|
| 1 | +# Testcontainers Java Contribution Research |
| 2 | + |
| 3 | +## Goal |
| 4 | + |
| 5 | +This note summarizes contribution candidates for a Spring / Java / Kotlin backend engineer targeting `testcontainers/testcontainers-java`. |
| 6 | + |
| 7 | +The focus is not only "what is easy", but "what will look strong to backend hiring teams while still having a realistic merge path". |
| 8 | + |
| 9 | +## Repository Health |
| 10 | + |
| 11 | +- Upstream repository: `testcontainers/testcontainers-java` |
| 12 | +- Local fork remote: `PreAgile/testcontainers-java` |
| 13 | +- Current local branch: `main` |
| 14 | +- Working tree state at review time: clean |
| 15 | + |
| 16 | +Signals that the repository is actively maintained: |
| 17 | + |
| 18 | +- `good first issue` labels are still in use |
| 19 | +- merged PRs exist in late 2025 and early 2026 |
| 20 | +- recent human-merged PRs include: |
| 21 | + - `#11498` merged on `2026-02-26` |
| 22 | + - `#11223` merged on `2026-03-02` |
| 23 | + |
| 24 | +## Local Codebase Areas That Matter |
| 25 | + |
| 26 | +Relevant local paths: |
| 27 | + |
| 28 | +- `docs/contributing.md` |
| 29 | +- `docs/contributing_docs.md` |
| 30 | +- `docs/modules/k6.md` |
| 31 | +- `modules/k6/src/test/java/org/testcontainers/k6/K6ContainerTests.java` |
| 32 | +- `modules/r2dbc/src/main/java/org/testcontainers/r2dbc/R2DBCDatabaseContainer.java` |
| 33 | +- `modules/postgresql/src/main/java/org/testcontainers/postgresql/PostgreSQLR2DBCDatabaseContainer.java` |
| 34 | +- `modules/mysql/src/main/java/org/testcontainers/mysql/MySQLR2DBCDatabaseContainer.java` |
| 35 | +- `modules/mongodb/src/main/java/org/testcontainers/mongodb/MongoDBContainer.java` |
| 36 | +- `modules/jdbc/src/main/java/org/testcontainers/jdbc/ContainerDatabaseDriver.java` |
| 37 | + |
| 38 | +## Maintainer Preference Pattern |
| 39 | + |
| 40 | +Observed pattern from recent merged PRs: |
| 41 | + |
| 42 | +- small scope |
| 43 | +- module-focused changes |
| 44 | +- tests included for code changes |
| 45 | +- low review burden |
| 46 | + |
| 47 | +Examples: |
| 48 | + |
| 49 | +- `#11498` changed 2 files and added a focused implementation plus test update |
| 50 | +- `#11223` changed 3 files with a targeted compatibility fix |
| 51 | + |
| 52 | +This strongly suggests that a first PR should avoid broad cross-cutting core changes. |
| 53 | + |
| 54 | +## Candidate Ranking |
| 55 | + |
| 56 | +### 1. Best first PR: `#8780` |
| 57 | + |
| 58 | +Title: |
| 59 | + |
| 60 | +- `[Enhancement]: Add example test to the K6 module docs` |
| 61 | + |
| 62 | +Why this is the safest first merge: |
| 63 | + |
| 64 | +- issue scope is clear |
| 65 | +- user pain is already documented in the issue body |
| 66 | +- implementation is mostly docs/example work |
| 67 | +- it does not require design negotiation across multiple modules |
| 68 | + |
| 69 | +Problem analysis: |
| 70 | + |
| 71 | +- `docs/modules/k6.md` currently includes a code snippet from `K6ContainerTests` |
| 72 | +- the docs show the API, but do not clearly present the "typical test shape" that users expect |
| 73 | +- the issue author explicitly compared it unfavorably to quickstart-style docs |
| 74 | + |
| 75 | +Root cause: |
| 76 | + |
| 77 | +- documentation exposes a snippet, but not an opinionated onboarding flow |
| 78 | +- discoverability is weaker than in other Testcontainers modules |
| 79 | + |
| 80 | +Why this helps a Spring/backend profile: |
| 81 | + |
| 82 | +- not as strong as a code PR, but excellent as a first merged foothold |
| 83 | +- shows you can read the repo conventions and contribute cleanly |
| 84 | + |
| 85 | +Recommended PR shape: |
| 86 | + |
| 87 | +- keep it to docs and possibly example snippet framing |
| 88 | +- do not over-design a new example project unless maintainers ask for it |
| 89 | +- align the K6 docs with the style of existing quickstart/test integration pages |
| 90 | + |
| 91 | +Merge probability: |
| 92 | + |
| 93 | +- high |
| 94 | + |
| 95 | +### 2. Best second PR: `#8797` |
| 96 | + |
| 97 | +Title: |
| 98 | + |
| 99 | +- `Add getR2dbcUrl helper method to JdbcDatabaseContainer` |
| 100 | + |
| 101 | +Why this is the best backend-signaling PR: |
| 102 | + |
| 103 | +- directly relevant to Spring, reactive stacks, database integration, and developer ergonomics |
| 104 | +- stronger hiring signal than a docs PR |
| 105 | +- maintainers already discussed acceptable design direction |
| 106 | + |
| 107 | +Problem analysis: |
| 108 | + |
| 109 | +- current R2DBC support exposes `getOptions(container)` methods |
| 110 | +- users who want a concrete R2DBC URL string must construct it manually |
| 111 | +- that hurts discoverability and ease of use |
| 112 | + |
| 113 | +Important maintainer guidance already found in the issue: |
| 114 | + |
| 115 | +- do not force this into a generic `JdbcDatabaseContainer` API immediately |
| 116 | +- preferred direction is static `getR2dbcUrl(container)` helpers on each `R2DBCDatabaseContainer` implementation |
| 117 | +- this was discussed explicitly by maintainers in the issue thread |
| 118 | + |
| 119 | +Existing failed attempt: |
| 120 | + |
| 121 | +- PR `#9569` attempted the feature |
| 122 | +- maintainer immediately asked for tests |
| 123 | +- contributor then got stuck on MSSQL-specific test issues |
| 124 | +- PR was eventually closed |
| 125 | + |
| 126 | +What that means: |
| 127 | + |
| 128 | +- the feature itself is not rejected |
| 129 | +- the implementation failed because the scope was too broad and the testing path got messy |
| 130 | + |
| 131 | +Best way to revive it: |
| 132 | + |
| 133 | +- comment on the issue first |
| 134 | +- confirm that the agreed static-method design is still welcome |
| 135 | +- explicitly propose a narrow first step |
| 136 | +- avoid broad "support everything at once" unless maintainers request it |
| 137 | + |
| 138 | +Safer implementation plan: |
| 139 | + |
| 140 | +1. Start with the stable implementations that are easiest to test cleanly |
| 141 | +2. Add tests alongside each helper |
| 142 | +3. Leave tricky drivers such as MSSQL for follow-up if needed |
| 143 | + |
| 144 | +Why this helps a Spring/backend profile: |
| 145 | + |
| 146 | +- strongest fit for Spring / Java backend positioning |
| 147 | +- good talking point for R2DBC, developer experience, and integration testing |
| 148 | + |
| 149 | +Merge probability: |
| 150 | + |
| 151 | +- medium to high if kept narrow |
| 152 | +- medium or lower if implemented across too many drivers at once |
| 153 | + |
| 154 | +### 3. High-value but risky: `#3066` |
| 155 | + |
| 156 | +Title: |
| 157 | + |
| 158 | +- `Support init scripts for MongoDBContainer without manually customizing the WaitStrategy` |
| 159 | + |
| 160 | +Why it is attractive: |
| 161 | + |
| 162 | +- strong backend signal |
| 163 | +- involves lifecycle, initialization order, replica set handling, and wait strategy design |
| 164 | + |
| 165 | +Problem analysis from issue and code: |
| 166 | + |
| 167 | +- placing scripts under `/docker-entrypoint-initdb.d` can trigger MongoDB restart behavior |
| 168 | +- local `MongoDBContainer` currently relies on `Wait.forLogMessage("(?i).*waiting for connections.*", 1)` |
| 169 | +- that does not robustly cover the init-script + replica-set startup sequence |
| 170 | + |
| 171 | +Root cause: |
| 172 | + |
| 173 | +- container readiness assumptions do not fully match Mongo init script behavior |
| 174 | +- `MongoDBContainer` also performs replica set initialization after start, which increases timing sensitivity |
| 175 | + |
| 176 | +Why it is risky as a first PR: |
| 177 | + |
| 178 | +- more design ambiguity |
| 179 | +- more regression risk |
| 180 | +- higher chance of maintainers asking for a reproducer-heavy discussion or broader design changes |
| 181 | + |
| 182 | +Merge probability: |
| 183 | + |
| 184 | +- medium at best for a newcomer first PR |
| 185 | + |
| 186 | +### 4. Not recommended first: `#1537` |
| 187 | + |
| 188 | +Title: |
| 189 | + |
| 190 | +- `Support hikari data-source-properties in ContainerDatabaseDriver` |
| 191 | + |
| 192 | +Why it looks easy but is not: |
| 193 | + |
| 194 | +- the surface request is "just pass `Properties info` through" |
| 195 | +- in practice it may affect: |
| 196 | + - JDBC property propagation |
| 197 | + - connection creation semantics |
| 198 | + - cache key behavior in `ContainerDatabaseDriver` |
| 199 | + |
| 200 | +Root cause: |
| 201 | + |
| 202 | +- `ContainerDatabaseDriver.connect(String url, Properties info)` does not currently flow those properties in a newcomer-safe way |
| 203 | +- any fix in this path touches JDBC core behavior, not just one module |
| 204 | + |
| 205 | +Why it is a poor first PR: |
| 206 | + |
| 207 | +- too core |
| 208 | +- too broad |
| 209 | +- too easy to create subtle regressions |
| 210 | + |
| 211 | +Merge probability: |
| 212 | + |
| 213 | +- low for a first contribution |
| 214 | + |
| 215 | +## Recommended Order |
| 216 | + |
| 217 | +1. `#8780` K6 docs/example PR |
| 218 | +2. `#8797` R2DBC helper PR |
| 219 | +3. `#3066` MongoDB wait-strategy/init-script PR |
| 220 | + |
| 221 | +## First PR Recommendation |
| 222 | + |
| 223 | +### Recommended first merged PR |
| 224 | + |
| 225 | +Issue: |
| 226 | + |
| 227 | +- `#8780` `[Enhancement]: Add example test to the K6 module docs` |
| 228 | + |
| 229 | +Link: |
| 230 | + |
| 231 | +- `https://github.com/testcontainers/testcontainers-java/issues/8780` |
| 232 | + |
| 233 | +Why this is the best first PR: |
| 234 | + |
| 235 | +- the issue scope is concrete |
| 236 | +- the user pain is already written clearly in the issue |
| 237 | +- it avoids deep design arguments |
| 238 | +- it matches the small-PR preference observed in recent merged work |
| 239 | + |
| 240 | +Problem analysis: |
| 241 | + |
| 242 | +- `docs/modules/k6.md` already exposes API usage through codeinclude |
| 243 | +- however, it still does not clearly show the "expected JUnit-style test shape" that users look for |
| 244 | +- the issue reporter explicitly compared the docs to quickstart-style guidance and found the current page harder to follow |
| 245 | + |
| 246 | +Root cause: |
| 247 | + |
| 248 | +- the docs expose a runnable snippet |
| 249 | +- but the onboarding flow is weaker than in more polished Testcontainers pages |
| 250 | +- the page answers "what API exists" better than "how should I write my first test" |
| 251 | + |
| 252 | +Recommended solution shape: |
| 253 | + |
| 254 | +1. Update `docs/modules/k6.md` |
| 255 | +2. Reuse the existing runnable snippet from `modules/k6/src/test/java/org/testcontainers/k6/K6ContainerTests.java` |
| 256 | +3. Add a short JUnit-style explanation section |
| 257 | +4. Keep the change limited to docs and snippet framing unless maintainers request more |
| 258 | + |
| 259 | +Why this should merge well: |
| 260 | + |
| 261 | +- it is easy to review |
| 262 | +- it improves user onboarding directly |
| 263 | +- it does not expand the public API |
| 264 | +- it does not create cross-module risk |
| 265 | + |
| 266 | +Estimated merge probability: |
| 267 | + |
| 268 | +- high, assuming the PR stays small and aligned with existing docs conventions |
| 269 | + |
| 270 | +## Maintainer Interaction Strategy |
| 271 | + |
| 272 | +### Who is most likely to matter |
| 273 | + |
| 274 | +Based on issue and PR review activity observed during research: |
| 275 | + |
| 276 | +- `eddumelendez` |
| 277 | +- `kiview` |
| 278 | + |
| 279 | +This does not mean directly targeting them personally. It means: |
| 280 | + |
| 281 | +- write the issue comment in a way that makes review easy for maintainers who already triage this area |
| 282 | +- keep scope explicit |
| 283 | +- make it obvious that the change is intentionally narrow |
| 284 | + |
| 285 | +### Best way to approach `#8780` |
| 286 | + |
| 287 | +Suggested issue comment: |
| 288 | + |
| 289 | +> I’d like to work on this. |
| 290 | +> My plan is to improve the K6 docs by adding a clearer JUnit-style example section based on the existing module test, while keeping the change small and aligned with the current docs/codeinclude style. |
| 291 | +
|
| 292 | +This is strong because: |
| 293 | + |
| 294 | +- it signals ownership |
| 295 | +- it shows you read the existing docs structure |
| 296 | +- it reduces reviewer uncertainty |
| 297 | + |
| 298 | +### Best PR description style for `#8780` |
| 299 | + |
| 300 | +Suggested PR structure: |
| 301 | + |
| 302 | +```md |
| 303 | +## Summary |
| 304 | +Improve the K6 module documentation with a clearer JUnit-style example test. |
| 305 | + |
| 306 | +## Problem |
| 307 | +The current docs show the API usage, but it is still hard to understand the expected test structure quickly. |
| 308 | + |
| 309 | +## Changes |
| 310 | +- add a clearer example section to the K6 docs |
| 311 | +- reuse the existing runnable test snippet |
| 312 | +- keep the documentation aligned with the current codeinclude style |
| 313 | + |
| 314 | +## Why |
| 315 | +This makes the module easier to adopt for users looking for a quickstart-like testing example. |
| 316 | +``` |
| 317 | + |
| 318 | +## Second PR Recommendation |
| 319 | + |
| 320 | +### Best first code PR after the docs PR |
| 321 | + |
| 322 | +Issue: |
| 323 | + |
| 324 | +- `#8797` `Add getR2dbcUrl helper method to JdbcDatabaseContainer` |
| 325 | + |
| 326 | +Link: |
| 327 | + |
| 328 | +- `https://github.com/testcontainers/testcontainers-java/issues/8797` |
| 329 | + |
| 330 | +Why this is the strongest backend-signaling follow-up: |
| 331 | + |
| 332 | +- aligns closely with Spring / reactive database usage |
| 333 | +- gives a better hiring signal than docs-only work |
| 334 | +- maintainers already discussed an acceptable design direction |
| 335 | + |
| 336 | +Important caution: |
| 337 | + |
| 338 | +- a previous attempt, PR `#9569`, was closed |
| 339 | +- the feature was not rejected |
| 340 | +- the attempt appears to have become too broad and got stuck in tests, especially around MSSQL |
| 341 | + |
| 342 | +Best way to revive it: |
| 343 | + |
| 344 | +Suggested issue comment: |
| 345 | + |
| 346 | +> I’d like to revive this using the static `getR2dbcUrl(container)` approach discussed above. |
| 347 | +> To keep the scope reviewable, I plan to start with the stable implementations first and include tests for each one, leaving any problematic driver-specific follow-up for a separate PR if needed. |
| 348 | +> Does that still match the expected direction? |
| 349 | +
|
| 350 | +Why this approach should improve merge odds: |
| 351 | + |
| 352 | +- it respects prior maintainer discussion |
| 353 | +- it shows you learned from the closed PR |
| 354 | +- it reduces fear of a large cross-driver change |
| 355 | + |
| 356 | +## Suggested PR Positioning |
| 357 | + |
| 358 | +### For `#8780` |
| 359 | + |
| 360 | +PR message should emphasize: |
| 361 | + |
| 362 | +- clearer onboarding |
| 363 | +- quickstart-style discoverability |
| 364 | +- alignment with existing docs conventions |
| 365 | + |
| 366 | +### For `#8797` |
| 367 | + |
| 368 | +PR message should emphasize: |
| 369 | + |
| 370 | +- narrowed scope |
| 371 | +- implementation following the already discussed issue design |
| 372 | +- tests added for each touched helper |
| 373 | +- no attempt to redesign the entire R2DBC abstraction |
| 374 | + |
| 375 | +Suggested issue comment before coding: |
| 376 | + |
| 377 | +> I would like to revive this using the static `getR2dbcUrl(container)` approach discussed above. |
| 378 | +> To keep the scope reviewable, I plan to add the helper with tests for the stable implementations first and leave any problematic drivers for a follow-up PR if needed. |
| 379 | +> Does that still match the expected direction? |
| 380 | +
|
| 381 | +## Practical Notes |
| 382 | + |
| 383 | +- For docs contributions, follow `docs/contributing_docs.md` |
| 384 | +- For code contributions, follow `docs/contributing.md` |
| 385 | +- Run formatting and tests before opening a PR |
| 386 | +- Keep each PR scoped to one issue and one clear problem |
| 387 | +- For a newcomer, "small and mergeable" is more valuable than "ambitious but stalled" |
0 commit comments