl-understanding-its-role-as-an-off-flavor-in-beer/).
[^oxford-beer]:
Oxford Companion to Beer via _Beer & Brewing_: "At low to moderate levels,
diacetyl can be perceived as a positive flavor characteristic in some ales
and stouts" but "particularly unwelcome in lager-style beers." Source:
[beerandbrewing.com — diacetyl](https://www.beerandbrewing.com/dictionary/48TDqQibPi).
[^yn-wiki]:
Wikipedia: "Yamada Nishiki (山田錦) is a short-grain Japanese rice famous
for its use in high-quality sake." Source:
[en.wikipedia.org/wiki/Yamada_Nishiki](https://en.wikipedia.org/wiki/Yamada_Nishiki).
[^yn-sakestreet]:
Sake Street: Yamadanishiki's large _shinpaku_ allows koji mold to penetrate
to the centre of the rice grain, making it "particularly suitable for
producing good koji." Source:
[sakestreet.com — What is Yamadanishiki?](https://sakestreet.com/en/media/what-is-yamadanishiki).
[^yn-sakeonline]:
Sake Online: "Steamed rice is added to make koji (rice malt) and yeast
starter, which promotes alcohol fermentation." Source:
[sakeonline.com.au — Types of Sake Rice: Yamada Nishiki](https://sakeonline.com.au/blogs/news/types-of-sake-rice-yamada-nishiki-and-its-characteristics).
[^anaerobic]:
Pai et al. (2022): "Breweries have recognized oxygen exclusion as a top
priority for the proper packaging and aging of beer… packaged beer is an
anaerobic environment." _Microbiology Spectrum._ Source:
[journals.asm.org](https://journals.asm.org/doi/10.1128/spectrum.02656-22).
[^packaging]:
Beer Production Processes (oboe.com): Professional packaging lines use
double CO_2 pre-evacuation cycles and closed transfers "so the beer moves in
a completely anaerobic environment." Source:
[oboe.com — Flavor Quality Control](https://oboe.com/learn/beer-production-processes-308lmf/flavor-quality-control-4).

276
pipeline/README.md
View File

@@ -1,13 +1,20 @@
# Biergarten Pipeline
A C++20 command-line pipeline that samples city records from local JSON, enriches each with Wikipedia context, and generates bilingual brewery names and descriptions via a local GGUF model or a deterministic mock.
A C++20 command-line pipeline that samples city records from local JSON,
enriches each with Wikipedia context, and generates bilingual brewery names and
descriptions via a local GGUF model or a deterministic mock.
> **This pipeline produces AI-generated data.** It is not a source of truth for
> brewing techniques, cultural representation, or local-language accuracy. See
> [ETHICS-AND-KNOWN-ISSUES.md](ETHICS-AND-KNOWN-ISSUES.md) for full
> documentation of limitations, hallucination patterns, and bias.
---
## Table of Contents
- [How It Fits The Main App](#how-it-fits-the-main-app)
- [Tech Stack](#tech-stack)
- [Quick Start](#quick-start)
- [Build](#build)
- [Model](#model)
- [Run](#run)
@@ -16,19 +23,20 @@ A C++20 command-line pipeline that samples city records from local JSON, enriche
- [Key Components](#key-components)
- [Runtime Behaviour](#runtime-behaviour)
- [Generated Output](#generated-output)
- [Language Generation Quality](#language-generation-quality)
- [Known Issues](#known-issues)
- [Tech Stack](#tech-stack)
- [Tested Hardware](#tested-hardware)
- [Fixture Strategy](#fixture-strategy)
- [Repo Layout](#repo-layout)
- [Code Tour](#code-tour)
- [Fixture Strategy](#fixture-strategy)
- [Next Steps](#next-steps)
---
## How It Fits The Main App
The pipeline is a data ingestion layer. It sits outside the web app runtime and produces seed records the app imports at startup or during a dedicated seed step.
The pipeline is a data ingestion layer. It sits outside the web app runtime and
produces seed records the app imports at startup or during a dedicated seed
step.
| Planned app area | Pipeline contribution |
| -------------------------------- | ------------------------------------------------------------------ |
@@ -39,35 +47,20 @@ The pipeline is a data ingestion layer. It sits outside the web app runtime and
---
## Tech Stack
## Quick Start
- C++20
- CMake 3.24+
- Boost.JSON, Boost.ProgramOptions, Boost.DI
- spdlog
- libcurl
- SQLite amalgamation fetched and compiled via CMake FetchContent
- llama.cpp
### Build
The build fetches Boost.DI, spdlog, llama.cpp, and SQLite via CMake. Metal is enabled on Apple Silicon; CUDA or HIP/ROCm is detected on Linux when the toolkit is present.
> **Code Style:** Modern C++20 throughout - RAII for ownership, `std::unique_ptr` for injected dependencies, `std::optional` for parse outcomes, `std::span` for read-only views over generated city data, structured bindings in pipeline loops. Formatting follows the Google C++ Style Guide via `.clang-format` with a narrow column limit and two-space indentation.
---
## Build
Requirements: C++20 compiler, CMake 3.24+, libcurl, Boost (JSON and ProgramOptions).
SQLite is fetched from the upstream amalgamation, so no system SQLite package is required.
Requirements: C++20 compiler, CMake 3.24+, libcurl, Boost (JSON and
ProgramOptions). SQLite is fetched from the upstream amalgamation, so no system
SQLite package is required.
```bash
cmake -S . -B build
cmake --build build
```
---
## Model
### Model
> Skip this step if you only need `--mocked`.
@@ -78,18 +71,18 @@ curl -L \
https://huggingface.co/bartowski/google_gemma-4-E4B-it-GGUF/resolve/main/google_gemma-4-E4B-it-Q6_K.gguf?download=true
```
---
### Run
## Run
Run from `build/` so the copied `locations.json` and `prompts/` are available. Each run also writes a fresh dated SQLite file such as `biergarten_seed_2026-04-19T15-30-45.123456Z.sqlite` into the working directory.
Run from `build/` so the copied `locations.json` and `prompts/` are available.
Each run also writes a fresh dated SQLite file such as
`biergarten_seed_2026-04-19T15-30-45.123456Z.sqlite` into the working directory.
```bash
./biergarten-pipeline --mocked
./biergarten-pipeline --model models/google_gemma-4-E4B-it-Q6_K.gguf --temperature 1.0 --top-p 0.95 --top-k 64 --n-ctx 8192 --seed -1
```
### CLI Flags
#### CLI Flags
| Flag | Purpose |
| --------------- | ------------------------------------------------------- |
@@ -102,9 +95,12 @@ Run from `build/` so the copied `locations.json` and `prompts/` are available. E
| `--seed` | Random seed. Default: `-1` (random at runtime). |
| `--help, -h` | Print usage and exit. |
`--mocked` and `--model` are mutually exclusive. Omitting both exits with an error before the pipeline starts. Sampling flags are ignored when `--mocked` is set.
`--mocked` and `--model` are mutually exclusive. Omitting both exits with an
error before the pipeline starts. Sampling flags are ignored when `--mocked` is
set.
The post-build step copies `prompts/` into `build/prompts/`. Rebuild after editing `prompts/system.md`.
The post-build step copies `prompts/` into `build/prompts/`. Rebuild after
editing `prompts/system.md`.
---
@@ -121,41 +117,58 @@ The post-build step copies `prompts/` into `build/prompts/`. Rebuild after editi
| Store | `SqliteExportService` writes each successful brewery into a fresh dated `.sqlite` database with normalized location and brewery tables. |
| Log | `spdlog` writes results and warnings to the console. |
If enrichment or generation fails for a city, that city is skipped and the pipeline continues.
If enrichment or generation fails for a city, that city is skipped and the
pipeline continues.
### Key Components
- `src/main.cc` - argument parsing and Boost.DI composition root.
- `JsonLoader` - validates curated location input.
- `WikipediaService` - queries Wikipedia extracts, caches results, returns empty context on failure.
- `LlamaGenerator` - formats prompts for Gemma 4, validates JSON output, retries malformed responses up to three times. If output looks truncated, the retry raises the token budget before trying again.
- `MockGenerator` - stable hash-based output so the same city input always produces the same brewery.
- `SqliteExportService` - creates a dated SQLite file per run and persists each successful brewery into normalized tables.
- Brewery payloads include English and local-language name and description fields.
- `src/main.cc` — argument parsing and Boost.DI composition root.
- `JsonLoader` — validates curated location input.
- `WikipediaService` — queries Wikipedia extracts, caches results, returns empty
context on failure.
- `LlamaGenerator` — formats prompts for Gemma 4, validates JSON output, retries
malformed responses up to three times. If output looks truncated, the retry
raises the token budget before trying again.
- `MockGenerator` — stable hash-based output so the same city input always
produces the same brewery.
- `SqliteExportService` — creates a dated SQLite file per run and persists each
successful brewery into normalized tables.
- Brewery payloads include English and local-language name and description
fields.
### Runtime Behaviour
`WikipediaService` queries city, country, and beer-related Wikipedia extracts using its configured lookup, then caches the first successful response per query string. The fetched extract text is included in the prompt as context for generation.
`WikipediaService` queries city, country, and beer-related Wikipedia extracts
using its configured lookup, then caches the first successful response per query
string. The fetched extract text is included in the prompt as context for
generation.
`GetLocationContext()` returns an empty string when the web client is unavailable or when lookup/parsing fails.
`GetLocationContext()` returns an empty string when the web client is
unavailable or when lookup/parsing fails.
`LlamaGenerator` validates model output as structured JSON. The retry path exists as a safety hatch for cases where the reasoning block consumes available token budget and compresses the JSON output space. All runs to date have produced valid output on the first pass; the path is kept for resilience.
`LlamaGenerator` validates model output as structured JSON. The retry path
exists as a safety hatch for cases where the reasoning block consumes available
token budget and compresses the JSON output space. All runs to date have
produced valid output on the first pass; the path is kept for resilience.
`MockGenerator` uses stable hashes for repeatable output in demos and Storybook runs.
`MockGenerator` uses stable hashes for repeatable output in demos and Storybook
runs.
### Process Flow - Activity Diagram
![An activity diagram](./diagrams/activity-diagram.svg)
![An activity diagram](./diagrams/current/output/activity.svg)
### Architectural Overview - Class Diagram
![A class diagram](./diagrams/class-diagram.svg)
![A class diagram](./diagrams/current/output/class.svg)
---
## Generated Output
Each successful run stores a `GeneratedBrewery` pair with the source location and a `BreweryResult` payload. The same generated records are also written to a fresh SQLite export file named with the current UTC timestamp.
Each successful run stores a `GeneratedBrewery` pair with the source location
and a `BreweryResult` payload. The same generated records are also written to a
fresh SQLite export file named with the current UTC timestamp.
| Field | Meaning |
| ------------------- | ------------------------------------------ |
@@ -164,7 +177,8 @@ Each successful run stores a `GeneratedBrewery` pair with the source location an
| `name_local` | Brewery name in the local language. |
| `description_local` | Brewery description in the local language. |
The log dump also includes city, country, state or province, ISO subdivision code, latitude, and longitude for each entry.
The log dump also includes city, country, state or province, ISO subdivision
code, latitude, and longitude for each entry.
### Consumer Data Shape
@@ -180,80 +194,25 @@ The log dump also includes city, country, state or province, ISO subdivision cod
---
## Language Generation Quality
## Tech Stack
The generation pipeline passes local language codes to the model to retrieve a translated `description_local`.
- C++20
- CMake 3.24+
- Boost.JSON, Boost.ProgramOptions, Boost.DI
- spdlog
- libcurl
- SQLite amalgamation fetched and compiled via CMake FetchContent
- llama.cpp
Output quality is reliable for high-resource languages such as French, though it may struggle with regional variants and idiomatic phrasing. This can be seen with these data points:
The build fetches Boost.DI, spdlog, llama.cpp, and SQLite via CMake. Metal is
enabled on Apple Silicon; CUDA or HIP/ROCm is detected on Linux when the toolkit
is present.
```json
[
{
"city": "Kinshasa",
"state_province": "Kinshasa",
"iso3166_2": "CD-KN",
"country": "Democratic Republic of the Congo",
"iso3166_1": "CD",
"latitude": -4.4419,
"longitude": 15.2663,
"local_languages": ["fr-CD", "ln"]
},
{
"city": "Paris",
"state_province": "Île-de-France",
"iso3166_2": "FR-IDF",
"country": "France",
"iso3166_1": "FR",
"latitude": 48.8566,
"longitude": 2.3522,
"local_languages": ["fr-FR"]
},
{
"city": "Abidjan",
"state_province": "Abidjan",
"iso3166_2": "CI-AB",
"country": "Ivory Coast",
"iso3166_1": "CI",
"latitude": 5.36,
"longitude": -4.0083,
"local_languages": ["fr-CI"]
},
{
"city": "Montreal",
"state_province": "Quebec",
"iso3166_2": "CA-QC",
"country": "Canada",
"iso3166_1": "CA",
"latitude": 45.5017,
"longitude": -73.5673,
"local_languages": ["fr-CA"]
},
{
"city": "Brussels",
"state_province": "Brussels-Capital Region",
"iso3166_2": "BE-BRU",
"country": "Belgium",
"iso3166_1": "BE",
"latitude": 50.8503,
"longitude": 4.3517,
"local_languages": ["fr-BE", "nl-BE"]
}
]
```
Output sample: [./out-sample/french-cities.example](out-sample/french-cities.example)
### Known Issues
#### Low-Resource Language Hallucination
For languages such as Welsh (Wales), Maori (Aotearoa/New Zealand), or Sicilian (Sicily, Italy), the model can generate text that looks syntactically plausible but is semantically incoherent. This comes from limited training-data coverage rather than prompt engineering.
#### Proposed Mitigations
- **Prevention via allowlist:** introduce a high-resource language allowlist. If a location's code is unlisted, skip `description_local` generation and fall back to English.
- **Upstream sanitization:** strip known low-resource language codes from the `locations.json` payload before generation.
- **Downstream flagging:** add a `description_local_confidence` column to the SQLite schema so downstream applications can filter or flag potentially hallucinated text by language tier.
> **Code Style:** Modern C++20 throughout — RAII for ownership,
> `std::unique_ptr` for injected dependencies, `std::optional` for parse
> outcomes, `std::span` for read-only views over generated city data, structured
> bindings in pipeline loops. Formatting follows the Google C++ Style Guide via
> `.clang-format` with a narrow column limit and two-space indentation.
---
@@ -283,62 +242,83 @@ For languages such as Welsh (Wales), Maori (Aotearoa/New Zealand), or Sicilian (
---
## Fixture Strategy
- `--mocked` for stable fixtures, repeatable screenshots, and Storybook runs.
- `--model` when geographically grounded content matters for demos.
- Keep `locations.json` structured enough to support discovery and future
filtering.
- Treat SQLite output as seed material for the app's brewery domain, not
production data.
---
## Repo Layout
| Path | Purpose |
| ---------------- | ---------------------------------------------- |
| ---------------------------- | -------------------------------------------------- |
| `includes/` | Public headers and shared models. |
| `src/` | Implementation files. |
| `locations.json` | Curated city input copied into the build tree. |
| `prompts/` | System prompt used by the model-backed path. |
| `diagrams/` | Architecture and pipeline diagrams. |
| `ETHICS-AND-KNOWN-ISSUES.md` | Ethics, bias, hallucination analysis, mitigations. |
---
## Code Tour
- `src/main.cc` - argument parsing and DI composition root.
- `src/biergarten_data_generator/` - orchestration, sampling, logging, and export.
- `src/services/wikipedia/` - enrichment service and cache.
- `src/services/sqlite/` - SQLite export implementation.
- `src/data_generation/llama/` - local inference, prompt loading, output validation.
- `src/data_generation/mock/` - deterministic fallback.
---
## Fixture Strategy
- `--mocked` for stable fixtures, repeatable screenshots, and Storybook runs.
- `--model` when geographically grounded content matters for demos.
- Keep `locations.json` structured enough to support discovery and future filtering.
- Treat SQLite output as seed material for the app's brewery domain, not production data.
- `src/main.cc` — argument parsing and DI composition root.
- `src/biergarten_data_generator/` — orchestration, sampling, logging, and
export.
- `src/services/wikipedia/` — enrichment service and cache.
- `src/services/sqlite/` — SQLite export implementation.
- `src/data_generation/llama/` — local inference, prompt loading, output
validation.
- `src/data_generation/mock/` — deterministic fallback.
---
## Next Steps
The pipeline currently produces city-aware brewery records and dated SQLite exports. The next passes add additional fixture types so the app can exercise the full brewery domain without live data.
The pipeline currently produces city-aware brewery records and dated SQLite
exports. The next passes add additional fixture types so the app can exercise
the full brewery domain without live data.
### Testing _(Very High Importance)_
### Testing — Very High Priority
- Unit test JSON validation and retry logic against malformed, truncated, and empty model outputs.
- Integration test the enrichment pipeline with missing context, short context, and fake context inputs.
- Adversarial context tests: feed plausible but geographically incorrect Wikipedia extracts and verify the model does not silently blend them with training data.
- Verify bilingual enrichment behaviour when only an English extract is available versus when both extracts are present.
- Confirm the retry path is reachable when the reasoning block consumes available token budget.
- Unit test JSON validation and retry logic against malformed, truncated, and
empty model outputs.
- Integration test the enrichment pipeline with missing context, short context,
and fake context inputs.
- Adversarial context tests: feed plausible but geographically incorrect
Wikipedia extracts and verify the model does not silently blend them with
training data.
- Verify bilingual enrichment behaviour when only an English extract is
available versus when both extracts are present.
- Confirm the retry path is reachable when the reasoning block consumes
available token budget.
### Beer Generation
Generate catalog entries with style, ABV, IBU, color, aroma notes, and food pairing hints. Link beers back to breweries and cities. Keep style coverage wide enough to exercise search, sort, and category filters.
Generate catalog entries with style, ABV, IBU, color, aroma notes, and food
pairing hints. Link beers back to breweries and cities. Keep style coverage wide
enough to exercise search, sort, and category filters.
### User Generation
Generate user profiles with stable names, bios, locale hints, and preference signals. Include stable IDs for downstream fixture joins. Keep output deterministic for screenshots while allowing larger randomized batches.
Generate user profiles with stable names, bios, locale hints, and preference
signals. Include stable IDs for downstream fixture joins. Keep output
deterministic for screenshots while allowing larger randomized batches.
### Check-In System
Produce timestamped check-in events between users and breweries. Use a J-curve activity profile - a small set of users accounts for most check-ins, the rest appear occasionally. Add bursty behaviour around weekends and travel periods.
Produce timestamped check-in events between users and breweries. Use a J-curve
activity profile — a small set of users accounts for most check-ins, the rest
appear occasionally. Add bursty behaviour around weekends and travel periods.
### Beer Ratings
Generate rating events with a strong positive skew and a long tail of lower scores. Avoid uniform distributions. Attach timestamps and user IDs so the app can compute averages, trends, and per-style comparisons.
Generate rating events with a strong positive skew and a long tail of lower
scores. Avoid uniform distributions. Attach timestamps and user IDs so the app
can compute averages, trends, and per-style comparisons.

1
pipeline/diagrams/activity-diagram.svg
View File

File diff suppressed because one or more lines are too long

34
pipeline/diagrams/biergarten-weizen-theme.puml Normal file
View File

@@ -0,0 +1,34 @@
skinparam shadowing false
skinparam backgroundColor #FCFCF7
skinparam defaultFontName "DM Sans"
skinparam defaultFontColor #14180C
skinparam titleFontName "Volkhov"
skinparam titleFontColor #14180C
skinparam ArrowColor #656F33
skinparam NoteBackgroundColor #DBEEDD
skinparam NoteFontColor #14180C
skinparam NoteBorderColor #4A5837
skinparam SwimlaneBorderColor #4A5837
skinparam SwimlaneBorderThickness 1
skinparam activityStartColor #EBECE3
skinparam activityEndColor #4A5837
skinparam activityStopColor #4A5837
skinparam ActivityBackgroundColor #EBECE3
skinparam ActivityBorderColor #4A5837
skinparam ActivityDiamondBackgroundColor #CBD2B5
skinparam ActivityDiamondBorderColor #4A5837
skinparam packageStyle rectangle
skinparam packageBackgroundColor #F1F3EA
skinparam packageBorderColor #4A5837
skinparam packageFontColor #14180C
skinparam classBackgroundColor #EBECE3
skinparam classBorderColor #4A5837
skinparam classFontColor #14180C
skinparam classAttributeFontColor #3F4724
skinparam classStereotypeFontColor #4A5837
skinparam interfaceBackgroundColor #DBEEDD
skinparam interfaceBorderColor #4A5837
skinparam interfaceFontColor #14180C
skinparam enumBackgroundColor #E4E6D8
skinparam enumBorderColor #4A5837
skinparam enumFontColor #14180C

1
pipeline/diagrams/class-diagram.svg
View File

File diff suppressed because one or more lines are too long

0
pipeline/diagrams/activity-diagram.puml → pipeline/diagrams/current/activity.puml
View File

0
pipeline/diagrams/class-diagram.puml → pipeline/diagrams/current/class.puml
View File

1
pipeline/diagrams/current/output/activity.svg Normal file
View File

File diff suppressed because one or more lines are too long

1
pipeline/diagrams/current/output/class.svg Normal file
View File

File diff suppressed because one or more lines are too long

360
pipeline/diagrams/planned/activity.puml Normal file
View File

@@ -0,0 +1,360 @@
@startuml biergarten_activity
!include ../biergarten-weizen-theme.puml
skinparam defaultFontSize 13
skinparam titleFontSize 20
title The Biergarten Data Pipeline — Activity Diagram
|Main|
start
:ParseArguments(argc, argv);
if (Invalid args?) then (yes)
:spdlog::error;
stop
else (no)
endif
:Init CurlGlobalState & LlamaBackendState;
:Build DI injector;
:Initialize SqliteExportService;
note right
Opens SQLite connection.
(Transactions are now managed
per-phase via batching).
end note
:Create BoundedChannel<LogEntry> log_ch;
:Spawn Log Worker thread;
note right
Log worker drains log_ch for the
entire pipeline lifetime.
All workers emit LogEntry structs
via PipelineLogger -- never spdlog directly.
end note
:BiergartenPipelineOrchestrator::Run();
|BiergartenPipelineOrchestrator::Run()|
fork
:JsonLoader::LoadBeerStyles("beer-styles.json");
:EnrichmentService::PreWarmBeerStyleCache(beer_styles);
fork again
:JsonLoader::LoadLocations("locations.json");
:EnrichmentService::PreWarmLocationCache(sampled_locations);
end fork
fork
:JsonLoader::LoadNamesByCountry("names-by-country.json");
fork again
:JsonLoader::LoadPersonas("personas.json");
end fork
' ═══════════════════════════════════════════
' PHASE 0 — USER GENERATION
' ═══════════════════════════════════════════
|Orchestrator|
:RunUserPhase(sampled_locations);
:Create BoundedChannels\n(loc_ch, exp_ch);
fork
|Orchestrator|
:Loop: Send Locations -> loc_ch;
:Close loc_ch;
note right
Producer closes loc_ch.
LLM Worker while loop
terminates on empty + closed.
end note
fork again
|LLM Worker|
while (loc_ch has items?) is (yes)
:Receive Location;
:GetLocationContextFromCache(location);
note right
Guaranteed cache hit from startup.
end note
:IPersonaSelectionStrategy::SelectPersona(\n personas_palette_);
note right
Guaranteed cache hit from startup.
Returns a Persona struct carrying
style_affinities, abv_range,
ibu_preference, checkin_weight.
end note
:NamesByCountry::SampleName(\n location.iso3166_1);
note right
Deterministic lookup -- no LLM involved.
Name selected from pre-keyed table
and passed into the generation prompt.
end note
:GenerateUser(enriched_city, persona, sampled_name)\nvia DataGenerator;
note right
LLM receives: EnrichedCity context + persona
description + sampled name. Generates
bio and preference signals grounded
in locale and persona.
end note
:PipelineLogger::Log(Info, UserGeneration,\n city, user_id, "llm");
:Send GeneratedUser -> exp_ch;
endwhile (no)
:Close exp_ch;
note right
Producer closes exp_ch.
SQLite Worker while loop
terminates on empty + closed.
end note
fork again
|SQLite Worker|
:BEGIN TRANSACTION;
while (exp_ch has items?) is (yes)
:Receive GeneratedUser;
:ProcessUser(user);
:PipelineLogger::Log(Info, UserGeneration,\n city, user_id, "sqlite");
:Append -> user_pool_;
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
endwhile (no)
:COMMIT (Final);
end fork
|Orchestrator|
:Join LLM Worker, SQLite Worker;
' ═══════════════════════════════════════════
' PHASE 1a — BREWERY GENERATION
' ═══════════════════════════════════════════
:RunBreweryPhase(sampled_locations);
:Create BoundedChannels\n(loc_ch, exp_ch);
fork
|Orchestrator|
:Loop: Sample User from user_pool_
and pair with Location;
:Send BreweryTask(Location, User) -> loc_ch;
:Close loc_ch;
fork again
|LLM Worker|
while (loc_ch has items?) is (yes)
:Receive BreweryTask(Location, User);
:GetLocationContextFromCache(task.location);
note right
Guaranteed cache hit from startup.
end note
:GenerateBrewery(enriched_city, context, task.user)\nvia DataGenerator;
note right
KV cache stays warm.
Brewery is linked to the sampled owner_user_id.
end note
:PipelineLogger::Log(Info,\n BreweryGeneration,\n city, brewery_id, "llm");
:Send GeneratedBrewery -> exp_ch;
endwhile (no)
:Close exp_ch;
fork again
|SQLite Worker|
:BEGIN TRANSACTION;
while (exp_ch has items?) is (yes)
:Receive GeneratedBrewery;
:ProcessBrewery(brewery);
:PipelineLogger::Log(Info,\n BreweryGeneration,\n city, brewery_id, "sqlite");
:Append -> brewery_pool_;
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
endwhile (no)
:COMMIT (Final);
end fork
|Orchestrator|
:Join LLM Worker, SQLite Worker;
note right
brewery_pool_ is now fully populated.
Phase 1b may begin.
end note
' ═══════════════════════════════════════════
' PHASE 1b — BEER GENERATION
' ═══════════════════════════════════════════
:RunBeerPhase();
:Create BoundedChannels\n(brew_ch, exp_ch);
fork
|Orchestrator|
:Loop: Send Breweries -> brew_ch;
:Close brew_ch;
fork again
|LLM Worker|
while (brew_ch has items?) is (yes)
:Receive GeneratedBrewery;
:IBeerSelectionStrategy::SelectStyles(\n brewery, beer_style_palette_);
while (For each selected BeerStyle?) is (remaining)
:GetStyleContextFromCache(style);
note right
Guaranteed cache hit from startup.
KV cache stays warm across all
beer generations -- system prompt
does not change within this phase.
end note
:GenerateBeer(brewery, style_context)\nvia DataGenerator;
:Attach GeneratedBeer to bundle;
endwhile (done)
:PipelineLogger::Log(Info,\n BeerGeneration,\n city, brewery_id, "llm");
:Send BeersBundle -> exp_ch;
endwhile (no)
:Close exp_ch;
fork again
|SQLite Worker|
:BEGIN TRANSACTION;
while (exp_ch has items?) is (yes)
:Receive BeersBundle;
while (For each beer in bundle?) is (remaining)
:Set beer.brewery_id from bundle;
:ProcessBeer(beer);
:Append -> beer_pool_;
endwhile (done)
:PipelineLogger::Log(Info,\n BeerGeneration,\n city, brewery_id, "sqlite");
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
endwhile (no)
:COMMIT (Final);
end fork
|Orchestrator|
:Join LLM Worker, SQLite Worker;
note right
Both brewery_pool_ and beer_pool_
are now completely populated.
Checkin and Follow phases may
now run in parallel.
end note
' ═══════════════════════════════════════════
' PHASE 2 — CHECKIN + FOLLOW GENERATION
' (parallel — both depend only on user_pool_
' and brewery_pool_ being fully populated)
' ═══════════════════════════════════════════
fork
|Orchestrator|
:RunCheckinPhase();
:ICheckinDistributionStrategy::\nAssignActivityWeights(user_pool_);
note right
Weights seeded from each user's
persona.checkin_weight. J-curve profile
emerges from persona distribution.
end note
:BEGIN TRANSACTION;
while (For each GeneratedUser in user_pool_?) is (remaining)
:CheckinsForUser(user, brewery_pool_.size());
while (For each checkin index?) is (remaining)
:TimestampFor(user, index);
:Select brewery from brewery_pool_;
:GenerateCheckin(user, brewery, timestamp)\nvia DataGenerator;
:ProcessCheckin(checkin);
:PipelineLogger::Log(Info, CheckinGeneration,\n nullopt, checkin_id, "sqlite");
:Append -> checkin_pool_;
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
endwhile (done)
endwhile (done)
:COMMIT (Final);
fork again
|Orchestrator|
:RunFollowPhase();
:IFollowGenerationStrategy::\nAssignFollowWeights(user_pool_);
note right
For RandomFollowStrategy, weights
are uniform. For ActivityWeightedFollowStrategy,
weights derived from user.activity_weight
so high-activity users attract more followers.
end note
:BEGIN TRANSACTION;
:IFollowGenerationStrategy::\nGenerateFollows(user_pool_);
note right
Self-follow constraint (follower_id != followed_id)
enforced here and at the DB schema level.
end note
while (For each GeneratedFollow?) is (remaining)
:ProcessFollow(follow);
:PipelineLogger::Log(Info, FollowGeneration,\n nullopt, follower_id, "sqlite");
:Append -> follow_pool_;
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
endwhile (done)
:COMMIT (Final);
end fork
|Orchestrator|
:Join CheckinPhase, FollowPhase;
note right
checkin_pool_ and follow_pool_
are now fully populated.
Rating phase may begin.
end note
' ═══════════════════════════════════════════
' PHASE 3 — RATING GENERATION
' ═══════════════════════════════════════════
:RunRatingPhase();
note right
Beer selection biased by
user.persona.style_affinities and abv_range.
Rating skew modulated per persona.
end note
:BEGIN TRANSACTION;
while (For each GeneratedCheckin in checkin_pool_?) is (remaining)
:Match brewery_id, select beer from beer_pool_\n(same brewery_id, biased by persona affinities);
if (Beer exists for brewery?) then (yes)
:GenerateRating(user, beer, checkin_id)\nvia DataGenerator;
:ProcessRating(rating);
:PipelineLogger::Log(Info, RatingGeneration,\n nullopt, rating_id, "sqlite");
if (Batch size reached?) then (yes)
:COMMIT & BEGIN;
else (no)
endif
else (no)
:PipelineLogger::Log(Warn, RatingGeneration,\n nullopt, brewery_id, "sqlite");
:Skip -- brewery has no beers;
endif
endwhile (done)
:COMMIT (Final);
' ═══════════════════════════════════════════
' TEARDOWN
' ═══════════════════════════════════════════
|Orchestrator|
:Finalize SqliteExportService;
note right
Safely closes the DB connection.
end note
:Close log_ch;
|Main|
:spdlog::info "Pipeline complete in X ms";
:Join Log Worker;
note right
Drain guarantees no LogEntry is
dropped at shutdown.
end note
stop
@enduml

559
pipeline/diagrams/planned/class.puml Normal file
View File

@@ -0,0 +1,559 @@
@startuml
' ==========================================
' CONFIGURATION & STYLING
' ==========================================
!include ../biergarten-weizen-theme.puml
skinparam classAttributeFontSize 9
skinparam defaultFontSize 25
skinparam titleFontSize 30
package "Domain: Models" {
class Location {
+ city : std::string
+ state_province : std::string
+ iso3166_2 : std::string
+ country : std::string
+ iso3166_1 : std::string
+ local_languages : std::vector<std::string>
+ latitude : double
+ longitude : double
}
class LocationContext {
+ text : std::string
+ completeness : Completeness
+ char_count : size_t
}
enum Completeness {
Full
Partial
Absent
}
class EnrichedCity {
+ location : Location
+ context : LocationContext
}
class BeerStyle {
+ name : std::string
+ description : std::string
+ min_abv : float
+ max_abv : float
+ min_ibu : int
+ max_ibu : int
}
class BreweryResult {
+ name_en : std::string
+ description_en : std::string
+ name_local : std::string
+ description_local : std::string
}
class BeerResult {
+ name_en : std::string
+ description_en : std::string
+ name_local : std::string
+ description_local : std::string
+ style : std::string
+ abv : float
+ ibu : int
}
class UserResult {
+ username : std::string
+ bio : std::string
+ activity_weight : float
}
class CheckinResult {
+ checked_in_at : std::string
+ note : std::string
}
class RatingResult {
+ score : float
+ note : std::string
}
class GenerationMetadata {
+ generation_id : uint64_t
+ generated_time : std::string
+ context_provided : bool
+ generated_with : std::string
}
class GeneratedBrewery {
+ brewery_id : uint64_t
+ location : Location
+ brewery : BreweryResult
+ context_completeness : LocationContext::Completeness
+ metadata : GenerationMetadata
}
class GeneratedBeer {
+ beer_id : uint64_t
+ brewery_id : uint64_t
+ location : Location
+ style : BeerStyle
+ beer : BeerResult
+ metadata : GenerationMetadata
}
class GeneratedUser {
+ user_id : uint64_t
+ location : Location
+ user : UserResult
+ metadata : GenerationMetadata
}
class GeneratedCheckin {
+ checkin_id : uint64_t
+ user_id : uint64_t
+ brewery_id : uint64_t
+ checkin : CheckinResult
+ metadata : GenerationMetadata
}
class GeneratedRating {
+ user_id : uint64_t
+ beer_id : uint64_t
+ checkin_id : uint64_t
+ rating : RatingResult
+ metadata : GenerationMetadata
}
class GeneratedFollow {
+ follower_id : uint64_t
+ followed_id : uint64_t
+ metadata : GenerationMetadata
}
class UserPersona {
+ name: std::string
+ description: std::string
+ style_affinities: std::vector<std::string>
}
LocationContext *-- Completeness
}
package "Domain: Application Configuration"{
class SamplingOptions {
+ temperature : float = 1.0F
+ top_p : float = 0.95F
+ top_k : uint32_t = 64
+ n_ctx : uint32_t = 8192
+ seed : int = -1
}
class GeneratorOptions {
+ model_path : std::filesystem::path
+ use_mocked : bool = false
+ sampling : SamplingOptions
}
class PipelineOptions {
+ output_path : std::filesystem::path
+ log_path : std::filesystem::path
}
class ApplicationOptions {
+ generator : GeneratorOptions
+ pipeline : PipelineOptions
}
' --- Domain Model Relationships ---
ApplicationOptions *-- GeneratorOptions
ApplicationOptions *-- PipelineOptions
GeneratorOptions *-- SamplingOptions
}
package "Domain: Policy" {
interface ContextStrategy <<interface>> {
+ QueriesFor(loc : const Location&) : std::vector<std::string>
+ MaxContextChars() : size_t
}
class BreweryContextStrategy {
+ QueriesFor(loc : const Location&) : std::vector<std::string>
+ MaxContextChars() : size_t
}
class BeerContextStrategy {
+ QueriesFor(loc : const Location&) : std::vector<std::string>
+ MaxContextChars() : size_t
}
interface SamplingStrategy <<interface>> {
+ Sample(locations : const std::vector<Location>&) : std::vector<Location>
}
class UniformSamplingStrategy {
- sample_size_ : size_t
+ Sample(locations : const std::vector<Location>&) : std::vector<Location>
}
interface BeerSelectionStrategy <<interface>> {
+ SelectStyles(brewery : const GeneratedBrewery&,\n palette : std::span<const BeerStyle>) : std::vector<BeerStyle>
}
class RandomBeerSelectionStrategy {
- rng_ : std::mt19937
- min_beers_ : size_t
- max_beers_ : size_t
+ SelectStyles(brewery : const GeneratedBrewery&,\n palette : std::span<const BeerStyle>) : std::vector<BeerStyle>
}
interface CheckinDistributionStrategy <<interface>> {
+ AssignActivityWeights(users : std::vector<GeneratedUser>&) : void
+ CheckinsForUser(user : const GeneratedUser&,\n brewery_count : size_t) : size_t
+ TimestampFor(user : const GeneratedUser&,\n index : size_t) : std::string
}
class JCurveCheckinStrategy {
- rng_ : std::mt19937
+ AssignActivityWeights(users : std::vector<GeneratedUser>&) : void
+ CheckinsForUser(user : const GeneratedUser&,\n brewery_count : size_t) : size_t
+ TimestampFor(user : const GeneratedUser&,\n index : size_t) : std::string
}
class RandomCheckinStrategy {
- rng_ : std::mt19937
- min_checkins_ : size_t
- max_checkins_ : size_t
+ AssignActivityWeights(users : std::vector<GeneratedUser>&) : void
+ CheckinsForUser(user : const GeneratedUser&,\n brewery_count : size_t) : size_t
+ TimestampFor(user : const GeneratedUser&,\n index : size_t) : std::string
}
interface FollowGenerationStrategy <<interface>> {
+ GenerateFollows(users : const std::vector<GeneratedUser>&) : std::vector<GeneratedFollow>
}
class RandomFollowStrategy {
- rng_ : std::mt19937
- min_follows_ : size_t
- max_follows_ : size_t
+ GenerateFollows(users : const std::vector<GeneratedUser>&) : std::vector<GeneratedFollow>
}
class ActivityWeightedFollowStrategy {
- rng_ : std::mt19937
- min_follows_ : size_t
- max_follows_ : size_t
+ GenerateFollows(users : const std::vector<GeneratedUser>&) : std::vector<GeneratedFollow>
}
}
package "Infrastructure: Logging" {
enum LogLevel {
Debug
Info
Warn
Error
}
enum PipelinePhase {
Startup
UserGeneration
BreweryAndBeerGeneration
CheckinGeneration
RatingGeneration
FollowGeneration
Teardown
}
class LogEntry {
+ timestamp : std::chrono::system_clock::time_point
+ level : LogLevel
+ phase : PipelinePhase
+ message : std::string
+ city : std::optional<std::string>
+ entity_id : std::optional<std::string>
+ worker : std::optional<std::string>
}
interface Logger <<interface>> {
+ Log(level, phase, message,\n city, entity_id, worker) : void
}
class PipelineLogger {
- log_ch_ : BoundedChannel<LogEntry>&
+ Log(level, phase, message,\n city, entity_id, worker) : void
}
class LogWorker {
- log_ch_ : BoundedChannel<LogEntry>&
+ Run() : void
- FormatTimestamp(tp) : std::string
- ToSpdlogLevel(level) : spdlog::level::level_enum
- ToString(phase) : std::string
}
' --- Logging Relationships ---
LogEntry *-- LogLevel
LogEntry *-- PipelinePhase
PipelineLogger ..> LogEntry : emits
LogWorker ..> LogEntry : consumes
}
package "Infrastructure: Pipeline Channel" {
class "BoundedChannel<T>" as BoundedChannel {
- queue_ : std::queue<T>
- mutex_ : std::mutex
- not_full_ : std::condition_variable
- not_empty_ : std::condition_variable
- capacity_ : size_t
- closed_ : bool
+ Send(item : T) : void
+ Receive() : std::optional<T>
+ Close() : void
}
}
package "Infrastructure: Data Preloading" {
interface DataPreloader <<interface>> {
+ LoadLocations(filepath : const std::filesystem::path&) : std::vector<Location>
+ LoadBeerStyles(filepath : const std::filesystem::path&) : std::vector<BeerStyle>
+ LoadPersonas(filepath : const std::filesystem::path&) : std::vector<Persona>
+ LoadNamesByCountry(filepath : const std::filesystem::path&) : NamesByCountry
}
class JsonLoader {
+ LoadLocations(filepath : const std::filesystem::path&) : std::vector<Location>
+ LoadBeerStyles(filepath : const std::filesystem::path&) : std::vector<BeerStyle>
+ LoadPersonas(filepath : const std::filesystem::path&) : std::vector<Persona>
+ LoadNamesByCountry(filepath : const std::filesystem::path&) : NamesByCountry
}
}
package "Infrastructure: Enrichment" {
interface EnrichmentService <<interface>> {
+ GetLocationContext(loc : const Location&,\n strategy : const ContextStrategy&) : LocationContext
}
class WikipediaService {
- client_ : std::unique_ptr<WebClient>
- extract_cache_ : std::unordered_map<std::string, std::string>
+ GetLocationContext(loc : const Location&,\n strategy : const ContextStrategy&) : LocationContext
- FetchExtract(query : std::string_view) : std::string
}
interface WebClient <<interface>> {
+ Get(url : const std::string&) : std::string
+ UrlEncode(value : const std::string&) : std::string
}
class CURLWebClient {
+ Get(url : const std::string&) : std::string
+ UrlEncode(value : const std::string&) : std::string
}
}
package "Infrastructure: Data Generation" {
interface DataGenerator <<interface>> {
+ GenerateBrewery(location : const Location&,\n context : const LocationContext&) : BreweryResult
+ GenerateBeer(brewery_id : uint64_t,\n location : const Location&,\n context : const LocationContext&,\n style : const BeerStyle&) : BeerResult
+ GenerateUser(location : const Location&) : UserResult
+ GenerateCheckin(user : const GeneratedUser&,\n brewery : const GeneratedBrewery&,\n timestamp : const std::string&) : CheckinResult
+ GenerateRating(user : const GeneratedUser&,\n beer : const GeneratedBeer&,\n checkin_id : uint64_t) : RatingResult
}
class MockGenerator {
+ GenerateBrewery(...) : BreweryResult
+ GenerateBeer(...) : BeerResult
+ GenerateUser(...) : UserResult
+ GenerateCheckin(...) : CheckinResult
+ GenerateRating(...) : RatingResult
- DeterministicHash(location : const Location&) : size_t
}
class LlamaGenerator {
- model_ : ModelHandle
- context_ : ContextHandle
- prompt_formatter_ : std::unique_ptr<PromptFormatter>
- rng_ : std::mt19937
+ GenerateBrewery(...) : BreweryResult
+ GenerateBeer(...) : BeerResult
+ GenerateUser(...) : UserResult
+ GenerateCheckin(...) : CheckinResult
+ GenerateRating(...) : RatingResult
- Load(opts : const GeneratorOptions&) : void
- Infer(system_prompt, user_prompt,\n max_tokens, grammar) : std::string
- ValidateModelArchitecture() : void
}
interface PromptFormatter <<interface>> {
+ Format(system_prompt : std::string_view,\n user_prompt : std::string_view) : std::string
+ ExpectedArchitecture() : std::string_view
}
class Gemma4JinjaPromptFormatter {
+ Format(...) : std::string
+ ExpectedArchitecture() : std::string_view
}
}
package "Infrastructure: Data Export" {
interface ExportService <<interface>> {
+ Initialize() : void
+ ProcessBrewery(brewery : const GeneratedBrewery&) : uint64_t
+ ProcessBeer(beer : const GeneratedBeer&) : uint64_t
+ ProcessUser(user : const GeneratedUser&) : uint64_t
+ ProcessCheckin(checkin : const GeneratedCheckin&) : uint64_t
+ ProcessRating(rating : const GeneratedRating&) : void
+ ProcessFollow(follow : const GeneratedFollow&) : void
+ Finalize() : void
}
class SqliteExportService {
- date_time_provider_ : std::unique_ptr<DateTimeProvider>
- db_handle_ : SqliteDatabaseHandle
- insert_location_stmt_ : SqliteStatementHandle
- insert_brewery_stmt_ : SqliteStatementHandle
- insert_beer_stmt_ : SqliteStatementHandle
- insert_user_stmt_ : SqliteStatementHandle
- insert_checkin_stmt_ : SqliteStatementHandle
- insert_rating_stmt_ : SqliteStatementHandle
- insert_follow_stmt_ : SqliteStatementHandle
- transaction_open_ : bool
- location_cache_ : std::unordered_map<std::string, uint64_t>
- brewery_cache_ : std::unordered_map<std::string, uint64_t>
+ Initialize() : void
+ ProcessBrewery(brewery : const GeneratedBrewery&) : uint64_t
+ ProcessBeer(beer : const GeneratedBeer&) : uint64_t
+ ProcessUser(user : const GeneratedUser&) : uint64_t
+ ProcessCheckin(checkin : const GeneratedCheckin&) : uint64_t
+ ProcessRating(rating : const GeneratedRating&) : void
+ ProcessFollow(follow : const GeneratedFollow&) : void
+ Finalize() : void
- InitializeSchema() : void
- PrepareStatements() : void
- RollbackAndCloseNoThrow() : void
- FinalizeStatements() : void
}
interface DateTimeProvider <<interface>> {
+ GetUtcTimestamp() : std::string
}
class SystemDateTimeProvider {
+ GetUtcTimestamp() : std::string
}
}
class BiergartenPipelineOrchestrator {
- preloader_ : std::unique_ptr<DataPreloader>
- enrichment_service_ : std::unique_ptr<EnrichmentService>
- generator_ : std::unique_ptr<DataGenerator>
- logger_ : std::unique_ptr<Logger>
- exporter_ : std::unique_ptr<ExportService>
- brewery_context_strategy_ : std::unique_ptr<ContextStrategy>
- sampling_strategy_ : std::unique_ptr<SamplingStrategy>
- beer_selection_strategy_ : std::unique_ptr<BeerSelectionStrategy>
- checkin_strategy_ : std::unique_ptr<CheckinDistributionStrategy>
- follow_strategy_ : std::unique_ptr<FollowGenerationStrategy>
- beer_style_palette_ : std::vector<BeerStyle>
- options_ : ApplicationOptions
--
- user_pool_ : std::vector<GeneratedUser>
- brewery_pool_ : std::vector<GeneratedBrewery>
- beer_pool_ : std::vector<GeneratedBeer>
- checkin_pool_ : std::vector<GeneratedCheckin>
- follow_pool_ : std::vector<GeneratedFollow>
--
+ Run() : bool
- RunUserPhase(locations : const std::vector<Location>&) : void
- RunBreweryAndBeerPhase(locations : const std::vector<Location>&) : void
- RunCheckinPhase() : void
- RunRatingPhase() : void
- RunFollowPhase() : void
}
' --- Orchestration Aggregations (Services & Strategies) ---
BiergartenPipelineOrchestrator *-- DataPreloader
BiergartenPipelineOrchestrator *-- EnrichmentService
BiergartenPipelineOrchestrator *-- DataGenerator
BiergartenPipelineOrchestrator *-- ExportService
BiergartenPipelineOrchestrator *-- CheckinDistributionStrategy
BiergartenPipelineOrchestrator *-- FollowGenerationStrategy
BiergartenPipelineOrchestrator *-- SamplingStrategy
BiergartenPipelineOrchestrator *-- BeerSelectionStrategy
BiergartenPipelineOrchestrator *-- ApplicationOptions
BiergartenPipelineOrchestrator *-- Logger
' --- Orchestration Aggregations (Data Pools) ---
BiergartenPipelineOrchestrator *-- "0..*" GeneratedUser : user_pool_
BiergartenPipelineOrchestrator *-- "0..*" GeneratedBrewery : brewery_pool_
BiergartenPipelineOrchestrator *-- "0..*" GeneratedBeer : beer_pool_
BiergartenPipelineOrchestrator *-- "0..*" GeneratedCheckin : checkin_pool_
BiergartenPipelineOrchestrator *-- "0..*" GeneratedFollow : follow_pool_
' --- Interfaces & Implementations ---
DataPreloader <|.. JsonLoader
Logger <|.. PipelineLogger
ContextStrategy <|.. BreweryContextStrategy
ContextStrategy <|.. BeerContextStrategy
SamplingStrategy <|.. UniformSamplingStrategy
BeerSelectionStrategy <|.. RandomBeerSelectionStrategy
CheckinDistributionStrategy <|.. JCurveCheckinStrategy
CheckinDistributionStrategy <|.. RandomCheckinStrategy
FollowGenerationStrategy <|.. RandomFollowStrategy
FollowGenerationStrategy <|.. ActivityWeightedFollowStrategy
EnrichmentService <|.. WikipediaService
WebClient <|.. CURLWebClient
DataGenerator <|.. MockGenerator
DataGenerator <|.. LlamaGenerator
PromptFormatter <|.. Gemma4JinjaPromptFormatter
ExportService <|.. SqliteExportService
DateTimeProvider <|.. SystemDateTimeProvider
' --- Service Compositions & Dependencies ---
WikipediaService *-- WebClient
WikipediaService ..> ContextStrategy
LlamaGenerator *-- PromptFormatter
LlamaGenerator ..> GeneratorOptions
SqliteExportService *-- DateTimeProvider
' --- Cross-Component Aggregations (Held References) ---
PipelineLogger o-- BoundedChannel : logs to
LogWorker o-- BoundedChannel : drains from
' --- Domain Containment ---
EnrichedCity *-- Location
EnrichedCity *-- LocationContext
GeneratedBrewery *-- Location
GeneratedBrewery *-- BreweryResult
GeneratedBrewery *-- GenerationMetadata
GeneratedBeer *-- Location
GeneratedBeer *-- BeerStyle
GeneratedBeer *-- BeerResult
GeneratedBeer *-- GenerationMetadata
GeneratedUser *-- Location
GeneratedUser *-- UserResult
GeneratedUser *-- GenerationMetadata
GeneratedCheckin *-- CheckinResult
GeneratedCheckin *-- GenerationMetadata
GeneratedRating *-- RatingResult
GeneratedRating *-- GenerationMetadata
GeneratedFollow *-- GenerationMetadata
@enduml

1
pipeline/diagrams/planned/output/biergarten_activity.svg Normal file
View File

File diff suppressed because one or more lines are too long

1
pipeline/diagrams/planned/output/class.svg Normal file
View File

File diff suppressed because one or more lines are too long