@playform/hook-dsh-normalize-quotes

The DeepSeek Harness Plugin Family for PlayForm.

[!NOTE]

The quote normalizer for model output - a DeepSeek Harness plugin that hooks the llm/stream waterfall (the interceptable wrapper around EVERY streaming model call, bound to the LlmRuntime) and normalizes the curly quote family in model output, live in the transcript: the eight typographic quote code points each map to their ASCII straight counterpart - U+2018/U+2019/U+201A/U+201B to the ASCII apostrophe, and U+201C/U+201D/U+201E/U+201F to the ASCII double quote.

A MAP flavor of the normalize family: the core's Quotes char-to-char table owns the substitution - no replacement config knob.

The @-sentence identity: Hook @ DSH @ Normalize @ Quotes.

The family's raw-write tool (registered by hook-dsh-normalize-dash) bypasses this flavor's transforms too - the exemption is family-wide: the core's Stream/Chunk + Stream/Block pass every raw-write call through by identity, so the tool's explicit normalize parameter (default false = verbatim) is the only normalization it applies.

Where It Fits

Family position (the @-sentence Hook @ DSH @ Normalize @ Quotes): a hook child of the plugin-dsh-factory service and the hook-dsh-core machinery; the second of the six stream normalizer siblings.

One of six siblings in the stream-normalization family, all built on the shared machinery - plugin-dsh-factory for State/Append (and the named Schema helper, shared: false, for the config), hook-dsh-core for the Quotes table and the generic ReplaceMap/Chunk/Block dispatch:

flavor	table	substitution
`hook-dsh-normalize-dash`	core `Dashes` class	`→ replacement` (default `-`)
hook-dsh-normalize-quotes (this bundle)	core `Quotes` MAP	curly → straight
`hook-dsh-normalize-ellipsis`	core `Ellipsis` class	U+2026 → `...`
`hook-dsh-normalize-spaces`	core `Spaces` class	unicode spaces → `" "`
`hook-dsh-normalize-invisible`	core `Invisible` class	removed (default `""`)
`hook-dsh-normalize-fullwidth`	core `Fullwidth` MAP	full-width → half-width

A non-manifest factory consumer: it injects ["pluginFactory"] and uses State (cell unwrap + shared Ledger/Enabled mappings + its own fields), Append and Journal (the shared P5 record on every N > 0 pass); the config is composed by the factory's standalone Schema helper with shared: false - the minimal block, no fs/observed dead fields.

It touches no files, so fs/write-intent and fs/observed never see it; it wraps the downstream result and always calls next(), so it composes with other llm/stream listeners regardless of registration order.

The DSH plugin family is the DeepSeek Harness plugin layer of the PlayForm ecosystem: TypeScript-first Source/ → Target/, the deterministic @playform build, prepublishOnly-only - the same conventions as every other @playform package.

Install

Terminal (registry)

pnpm add @playform/hook-dsh-normalize-quotes

Remote / dependency install (auto-activation): pnpm add @playform/hook-dsh-normalize-quotes in a profile dir (+ the package name in dsh.profile.bundles, or dsh plugin add) - installed automatically and activated at the next host start.
Local git clone: clone → pnpm install inside the bundle (its pnpm-workspace.yaml pins packages: [.] so the install lands in the bundle, not the workspace root) → build → dsh plugin --profile <name> add <this directory>.
Tarball / npm: dsh plugin --profile <name> add <tarball-or-spec> - the package ships built Target/ (minified, .d.ts twins), so no build runs at install time.

Verify with the hook-dsh-normalize-quotes: activated (...) ledger line.

Usage

The bundle is configured through its cordis.patch.yml row (or the profile's dsh.bundle manifest) - the full config table is in The Config:

- insert:
      - id: hook-dsh-normalize-quotes
        name: "@playform/hook-dsh-normalize-quotes"
        config:
            log: true
            logFile: ~/.dsh/hook-dsh-normalize-quotes.log
            normalizeReasoning: true

The Problem

Smart-quote processors and code editors emit typographic quotes; every parser, shell and diff understands the straight ASCII ones.

Left in model output, a curly quote is a silent correctness hazard - in code blocks, commands and file paths it is simply the wrong character.

This flavor makes the straight form the one that reaches the transcript.

How It Works

The stream pipeline

  llm/stream waterfall (options, next)     the interceptable wrapper around
       │                                   EVERY streaming model call
       ▼  next() called FIRST, always - options never touched
  Normalize(upstream, state)               the async generator
       │   for await (chunk of upstream)
       ▼
  CoreChunk(chunk, transform, reasoning, toolArgs, raw)  the core's per-chunk dispatch
       │
       ├─ text-delta ────────► ReplaceMap(text, Quotes)     rewrite the text field
       ├─ reasoning-delta ───► same, when normalizeReasoning (default ON)
       ├─ block-end ─────────► the assembled block's text fields -
       │                       TextBlock.text / ReasoningBlock.text (plus a
       │                       runtime `thinking` string field) - the deltas
       │                       AND the block must agree, or consumers see
       │                       inconsistencies
       ├─ tool-call-delta ───► ReplaceMap(argumentsDelta, Quotes) when
       │                       normalizeToolArguments (IMPLEMENTED, default
       │                       OFF - execution-critical raw JSON, the user's
       │                       accepted risk; the example patch turns it
       │                       on) - the assembled ToolCallBlock.arguments
       │                       follows the same flag via block-end
       │                       The gate is THREE-WAY with the flag on: a
       │                       delta/block whose `name` is "edit", "raw-write"
       │                       or "normalize-file" passes through BY IDENTITY
       │                       (the edit tool's `old_string` must match the real
       │                       file bytes; the raw-write tool owns its content
       │                       normalization; the normalize-file tool's arguments
       │                       carry a file path - a normalized dash inside a
       │                       filename would corrupt the target),
       │                       and a call whose arguments open with the
       │                       `{"__normalize":false` marker (FIRST key,
       │                       tracked per call id) passes through
       │                       UNNORMALIZED with the marker entry stripped,
       │                       so the executed call carries no unknown key
       └─ block-start / usage / finish
                                ──► PASSTHROUGH BY IDENTITY, ALWAYS
                                    (usage/finish ordering is the
                                     adapter contract)
       │   count === 0 → original chunk BY IDENTITY; rewritten → shallow copy
       ▼
  yield ──► downstream consumers = the live UI + the durable transcript
       │    (order preserved, no buffering; upstream throws propagate)
       ▼  normal loop completion, Count > 0
  Factory.Append ──► `hook-dsh-normalize-quotes: normalized N quote char(s) in one stream`

The transform

Exactly the core's Quotes map (@playform/hook-dsh-core's Normalize/Quotes), applied per text segment through the core's ReplaceMap - eight entries, each typographic quote code point to its ASCII straight counterpart:

Code point	Character (by name)	ASCII
U+2018	left single quotation mark	`'` (U+0027)
U+2019	right single quotation mark	`'`
U+201A	single low-9 quotation mark	`'`
U+201B	single high-reversed-9 quotation mark	`'`
U+201C	left double quotation mark	`"` (U+0022)
U+201D	right double quotation mark	`"`
U+201E	double low-9 quotation mark	`"`
U+201F	double high-reversed-9 quotation mark	`"`

No context rules - one character in, its straight counterpart out.
Chunk-boundary-safe: single-character substitution, no lookahead - per-chunk application can never disagree with whole-text application.
The mapping values are returned through a function replacer, so they are inserted literally (no $-pattern interpretation).
Replaced characters are counted per stream for the ledger line.

The Pitfalls

A dependency alone is NOT activation - the package must be in dsh.profile.bundles (or the patch row inserted), or no listener registers and no ledger exists.
The map is quote-only - the em dash U+2014 and the rest of the dash family are the dashes flavor's class; a quotes-only pass leaves them untouched - run the flavors the session's characters call for.
The stream gate rewrites model output only - it never touches disk; files already on disk are the normalize-file tool's job.
The count is per replaced code point - N counts matched characters; N = 0 and an upstream throw both write no count line and no journal record.

The Config

Field	Type	Default	Volatile	Meaning
`log`	boolean	`true`	yes	write the durable ledger file
`logFile`	string	`~/.dsh/hook-dsh-normalize-quotes.log`	yes	the quotes ledger (separate from the family's logs)
`normalizeReasoning`	boolean	`true`	no	normalize reasoning deltas and the assembled reasoning block too
`normalizeToolArguments`	boolean	`false`	no	IMPLEMENTED (default OFF): rewrite the tool-call argumentsDelta and the assembled ToolCallBlock.arguments when on (with three exemptions: `edit`, `raw-write` and `normalize-file` calls pass through by identity (the edit tool's `old_string` must match the real file bytes; the raw-write tool owns its content normalization through its explicit `normalize` parameter; the normalize-file tool's arguments carry a file path - a normalized dash inside a filename would corrupt the target); and a call whose arguments open with the `{"__normalize":false` first-key marker passes through unnormalized, the marker entry stripped) - execution-critical raw JSON, the user's accepted risk; the example patch turns it on

There is no replacement field (MAP flavor).

Volatile cells commit without remounting the plugin; the factory's State builder unwraps them defensively.

In Action

One stream, one transformation.

The model emits smart-quoted prose; the live UI and the transcript receive the straight ASCII forms:

text-delta in (what the model wrote):

  ‘She said “let’s ship it” — today’s the day’

text-delta out (what reaches the transcript):

  'She said "let's ship it" — today's the day'

Every typographic quote in the incoming text - the four single forms and the four double forms - becomes its ASCII counterpart; straight quotes that were already there are untouched (the count only counts replacements). The sample carries six of the eight forms; the em dash U+2014 is the dashes flavor's character, not the quotes map's - with only the quotes flavor it passes through untouched.

In code blocks, commands and file paths this is the difference between a command that runs and one that fails for no visible reason.

When a stream finishes normally with replacements made, the ledger gets the count line:

hook-dsh-normalize-quotes: normalized 6 quote char(s) in one stream

(The same pass runs over reasoning deltas when normalizeReasoning is on, and over tool-call arguments when the example patch enables normalizeToolArguments - with the edit, raw-write and normalize-file names exempt and the raw-marker calls passing through unnormalized.)

The Ledger

Two lines, both written through the factory's Append (the hook-dsh-normalize-quotes: prefix is the logger's <State.Module>:; the durable file line is [<ISO>] <message>):

hook-dsh-normalize-quotes: activated (reasoning=on, toolArgs=off, logFile=~/.dsh/hook-dsh-normalize-quotes.log)
hook-dsh-normalize-quotes: normalized N quote char(s) in one stream

The activation line is written by apply(); the count line only follows a normal stream completion and only when N > 0 (a thrown-away stream writes no ledger line).

Every N > 0 pass also journals one normalized record into the shared package_governance v2 domain - event normalized, path = the empty stream-level path, detail byte-identical to the count line - best-effort: with no storage facility the record buffers or drops, and the human ledger stays the complete record.

The two release groups (CLASSIC vs EFFECT-TS)

Two published groups, one contract: this package is the CLASSIC build - plain TypeScript, zero framework dependencies: import ... from "@playform/hook-dsh-normalize-quotes" resolves to Target/. The EFFECT-TS implementation of the same contract is the separate package @playform/ets-hook-dsh-normalize-quotes - effect-backed, built on the shared @playform/ets-base-dsh base with effect v4.0.2 as its runtime dependency. Install the group you run and never mix the groups in one graph (the effect-ts Update returns Effect envelopes, not the classic dispatch envelope).

License 📜

CC0-1.0.

Install

Source

Readme

@playform/hook-dsh-normalize-quotes

Where It Fits

Install

Usage

The Problem

How It Works

The transform

The Pitfalls

The Config

In Action

The Ledger

The two release groups (CLASSIC vs EFFECT-TS)

License 📜