localenv: support TOML multiline strings during merge - #6445

Two changes to the four `golangci-lint run` invocations in
`Taskfile.yml`:

- Add `-j=4` to cap analyzer concurrency.
- Add `--allow-parallel-runners` and drop the per-task `TMPDIR`
workaround introduced in databricks#5050.

`./task lint`, fresh caches, same `tools/go.mod`:

| | Cold wall | Cold CPU (user+sys) | Warm wall |
|---|---|---|---|
| origin/main | 185s | 2282s | 4.1s |
| this PR | 76s | 327s | 4.0s |

2.4× faster cold wall, 7× less CPU. Warm runs are cache-bound and
unchanged.

## `-j=4`

golangci-lint scales poorly past 4–6 workers (upstream
[databricks#5149](golangci/golangci-lint#5149)). With
three sub-lints running in parallel under `lint-go`, the default
`-j=GOMAXPROCS` puts ~48 nominal workers on 16 cores — most of the 2282s
on origin/main is kernel scheduling contention. Capping at 4 lets the
three modules cooperate.

## `--allow-parallel-runners`

By default golangci-lint takes a system-wide lock and refuses to start
if another copy is already running. The per-task `TMPDIR` override
existed to give each invocation its own copy of that lock file so it
didn't collide; `--allow-parallel-runners` tells golangci-lint not to
take the lock in the first place. Same outcome, no `TMPDIR` plumbing.

Parallel runs happen when `lint-go` fans out to the three module
sub-lints, and when multiple worktrees lint at once.

Safe to skip the lock because golangci-lint v2.12.0 added
content-addressable cache keys ([PR
databricks#6445](golangci/golangci-lint#6445)). Before
that, two worktrees produced separate cache entries (keyed by absolute
path) — no sharing, but also no contention. Now the entries are shared
by content, so concurrent runs read and write the same files; the cache
uses OS-level file locks with atomic write semantics to keep that safe.
`origin/main` already pins v2.12.2.

## Trade-off

Two `task lint` runs in the *same* worktree no longer fail-fast with
`"parallel golangci-lint is running"` — they race instead, harmlessly
but wastefully (duplicate analysis work). Minor.

## Test plan

- [x] `task lint` clean
- [x] `task lint-q` clean
- [x] Parallel `task lint` across two worktrees both succeed
- [ ] CI lint job passes

This pull request and its description were written by Isaac.

## Release v1.15.0

### CLI

 * When `uv python install` fails, `databricks environments setup-local` now falls back to a compatible Python interpreter already installed on the machine. ([#6457](#6457))
 * Allow `databricks environments setup-local` to update `pyproject.toml` files containing TOML multi-line strings. ([#6445](#6445))

### Bundles

 * Before committing the automatic terraform→direct migration, run a deployment plan against the converted state; if the plan fails the migration is abandoned. ([#6486](#6486))
 * The `dbt-sql` bundle template now uses Databricks Runtime 16.4 LTS (up from 15.4 LTS) for classic (non-serverless) compute. ([#6418](#6418))
 * Fixed the direct engine silently ignoring edits to duration and timestamp fields, such as a Lakebase endpoint's `suspend_timeout_duration`. Such a change planned `0 to change` and was never applied. ([#6377](#6377))
 * Fixed `$${...}` not escaping a literal `${...}` on the direct engine, which failed with an `invalid dependency` error. ([#6484](#6484), [#6489](#6489))
 * direct: Fix deploying an update to `postgres_projects.default_endpoint_settings`. ([#6440](#6440))
 * direct: Fix deploying an update to `postgres_endpoints.settings.pg_settings`. ([#6441](#6441))
 * direct: Fix deploying an update to `expire_time`, `ttl` or `suspend_timeout_duration` on Lakebase resources. ([#6443](#6443))
 * Added PyDABs (Python) support for catalogs: `Resources.add_catalog` and the `catalog_mutator` decorator. ([#6408](#6408))
 * Bundle templates now use serverless [environment version 5](https://docs.databricks.com/aws/en/release-notes/serverless/environment-version/five), which offers better performance, and `databricks-connect` 16.4. ([#6378](#6378))
 * Fixed a job with a `table_update` trigger never converging on the direct engine. ([#6442](#6442))

### Dependency Updates

 * Bump Go toolchain to 1.26.8. ([#6476](#6476))

## Release v1.15.0

### CLI

 * When `uv python install` fails, `databricks environments setup-local` now falls back to a compatible Python interpreter already installed on the machine. ([#6457](#6457))
 * Allow `databricks environments setup-local` to update `pyproject.toml` files containing TOML multi-line strings. ([#6445](#6445))

### Bundles

 * Before committing the automatic terraform→direct migration, run a deployment plan against the converted state; if the plan fails the migration is abandoned. ([#6486](#6486))
 * The `dbt-sql` bundle template now uses Databricks Runtime 16.4 LTS (up from 15.4 LTS) for classic (non-serverless) compute. ([#6418](#6418))
 * Fixed the direct engine silently ignoring edits to duration and timestamp fields, such as a Lakebase endpoint's `suspend_timeout_duration`. Such a change planned `0 to change` and was never applied. ([#6377](#6377))
 * Fixed `$${...}` not escaping a literal `${...}` on the direct engine, which failed with an `invalid dependency` error. ([#6484](#6484), [#6489](#6489))
 * Remove forward_user_access_token from update_mask for Apps because it's not supported. Fixes regression in 1.14.1. ([#6510](#6510))
 * direct: Fix deploying an update to `postgres_projects.default_endpoint_settings`. ([#6440](#6440))
 * direct: Fix deploying an update to `postgres_endpoints.settings.pg_settings`. ([#6441](#6441))
 * direct: Fix deploying an update to `expire_time`, `ttl` or `suspend_timeout_duration` on Lakebase resources. ([#6443](#6443))
 * Added PyDABs (Python) support for catalogs: `Resources.add_catalog` and the `catalog_mutator` decorator. ([#6408](#6408))
 * Bundle templates now use serverless [environment version 5](https://docs.databricks.com/aws/en/release-notes/serverless/environment-version/five), which offers better performance, and `databricks-connect` 16.4. ([#6378](#6378))
 * Fixed a job with a `table_update` trigger never converging on the direct engine. ([#6442](#6442))

### Dependency Updates

 * Bump Go toolchain to 1.26.8. ([#6476](#6476))

## Changes
Allow the setup-local formatting-preserving merge to handle TOML
multi-line strings without treating their contents as project structure.

## Why
Valid pyproject.toml files commonly use multi-line project descriptions.
The previous conservative guard rejected every such file, including
strings unrelated to managed fields. Tracks
[DECO-28330](https://databricks.atlassian.net/browse/DECO-28330).

## Tests
- GOTOOLCHAIN=local go test ./libs/localenv ./cmd/environments -count=1
- Changed Go package lint: 0 issues
- Full fmt/checks could not download ruff and golang.org/x/net because
package network connections were refused; PR CI will run them

_This PR was written with Codex._

[DECO-28330]:
https://databricks.atlassian.net/browse/DECO-28330?atlOrigin=eyJpIjoiNWRkNTljNzYxNjVmNDY3MDlhMDU5Y2ZhYzA5YTRkZjUiLCJwIjoiZ2l0aHViLWNvbS1KU1cifQ

## Release v1.15.0

### CLI

 * When `uv python install` fails, `databricks environments setup-local` now falls back to a compatible Python interpreter already installed on the machine. ([#6457](#6457))
 * Allow `databricks environments setup-local` to update `pyproject.toml` files containing TOML multi-line strings. ([#6445](#6445))

### Bundles

 * Before committing the automatic terraform→direct migration, run a deployment plan against the converted state; if the plan fails the migration is abandoned. ([#6486](#6486))
 * The `dbt-sql` bundle template now uses Databricks Runtime 16.4 LTS (up from 15.4 LTS) for classic (non-serverless) compute. ([#6418](#6418))
 * Fixed the direct engine silently ignoring edits to duration and timestamp fields, such as a Lakebase endpoint's `suspend_timeout_duration`. Such a change planned `0 to change` and was never applied. ([#6377](#6377))
 * Fixed `$${...}` not escaping a literal `${...}` on the direct engine, which failed with an `invalid dependency` error. ([#6484](#6484), [#6489](#6489))
 * direct: Fix deploying an update to `postgres_projects.default_endpoint_settings`. ([#6440](#6440))
 * direct: Fix deploying an update to `postgres_endpoints.settings.pg_settings`. ([#6441](#6441))
 * direct: Fix deploying an update to `expire_time`, `ttl` or `suspend_timeout_duration` on Lakebase resources. ([#6443](#6443))
 * Added PyDABs (Python) support for catalogs: `Resources.add_catalog` and the `catalog_mutator` decorator. ([#6408](#6408))
 * Bundle templates now use serverless [environment version 5](https://docs.databricks.com/aws/en/release-notes/serverless/environment-version/five), which offers better performance, and `databricks-connect` 16.4. ([#6378](#6378))
 * Fixed a job with a `table_update` trigger never converging on the direct engine. ([#6442](#6442))

### Dependency Updates

 * Bump Go toolchain to 1.26.8. ([#6476](#6476))

## Release v1.15.0

### CLI

 * When `uv python install` fails, `databricks environments setup-local` now falls back to a compatible Python interpreter already installed on the machine. ([#6457](#6457))
 * Allow `databricks environments setup-local` to update `pyproject.toml` files containing TOML multi-line strings. ([#6445](#6445))

### Bundles

 * Before committing the automatic terraform→direct migration, run a deployment plan against the converted state; if the plan fails the migration is abandoned. ([#6486](#6486))
 * The `dbt-sql` bundle template now uses Databricks Runtime 16.4 LTS (up from 15.4 LTS) for classic (non-serverless) compute. ([#6418](#6418))
 * Fixed the direct engine silently ignoring edits to duration and timestamp fields, such as a Lakebase endpoint's `suspend_timeout_duration`. Such a change planned `0 to change` and was never applied. ([#6377](#6377))
 * Fixed `$${...}` not escaping a literal `${...}` on the direct engine, which failed with an `invalid dependency` error. ([#6484](#6484), [#6489](#6489))
 * Remove forward_user_access_token from update_mask for Apps because it's not supported. Fixes regression in 1.14.1. ([#6510](#6510))
 * direct: Fix deploying an update to `postgres_projects.default_endpoint_settings`. ([#6440](#6440))
 * direct: Fix deploying an update to `postgres_endpoints.settings.pg_settings`. ([#6441](#6441))
 * direct: Fix deploying an update to `expire_time`, `ttl` or `suspend_timeout_duration` on Lakebase resources. ([#6443](#6443))
 * Added PyDABs (Python) support for catalogs: `Resources.add_catalog` and the `catalog_mutator` decorator. ([#6408](#6408))
 * Bundle templates now use serverless [environment version 5](https://docs.databricks.com/aws/en/release-notes/serverless/environment-version/five), which offers better performance, and `databricks-connect` 16.4. ([#6378](#6378))
 * Fixed a job with a `table_update` trigger never converging on the direct engine. ([#6442](#6442))

### Dependency Updates

 * Bump Go toolchain to 1.26.8. ([#6476](#6476))