diff --git a/.gitignore b/.gitignore
index d86052c..1ea055c 100644
--- a/.gitignore
+++ b/.gitignore
@@ -18,7 +18,6 @@ dist/
 .build/
 
 ### Rust build artefacts ###
-parser/target/
 
 ### Assembled Pages artifact ###
 _site/
diff --git a/README.md b/README.md
index 883ccf6..02c569d 100644
--- a/README.md
+++ b/README.md
@@ -34,7 +34,7 @@ docs/                 architecture, data pipeline, deployment
 The dataset id is one identifier end to end:
 
 ```
-data/2017-old/ → parser/configs/2017-old.yml → db/2017-old.db.gz → /thptqg/2017-old/
+data/2017-old/ → go-parser/configs/2017-old.yml → db/2017-old.db.gz → /thptqg/2017-old/
 ```
 
 ## Build
@@ -53,7 +53,7 @@ Pushing to `main` runs the same steps in
 ## Adding a dataset
 
 1. Put the Excel files in `data/<id>/`
-2. Add `parser/configs/<id>.yml` — sheet mode, column indices, validation
+2. Add `go-parser/configs/<id>.yml` — sheet mode, column indices, validation
    guards. No SQL; the schema is canonical.
 3. Add an entry to `DATASETS` in `src/datasets.js`
 
diff --git a/docs/data-pipeline.md b/docs/data-pipeline.md
index ef3e1d6..bbfe48d 100644
--- a/docs/data-pipeline.md
+++ b/docs/data-pipeline.md
@@ -2,9 +2,9 @@
 
 From raw Excel files to a compressed SQLite file the browser can load.
 
-One Rust binary (`parser/`) builds every dataset. What differs per dataset is
+One Rust binary (`go-parser/`) builds every dataset. What differs per dataset is
 parse rules only — sheet strategy, column layout, validation guards — declared
-in `parser/configs/<id>.yml`. The table shape, the INSERT and the subject
+in `go-parser/configs/<id>.yml`. The table shape, the INSERT and the subject
 regexes are canonical and live in `go-parser/internal/schema/schema.go`.
 
 ## Sources
@@ -19,7 +19,7 @@ regexes are canonical and live in `go-parser/internal/schema/schema.go`.
 Only `2017` can be re-fetched:
 
 ```bash
-node parser/scripts/crawl-baotintuc.js
+node go-parser/scripts/crawl-baotintuc.js
 ```
 
 Idempotent — skips files already present, saves to `data/2017/`. Source article:
@@ -116,34 +116,44 @@ silently drops 13,720 students** (Hanoi +7,275, HCM +6,445). That is what
 
 ## Verifying a rebuild
 
-`parser/scripts/db-stats.js` dumps row counts, per-column non-NULL counts, file
-size and a deterministic student sample as JSON.
-`parser/scripts/verify-parity.js` diffs two such files and exits non-zero on any
-mismatch.
+`npm run build:db` verifies itself: each database's row count must match the
+figure in the table above, and each `.db.gz` must be at least 90% of its usual
+size, or the build fails rather than publishing. That guard is the reason a
+truncated dataset cannot reach the site with a green pipeline.
+
+For a deeper check, `go-parser/scripts/differential-parity.mjs` compares two sets
+of databases field-by-field — row counts, per-column non-NULL counts, a
+full-table SHA-256 over every row ordered by `so_bao_danh`, schema metadata, and
+build stdout:
 
 ```bash
-node parser/scripts/db-stats.js 2016=<path>.db … > current.json
-node parser/scripts/verify-parity.js plans/reports/parser-parity-baseline.json current.json
+node go-parser/scripts/differential-parity.mjs \
+  --rust /path/to/a-{id}.db --go /path/to/b-{id}.db
 ```
 
-The committed baseline was built with the pre-refactor code and cannot be
-regenerated — the two old crates no longer exist. Both scripts use the built-in
-`node:sqlite`, so they need no dependencies.
+It exits non-zero on any mismatch and fails loudly if a dataset is missing rather
+than skipping it. Written for the Rust-to-Go migration, it works for any two
+builds. Uses the built-in `node:sqlite`, so it needs no dependencies.
+
+`go-parser/internal/reader` additionally carries a frozen oracle of per-file
+cell-dump hashes covering all 299 inputs; `npm run test:go` fails if any single
+cell of any input file reads differently.
 
 ## Refreshing the 2017 data
 
 ```bash
 rm data/2017/*.xls
-node parser/scripts/crawl-baotintuc.js
+node go-parser/scripts/crawl-baotintuc.js
 node go-parser/scripts/build-db.js 2017
 ```
 
-Then re-run the parity check above and confirm the row count still matches.
+The row-count guard in `build:db` confirms the rebuild matches the expected total.
 
-## Legacy scripts
+## Removed scripts
 
-`parser/scripts/check-duplicates.js` and `diff-datasets.js` are one-off audits
-that were already broken before the repo was unified — a hardcoded Windows path
-in one, an undeclared `better-sqlite3` dependency and stale paths in the other.
-Each carries a comment saying so. For comparing two builds, use `db-stats.js`
-plus `verify-parity.js` instead.
+`check-duplicates.js`, `diff-datasets.js`, `db-stats.js` and `verify-parity.js`
+were dropped with the Rust parser. The first two had been broken since before the
+repo was unified (a hardcoded Windows path in one, an undeclared
+`better-sqlite3` dependency in the other) and neither had any automated caller.
+The latter two are superseded by `differential-parity.mjs`, which compares more
+and cannot silently skip a dataset.
diff --git a/docs/deployment-guide.md b/docs/deployment-guide.md
index 65af318..80710bc 100644
--- a/docs/deployment-guide.md
+++ b/docs/deployment-guide.md
@@ -56,7 +56,7 @@ up as a blank page with 404s on `/assets/...`.
 ## Adding a dataset
 
 1. Put the Excel files in `data/<id>/`
-2. Add `parser/configs/<id>.yml` with the parse rules — sheet mode, column
+2. Add `go-parser/configs/<id>.yml` with the parse rules — sheet mode, column
    indices, SBD validation, header tokens, blank-row stripping. No SQL: the
    schema is canonical and lives in `go-parser/internal/schema/schema.go`
 3. Add an entry to `DATASETS` in `src/datasets.js`
diff --git a/docs/system-architecture.md b/docs/system-architecture.md
index e525488..ab5040d 100644
--- a/docs/system-architecture.md
+++ b/docs/system-architecture.md
@@ -31,7 +31,7 @@ data/<id>/*.xls(x)
 One identifier ties the whole pipeline together:
 
 ```
-data/2017-old/  →  parser/configs/2017-old.yml  →  db/2017-old.db.gz  →  /thptqg/2017-old/
+data/2017-old/  →  go-parser/configs/2017-old.yml  →  db/2017-old.db.gz  →  /thptqg/2017-old/
 ```
 
 `src/datasets.js` declares the four ids once. The frontend, the database build
diff --git a/eslint.config.js b/eslint.config.js
index 097adc2..7ce9419 100644
--- a/eslint.config.js
+++ b/eslint.config.js
@@ -28,7 +28,7 @@ export default defineConfig([
   },
   {
     // Node-executed files (Vite config, parser tooling) run with Node globals.
-    files: ['vite.config.js', 'scripts/**/*.js', 'go-parser/scripts/**/*.js', 'parser/scripts/**/*.js'],
+    files: ['vite.config.js', 'scripts/**/*.js', 'go-parser/scripts/**/*.js'],
     languageOptions: {
       globals: { ...globals.node },
     },
diff --git a/go-parser/README.md b/go-parser/README.md
new file mode 100644
index 0000000..9d25b24
--- /dev/null
+++ b/go-parser/README.md
@@ -0,0 +1,67 @@
+# go-parser
+
+Reads the `.xls`/`.xlsx` source spreadsheets in `data/` and writes one SQLite
+database per dataset.
+
+```bash
+npm run build:go    # compile go-parser/bin/xlsxread
+npm run build:db    # build + verify + gzip all four datasets
+npm run test:go     # unit tests + the 299-file reader-fidelity suite
+```
+
+```
+xlsxread build --schema go-parser/configs/<id>.yml --input data/<id> --output <db>
+xlsxread audit --schema go-parser/configs/<id>.yml --input data/<id> --db <db>
+```
+
+## Layout
+
+| path | role |
+|---|---|
+| `internal/reader` | spreadsheet reading; the only place that knows about file formats |
+| `internal/ingest` | dataset policy — sheet selection, header skipping, blank rows, the build loop, and the 2016 per-sheet format detection |
+| `internal/transform` | `ToAscii`, score-regex parsing, row validation |
+| `internal/schema` | the canonical 22-column table: DDL, INSERT, subject regexes |
+| `internal/config` | per-dataset YAML parse rules |
+| `internal/writer` | SQLite lifecycle and the stats block |
+| `internal/audit` | source-vs-database SBD comparison |
+
+The reader deliberately knows nothing about datasets: it reports every sheet and
+every row verbatim. All policy lives in `ingest`. That split is what made the
+reader independently verifiable against a hash oracle.
+
+## Provenance
+
+This is a port of a Rust crate that lived at `parser/` until the Go
+implementation reached full parity. Source comments cite the original by file
+and line (`parser/src/transform.rs:56` and similar); those paths resolve at the
+tag **`pre-go-parser-removal`**, the last commit containing the Rust code.
+
+The port was gated on a field-by-field comparison of both implementations across
+all four datasets — 3,265,641 rows with identical full-table SHA-256, identical
+per-column non-NULL counts, identical schema metadata and identical build
+stdout. `scripts/differential-parity.mjs` is that comparator and still runs
+against any two sets of databases.
+
+Behaviour was matched bug-for-bug, deliberately. Several quirks look like
+defects and are load-bearing for the published data:
+
+- a parsed score of `0` becomes NULL in the 2016 separate-scores layout,
+  replicating a JavaScript falsy check;
+- `ToAscii` strips combining marks in the literal range U+0300–U+036F rather
+  than by Unicode category, which is narrower;
+- gender is a two-value allowlist, and anything else becomes NULL;
+- `diem_thi` is read untrimmed while the other three fields are trimmed;
+- the `"SINH "` header token carries a trailing space.
+
+Each has a test naming it, so none can be tidied away by accident.
+
+## Verification
+
+`testdata/reader-fidelity-hashes.tsv` holds a SHA-256 per input file over a
+canonical dump of every cell of every sheet. It is **frozen**: it was produced
+by the Rust reader, which no longer exists, so it cannot be regenerated. It
+still fails if any single cell of any of the 299 files reads differently.
+
+`npm run build:db` refuses to publish a database whose row count does not match
+the known figure, or whose artifact is under 90% of its usual size.
diff --git a/parser/configs/2016.yml b/go-parser/configs/2016.yml
similarity index 100%
rename from parser/configs/2016.yml
rename to go-parser/configs/2016.yml
diff --git a/parser/configs/2017-old.yml b/go-parser/configs/2017-old.yml
similarity index 100%
rename from parser/configs/2017-old.yml
rename to go-parser/configs/2017-old.yml
diff --git a/parser/configs/2017-old2.yml b/go-parser/configs/2017-old2.yml
similarity index 100%
rename from parser/configs/2017-old2.yml
rename to go-parser/configs/2017-old2.yml
diff --git a/parser/configs/2017.yml b/go-parser/configs/2017.yml
similarity index 100%
rename from parser/configs/2017.yml
rename to go-parser/configs/2017.yml
diff --git a/go-parser/internal/config/config_test.go b/go-parser/internal/config/config_test.go
index 8a88ad5..c3a1a1c 100644
--- a/go-parser/internal/config/config_test.go
+++ b/go-parser/internal/config/config_test.go
@@ -155,7 +155,7 @@ func TestLoadRealConfigs(t *testing.T) {
 	}
 	for id, w := range want {
 		t.Run(id, func(t *testing.T) {
-			cfg, err := Load(filepath.Join(root, "parser", "configs", id+".yml"))
+			cfg, err := Load(filepath.Join(root, "go-parser", "configs", id+".yml"))
 			if err != nil {
 				t.Fatalf("load: %v", err)
 			}
@@ -217,7 +217,7 @@ func repoRoot(t *testing.T) string {
 		t.Fatalf("getwd: %v", err)
 	}
 	for i := 0; i < 6; i++ {
-		if _, err := os.Stat(filepath.Join(dir, "parser", "configs")); err == nil {
+		if _, err := os.Stat(filepath.Join(dir, "go-parser", "configs")); err == nil {
 			return dir
 		}
 		dir = filepath.Dir(dir)
diff --git a/go-parser/scripts/build-db.js b/go-parser/scripts/build-db.js
index 874bbe9..f3ac9da 100644
--- a/go-parser/scripts/build-db.js
+++ b/go-parser/scripts/build-db.js
@@ -81,7 +81,7 @@ for (const id of targets) {
     [
       "build",
       "--schema",
-      resolve(ROOT, `parser/configs/${id}.yml`),
+      resolve(ROOT, `go-parser/configs/${id}.yml`),
       "--input",
       resolve(ROOT, `data/${id}`),
       "--output",
diff --git a/parser/scripts/crawl-baotintuc.js b/go-parser/scripts/crawl-baotintuc.js
similarity index 100%
rename from parser/scripts/crawl-baotintuc.js
rename to go-parser/scripts/crawl-baotintuc.js
diff --git a/go-parser/scripts/regen-fidelity-hashes.sh b/go-parser/scripts/regen-fidelity-hashes.sh
deleted file mode 100755
index c26d5d2..0000000
--- a/go-parser/scripts/regen-fidelity-hashes.sh
+++ /dev/null
@@ -1,34 +0,0 @@
-#!/usr/bin/env bash
-# Regenerates the reader-fidelity oracle from the Rust/calamine ground truth.
-#
-# Emits SHA-256 per input file over the canonical cell dump. Only hashes are
-# committed: the dumps are real student PII, and parser/tests/fixtures/README.md
-# establishes that fixtures in this repo carry synthetic data only.
-#
-# Requires the Rust parser to still build. Run from the repo root.
-set -euo pipefail
-
-OUT=go-parser/testdata/reader-fidelity-hashes.tsv
-CANON='BEGIN{OFS="\t"}
-  $1=="FILE"{next}
-  $1=="SHEETCOUNT"{print;next}
-  $1=="SHEET"{print $1,$2,$3,$4,$5;next}
-  $1=="ROW"{print;next}
-  $1=="CELL"{print $1,$2,$3,$4,$7;next}'
-
-{
-  echo "# Canonical cell-dump SHA-256 per input file, produced by the Rust/calamine"
-  echo "# ground truth (parser/examples/dump_cells.rs). The Go reader must reproduce"
-  echo "# each hash exactly. Hashes only - the dumps themselves are real student PII"
-  echo "# and are never committed, per parser/tests/fixtures/README.md."
-  echo "# Regenerate: go-parser/scripts/regen-fidelity-hashes.sh"
-} > "$OUT"
-
-for f in data/2016/* data/2017/* data/2017-old/* data/2017-old2/*; do
-  h=$(cargo run --release --quiet --manifest-path parser/Cargo.toml \
-        --example dump_cells -- "$f" /dev/stdout \
-      | awk -F'\t' "$CANON" | sha256sum | cut -d' ' -f1)
-  printf '%s\t%s\n' "$f" "$h" >> "$OUT"
-done
-
-echo "wrote $(grep -cv '^#' "$OUT") hashes to $OUT"
diff --git a/go-parser/testdata/reader-fidelity-hashes.tsv b/go-parser/testdata/reader-fidelity-hashes.tsv
index 72d1d5c..8c9ce2f 100644
--- a/go-parser/testdata/reader-fidelity-hashes.tsv
+++ b/go-parser/testdata/reader-fidelity-hashes.tsv
@@ -2,7 +2,12 @@
 # ground truth (parser/examples/dump_cells.rs). The Go reader must reproduce
 # each hash exactly. Hashes only - the dumps themselves are real student PII
 # and are never committed, per parser/tests/fixtures/README.md.
-# Regenerate: go-parser/scripts/regen-fidelity-hashes.sh
+# FROZEN. These hashes were produced by the Rust/calamine parser, which has
+# since been removed, so they cannot be regenerated. They remain a real
+# regression guard: any change to the Go reader that alters a single cell of any
+# of the 299 input files fails the suite. If the inputs themselves ever change,
+# the affected rows must be re-derived deliberately and reviewed, not refreshed
+# in bulk.
 data/2016/023718c7d3cf7ace3a7116fabb12bd9cdhyduoccantho-1468920829104.xlsx	fc3afd8da9ed28b0fc177192fab7b19a06535cb45d86841835d8e8828fc61c1c
 data/2016/08cdeb3636dcb2adaef829d62968274atravinh-1468902443859.xlsx	71f0e61918c63b95f08a16d502f5e9f2054e3c830bd83c725f1285bb1aabead1
 data/2016/0b583c9875443e65ffa449d5ca76fae8ninhbinh-1468899778557.xlsx	b0c2cfb9ab419442f151be5a014439f7f477fa7b31a9ae3f37b83073276bdb59
diff --git a/parser/Cargo.lock b/parser/Cargo.lock
deleted file mode 100644
index bcd7270..0000000
--- a/parser/Cargo.lock
+++ /dev/null
@@ -1,1131 +0,0 @@
-# This file is automatically @generated by Cargo.
-# It is not intended for manual editing.
-version = 4
-
-[[package]]
-name = "adler2"
-version = "2.0.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "320119579fcad9c21884f5c4861d16174d0e06250625266f50fe6898340abefa"
-
-[[package]]
-name = "aes"
-version = "0.8.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b169f7a6d4742236a0a00c541b845991d0ac43e546831af1249753ab4c3aa3a0"
-dependencies = [
- "cfg-if",
- "cipher",
- "cpufeatures",
-]
-
-[[package]]
-name = "ahash"
-version = "0.8.12"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5a15f179cd60c4584b8a8c596927aadc462e27f2ca70c04e0071964a73ba7a75"
-dependencies = [
- "cfg-if",
- "once_cell",
- "version_check",
- "zerocopy",
-]
-
-[[package]]
-name = "aho-corasick"
-version = "1.1.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ddd31a130427c27518df266943a5308ed92d4b226cc639f5a8f1002816174301"
-dependencies = [
- "memchr",
-]
-
-[[package]]
-name = "anstream"
-version = "1.0.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "824a212faf96e9acacdbd09febd34438f8f711fb84e09a8916013cd7815ca28d"
-dependencies = [
- "anstyle",
- "anstyle-parse",
- "anstyle-query",
- "anstyle-wincon",
- "colorchoice",
- "is_terminal_polyfill",
- "utf8parse",
-]
-
-[[package]]
-name = "anstyle"
-version = "1.0.14"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "940b3a0ca603d1eade50a4846a2afffd5ef57a9feac2c0e2ec2e14f9ead76000"
-
-[[package]]
-name = "anstyle-parse"
-version = "1.0.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "52ce7f38b242319f7cabaa6813055467063ecdc9d355bbb4ce0c68908cd8130e"
-dependencies = [
- "utf8parse",
-]
-
-[[package]]
-name = "anstyle-query"
-version = "1.1.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "40c48f72fd53cd289104fc64099abca73db4166ad86ea0b4341abe65af83dadc"
-dependencies = [
- "windows-sys",
-]
-
-[[package]]
-name = "anstyle-wincon"
-version = "3.0.11"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "291e6a250ff86cd4a820112fb8898808a366d8f9f58ce16d1f538353ad55747d"
-dependencies = [
- "anstyle",
- "once_cell_polyfill",
- "windows-sys",
-]
-
-[[package]]
-name = "anyhow"
-version = "1.0.102"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7f202df86484c868dbad7eaa557ef785d5c66295e41b460ef922eca0723b842c"
-
-[[package]]
-name = "arbitrary"
-version = "1.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c3d036a3c4ab069c7b410a2ce876bd74808d2d0888a82667669f8e783a898bf1"
-dependencies = [
- "derive_arbitrary",
-]
-
-[[package]]
-name = "bitflags"
-version = "2.11.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c4512299f36f043ab09a583e57bceb5a5aab7a73db1805848e8fef3c9e8c78b3"
-
-[[package]]
-name = "block-buffer"
-version = "0.10.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3078c7629b62d3f0439517fa394996acacc5cbc91c5a20d8c658e77abd503a71"
-dependencies = [
- "generic-array",
-]
-
-[[package]]
-name = "bumpalo"
-version = "3.20.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5d20789868f4b01b2f2caec9f5c4e0213b41e3e5702a50157d699ae31ced2fcb"
-
-[[package]]
-name = "byteorder"
-version = "1.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1fd0f2584146f6f2ef48085050886acf353beff7305ebd1ae69500e27c67f64b"
-
-[[package]]
-name = "bzip2"
-version = "0.5.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "49ecfb22d906f800d4fe833b6282cf4dc1c298f5057ca0b5445e5c209735ca47"
-dependencies = [
- "bzip2-sys",
-]
-
-[[package]]
-name = "bzip2-sys"
-version = "0.1.13+1.0.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "225bff33b2141874fe80d71e07d6eec4f85c5c216453dd96388240f96e1acc14"
-dependencies = [
- "cc",
- "pkg-config",
-]
-
-[[package]]
-name = "calamine"
-version = "0.26.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "138646b9af2c5d7f1804ea4bf93afc597737d2bd4f7341d67c48b03316976eb1"
-dependencies = [
- "byteorder",
- "codepage",
- "encoding_rs",
- "log",
- "quick-xml",
- "serde",
- "zip",
-]
-
-[[package]]
-name = "cc"
-version = "1.2.62"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a1dce859f0832a7d088c4f1119888ab94ef4b5d6795d1ce05afb7fe159d79f98"
-dependencies = [
- "find-msvc-tools",
- "jobserver",
- "libc",
- "shlex",
-]
-
-[[package]]
-name = "cfg-if"
-version = "1.0.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9330f8b2ff13f34540b44e946ef35111825727b38d33286ef986142615121801"
-
-[[package]]
-name = "cipher"
-version = "0.4.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "773f3b9af64447d2ce9850330c473515014aa235e6a783b02db81ff39e4a3dad"
-dependencies = [
- "crypto-common",
- "inout",
-]
-
-[[package]]
-name = "clap"
-version = "4.6.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1ddb117e43bbf7dacf0a4190fef4d345b9bad68dfc649cb349e7d17d28428e51"
-dependencies = [
- "clap_builder",
- "clap_derive",
-]
-
-[[package]]
-name = "clap_builder"
-version = "4.6.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "714a53001bf66416adb0e2ef5ac857140e7dc3a0c48fb28b2f10762fc4b5069f"
-dependencies = [
- "anstream",
- "anstyle",
- "clap_lex",
- "strsim",
-]
-
-[[package]]
-name = "clap_derive"
-version = "4.6.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f2ce8604710f6733aa641a2b3731eaa1e8b3d9973d5e3565da11800813f997a9"
-dependencies = [
- "heck",
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "clap_lex"
-version = "1.1.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c8d4a3bb8b1e0c1050499d1815f5ab16d04f0959b233085fb31653fbfc9d98f9"
-
-[[package]]
-name = "codepage"
-version = "0.1.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "48f68d061bc2828ae826206326e61251aca94c1e4a5305cf52d9138639c918b4"
-dependencies = [
- "encoding_rs",
-]
-
-[[package]]
-name = "colorchoice"
-version = "1.0.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1d07550c9036bf2ae0c684c4297d503f838287c83c53686d05370d0e139ae570"
-
-[[package]]
-name = "constant_time_eq"
-version = "0.3.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7c74b8349d32d297c9134b8c88677813a227df8f779daa29bfc29c183fe3dca6"
-
-[[package]]
-name = "cpufeatures"
-version = "0.2.17"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "59ed5838eebb26a2bb2e58f6d5b5316989ae9d08bab10e0e6d103e656d1b0280"
-dependencies = [
- "libc",
-]
-
-[[package]]
-name = "crc"
-version = "3.4.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5eb8a2a1cd12ab0d987a5d5e825195d372001a4094a0376319d5a0ad71c1ba0d"
-dependencies = [
- "crc-catalog",
-]
-
-[[package]]
-name = "crc-catalog"
-version = "2.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "217698eaf96b4a3f0bc4f3662aaa55bdf913cd54d7204591faa790070c6d0853"
-
-[[package]]
-name = "crc32fast"
-version = "1.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9481c1c90cbf2ac953f07c8d4a58aa3945c425b7185c9154d67a65e4230da511"
-dependencies = [
- "cfg-if",
-]
-
-[[package]]
-name = "crossbeam-utils"
-version = "0.8.21"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d0a5c400df2834b80a4c3327b3aad3a4c4cd4de0629063962b03235697506a28"
-
-[[package]]
-name = "crypto-common"
-version = "0.1.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "78c8292055d1c1df0cce5d180393dc8cce0abec0a7102adb6c7b1eef6016d60a"
-dependencies = [
- "generic-array",
- "typenum",
-]
-
-[[package]]
-name = "deflate64"
-version = "0.1.12"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ac6b926516df9c60bfa16e107b21086399f8285a44ca9711344b9e553c5146e2"
-
-[[package]]
-name = "deranged"
-version = "0.5.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7cd812cc2bc1d69d4764bd80df88b4317eaef9e773c75226407d9bc0876b211c"
-dependencies = [
- "powerfmt",
-]
-
-[[package]]
-name = "derive_arbitrary"
-version = "1.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1e567bd82dcff979e4b03460c307b3cdc9e96fde3d73bed1496d2bc75d9dd62a"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "digest"
-version = "0.10.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9ed9a281f7bc9b7576e61468ba615a66a5c8cfdff42420a70aa82701a3b1e292"
-dependencies = [
- "block-buffer",
- "crypto-common",
- "subtle",
-]
-
-[[package]]
-name = "displaydoc"
-version = "0.2.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "97369cbbc041bc366949bc74d34658d6cda5621039731c6310521892a3a20ae0"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "encoding_rs"
-version = "0.8.35"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "75030f3c4f45dafd7586dd6780965a8c7e8e285a5ecb86713e63a79c5b2766f3"
-dependencies = [
- "cfg-if",
-]
-
-[[package]]
-name = "equivalent"
-version = "1.0.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "877a4ace8713b0bcf2a4e7eec82529c029f1d0619886d18145fea96c3ffe5c0f"
-
-[[package]]
-name = "fallible-iterator"
-version = "0.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2acce4a10f12dc2fb14a218589d4f1f62ef011b2d0cc4b3cb1bba8e94da14649"
-
-[[package]]
-name = "fallible-streaming-iterator"
-version = "0.1.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7360491ce676a36bf9bb3c56c1aa791658183a54d2744120f27285738d90465a"
-
-[[package]]
-name = "find-msvc-tools"
-version = "0.1.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5baebc0774151f905a1a2cc41989300b1e6fbb29aff0ceffa1064fdd3088d582"
-
-[[package]]
-name = "flate2"
-version = "1.1.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "843fba2746e448b37e26a819579957415c8cef339bf08564fe8b7ddbd959573c"
-dependencies = [
- "crc32fast",
- "miniz_oxide",
-]
-
-[[package]]
-name = "generic-array"
-version = "0.14.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "85649ca51fd72272d7821adaf274ad91c288277713d9c18820d8499a7ff69e9a"
-dependencies = [
- "typenum",
- "version_check",
-]
-
-[[package]]
-name = "getrandom"
-version = "0.3.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "899def5c37c4fd7b2664648c28120ecec138e4d395b459e5ca34f9cce2dd77fd"
-dependencies = [
- "cfg-if",
- "js-sys",
- "libc",
- "r-efi",
- "wasip2",
- "wasm-bindgen",
-]
-
-[[package]]
-name = "glob"
-version = "0.3.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0cc23270f6e1808e30a928bdc84dea0b9b4136a8bc82338574f23baf47bbd280"
-
-[[package]]
-name = "hashbrown"
-version = "0.14.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e5274423e17b7c9fc20b6e7e208532f9b19825d82dfd615708b70edd83df41f1"
-dependencies = [
- "ahash",
-]
-
-[[package]]
-name = "hashbrown"
-version = "0.17.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ed5909b6e89a2db4456e54cd5f673791d7eca6732202bbf2a9cc504fe2f9b84a"
-
-[[package]]
-name = "hashlink"
-version = "0.9.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6ba4ff7128dee98c7dc9794b6a411377e1404dba1c97deb8d1a55297bd25d8af"
-dependencies = [
- "hashbrown 0.14.5",
-]
-
-[[package]]
-name = "heck"
-version = "0.5.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
-
-[[package]]
-name = "hmac"
-version = "0.12.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6c49c37c09c17a53d937dfbb742eb3a961d65a994e6bcdcf37e7399d0cc8ab5e"
-dependencies = [
- "digest",
-]
-
-[[package]]
-name = "indexmap"
-version = "2.14.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d466e9454f08e4a911e14806c24e16fba1b4c121d1ea474396f396069cf949d9"
-dependencies = [
- "equivalent",
- "hashbrown 0.17.1",
-]
-
-[[package]]
-name = "inout"
-version = "0.1.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "879f10e63c20629ecabbb64a8010319738c66a5cd0c29b02d63d272b03751d01"
-dependencies = [
- "generic-array",
-]
-
-[[package]]
-name = "is_terminal_polyfill"
-version = "1.70.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a6cb138bb79a146c1bd460005623e142ef0181e3d0219cb493e02f7d08a35695"
-
-[[package]]
-name = "itoa"
-version = "1.0.18"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8f42a60cbdf9a97f5d2305f08a87dc4e09308d1276d28c869c684d7777685682"
-
-[[package]]
-name = "jobserver"
-version = "0.1.34"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9afb3de4395d6b3e67a780b6de64b51c978ecf11cb9a462c66be7d4ca9039d33"
-dependencies = [
- "getrandom",
- "libc",
-]
-
-[[package]]
-name = "js-sys"
-version = "0.3.98"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "67df7112613f8bfd9150013a0314e196f4800d3201ae742489d999db2f979f08"
-dependencies = [
- "cfg-if",
- "once_cell",
- "wasm-bindgen",
-]
-
-[[package]]
-name = "libc"
-version = "0.2.186"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "68ab91017fe16c622486840e4c83c9a37afeff978bd239b5293d61ece587de66"
-
-[[package]]
-name = "libsqlite3-sys"
-version = "0.30.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2e99fb7a497b1e3339bc746195567ed8d3e24945ecd636e3619d20b9de9e9149"
-dependencies = [
- "cc",
- "pkg-config",
- "vcpkg",
-]
-
-[[package]]
-name = "log"
-version = "0.4.29"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5e5032e24019045c762d3c0f28f5b6b8bbf38563a65908389bf7978758920897"
-
-[[package]]
-name = "lzma-rs"
-version = "0.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "297e814c836ae64db86b36cf2a557ba54368d03f6afcd7d947c266692f71115e"
-dependencies = [
- "byteorder",
- "crc",
-]
-
-[[package]]
-name = "lzma-sys"
-version = "0.1.20"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5fda04ab3764e6cde78b9974eec4f779acaba7c4e84b36eca3cf77c581b85d27"
-dependencies = [
- "cc",
- "libc",
- "pkg-config",
-]
-
-[[package]]
-name = "memchr"
-version = "2.8.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f8ca58f447f06ed17d5fc4043ce1b10dd205e060fb3ce5b979b8ed8e59ff3f79"
-
-[[package]]
-name = "miniz_oxide"
-version = "0.8.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1fa76a2c86f704bdb222d66965fb3d63269ce38518b83cb0575fca855ebb6316"
-dependencies = [
- "adler2",
- "simd-adler32",
-]
-
-[[package]]
-name = "num-conv"
-version = "0.2.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "521739c6d2bac4aa25192232afe6841231376b2b26d4d9fae5ecf8ca5772e441"
-
-[[package]]
-name = "once_cell"
-version = "1.21.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9f7c3e4beb33f85d45ae3e3a1792185706c8e16d043238c593331cc7cd313b50"
-
-[[package]]
-name = "once_cell_polyfill"
-version = "1.70.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "384b8ab6d37215f3c5301a95a4accb5d64aa607f1fcb26a11b5303878451b4fe"
-
-[[package]]
-name = "pbkdf2"
-version = "0.12.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f8ed6a7761f76e3b9f92dfb0a60a6a6477c61024b775147ff0973a02653abaf2"
-dependencies = [
- "digest",
- "hmac",
-]
-
-[[package]]
-name = "pkg-config"
-version = "0.3.33"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "19f132c84eca552bf34cab8ec81f1c1dcc229b811638f9d283dceabe58c5569e"
-
-[[package]]
-name = "powerfmt"
-version = "0.2.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "439ee305def115ba05938db6eb1644ff94165c5ab5e9420d1c1bcedbba909391"
-
-[[package]]
-name = "proc-macro2"
-version = "1.0.106"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8fd00f0bb2e90d81d1044c2b32617f68fcb9fa3bb7640c23e9c748e53fb30934"
-dependencies = [
- "unicode-ident",
-]
-
-[[package]]
-name = "quick-xml"
-version = "0.31.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1004a344b30a54e2ee58d66a71b32d2db2feb0a31f9a2d302bf0536f15de2a33"
-dependencies = [
- "encoding_rs",
- "memchr",
-]
-
-[[package]]
-name = "quote"
-version = "1.0.45"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "41f2619966050689382d2b44f664f4bc593e129785a36d6ee376ddf37259b924"
-dependencies = [
- "proc-macro2",
-]
-
-[[package]]
-name = "r-efi"
-version = "5.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "69cdb34c158ceb288df11e18b4bd39de994f6657d83847bdffdbd7f346754b0f"
-
-[[package]]
-name = "regex"
-version = "1.12.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e10754a14b9137dd7b1e3e5b0493cc9171fdd105e0ab477f51b72e7f3ac0e276"
-dependencies = [
- "aho-corasick",
- "memchr",
- "regex-automata",
- "regex-syntax",
-]
-
-[[package]]
-name = "regex-automata"
-version = "0.4.14"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6e1dd4122fc1595e8162618945476892eefca7b88c52820e74af6262213cae8f"
-dependencies = [
- "aho-corasick",
- "memchr",
- "regex-syntax",
-]
-
-[[package]]
-name = "regex-syntax"
-version = "0.8.10"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "dc897dd8d9e8bd1ed8cdad82b5966c3e0ecae09fb1907d58efaa013543185d0a"
-
-[[package]]
-name = "rusqlite"
-version = "0.32.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7753b721174eb8ff87a9a0e799e2d7bc3749323e773db92e0984debb00019d6e"
-dependencies = [
- "bitflags",
- "fallible-iterator",
- "fallible-streaming-iterator",
- "hashlink",
- "libsqlite3-sys",
- "smallvec",
-]
-
-[[package]]
-name = "rustversion"
-version = "1.0.22"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b39cdef0fa800fc44525c84ccb54a029961a8215f9619753635a9c0d2538d46d"
-
-[[package]]
-name = "ryu"
-version = "1.0.23"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9774ba4a74de5f7b1c1451ed6cd5285a32eddb5cccb8cc655a4e50009e06477f"
-
-[[package]]
-name = "serde"
-version = "1.0.228"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9a8e94ea7f378bd32cbbd37198a4a91436180c5bb472411e48b5ec2e2124ae9e"
-dependencies = [
- "serde_core",
- "serde_derive",
-]
-
-[[package]]
-name = "serde_core"
-version = "1.0.228"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "41d385c7d4ca58e59fc732af25c3983b67ac852c1a25000afe1175de458b67ad"
-dependencies = [
- "serde_derive",
-]
-
-[[package]]
-name = "serde_derive"
-version = "1.0.228"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d540f220d3187173da220f885ab66608367b6574e925011a9353e4badda91d79"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "serde_yaml"
-version = "0.9.34+deprecated"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6a8b1a1a2ebf674015cc02edccce75287f1a0130d394307b36743c2f5d504b47"
-dependencies = [
- "indexmap",
- "itoa",
- "ryu",
- "serde",
- "unsafe-libyaml",
-]
-
-[[package]]
-name = "sha1"
-version = "0.10.6"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e3bf829a2d51ab4a5ddf1352d8470c140cadc8301b2ae1789db023f01cedd6ba"
-dependencies = [
- "cfg-if",
- "cpufeatures",
- "digest",
-]
-
-[[package]]
-name = "shlex"
-version = "1.3.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0fda2ff0d084019ba4d7c6f371c95d8fd75ce3524c3cb8fb653a3023f6323e64"
-
-[[package]]
-name = "simd-adler32"
-version = "0.3.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "703d5c7ef118737c72f1af64ad2f6f8c5e1921f818cdcb97b8fe6fc69bf66214"
-
-[[package]]
-name = "smallvec"
-version = "1.15.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "67b1b7a3b5fe4f1376887184045fcf45c69e92af734b7aaddc05fb777b6fbd03"
-
-[[package]]
-name = "strsim"
-version = "0.11.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7da8b5736845d9f2fcb837ea5d9e2628564b3b043a70948a3f0b778838c5fb4f"
-
-[[package]]
-name = "subtle"
-version = "2.6.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "13c2bddecc57b384dee18652358fb23172facb8a2c51ccc10d74c157bdea3292"
-
-[[package]]
-name = "syn"
-version = "2.0.117"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e665b8803e7b1d2a727f4023456bbbbe74da67099c585258af0ad9c5013b9b99"
-dependencies = [
- "proc-macro2",
- "quote",
- "unicode-ident",
-]
-
-[[package]]
-name = "thiserror"
-version = "1.0.69"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b6aaf5339b578ea85b50e080feb250a3e8ae8cfcdff9a461c9ec2904bc923f52"
-dependencies = [
- "thiserror-impl 1.0.69",
-]
-
-[[package]]
-name = "thiserror"
-version = "2.0.18"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4288b5bcbc7920c07a1149a35cf9590a2aa808e0bc1eafaade0b80947865fbc4"
-dependencies = [
- "thiserror-impl 2.0.18",
-]
-
-[[package]]
-name = "thiserror-impl"
-version = "1.0.69"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4fee6c4efc90059e10f81e6d42c60a18f76588c3d74cb83a0b242a2b6c7504c1"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "thiserror-impl"
-version = "2.0.18"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ebc4ee7f67670e9b64d05fa4253e753e016c6c95ff35b89b7941d6b856dec1d5"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "time"
-version = "0.3.47"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "743bd48c283afc0388f9b8827b976905fb217ad9e647fae3a379a9283c4def2c"
-dependencies = [
- "deranged",
- "num-conv",
- "powerfmt",
- "serde_core",
- "time-core",
-]
-
-[[package]]
-name = "time-core"
-version = "0.1.8"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "7694e1cfe791f8d31026952abf09c69ca6f6fa4e1a1229e18988f06a04a12dca"
-
-[[package]]
-name = "tinyvec"
-version = "1.11.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "3e61e67053d25a4e82c844e8424039d9745781b3fc4f32b8d55ed50f5f667ef3"
-dependencies = [
- "tinyvec_macros",
-]
-
-[[package]]
-name = "tinyvec_macros"
-version = "0.1.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
-
-[[package]]
-name = "typenum"
-version = "1.20.0"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "40ce102ab67701b8526c123c1bab5cbe42d7040ccfd0f64af1a385808d2f43de"
-
-[[package]]
-name = "unicode-ident"
-version = "1.0.24"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e6e4313cd5fcd3dad5cafa179702e2b244f760991f45397d14d4ebf38247da75"
-
-[[package]]
-name = "unicode-normalization"
-version = "0.1.25"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5fd4f6878c9cb28d874b009da9e8d183b5abc80117c40bbd187a1fde336be6e8"
-dependencies = [
- "tinyvec",
-]
-
-[[package]]
-name = "unsafe-libyaml"
-version = "0.2.11"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "673aac59facbab8a9007c7f6108d11f63b603f7cabff99fabf650fea5c32b861"
-
-[[package]]
-name = "utf8parse"
-version = "0.2.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
-
-[[package]]
-name = "vcpkg"
-version = "0.2.15"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "accd4ea62f7bb7a82fe23066fb0957d48ef677f6eeb8215f372f52e48bb32426"
-
-[[package]]
-name = "version_check"
-version = "0.9.5"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0b928f33d975fc6ad9f86c8f283853ad26bdd5b10b7f1542aa2fa15e2289105a"
-
-[[package]]
-name = "wasip2"
-version = "1.0.3+wasi-0.2.9"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "20064672db26d7cdc89c7798c48a0fdfac8213434a1186e5ef29fd560ae223d6"
-dependencies = [
- "wit-bindgen",
-]
-
-[[package]]
-name = "wasm-bindgen"
-version = "0.2.121"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "49ace1d07c165b0864824eee619580c4689389afa9dc9ed3a4c75040d82e6790"
-dependencies = [
- "cfg-if",
- "once_cell",
- "rustversion",
- "wasm-bindgen-macro",
- "wasm-bindgen-shared",
-]
-
-[[package]]
-name = "wasm-bindgen-macro"
-version = "0.2.121"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8e68e6f4afd367a562002c05637acb8578ff2dea1943df76afb9e83d177c8578"
-dependencies = [
- "quote",
- "wasm-bindgen-macro-support",
-]
-
-[[package]]
-name = "wasm-bindgen-macro-support"
-version = "0.2.121"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d95a9ec35c64b2a7cb35d3fead40c4238d0940c86d107136999567a4703259f2"
-dependencies = [
- "bumpalo",
- "proc-macro2",
- "quote",
- "syn",
- "wasm-bindgen-shared",
-]
-
-[[package]]
-name = "wasm-bindgen-shared"
-version = "0.2.121"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c4e0100b01e9f0d03189a92b96772a1fb998639d981193d7dbab487302513441"
-dependencies = [
- "unicode-ident",
-]
-
-[[package]]
-name = "windows-link"
-version = "0.2.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f0805222e57f7521d6a62e36fa9163bc891acd422f971defe97d64e70d0a4fe5"
-
-[[package]]
-name = "windows-sys"
-version = "0.61.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ae137229bcbd6cdf0f7b80a31df61766145077ddf49416a728b02cb3921ff3fc"
-dependencies = [
- "windows-link",
-]
-
-[[package]]
-name = "wit-bindgen"
-version = "0.57.1"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "1ebf944e87a7c253233ad6766e082e3cd714b5d03812acc24c318f549614536e"
-
-[[package]]
-name = "xlsxread"
-version = "0.1.0"
-dependencies = [
- "anyhow",
- "calamine",
- "clap",
- "glob",
- "regex",
- "rusqlite",
- "serde",
- "serde_yaml",
- "thiserror 1.0.69",
- "unicode-normalization",
- "zip",
-]
-
-[[package]]
-name = "xz2"
-version = "0.1.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "388c44dc09d76f1536602ead6d325eb532f5c122f17782bd57fb47baeeb767e2"
-dependencies = [
- "lzma-sys",
-]
-
-[[package]]
-name = "zerocopy"
-version = "0.8.48"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "eed437bf9d6692032087e337407a86f04cd8d6a16a37199ed57949d415bd68e9"
-dependencies = [
- "zerocopy-derive",
-]
-
-[[package]]
-name = "zerocopy-derive"
-version = "0.8.48"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "70e3cd084b1788766f53af483dd21f93881ff30d7320490ec3ef7526d203bad4"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "zeroize"
-version = "1.8.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b97154e67e32c85465826e8bcc1c59429aaaf107c1e4a9e53c8d8ccd5eff88d0"
-dependencies = [
- "zeroize_derive",
-]
-
-[[package]]
-name = "zeroize_derive"
-version = "1.4.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "85a5b4158499876c763cb03bc4e49185d3cccbabb15b33c627f7884f43db852e"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
-
-[[package]]
-name = "zip"
-version = "2.4.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "fabe6324e908f85a1c52063ce7aa26b68dcb7eb6dbc83a2d148403c9bc3eba50"
-dependencies = [
- "aes",
- "arbitrary",
- "bzip2",
- "constant_time_eq",
- "crc32fast",
- "crossbeam-utils",
- "deflate64",
- "displaydoc",
- "flate2",
- "getrandom",
- "hmac",
- "indexmap",
- "lzma-rs",
- "memchr",
- "pbkdf2",
- "sha1",
- "thiserror 2.0.18",
- "time",
- "xz2",
- "zeroize",
- "zopfli",
- "zstd",
-]
-
-[[package]]
-name = "zopfli"
-version = "0.8.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f05cd8797d63865425ff89b5c4a48804f35ba0ce8d125800027ad6017d2b5249"
-dependencies = [
- "bumpalo",
- "crc32fast",
- "log",
- "simd-adler32",
-]
-
-[[package]]
-name = "zstd"
-version = "0.13.3"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e91ee311a569c327171651566e07972200e76fcfe2242a4fa446149a3881c08a"
-dependencies = [
- "zstd-safe",
-]
-
-[[package]]
-name = "zstd-safe"
-version = "7.2.4"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8f49c4d5f0abb602a93fb8736af2a4f4dd9512e36f7f570d66e65ff867ed3b9d"
-dependencies = [
- "zstd-sys",
-]
-
-[[package]]
-name = "zstd-sys"
-version = "2.0.16+zstd.1.5.7"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "91e19ebc2adc8f83e43039e79776e3fda8ca919132d68a1fed6a5faca2683748"
-dependencies = [
- "cc",
- "pkg-config",
-]
diff --git a/parser/Cargo.toml b/parser/Cargo.toml
deleted file mode 100644
index ec0ef94..0000000
--- a/parser/Cargo.toml
+++ /dev/null
@@ -1,30 +0,0 @@
-[package]
-name = "xlsxread"
-version = "0.1.0"
-edition = "2021"
-description = "Rust CLI replacing SheetJS xlsx build scripts for thptqg2017/thptqg2016"
-
-[dependencies]
-calamine       = "0.26"
-rusqlite       = { version = "0.32", features = ["bundled"] }
-clap           = { version = "4", features = ["derive"] }
-serde          = { version = "1", features = ["derive"] }
-serde_yaml     = "0.9"   # deprecated upstream but stable; this crate is deleted at cutover
-regex          = "1"
-unicode-normalization = "0.1"
-thiserror      = "1"
-anyhow         = "1"
-glob           = "0.3"
-
-# zip is already a transitive dep of calamine; pin explicitly so tests can use it
-[dev-dependencies]
-zip = "2"
-rusqlite = { version = "0.32", features = ["bundled"] }
-
-[[bin]]
-name = "xlsxread"
-path = "src/main.rs"
-
-[[test]]
-name = "golden"
-path = "tests/golden.rs"
diff --git a/parser/examples/dump_cells.rs b/parser/examples/dump_cells.rs
deleted file mode 100644
index b6d15ca..0000000
--- a/parser/examples/dump_cells.rs
+++ /dev/null
@@ -1,106 +0,0 @@
-//! Ground-truth cell dumper for the Go reader fidelity gate.
-//!
-//! Emits a canonical, reader-agnostic text rendering of every sheet and every
-//! cell of a spreadsheet exactly as calamine sees it. The Go reader must
-//! reproduce this byte-for-byte; the two dumps are compared by hash.
-//!
-//! Deliberately dumps the RAW used range: every sheet, every row, header rows
-//! included, no config applied. This gate is about cell fidelity, not build
-//! semantics — sheet selection and header skipping are exercised later.
-//!
-//! Usage: cargo run --release --example dump_cells -- <spreadsheet> [out-file]
-//! With no out-file the dump goes to stdout.
-
-use std::env;
-use std::fs::File;
-use std::io::{self, BufWriter, Write};
-
-use calamine::{open_workbook_auto, Data, Reader, Sheets};
-
-/// Escapes the field separators so a cell value can never break the line format.
-fn escape(s: &str) -> String {
-    let mut out = String::with_capacity(s.len());
-    for ch in s.chars() {
-        match ch {
-            '\\' => out.push_str("\\\\"),
-            '\t' => out.push_str("\\t"),
-            '\n' => out.push_str("\\n"),
-            '\r' => out.push_str("\\r"),
-            _ => out.push(ch),
-        }
-    }
-    out
-}
-
-/// Discriminates the calamine variant so a Go port can be checked against the
-/// actual type, not just the rendered string.
-fn kind(d: &Data) -> &'static str {
-    match d {
-        Data::Empty => "empty",
-        Data::String(_) => "str",
-        Data::Float(_) => "float",
-        Data::Int(_) => "int",
-        Data::Bool(_) => "bool",
-        Data::Error(_) => "err",
-        Data::DateTime(_) => "datetime",
-        Data::DateTimeIso(_) => "datetimeiso",
-        Data::DurationIso(_) => "durationiso",
-    }
-}
-
-fn main() -> Result<(), Box<dyn std::error::Error>> {
-    let args: Vec<String> = env::args().collect();
-    if args.len() < 2 {
-        eprintln!("usage: dump_cells <spreadsheet> [out-file]");
-        std::process::exit(2);
-    }
-    let path = &args[1];
-
-    let mut out: Box<dyn Write> = match args.get(2) {
-        Some(p) => Box::new(BufWriter::new(File::create(p)?)),
-        None => Box::new(BufWriter::new(io::stdout())),
-    };
-
-    let mut workbook: Sheets<_> = open_workbook_auto(path)?;
-    let sheet_names: Vec<String> = workbook.sheet_names().to_vec();
-
-    writeln!(out, "FILE\t{}", escape(path))?;
-    writeln!(out, "SHEETCOUNT\t{}", sheet_names.len())?;
-
-    for (idx, name) in sheet_names.iter().enumerate() {
-        let range = workbook.worksheet_range(name)?;
-        // start() is the used-range origin — the key question for any Go reader,
-        // which may index absolutely from A1 instead.
-        let (srow, scol) = range.start().unwrap_or((0, 0));
-        writeln!(
-            out,
-            "SHEET\t{}\t{}\t{}\t{}\t{}\t{}",
-            idx,
-            escape(name),
-            range.height(),
-            range.width(),
-            srow,
-            scol
-        )?;
-
-        for (r, row) in range.rows().enumerate() {
-            writeln!(out, "ROW\t{}\t{}\t{}", idx, r, row.len())?;
-            for (c, cell) in row.iter().enumerate() {
-                let s = cell.to_string();
-                writeln!(
-                    out,
-                    "CELL\t{}\t{}\t{}\t{}\t{}\t{}",
-                    idx,
-                    r,
-                    c,
-                    kind(cell),
-                    if matches!(cell, Data::Empty) { 1 } else { 0 },
-                    escape(&s)
-                )?;
-            }
-        }
-    }
-
-    out.flush()?;
-    Ok(())
-}
diff --git a/parser/examples/scan_kinds.rs b/parser/examples/scan_kinds.rs
deleted file mode 100644
index fc06a05..0000000
--- a/parser/examples/scan_kinds.rs
+++ /dev/null
@@ -1,76 +0,0 @@
-//! Corpus-wide cell-kind and sheet-geometry scan, for the Go reader fidelity gate.
-//!
-//! For every file given, prints one TSV line per sheet and one summary line per
-//! file. Aggregates only — never materialises the cell text — so the whole
-//! 418 MB corpus can be scanned quickly.
-//!
-//! The point is to find out which calamine `Data` variants actually occur in
-//! real inputs. `DateTime` and `Float` are the variants whose rendering differs
-//! between readers; if they never appear, the divergence risk is theoretical.
-//!
-//! Usage: cargo run --release --example scan_kinds -- <file>...
-
-use std::env;
-
-use calamine::{open_workbook_auto, Data, Reader, Sheets};
-
-fn main() {
-    let args: Vec<String> = env::args().skip(1).collect();
-    if args.is_empty() {
-        eprintln!("usage: scan_kinds <file>...");
-        std::process::exit(2);
-    }
-
-    println!("#TYPE\tpath\tsheet_idx\tname\theight\twidth\tstart_row\tstart_col\tempty\tstr\tfloat\tint\tbool\tdatetime\tdtiso\tduriso\terr");
-
-    for path in &args {
-        let mut wb: Sheets<_> = match open_workbook_auto(path) {
-            Ok(w) => w,
-            Err(e) => {
-                println!("ERR\t{path}\t{e}");
-                continue;
-            }
-        };
-
-        let names: Vec<String> = wb.sheet_names().to_vec();
-        for (idx, name) in names.iter().enumerate() {
-            let range = match wb.worksheet_range(name) {
-                Ok(r) => r,
-                Err(e) => {
-                    println!("SHEETERR\t{path}\t{idx}\t{e}");
-                    continue;
-                }
-            };
-            let (sr, sc) = range.start().unwrap_or((0, 0));
-            let mut k = [0usize; 9]; // empty,str,float,int,bool,datetime,dtiso,duriso,err
-            for row in range.rows() {
-                for cell in row {
-                    let i = match cell {
-                        Data::Empty => 0,
-                        Data::String(_) => 1,
-                        Data::Float(_) => 2,
-                        Data::Int(_) => 3,
-                        Data::Bool(_) => 4,
-                        Data::DateTime(_) => 5,
-                        Data::DateTimeIso(_) => 6,
-                        Data::DurationIso(_) => 7,
-                        Data::Error(_) => 8,
-                    };
-                    k[i] += 1;
-                }
-            }
-            println!(
-                "SHEET\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}\t{}",
-                path,
-                idx,
-                name.replace('\t', " "),
-                range.height(),
-                range.width(),
-                sr,
-                sc,
-                k[0], k[1], k[2], k[3], k[4], k[5], k[6], k[7], k[8]
-            );
-        }
-        println!("FILE\t{}\t{}", path, names.len());
-    }
-}
diff --git a/parser/scripts/build-db.js b/parser/scripts/build-db.js
deleted file mode 100644
index 5bdd991..0000000
--- a/parser/scripts/build-db.js
+++ /dev/null
@@ -1,66 +0,0 @@
-#!/usr/bin/env node
-/**
- * Build the SQLite database for one or all datasets, then gzip it.
- *
- * Replaces the six per-dataset npm scripts the two old projects carried. The
- * dataset list comes from src/datasets.js so it is written in exactly one place.
- *
- * Output goes to .build/public/db/ — the directory Vite copies as its publicDir.
- * Only the .gz survives: shipping a 100+ MB uncompressed database is made
- * structurally impossible rather than left to a cleanup step.
- *
- * Usage:
- *   node parser/scripts/build-db.js            # all four datasets
- *   node parser/scripts/build-db.js 2017-old   # just one
- */
-
-import { execFileSync } from "node:child_process";
-import { mkdirSync, rmSync, existsSync } from "node:fs";
-import { dirname, resolve } from "node:path";
-import { fileURLToPath } from "node:url";
-
-import { DATASET_IDS } from "../../src/datasets.js";
-
-const ROOT = resolve(dirname(fileURLToPath(import.meta.url)), "../..");
-const BIN = resolve(ROOT, "parser/target/release/xlsxread");
-const OUT_DIR = resolve(ROOT, ".build/public/db");
-
-const requested = process.argv.slice(2);
-const unknown = requested.filter((id) => !DATASET_IDS.includes(id));
-if (unknown.length) {
-  console.error(`unknown dataset(s): ${unknown.join(", ")}`);
-  console.error(`known: ${DATASET_IDS.join(", ")}`);
-  process.exit(2);
-}
-const targets = requested.length ? requested : DATASET_IDS;
-
-if (!existsSync(BIN)) {
-  console.error(`parser binary not found at ${BIN}`);
-  console.error("run: npm run build:rust");
-  process.exit(1);
-}
-
-mkdirSync(OUT_DIR, { recursive: true });
-
-for (const id of targets) {
-  const db = resolve(OUT_DIR, `${id}.db`);
-
-  execFileSync(
-    BIN,
-    [
-      "build",
-      "--schema",
-      resolve(ROOT, `parser/configs/${id}.yml`),
-      "--input",
-      resolve(ROOT, `data/${id}`),
-      "--output",
-      db,
-    ],
-    { stdio: "inherit" },
-  );
-
-  // -9 without -k: the raw .db must not reach the published artifact.
-  rmSync(`${db}.gz`, { force: true });
-  execFileSync("gzip", ["-9", db], { stdio: "inherit" });
-  console.log(`  → db/${id}.db.gz\n`);
-}
diff --git a/parser/scripts/check-duplicates.js b/parser/scripts/check-duplicates.js
deleted file mode 100644
index c0b4a53..0000000
--- a/parser/scripts/check-duplicates.js
+++ /dev/null
@@ -1,30 +0,0 @@
-// One-off audit: detect content-identical Excel files via md5.
-//
-// BROKEN as committed — `dirs` below is a hardcoded Windows path from the
-// original author's machine. Predates the repo unification; left unchanged
-// rather than half-fixed. Point `dirs` at data/<dataset> to use it.
-import crypto from "crypto";
-import fs from "fs";
-import path from "path";
-
-const dirs = ["D:/tiennm99/thptqg2017/data"];
-
-const byHash = {};
-for (const d of dirs) {
-  for (const f of fs.readdirSync(d)) {
-    const full = path.join(d, f);
-    if (!fs.statSync(full).isFile() || !f.endsWith(".xlsx")) continue;
-    const h = crypto.createHash("md5").update(fs.readFileSync(full)).digest("hex");
-    (byHash[h] ||= []).push(full);
-  }
-}
-
-const total = Object.values(byHash).reduce((s, a) => s + a.length, 0);
-const dupes = Object.entries(byHash).filter(([, a]) => a.length > 1);
-console.log(`Total files: ${total}`);
-console.log(`Unique by md5: ${Object.keys(byHash).length}`);
-console.log(`Duplicate groups: ${dupes.length}`);
-for (const [h, a] of dupes) {
-  console.log(`  ${h.slice(0, 12)}:`);
-  a.forEach((p) => console.log(`    ${p}`));
-}
diff --git a/parser/scripts/db-stats.js b/parser/scripts/db-stats.js
deleted file mode 100644
index 8582821..0000000
--- a/parser/scripts/db-stats.js
+++ /dev/null
@@ -1,79 +0,0 @@
-#!/usr/bin/env node
-/**
- * Dump per-dataset statistics from built SQLite databases as JSON.
- *
- * Used twice: once against the pre-refactor databases to capture a baseline,
- * and again after the schema unification. Comparing the two outputs is what
- * proves no score data was lost or invented.
- *
- * Schema-agnostic on purpose — columns come from PRAGMA table_info, so the same
- * script runs against the old 18/20-column tables and the new 22-column one.
- *
- * Usage:
- *   node db-stats.js <label>=<db-path> [<label>=<db-path> ...] > stats.json
- */
-
-import { DatabaseSync } from "node:sqlite";
-import { statSync } from "node:fs";
-
-// Deterministic value-level sample: every SBD ending in these digits.
-// Re-running against a rebuilt DB compares the same students.
-const SAMPLE_SUFFIX = "0000";
-
-function collect(dbPath) {
-  const db = new DatabaseSync(dbPath, { readOnly: true });
-
-  const columns = db
-    .prepare("PRAGMA table_info(student)")
-    .all()
-    .map((c) => c.name);
-
-  const rowCount = db.prepare("SELECT COUNT(*) AS c FROM student").get().c;
-
-  // One pass over the table counting non-NULLs for every column at once.
-  const sums = columns
-    .map((c) => `SUM(CASE WHEN "${c}" IS NOT NULL THEN 1 ELSE 0 END) AS "${c}"`)
-    .join(", ");
-  const nonNull = db.prepare(`SELECT ${sums} FROM student`).get();
-
-  const sample = db
-    .prepare(
-      `SELECT * FROM student WHERE so_bao_danh LIKE '%${SAMPLE_SUFFIX}'
-       ORDER BY so_bao_danh`,
-    )
-    .all();
-
-  db.close();
-
-  return {
-    rowCount,
-    columns,
-    nonNull: Object.fromEntries(columns.map((c) => [c, Number(nonNull[c])])),
-    sizeBytes: statSync(dbPath).size,
-    sampleCount: sample.length,
-    // Store the sample keyed by SBD so a later diff can report which student
-    // and which field changed, not just that something did.
-    sample: Object.fromEntries(sample.map((r) => [r.so_bao_danh, r])),
-  };
-}
-
-const args = process.argv.slice(2);
-if (args.length === 0) {
-  console.error("usage: db-stats.js <label>=<db-path> [...]");
-  process.exit(2);
-}
-
-const out = {};
-for (const arg of args) {
-  const idx = arg.indexOf("=");
-  if (idx === -1) {
-    console.error(`bad argument (expected label=path): ${arg}`);
-    process.exit(2);
-  }
-  const label = arg.slice(0, idx);
-  const path = arg.slice(idx + 1);
-  process.stderr.write(`collecting ${label} from ${path}\n`);
-  out[label] = collect(path);
-}
-
-process.stdout.write(JSON.stringify(out, null, 2) + "\n");
diff --git a/parser/scripts/diff-datasets.js b/parser/scripts/diff-datasets.js
deleted file mode 100644
index affcae8..0000000
--- a/parser/scripts/diff-datasets.js
+++ /dev/null
@@ -1,120 +0,0 @@
-// Compare two builds of the 2017 database, column by column.
-//
-// BROKEN as committed, for two reasons that both predate the repo unification:
-//   - `better-sqlite3` is imported but declared in no package.json
-//   - the public/ and backup/ paths it reads no longer exist
-//
-// Left unchanged rather than half-fixed. For a working comparison of two
-// builds use db-stats.js + verify-parity.js, which need no dependencies.
-import Database from "better-sqlite3";
-
-const NEW = new Database("public/thptqg2017.db", { readonly: true });
-const OLD = new Database("backup/thptqg2017.old.db", { readonly: true });
-
-const SCORE_COLS = [
-  "toan", "ngu_van", "vat_ly", "hoa_hoc", "sinh_hoc", "khtn",
-  "lich_su", "dia_ly", "gdcd", "khxh",
-  "tieng_anh", "tieng_phap", "tieng_nga", "tieng_trung",
-];
-
-// Old schema may lack some newer columns — detect and project a subset
-const oldCols = new Set(
-  OLD.prepare("PRAGMA table_info(student)").all().map((c) => c.name),
-);
-const commonCols = SCORE_COLS.filter((c) => oldCols.has(c));
-
-const newCount = NEW.prepare("SELECT COUNT(*) c FROM student").get().c;
-const oldCount = OLD.prepare("SELECT COUNT(*) c FROM student").get().c;
-console.log(`\n=== Row counts ===`);
-console.log(`  new: ${newCount}`);
-console.log(`  old: ${oldCount}`);
-console.log(`  diff: ${newCount - oldCount >= 0 ? "+" : ""}${newCount - oldCount}`);
-
-// Build SBD sets (memory: 861k strings ~50 MB, fine)
-const newSbd = new Set(
-  NEW.prepare("SELECT so_bao_danh FROM student").all().map((r) => r.so_bao_danh),
-);
-const oldSbd = new Set(
-  OLD.prepare("SELECT so_bao_danh FROM student").all().map((r) => r.so_bao_danh),
-);
-
-let newOnly = 0, oldOnly = 0, common = 0;
-for (const s of newSbd) (oldSbd.has(s) ? common++ : newOnly++);
-for (const s of oldSbd) if (!newSbd.has(s)) oldOnly++;
-
-console.log(`\n=== SBD membership ===`);
-console.log(`  new-only (added):     ${newOnly}`);
-console.log(`  old-only (removed):   ${oldOnly}`);
-console.log(`  common:               ${common}`);
-
-// Sample 5 new-only SBDs to peek
-const sampleNewOnly = [];
-for (const s of newSbd) {
-  if (!oldSbd.has(s)) sampleNewOnly.push(s);
-  if (sampleNewOnly.length >= 5) break;
-}
-if (sampleNewOnly.length) {
-  const rows = NEW.prepare(
-    `SELECT so_bao_danh, ho_ten FROM student WHERE so_bao_danh IN (${sampleNewOnly.map(() => "?").join(",")})`,
-  ).all(...sampleNewOnly);
-  console.log(`  sample new-only rows:`);
-  rows.forEach((r) => console.log(`    ${r.so_bao_danh}  ${r.ho_ten}`));
-}
-
-// Score comparison on common SBDs
-console.log(`\n=== Score comparison (common SBDs) ===`);
-console.log(`  columns compared: ${commonCols.join(", ")}`);
-
-const newStmt = NEW.prepare(
-  `SELECT so_bao_danh, ${commonCols.join(",")} FROM student`,
-);
-const oldStmt = OLD.prepare(
-  `SELECT so_bao_danh, ${commonCols.join(",")} FROM student`,
-);
-
-// Build old map
-const oldMap = new Map();
-for (const r of oldStmt.iterate()) oldMap.set(r.so_bao_danh, r);
-
-let identical = 0,
-  differ = 0;
-const colChanges = Object.fromEntries(commonCols.map((c) => [c, { changed: 0, newNull: 0, oldNull: 0 }]));
-const sampleDiffs = [];
-
-for (const nr of newStmt.iterate()) {
-  const or = oldMap.get(nr.so_bao_danh);
-  if (!or) continue;
-  let rowDiffers = false;
-  const rowChanges = [];
-  for (const c of commonCols) {
-    const a = nr[c], b = or[c];
-    if (a === b) continue;
-    // treat very-close floats as equal (xlsx may emit 5.5 vs 5.50)
-    if (a !== null && b !== null && Math.abs(a - b) < 1e-9) continue;
-    rowDiffers = true;
-    colChanges[c].changed++;
-    if (a === null) colChanges[c].newNull++;
-    if (b === null) colChanges[c].oldNull++;
-    rowChanges.push(`${c}: ${b} → ${a}`);
-  }
-  if (rowDiffers) {
-    differ++;
-    if (sampleDiffs.length < 5) sampleDiffs.push({ sbd: nr.so_bao_danh, changes: rowChanges });
-  } else {
-    identical++;
-  }
-}
-
-console.log(`  identical:  ${identical}`);
-console.log(`  differing:  ${differ}`);
-console.log(`  per-column changes:`);
-for (const [c, info] of Object.entries(colChanges)) {
-  if (info.changed === 0) continue;
-  console.log(`    ${c.padEnd(14)} changed=${info.changed}  (new-null=${info.newNull}, old-null=${info.oldNull})`);
-}
-if (sampleDiffs.length) {
-  console.log(`  sample diffs:`);
-  sampleDiffs.forEach((d) => {
-    console.log(`    SBD ${d.sbd}: ${d.changes.slice(0, 4).join("; ")}${d.changes.length > 4 ? " ..." : ""}`);
-  });
-}
diff --git a/parser/scripts/verify-parity.js b/parser/scripts/verify-parity.js
deleted file mode 100644
index b2b6455..0000000
--- a/parser/scripts/verify-parity.js
+++ /dev/null
@@ -1,157 +0,0 @@
-#!/usr/bin/env node
-/**
- * Compare rebuilt databases against the pre-refactor parity baseline.
- *
- * The schema deliberately changed shape, so a whole-file hash is meaningless.
- * What must hold instead:
- *
- *   1. Row count per dataset is unchanged.
- *   2. Every column that existed before has the same non-NULL count.
- *   3. Every column newly added to a dataset has a non-NULL count of exactly 0.
- *      This is the check that catches the union-regex risk — if the 16-pattern
- *      map starts matching text the narrower per-year map ignored, it shows up
- *      here rather than silently corrupting the dataset.
- *   4. A deterministic sample of students is identical field by field.
- *
- * Exits non-zero on any mismatch.
- *
- * Usage:
- *   node verify-parity.js <baseline.json> <current.json>
- */
-
-import { readFileSync } from "node:fs";
-
-/**
- * Foreign-language scores the pre-refactor configs silently discarded.
- *
- * The 2016 config listed 12 subject regexes and the 2017 configs listed 14;
- * neither list was complete. Candidates could sit German, Japanese and Russian
- * in both exam years, so every one of these students previously ended up with
- * no foreign-language score at all.
- *
- * Unifying to the canonical 16 patterns recovers them. Verified real, not
- * spurious matches: across all four datasets every student holds either zero
- * or exactly one foreign language — never two — and each affected student had
- * all language columns NULL beforehand.
- *
- * These exact counts are approved. Any other newly-populated column, or any
- * drift in these numbers, still fails the gate.
- */
-const APPROVED_RECOVERY = {
-  "2016": { tieng_nga: 182 },
-  "2017": { tieng_duc: 93, tieng_nhat: 512 },
-  "2017-old": { tieng_duc: 85, tieng_nhat: 484 },
-  "2017-old2": { tieng_duc: 22, tieng_nhat: 313 },
-};
-
-const [baselinePath, currentPath] = process.argv.slice(2);
-if (!baselinePath || !currentPath) {
-  console.error("usage: verify-parity.js <baseline.json> <current.json>");
-  process.exit(2);
-}
-
-const baseline = JSON.parse(readFileSync(baselinePath, "utf8"));
-const current = JSON.parse(readFileSync(currentPath, "utf8"));
-
-const failures = [];
-const notes = [];
-
-for (const dataset of Object.keys(baseline)) {
-  const b = baseline[dataset];
-  const c = current[dataset];
-
-  if (!c) {
-    failures.push(`${dataset}: missing from current stats`);
-    continue;
-  }
-
-  // 1. Row count
-  if (b.rowCount !== c.rowCount) {
-    failures.push(
-      `${dataset}: row count ${b.rowCount} → ${c.rowCount} (${c.rowCount - b.rowCount >= 0 ? "+" : ""}${c.rowCount - b.rowCount})`,
-    );
-  }
-
-  // 2. Pre-existing columns keep their non-NULL counts
-  for (const col of b.columns) {
-    if (!(col in c.nonNull)) {
-      failures.push(`${dataset}.${col}: column disappeared from schema`);
-      continue;
-    }
-    if (b.nonNull[col] !== c.nonNull[col]) {
-      failures.push(
-        `${dataset}.${col}: non-NULL ${b.nonNull[col]} → ${c.nonNull[col]} (${c.nonNull[col] - b.nonNull[col] >= 0 ? "+" : ""}${c.nonNull[col] - b.nonNull[col]})`,
-      );
-    }
-  }
-
-  // 3. Newly added columns must be NULL — except the approved recoveries,
-  //    which must match their approved count exactly.
-  const approved = APPROVED_RECOVERY[dataset] ?? {};
-  const added = c.columns.filter((col) => !b.columns.includes(col));
-  const recovered = [];
-  for (const col of added) {
-    const expected = approved[col] ?? 0;
-    if (c.nonNull[col] !== expected) {
-      failures.push(
-        expected === 0
-          ? `${dataset}.${col}: new column has ${c.nonNull[col]} non-NULL values, expected 0`
-          : `${dataset}.${col}: recovered ${c.nonNull[col]} values, approved count is ${expected}`,
-      );
-    } else if (expected > 0) {
-      recovered.push(`${col}=${expected}`);
-    }
-  }
-  // An approved recovery that vanished means the union patterns regressed.
-  for (const [col, expected] of Object.entries(approved)) {
-    if (!added.includes(col)) {
-      failures.push(
-        `${dataset}.${col}: expected ${expected} recovered values but column is not new`,
-      );
-    }
-  }
-  if (added.length) {
-    const nulls = added.length - recovered.length;
-    notes.push(
-      `${dataset}: +${added.length} new columns (${nulls} all-NULL as expected` +
-        (recovered.length ? `, recovered ${recovered.join(", ")}` : "") +
-        ")",
-    );
-  }
-
-  // 4. Deterministic sample compared field by field
-  if (b.sampleCount !== c.sampleCount) {
-    failures.push(
-      `${dataset}: sample size ${b.sampleCount} → ${c.sampleCount}`,
-    );
-  }
-  for (const [sbd, bRow] of Object.entries(b.sample)) {
-    const cRow = c.sample[sbd];
-    if (!cRow) {
-      failures.push(`${dataset}: sampled student ${sbd} missing after rebuild`);
-      continue;
-    }
-    for (const [field, bVal] of Object.entries(bRow)) {
-      if (cRow[field] !== bVal) {
-        failures.push(
-          `${dataset}: student ${sbd} field ${field}: ${JSON.stringify(bVal)} → ${JSON.stringify(cRow[field])}`,
-        );
-      }
-    }
-  }
-
-  const sizeDelta = ((c.sizeBytes - b.sizeBytes) / b.sizeBytes) * 100;
-  notes.push(
-    `${dataset}: ${c.rowCount} rows, ${b.columns.length} → ${c.columns.length} cols, size ${sizeDelta >= 0 ? "+" : ""}${sizeDelta.toFixed(1)}%`,
-  );
-}
-
-for (const n of notes) console.log(`  ${n}`);
-
-if (failures.length) {
-  console.error(`\nPARITY FAILED — ${failures.length} mismatch(es):\n`);
-  for (const f of failures) console.error(`  ✗ ${f}`);
-  process.exit(1);
-}
-
-console.log("\nPARITY OK — row counts, per-column non-NULL counts and samples all match.");
diff --git a/parser/src/audit.rs b/parser/src/audit.rs
deleted file mode 100644
index c0bd9b4..0000000
--- a/parser/src/audit.rs
+++ /dev/null
@@ -1,182 +0,0 @@
-/// Audit subcommand: replicates audit-row-counts.js exactly.
-///
-/// Reads all .xlsx files from the input directory (sheet 0 only, matching the
-/// JS script's behaviour at audit-row-counts.js:33), collects distinct SBDs
-/// into a HashSet, then queries `SELECT COUNT(*) FROM student` from the DB.
-/// Prints the same lines as audit-row-counts.js:54-62 and exits 0 on match,
-/// 1 on mismatch.
-use std::collections::HashSet;
-use std::path::Path;
-
-use calamine::{open_workbook_auto, Data, Reader};
-
-use crate::config::DatasetConfig;
-use crate::error::BuildError;
-use crate::reader::is_header_row;
-
-// ---------------------------------------------------------------------------
-// Audit result
-// ---------------------------------------------------------------------------
-
-pub struct AuditResult {
-    pub total_data_rows: u64,
-    pub both_empty: u64,
-    pub empty_name: u64,
-    pub empty_sbd: u64,
-    pub distinct_sbds: usize,
-    pub db_count: i64,
-    pub matched: bool,
-}
-
-// ---------------------------------------------------------------------------
-// Main audit logic
-// ---------------------------------------------------------------------------
-
-/// Collect distinct SBDs from all xlsx files in `input_dir`, query `db_path`,
-/// print the audit report and return the result.
-///
-/// The JS script reads only sheet 0 for every file (audit-row-counts.js:33).
-/// Unlike build-database.js, the audit script does NOT iterate all sheets.
-pub fn run_audit(
-    input_dir: &Path,
-    db_path: &Path,
-    cfg: &DatasetConfig,
-) -> Result<AuditResult, BuildError> {
-    // Collect .xlsx files (audit-row-counts.js only checks .xlsx — line 15)
-    let mut files: Vec<std::path::PathBuf> = std::fs::read_dir(input_dir)
-        .map_err(|e| BuildError::Io {
-            path: input_dir.display().to_string(),
-            source: e,
-        })?
-        .filter_map(|e| e.ok())
-        .map(|e| e.path())
-        .filter(|p| {
-            p.is_file()
-                && p.extension()
-                    .and_then(|e| e.to_str())
-                    .map(|e| e.eq_ignore_ascii_case("xlsx"))
-                    .unwrap_or(false)
-        })
-        .collect();
-    files.sort();
-
-    let mut all_sbd: HashSet<String> = HashSet::new();
-    let mut total_data_rows: u64 = 0;
-    let mut empty_name: u64 = 0;
-    let mut empty_sbd: u64 = 0;
-    let mut both_empty: u64 = 0;
-
-    for file in &files {
-        let path_str = file.display().to_string();
-        let mut workbook = open_workbook_auto(file).map_err(|e| BuildError::Calamine {
-            path: path_str.clone(),
-            source: e,
-        })?;
-
-        let sheet_names = workbook.sheet_names().to_vec();
-        if sheet_names.is_empty() {
-            continue;
-        }
-
-        // audit-row-counts.js reads only sheet 0 (line 33: wb.SheetNames[0])
-        let range =
-            workbook
-                .worksheet_range(&sheet_names[0])
-                .map_err(|e| BuildError::Calamine {
-                    path: path_str.clone(),
-                    source: e,
-                })?;
-
-        let mut first_row = true;
-        for raw in range.rows() {
-            let row: Vec<Data> = raw.to_vec();
-
-            // Skip header row on first row only
-            if first_row {
-                first_row = false;
-                if is_header_row(&row, &cfg.header) {
-                    continue;
-                }
-            }
-
-            total_data_rows += 1;
-
-            // For format_detection configs the audit uses positional defaults (col 0 = SBD,
-            // col 1 = HO_TEN) because the audit is best-effort and mirrors the JS script
-            // which also uses a fixed column assumption (audit-row-counts.js:33–36).
-            let (ho_ten_col, sbd_col) = cfg
-                .columns
-                .as_ref()
-                .map(|c| (c.ho_ten, c.so_bao_danh))
-                .unwrap_or((1, 0));
-            let ho_ten = row
-                .get(ho_ten_col)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-            let sbd = row
-                .get(sbd_col)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-
-            if ho_ten.is_empty() && sbd.is_empty() {
-                both_empty += 1;
-                continue;
-            }
-            if ho_ten.is_empty() {
-                empty_name += 1;
-            }
-            if sbd.is_empty() {
-                empty_sbd += 1;
-            }
-            if !sbd.is_empty() {
-                all_sbd.insert(sbd);
-            }
-        }
-    }
-
-    // Query DB count
-    let conn =
-        rusqlite::Connection::open_with_flags(db_path, rusqlite::OpenFlags::SQLITE_OPEN_READ_ONLY)?;
-    let db_count: i64 = conn.query_row("SELECT COUNT(*) FROM student", [], |row| row.get(0))?;
-
-    let distinct_sbds = all_sbd.len();
-    let matched = distinct_sbds as i64 == db_count;
-
-    Ok(AuditResult {
-        total_data_rows,
-        both_empty,
-        empty_name,
-        empty_sbd,
-        distinct_sbds,
-        db_count,
-        matched,
-    })
-}
-
-// ---------------------------------------------------------------------------
-// Print audit report — mirrors audit-row-counts.js:54-62 exactly
-// ---------------------------------------------------------------------------
-
-pub fn print_audit_report(r: &AuditResult) {
-    println!("=== Source vs DB ===");
-    println!(
-        "Source: total data rows across all files: {}",
-        r.total_data_rows
-    );
-    println!(
-        "Source: rows with empty name AND sbd (skipped): {}",
-        r.both_empty
-    );
-    println!("Source: rows with missing name only: {}", r.empty_name);
-    println!("Source: rows with missing sbd only: {}", r.empty_sbd);
-    println!("Source: distinct SBDs: {}", r.distinct_sbds);
-    println!("DB:     row count: {}", r.db_count);
-    println!(
-        "Match:  {}",
-        if r.matched {
-            "YES — all unique SBDs accounted for".to_string()
-        } else {
-            format!("NO — gap of {}", r.distinct_sbds as i64 - r.db_count)
-        }
-    );
-}
diff --git a/parser/src/cli.rs b/parser/src/cli.rs
deleted file mode 100644
index bce8285..0000000
--- a/parser/src/cli.rs
+++ /dev/null
@@ -1,48 +0,0 @@
-/// CLI argument structs via clap derive.
-use std::path::PathBuf;
-
-use clap::{Parser, Subcommand};
-
-#[derive(Parser)]
-#[command(
-    name = "xlsxread",
-    version,
-    about = "Read .xls/.xlsx files and build SQLite databases for thptqg datasets"
-)]
-pub struct Cli {
-    #[command(subcommand)]
-    pub cmd: Cmd,
-}
-
-#[derive(Subcommand)]
-pub enum Cmd {
-    /// Read input spreadsheets and write a SQLite database
-    Build {
-        /// Path to the dataset TOML config file
-        #[arg(long)]
-        schema: PathBuf,
-
-        /// Directory containing the .xls / .xlsx source files
-        #[arg(long)]
-        input: PathBuf,
-
-        /// Output SQLite database path
-        #[arg(long)]
-        output: PathBuf,
-    },
-
-    /// Audit: compare distinct SBD count from xlsx files vs DB row count
-    Audit {
-        /// Path to the dataset TOML config file
-        #[arg(long)]
-        schema: PathBuf,
-
-        /// Directory containing the .xlsx source files
-        #[arg(long)]
-        input: PathBuf,
-
-        /// SQLite database to compare against
-        #[arg(long)]
-        db: PathBuf,
-    },
-}
diff --git a/parser/src/config.rs b/parser/src/config.rs
deleted file mode 100644
index bafc5ef..0000000
--- a/parser/src/config.rs
+++ /dev/null
@@ -1,169 +0,0 @@
-use std::fs;
-use std::path::Path;
-
-use serde::Deserialize;
-
-use crate::error::BuildError;
-
-// ---------------------------------------------------------------------------
-// Top-level dataset configuration loaded from a .yml file
-// ---------------------------------------------------------------------------
-
-/// Per-dataset parse rules.
-///
-/// Deliberately carries no SQL. The table shape, the INSERT and the subject
-/// regexes are identical for every dataset and live in `crate::schema` — keeping
-/// them here meant four copies of the same DDL, which is how the 2016 and 2017
-/// schemas drifted apart.
-#[derive(Debug, Deserialize, Clone)]
-#[serde(deny_unknown_fields)]
-pub struct DatasetConfig {
-    pub reader: ReaderCfg,
-    /// Fixed column indices. Optional when format_detection handles per-file mapping.
-    #[serde(default)]
-    pub columns: Option<ColumnMap>,
-    pub validation: ValidationCfg,
-    pub header: HeaderCfg,
-    /// When set to "thptqg2016", enables per-file format auto-detection.
-    /// Each file's header row is inspected at runtime to choose the right
-    /// column layout (separate-scores / mapped / default-positional).
-    #[serde(default)]
-    pub format_detection: Option<String>,
-}
-
-#[derive(Debug, Deserialize, Clone)]
-pub struct ReaderCfg {
-    /// "all" → iterate every sheet (handles HCM/HN overflow); "first" → sheet 0 only
-    pub sheet_mode: SheetMode,
-    /// If true, skip rows where every cell is empty/null before counting (data-old2 quirk)
-    pub strip_blank_rows: bool,
-}
-
-#[derive(Debug, Deserialize, Clone, PartialEq, Eq)]
-#[serde(rename_all = "lowercase")]
-pub enum SheetMode {
-    All,
-    First,
-}
-
-/// Zero-indexed column positions in the source spreadsheet row.
-/// Used by thptqg2017 configs. thptqg2016 uses runtime format detection instead.
-#[derive(Debug, Deserialize, Clone)]
-pub struct ColumnMap {
-    pub ho_ten: usize,
-    pub ngay_sinh: usize,
-    pub so_bao_danh: usize,
-    pub diem_thi: usize,
-}
-
-#[derive(Debug, Deserialize, Clone)]
-pub struct ValidationCfg {
-    /// build-database-old.js / -old2.js require soBaoDanh to match ^\d+$
-    pub require_numeric_sbd: bool,
-    pub require_nonempty_name: bool,
-    pub require_nonempty_sbd: bool,
-}
-
-#[derive(Debug, Deserialize, Clone)]
-pub struct HeaderCfg {
-    /// Tokens to match against row[0].to_uppercase() to detect a header row
-    pub tokens: Vec<String>,
-}
-
-// ---------------------------------------------------------------------------
-// Loader
-// ---------------------------------------------------------------------------
-
-pub fn load_config(path: &Path) -> Result<DatasetConfig, BuildError> {
-    let text = fs::read_to_string(path).map_err(|e| BuildError::Io {
-        path: path.display().to_string(),
-        source: e,
-    })?;
-    let cfg: DatasetConfig = serde_yaml::from_str(&text)?;
-    Ok(cfg)
-}
-
-// ---------------------------------------------------------------------------
-// Unit tests
-// ---------------------------------------------------------------------------
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    const SAMPLE_YAML: &str = r#"
-reader:
-  sheet_mode: all
-  strip_blank_rows: false
-
-columns:
-  ho_ten: 0
-  ngay_sinh: 1
-  so_bao_danh: 2
-  diem_thi: 3
-
-validation:
-  require_numeric_sbd: false
-  require_nonempty_name: true
-  require_nonempty_sbd: true
-
-header:
-  tokens: ["HO_TEN", "HỌ TÊN", "STT"]
-"#;
-
-    #[test]
-    fn config_round_trip() {
-        let cfg: DatasetConfig = serde_yaml::from_str(SAMPLE_YAML).expect("parse failed");
-        assert_eq!(cfg.reader.sheet_mode, SheetMode::All);
-        assert!(!cfg.reader.strip_blank_rows);
-        let cols = cfg.columns.as_ref().unwrap();
-        assert_eq!(cols.ho_ten, 0);
-        assert_eq!(cols.diem_thi, 3);
-        assert!(!cfg.validation.require_numeric_sbd);
-        assert!(cfg.validation.require_nonempty_name);
-        assert_eq!(cfg.header.tokens.len(), 3);
-        assert!(cfg.format_detection.is_none());
-    }
-
-    /// A config carrying leftover SQL sections must be rejected rather than
-    /// silently ignored — otherwise a stale [schema] block would look effective
-    /// while `crate::schema` was actually driving the build.
-    #[test]
-    fn config_rejects_leftover_sql_sections() {
-        let with_ddl = format!(
-            "{SAMPLE_YAML}\nschema:\n  ddl: \"CREATE TABLE student (so_bao_danh TEXT);\"\n"
-        );
-        assert!(serde_yaml::from_str::<DatasetConfig>(&with_ddl).is_err());
-    }
-
-    #[test]
-    fn config_first_sheet_mode() {
-        let yaml_str = SAMPLE_YAML.replace("sheet_mode: all", "sheet_mode: first");
-        let cfg: DatasetConfig = serde_yaml::from_str(&yaml_str).expect("parse failed");
-        assert_eq!(cfg.reader.sheet_mode, SheetMode::First);
-    }
-
-    #[test]
-    fn config_format_detection_field() {
-        // Configs without a `columns:` mapping and with format_detection:
-        // thptqg2016 parse correctly
-        let yaml_str = r#"
-format_detection: thptqg2016
-
-reader:
-  sheet_mode: all
-  strip_blank_rows: false
-
-validation:
-  require_numeric_sbd: false
-  require_nonempty_name: true
-  require_nonempty_sbd: true
-
-header:
-  tokens: ["SBD", "SOBAODANH", "STT"]
-"#;
-        let cfg: DatasetConfig = serde_yaml::from_str(yaml_str).expect("parse failed");
-        assert_eq!(cfg.format_detection.as_deref(), Some("thptqg2016"));
-        assert!(cfg.columns.is_none());
-    }
-}
diff --git a/parser/src/error.rs b/parser/src/error.rs
deleted file mode 100644
index e700ae2..0000000
--- a/parser/src/error.rs
+++ /dev/null
@@ -1,34 +0,0 @@
-use thiserror::Error;
-
-#[derive(Debug, Error)]
-pub enum BuildError {
-    #[error("I/O error for {path}: {source}")]
-    Io {
-        path: String,
-        #[source]
-        source: std::io::Error,
-    },
-
-    #[error("Calamine error for {path}: {source}")]
-    Calamine {
-        path: String,
-        #[source]
-        source: calamine::Error,
-    },
-
-    #[error("SQLite error: {0}")]
-    Sqlite(#[from] rusqlite::Error),
-
-    #[error("Config parse error: {0}")]
-    Config(#[from] serde_yaml::Error),
-
-    #[error("Regex compile error for pattern '{pattern}': {source}")]
-    Regex {
-        pattern: String,
-        #[source]
-        source: regex::Error,
-    },
-
-    #[error("Schema has no sheets in file: {0}")]
-    NoSheets(String),
-}
diff --git a/parser/src/format_detect_2016.rs b/parser/src/format_detect_2016.rs
deleted file mode 100644
index fd62f8c..0000000
--- a/parser/src/format_detect_2016.rs
+++ /dev/null
@@ -1,548 +0,0 @@
-/// Per-file format auto-detection for the thptqg2016 dataset.
-///
-/// Translates `detectFormat` from scripts/build-database.js (lines 63–87) and the
-/// three row-processing functions (lines 90–146) into Rust.
-///
-/// The JS source has three formats:
-///
-/// 1. `separate-scores`  — header row[0]=="SBD" && row[2]=="TOAN"
-///    Columns: SBD(0) HOTEN(1) TOAN(2) VAN(3) LY(4) HOA(5) SINH(6) SU(7) DIA(8)
-///    NGOAINGUTN(9) NGOAINGUTL(10) NGOAINGU-total(11)
-///    → maps col 11 → tieng_anh; no ngay_sinh / ten_cum_thi / gioi_tinh / DIEM_THI
-///    → JS: build-database.js:90–116  (processSeparateScoresRow)
-///
-/// 2. `mapped`           — header row has SOBAODANH|SBD and DIEM_THI columns
-///    → dynamic column indices built from header names
-///    → JS: build-database.js:119–146 (processMappedRow with map from detectFormat)
-///
-/// 3. `default`          — no recognised header; positional 6-col layout
-///    SBD(0) HO_TEN(1) NGAY_SINH(2) TEN_CUMTHI(3) GIOI_TINH(4) DIEM_THI(5)
-///    → JS: build-database.js:149–151 DEFAULT_MAP + processMappedRow
-///
-/// The `build-database.js` citations throughout this file refer to the Node
-/// script this parser replaced, in the original standalone thptqg2016 repo.
-/// That file no longer exists here; the references are kept because they
-/// explain why several of the rules below look arbitrary.
-use calamine::Data;
-
-use crate::transform::{parse_scores, to_ascii, CompiledPatterns, ParsedRow};
-
-// ---------------------------------------------------------------------------
-// Known header tokens (mirrors JS KNOWN_HEADERS set, build-database.js:50–54)
-// ---------------------------------------------------------------------------
-
-/// Upper-cased strings that identify a header row's first cell.
-/// Mirrors `KNOWN_HEADERS` in build-database.js:50–54.
-const KNOWN_HEADERS: &[&str] = &[
-    "SOBAODANH",
-    "SBD",
-    "HO_TEN",
-    "HOTEN",
-    "HỌ TÊN",
-    "NGAY_SINH",
-    "TEN_CUMTHI",
-    "GIOI_TINH",
-    "DIEM_THI",
-    "STT",
-    "TOAN",
-    "VAN",
-    "LY",
-    "HOA",
-    "SINH ",
-    "SU",
-    "DIA",
-];
-
-/// Returns true when `row[0]` (uppercased, trimmed) is in the known-headers set.
-/// Mirrors `isHeaderRow` at build-database.js:56–60.
-pub fn is_header_row_2016(row: &[Data]) -> bool {
-    if row.len() < 2 {
-        return false;
-    }
-    let first = row[0].to_string().trim().to_uppercase();
-    KNOWN_HEADERS.contains(&first.as_str())
-}
-
-// ---------------------------------------------------------------------------
-// Detected per-file format
-// ---------------------------------------------------------------------------
-
-/// The three layouts the thptqg2016 dataset uses, detected per file.
-#[derive(Debug, Clone)]
-pub enum DetectedFormat {
-    /// SBD/HOTEN/TOAN/VAN/LY/HOA/SINH/SU/DIA/NGOAINGUTN/NGOAINGUTL/NGOAINGU columns.
-    /// Corresponds to the dhhanghai-style files. build-database.js:67–68.
-    SeparateScores,
-    /// Header present with SOBAODANH|SBD and DIEM_THI; dynamic column indices.
-    /// build-database.js:70–86.
-    Mapped {
-        sbd: usize,
-        ho_ten: usize,
-        ngay_sinh: Option<usize>,
-        ten_cum_thi: Option<usize>,
-        gioi_tinh: Option<usize>,
-        diem_thi: usize,
-    },
-    /// No recognised header; standard 6-column positional layout.
-    /// build-database.js:149–151 DEFAULT_MAP.
-    Default,
-}
-
-/// Inspect a header row and decide which format applies.
-/// Returns `None` when the header is present but unrecognised (treated as Default).
-///
-/// Mirrors `detectFormat` at build-database.js:63–87.
-pub fn detect_format(header_row: &[Data]) -> DetectedFormat {
-    let cols: Vec<String> = header_row
-        .iter()
-        .map(|c| c.to_string().trim().to_uppercase())
-        .collect();
-
-    // Format 1: SBD in col 0 AND TOAN in col 2 → separate-scores
-    // build-database.js:68: if (cols[0] === "SBD" && cols[2] === "TOAN")
-    if cols.first().map(|s| s.as_str()) == Some("SBD")
-        && cols.get(2).map(|s| s.as_str()) == Some("TOAN")
-    {
-        return DetectedFormat::SeparateScores;
-    }
-
-    // Format 2: build column index map — check for SOBAODANH|SBD and DIEM_THI
-    // build-database.js:70–86
-    let mut sbd_idx: Option<usize> = None;
-    let mut ho_ten_idx: Option<usize> = None;
-    let mut ngay_sinh_idx: Option<usize> = None;
-    let mut ten_cum_thi_idx: Option<usize> = None;
-    let mut gioi_tinh_idx: Option<usize> = None;
-    let mut diem_thi_idx: Option<usize> = None;
-
-    for (i, c) in cols.iter().enumerate() {
-        match c.as_str() {
-            "SOBAODANH" | "SBD" => sbd_idx = Some(i),
-            "HO_TEN" | "HOTEN" | "HỌ TÊN" => ho_ten_idx = Some(i),
-            "NGAY_SINH" => ngay_sinh_idx = Some(i),
-            "TEN_CUMTHI" => ten_cum_thi_idx = Some(i),
-            "GIOI_TINH" => gioi_tinh_idx = Some(i),
-            "DIEM_THI" => diem_thi_idx = Some(i),
-            _ => {}
-        }
-    }
-
-    // build-database.js:82–84: if (map.sbd !== undefined && map.diem_thi !== undefined)
-    if let (Some(sbd), Some(diem_thi)) = (sbd_idx, diem_thi_idx) {
-        let ho_ten = ho_ten_idx.unwrap_or(1); // fallback: col 1 (present in all known files)
-        return DetectedFormat::Mapped {
-            sbd,
-            ho_ten,
-            ngay_sinh: ngay_sinh_idx,
-            ten_cum_thi: ten_cum_thi_idx,
-            gioi_tinh: gioi_tinh_idx,
-            diem_thi,
-        };
-    }
-
-    // Unrecognised header (or no header) → positional default
-    DetectedFormat::Default
-}
-
-// ---------------------------------------------------------------------------
-// Row processors
-// ---------------------------------------------------------------------------
-
-/// Cell accessor helper.
-fn cell_str(row: &[Data], idx: usize) -> String {
-    row.get(idx)
-        .map(|c| c.to_string().trim().to_owned())
-        .unwrap_or_default()
-}
-
-/// Parse a cell that should hold a float score; returns None for blank/non-numeric.
-/// Mirrors `parseFloat(row[N]) || null` in JS.
-fn parse_float_cell(row: &[Data], idx: usize) -> Option<f64> {
-    let s = cell_str(row, idx);
-    if s.is_empty() {
-        return None;
-    }
-    s.parse::<f64>().ok().filter(|v| v.is_finite() && *v != 0.0)
-}
-
-/// Process a row in the `separate-scores` format.
-///
-/// Column layout (build-database.js:90–116 `processSeparateScoresRow`):
-///   0=SBD  1=HOTEN  2=TOAN  3=VAN  4=LY  5=HOA  6=SINH  7=SU  8=DIA
-///   9=NGOAINGUTN  10=NGOAINGUTL  11=NGOAINGU(total→tieng_anh)
-///
-/// tieng_phap / tieng_duc / tieng_nhat / tieng_trung all → None
-/// ngay_sinh / ten_cum_thi / gioi_tinh all → None (not in this format)
-pub fn process_separate_scores_row(row: &[Data], patterns: &CompiledPatterns) -> Option<ParsedRow> {
-    let sbd = cell_str(row, 0);
-    let ho_ten = cell_str(row, 1);
-    if sbd.is_empty() || ho_ten.is_empty() {
-        return None;
-    }
-
-    let ho_ten_ascii = to_ascii(&ho_ten);
-
-    // build-database.js:102–114: explicit per-column score mapping
-    let mut scores = std::collections::HashMap::new();
-    macro_rules! add_score {
-        ($field:expr, $idx:expr) => {
-            if let Some(v) = parse_float_cell(row, $idx) {
-                scores.insert($field.to_string(), v);
-            }
-        };
-    }
-    add_score!("toan", 2);
-    add_score!("ngu_van", 3);
-    add_score!("vat_ly", 4);
-    add_score!("hoa_hoc", 5);
-    add_score!("sinh_hoc", 6);
-    add_score!("lich_su", 7);
-    add_score!("dia_ly", 8);
-    // col 11 = NGOAINGU total → tieng_anh (build-database.js:110–111)
-    add_score!("tieng_anh", 11);
-
-    // Suppress unused-variable warning; patterns not used in this path (no DIEM_THI string)
-    let _ = patterns;
-
-    Some(ParsedRow {
-        so_bao_danh: sbd,
-        ho_ten,
-        ho_ten_ascii,
-        ngay_sinh: None,
-        ten_cum_thi: None,
-        gioi_tinh: None,
-        scores,
-    })
-}
-
-/// Process a row in the `mapped` format (header-derived column indices).
-///
-/// Mirrors `processMappedRow` at build-database.js:119–146.
-/// Gender is normalised: only "Nam" or "Nữ" are kept; everything else → None.
-/// (build-database.js:132: `(rawGioiTinh === "Nam" || rawGioiTinh === "Nữ") ? rawGioiTinh : null`)
-// Mirrors the JS column map one-for-one; grouping the indices into a struct
-// would obscure that correspondence for no benefit.
-#[allow(clippy::too_many_arguments)]
-pub fn process_mapped_row(
-    row: &[Data],
-    sbd_idx: usize,
-    ho_ten_idx: usize,
-    ngay_sinh_idx: Option<usize>,
-    ten_cum_thi_idx: Option<usize>,
-    gioi_tinh_idx: Option<usize>,
-    diem_thi_idx: usize,
-    patterns: &CompiledPatterns,
-) -> Option<ParsedRow> {
-    let sbd = cell_str(row, sbd_idx);
-    let ho_ten = cell_str(row, ho_ten_idx);
-    if sbd.is_empty() || ho_ten.is_empty() {
-        return None;
-    }
-
-    // Skip rows where SBD or HO_TEN are themselves header tokens (leaked header rows).
-    // build-database.js:125–126: KNOWN_HEADERS.has(sbdUpper) || KNOWN_HEADERS.has(hoTenUpper)
-    let sbd_upper = sbd.to_uppercase();
-    let ho_ten_upper = ho_ten.to_uppercase();
-    if KNOWN_HEADERS.contains(&sbd_upper.as_str())
-        || KNOWN_HEADERS.contains(&ho_ten_upper.as_str())
-    {
-        return None;
-    }
-
-    let ho_ten_ascii = to_ascii(&ho_ten);
-
-    let ngay_sinh = ngay_sinh_idx
-        .map(|i| cell_str(row, i))
-        .filter(|s| !s.is_empty());
-
-    let ten_cum_thi = ten_cum_thi_idx
-        .map(|i| cell_str(row, i))
-        .filter(|s| !s.is_empty());
-
-    // build-database.js:130–132: normalise gender
-    let gioi_tinh = gioi_tinh_idx
-        .map(|i| cell_str(row, i))
-        .and_then(|s| {
-            if s == "Nam" || s == "Nữ" {
-                Some(s)
-            } else {
-                None
-            }
-        });
-
-    let diem_thi = row
-        .get(diem_thi_idx)
-        .map(|c| c.to_string())
-        .unwrap_or_default();
-
-    let scores = parse_scores(&diem_thi, patterns);
-
-    Some(ParsedRow {
-        so_bao_danh: sbd,
-        ho_ten,
-        ho_ten_ascii,
-        ngay_sinh,
-        ten_cum_thi,
-        gioi_tinh,
-        scores,
-    })
-}
-
-/// Process a row using the default positional 6-column layout.
-///
-/// Column order: SBD(0) HO_TEN(1) NGAY_SINH(2) TEN_CUMTHI(3) GIOI_TINH(4) DIEM_THI(5)
-/// Mirrors `processMappedRow(row, DEFAULT_MAP)` at build-database.js:233–236.
-pub fn process_default_row(row: &[Data], patterns: &CompiledPatterns) -> Option<ParsedRow> {
-    process_mapped_row(
-        row,
-        0, // sbd
-        1, // ho_ten
-        Some(2),
-        Some(3),
-        Some(4),
-        5, // diem_thi
-        patterns,
-    )
-}
-
-/// Dispatch a data row through the correct processor for the detected format.
-///
-/// Returns `None` when the row is empty/invalid and should be skipped.
-pub fn process_row_2016(
-    row: &[Data],
-    fmt: &DetectedFormat,
-    patterns: &CompiledPatterns,
-) -> Option<ParsedRow> {
-    match fmt {
-        DetectedFormat::SeparateScores => process_separate_scores_row(row, patterns),
-        DetectedFormat::Mapped {
-            sbd,
-            ho_ten,
-            ngay_sinh,
-            ten_cum_thi,
-            gioi_tinh,
-            diem_thi,
-        } => process_mapped_row(
-            row,
-            *sbd,
-            *ho_ten,
-            *ngay_sinh,
-            *ten_cum_thi,
-            *gioi_tinh,
-            *diem_thi,
-            patterns,
-        ),
-        DetectedFormat::Default => process_default_row(row, patterns),
-    }
-}
-
-// ---------------------------------------------------------------------------
-// Unit tests — 3 detection branches + key processing cases
-// ---------------------------------------------------------------------------
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    fn s(v: &str) -> Data {
-        Data::String(v.to_string())
-    }
-
-    fn make_patterns() -> CompiledPatterns {
-        CompiledPatterns::new().unwrap()
-    }
-
-    // --- detect_format: branch 1 — separate-scores ---
-
-    #[test]
-    fn detect_separate_scores() {
-        let header = vec![s("SBD"), s("HOTEN"), s("TOAN"), s("VAN")];
-        match detect_format(&header) {
-            DetectedFormat::SeparateScores => {}
-            other => panic!("expected SeparateScores, got {other:?}"),
-        }
-    }
-
-    // --- detect_format: branch 2 — mapped ---
-
-    #[test]
-    fn detect_mapped_with_named_cols() {
-        let header = vec![
-            s("SOBAODANH"),
-            s("HO_TEN"),
-            s("NGAY_SINH"),
-            s("TEN_CUMTHI"),
-            s("GIOI_TINH"),
-            s("DIEM_THI"),
-        ];
-        match detect_format(&header) {
-            DetectedFormat::Mapped {
-                sbd,
-                ho_ten,
-                ngay_sinh,
-                ten_cum_thi,
-                gioi_tinh,
-                diem_thi,
-            } => {
-                assert_eq!(sbd, 0);
-                assert_eq!(ho_ten, 1);
-                assert_eq!(ngay_sinh, Some(2));
-                assert_eq!(ten_cum_thi, Some(3));
-                assert_eq!(gioi_tinh, Some(4));
-                assert_eq!(diem_thi, 5);
-            }
-            other => panic!("expected Mapped, got {other:?}"),
-        }
-    }
-
-    #[test]
-    fn detect_mapped_sbd_variant() {
-        // "SBD" (not "SOBAODANH") + DIEM_THI at different positions
-        let header = vec![s("STT"), s("SBD"), s("HOTEN"), s("DIEM_THI")];
-        match detect_format(&header) {
-            DetectedFormat::Mapped { sbd, diem_thi, .. } => {
-                assert_eq!(sbd, 1);
-                assert_eq!(diem_thi, 3);
-            }
-            other => panic!("expected Mapped, got {other:?}"),
-        }
-    }
-
-    // --- detect_format: branch 3 — default ---
-
-    #[test]
-    fn detect_default_when_no_header() {
-        // A data row: positional default used
-        let data_row = vec![
-            s("12345678"),
-            s("Nguyễn Văn A"),
-            s("01/01/2000"),
-            s("TP HCM"),
-            s("Nam"),
-            s("Toán: 8.5"),
-        ];
-        // Default is returned when there is no recognised header
-        match detect_format(&data_row) {
-            DetectedFormat::Default => {}
-            other => panic!("expected Default, got {other:?}"),
-        }
-    }
-
-    // --- process_separate_scores_row ---
-
-    #[test]
-    fn separate_scores_basic() {
-        let p = make_patterns();
-        // 12 columns: SBD HOTEN TOAN VAN LY HOA SINH SU DIA NGUTN NGUTL NGUTOTAL
-        let row = vec![
-            s("TP001"),
-            s("Nguyễn Thị Lan"),
-            Data::Float(8.0),
-            Data::Float(7.5),
-            Data::Float(9.0),
-            Data::Float(6.5),
-            Data::Float(5.0),
-            Data::Float(4.5),
-            Data::Float(8.0),
-            Data::Empty,
-            Data::Empty,
-            Data::Float(7.0), // col 11 → tieng_anh
-        ];
-        let row = process_separate_scores_row(&row, &p).expect("should parse");
-        assert_eq!(row.so_bao_danh, "TP001");
-        assert_eq!(row.ho_ten, "Nguyễn Thị Lan");
-        assert_eq!(row.ho_ten_ascii, "nguyen thi lan");
-        assert_eq!(row.scores.get("toan"), Some(&8.0));
-        assert_eq!(row.scores.get("tieng_anh"), Some(&7.0));
-        assert!(row.ngay_sinh.is_none());
-        assert!(row.ten_cum_thi.is_none());
-        assert!(row.gioi_tinh.is_none());
-    }
-
-    #[test]
-    fn separate_scores_skips_empty_sbd() {
-        let p = make_patterns();
-        let row = vec![s(""), s("Nguyễn Văn A"), Data::Float(5.0)];
-        assert!(process_separate_scores_row(&row, &p).is_none());
-    }
-
-    // --- process_mapped_row ---
-
-    #[test]
-    fn mapped_row_full_fields() {
-        let p = make_patterns();
-        let row = vec![
-            s("HCM001"),
-            s("Trần Thị Bích"),
-            s("15/3/1999"),
-            s("Cụm thi HCM"),
-            s("Nữ"),
-            s("Toán: 9.0 Ngữ văn: 8.5 Tiếng Anh: 7.75"),
-        ];
-        let parsed = process_mapped_row(&row, 0, 1, Some(2), Some(3), Some(4), 5, &p)
-            .expect("should parse");
-        assert_eq!(parsed.so_bao_danh, "HCM001");
-        assert_eq!(parsed.ngay_sinh.as_deref(), Some("15/3/1999"));
-        assert_eq!(parsed.ten_cum_thi.as_deref(), Some("Cụm thi HCM"));
-        assert_eq!(parsed.gioi_tinh.as_deref(), Some("Nữ"));
-        assert_eq!(parsed.scores.get("toan"), Some(&9.0));
-        assert_eq!(parsed.scores.get("ngu_van"), Some(&8.5));
-        assert_eq!(parsed.scores.get("tieng_anh"), Some(&7.75));
-    }
-
-    #[test]
-    fn mapped_row_gender_normalisation() {
-        let p = make_patterns();
-        // Gender "Unknown" → None
-        let row = vec![s("ABC"), s("Nguyen Van A"), s(""), s(""), s("Unknown"), s("")];
-        let parsed =
-            process_mapped_row(&row, 0, 1, Some(2), Some(3), Some(4), 5, &p).expect("should parse");
-        assert!(parsed.gioi_tinh.is_none());
-
-        // "Nam" passes through
-        let row2 = vec![s("ABC"), s("Nguyen Van A"), s(""), s(""), s("Nam"), s("")];
-        let parsed2 =
-            process_mapped_row(&row2, 0, 1, Some(2), Some(3), Some(4), 5, &p).expect("should parse");
-        assert_eq!(parsed2.gioi_tinh.as_deref(), Some("Nam"));
-    }
-
-    #[test]
-    fn mapped_row_skips_leaked_header() {
-        let p = make_patterns();
-        // A leaked header row — SBD cell contains "SOBAODANH"
-        let row = vec![s("SOBAODANH"), s("HO_TEN"), s("NGAY_SINH"), s(""), s(""), s("")];
-        assert!(process_mapped_row(&row, 0, 1, Some(2), Some(3), Some(4), 5, &p).is_none());
-    }
-
-    // --- process_default_row ---
-
-    #[test]
-    fn default_row_positional() {
-        let p = make_patterns();
-        let row = vec![
-            s("DN001"),
-            s("Lê Văn Long"),
-            s("10/5/1998"),
-            s("Cụm Đà Nẵng"),
-            s("Nam"),
-            s("Tiếng Đức: 6.25"),
-        ];
-        let parsed = process_default_row(&row, &p).expect("should parse");
-        assert_eq!(parsed.so_bao_danh, "DN001");
-        assert_eq!(parsed.ho_ten_ascii, "le van long");
-        assert_eq!(parsed.ngay_sinh.as_deref(), Some("10/5/1998"));
-        assert_eq!(parsed.ten_cum_thi.as_deref(), Some("Cụm Đà Nẵng"));
-        assert_eq!(parsed.gioi_tinh.as_deref(), Some("Nam"));
-        assert_eq!(parsed.scores.get("tieng_duc"), Some(&6.25));
-    }
-
-    // --- is_header_row_2016 ---
-
-    #[test]
-    fn header_row_detection_2016() {
-        assert!(is_header_row_2016(&[s("SOBAODANH"), s("HO_TEN")]));
-        assert!(is_header_row_2016(&[s("SBD"), s("HOTEN"), s("TOAN")]));
-        assert!(!is_header_row_2016(&[s("12345678"), s("Nguyen Van A")]));
-        assert!(!is_header_row_2016(&[s("SBD")])); // too short (< 2 cells)
-    }
-}
diff --git a/parser/src/lib.rs b/parser/src/lib.rs
deleted file mode 100644
index ce7f5e4..0000000
--- a/parser/src/lib.rs
+++ /dev/null
@@ -1,11 +0,0 @@
-/// Public library interface for integration tests.
-/// The binary entry point is src/main.rs; this file re-exports the internal
-/// modules so tests/golden.rs can call them without going through the CLI.
-pub mod audit;
-pub mod config;
-pub mod error;
-pub mod format_detect_2016;
-pub mod reader;
-pub mod schema;
-pub mod transform;
-pub mod writer;
diff --git a/parser/src/main.rs b/parser/src/main.rs
deleted file mode 100644
index fd8b342..0000000
--- a/parser/src/main.rs
+++ /dev/null
@@ -1,377 +0,0 @@
-/// xlsxread — Rust CLI replacing the SheetJS xlsx build scripts.
-///
-/// Subcommands:
-///   build  — read .xls/.xlsx files → write SQLite DB
-///   audit  — compare distinct SBD count from xlsx vs DB row count
-///
-/// Library modules are declared in lib.rs; main.rs only adds the CLI layer.
-///
-/// When config contains `format_detection = "thptqg2016"` the build subcommand
-/// uses per-file header inspection to pick the right column layout, replicating
-/// the `detectFormat` logic from scripts/build-database.js (lines 63–87).
-mod cli;
-
-use std::path::Path;
-
-use anyhow::{Context, Result};
-use calamine::Data;
-use clap::Parser;
-
-use cli::{Cli, Cmd};
-use xlsxread::audit;
-use xlsxread::config::load_config;
-use xlsxread::format_detect_2016::{
-    detect_format, is_header_row_2016, process_row_2016, DetectedFormat,
-};
-use xlsxread::reader::{is_all_blank, process_file};
-use xlsxread::transform::{validate_row, CompiledPatterns, SkipReason};
-use xlsxread::writer::{finish_db, insert_row, open_db};
-
-fn main() -> Result<()> {
-    let cli = Cli::parse();
-
-    match cli.cmd {
-        Cmd::Build {
-            schema,
-            input,
-            output,
-        } => {
-            run_build(&schema, &input, &output)?;
-        }
-        Cmd::Audit { schema, input, db } => {
-            let cfg = load_config(&schema)
-                .with_context(|| format!("Failed to load config: {}", schema.display()))?;
-            let result = audit::run_audit(&input, &db, &cfg).with_context(|| "Audit failed")?;
-            audit::print_audit_report(&result);
-            if !result.matched {
-                std::process::exit(1);
-            }
-        }
-    }
-
-    Ok(())
-}
-
-// ---------------------------------------------------------------------------
-// Build subcommand — dispatches to thptqg2016 or standard path
-// ---------------------------------------------------------------------------
-
-fn run_build(schema_path: &Path, input_dir: &Path, output_path: &Path) -> Result<()> {
-    let cfg = load_config(schema_path)
-        .with_context(|| format!("Failed to load config: {}", schema_path.display()))?;
-
-    if cfg.format_detection.as_deref() == Some("thptqg2016") {
-        run_build_2016(&cfg, input_dir, output_path)
-    } else {
-        run_build_standard(&cfg, input_dir, output_path)
-    }
-}
-
-// ---------------------------------------------------------------------------
-// Standard build path (thptqg2017 and similar fixed-column configs)
-// ---------------------------------------------------------------------------
-
-fn run_build_standard(
-    cfg: &xlsxread::config::DatasetConfig,
-    input_dir: &Path,
-    output_path: &Path,
-) -> Result<()> {
-    let patterns =
-        CompiledPatterns::new().with_context(|| "Failed to compile score regexes")?;
-
-    let mut files: Vec<std::path::PathBuf> = std::fs::read_dir(input_dir)
-        .with_context(|| format!("Cannot read input dir: {}", input_dir.display()))?
-        .filter_map(|e| e.ok())
-        .map(|e| e.path())
-        .filter(|p| {
-            p.is_file()
-                && p.extension()
-                    .and_then(|e| e.to_str())
-                    .map(|e| {
-                        let lower = e.to_lowercase();
-                        lower == "xls" || lower == "xlsx"
-                    })
-                    .unwrap_or(false)
-        })
-        .collect();
-    files.sort();
-
-    let dataset_label = input_dir
-        .file_name()
-        .and_then(|n| n.to_str())
-        .unwrap_or("data");
-
-    println!(
-        "[build] {dataset_label}/ → {}  ({} files)",
-        output_path.display(),
-        files.len()
-    );
-
-    let conn = open_db(output_path)
-        .with_context(|| format!("Failed to open DB: {}", output_path.display()))?;
-
-    let mut total_source_rows: u64 = 0;
-    let mut total_skipped: u64 = 0;
-    let mut total_errors: u64 = 0;
-
-    let is_old2 = dataset_label.contains("old2");
-    let strip_blank = cfg.reader.strip_blank_rows;
-
-    conn.execute_batch("BEGIN")?;
-
-    for file in &files {
-        let base = file
-            .file_name()
-            .and_then(|n| n.to_str())
-            .unwrap_or("?")
-            .to_owned();
-        let mut file_rows: u64 = 0;
-        let mut file_skipped: u64 = 0;
-        let mut file_errors: u64 = 0;
-
-        let process_result = process_file(file, cfg, |_sheet_idx, raw| {
-            let all_blank = is_all_blank(raw);
-            if strip_blank && all_blank {
-                return;
-            }
-
-            total_source_rows += 1;
-
-            let ho_ten = raw
-                .get(cfg.columns.as_ref().unwrap().ho_ten)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-            let so_bao_danh = raw
-                .get(cfg.columns.as_ref().unwrap().so_bao_danh)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-
-            match validate_row(&ho_ten, &so_bao_danh, &cfg.validation, strip_blank, all_blank) {
-                Err(SkipReason::BlankRow) => {}
-                Err(_) => {
-                    file_skipped += 1;
-                    return;
-                }
-                Ok(()) => {}
-            }
-
-            let parsed = xlsxread::transform::transform_row(raw, cfg, &patterns);
-
-            match insert_row(&conn, &parsed) {
-                Ok(()) => file_rows += 1,
-                Err(e) => {
-                    file_errors += 1;
-                    if total_errors + file_errors <= 5 {
-                        eprintln!("  [warn] {base}: {e}");
-                    }
-                }
-            }
-        });
-
-        match process_result {
-            Ok(_) => {}
-            Err(e) => {
-                eprintln!("  [error] {base}: {e}");
-                file_errors += 1;
-            }
-        }
-
-        total_skipped += file_skipped;
-        total_errors += file_errors;
-        println!("  {base}: {file_rows} rows");
-    }
-
-    conn.execute_batch("COMMIT")?;
-
-    finish_db(
-        &conn,
-        output_path,
-        total_source_rows,
-        total_skipped,
-        total_errors,
-        dataset_label,
-        files.len(),
-        is_old2,
-    )
-    .with_context(|| "Failed to finalise DB")?;
-
-    Ok(())
-}
-
-// ---------------------------------------------------------------------------
-// thptqg2016 build path — per-file format detection
-// ---------------------------------------------------------------------------
-
-/// Build the thptqg2016 database.
-///
-/// Each file is processed independently: the first row is inspected to determine
-/// which of the three column layouts applies (separate-scores / mapped / default).
-/// This mirrors `detectFormat` in scripts/build-database.js lines 63–87, called
-/// once per file inside the file loop at build-database.js:218–219.
-fn run_build_2016(
-    cfg: &xlsxread::config::DatasetConfig,
-    input_dir: &Path,
-    output_path: &Path,
-) -> Result<()> {
-    let patterns =
-        CompiledPatterns::new().with_context(|| "Failed to compile score regexes")?;
-
-    let mut files: Vec<std::path::PathBuf> = std::fs::read_dir(input_dir)
-        .with_context(|| format!("Cannot read input dir: {}", input_dir.display()))?
-        .filter_map(|e| e.ok())
-        .map(|e| e.path())
-        .filter(|p| {
-            p.is_file()
-                && p.extension()
-                    .and_then(|e| e.to_str())
-                    .map(|e| {
-                        let lower = e.to_lowercase();
-                        lower == "xls" || lower == "xlsx"
-                    })
-                    .unwrap_or(false)
-        })
-        .collect();
-    files.sort();
-
-    let dataset_label = input_dir
-        .file_name()
-        .and_then(|n| n.to_str())
-        .unwrap_or("data");
-
-    println!(
-        "[build:2016] {dataset_label}/ → {}  ({} files)",
-        output_path.display(),
-        files.len()
-    );
-
-    let conn = open_db(output_path)
-        .with_context(|| format!("Failed to open DB: {}", output_path.display()))?;
-
-    let mut total_source_rows: u64 = 0;
-    let total_skipped: u64 = 0;
-    let mut total_errors: u64 = 0;
-
-    conn.execute_batch("BEGIN")?;
-
-    for file in &files {
-        let base = file
-            .file_name()
-            .and_then(|n| n.to_str())
-            .unwrap_or("?")
-            .to_owned();
-
-        match process_file_2016(
-            file,
-            cfg,
-            &patterns,
-            &conn,
-            &base,
-            &mut total_source_rows,
-            &mut total_errors,
-        ) {
-            Ok(file_rows) => {
-                println!("  {base}: {file_rows} rows");
-            }
-            Err(e) => {
-                eprintln!("  [error] {base}: {e}");
-                total_errors += 1;
-            }
-        }
-    }
-
-    conn.execute_batch("COMMIT")?;
-
-    finish_db(
-        &conn,
-        output_path,
-        total_source_rows,
-        total_skipped,
-        total_errors,
-        dataset_label,
-        files.len(),
-        false,
-    )
-    .with_context(|| "Failed to finalise DB")?;
-
-    Ok(())
-}
-
-/// Process one file in the thptqg2016 format-detection path.
-///
-/// Reads the file, uses the first row to detect the column layout, then processes
-/// all subsequent data rows. Returns the count of successfully inserted rows.
-fn process_file_2016(
-    file: &Path,
-    cfg: &xlsxread::config::DatasetConfig,
-    patterns: &CompiledPatterns,
-    conn: &rusqlite::Connection,
-    base: &str,
-    total_source_rows: &mut u64,
-    total_errors: &mut u64,
-) -> Result<u64> {
-    use calamine::{open_workbook_auto, Reader, Sheets};
-
-    let path_str = file.display().to_string();
-    let mut workbook: Sheets<_> =
-        open_workbook_auto(file).with_context(|| format!("Cannot open {path_str}"))?;
-
-    let sheet_names: Vec<String> = workbook.sheet_names().to_vec();
-    if sheet_names.is_empty() {
-        return Ok(0);
-    }
-
-    // Sheet selection: thptqg2016 data/ has all-sheets mode to handle
-    // HCM/HN overflow (same reason as thptqg2017 data/).
-    let sheets_to_read: Vec<String> = match cfg.reader.sheet_mode {
-        xlsxread::config::SheetMode::All => sheet_names.clone(),
-        xlsxread::config::SheetMode::First => vec![sheet_names[0].clone()],
-    };
-
-    let mut file_rows: u64 = 0;
-
-    for sheet_name in &sheets_to_read {
-        let range = workbook
-            .worksheet_range(sheet_name)
-            .with_context(|| format!("Cannot read sheet {sheet_name} in {path_str}"))?;
-
-        let rows: Vec<Vec<Data>> = range.rows().map(|r| r.to_vec()).collect();
-        if rows.is_empty() {
-            continue;
-        }
-
-        // Detect format from first row, then determine start index.
-        // Mirrors build-database.js:215–220: isHeaderRow check + detectFormat.
-        let (fmt, start_idx) = if is_header_row_2016(&rows[0]) {
-            (detect_format(&rows[0]), 1)
-        } else {
-            (DetectedFormat::Default, 0)
-        };
-
-        for row in rows.iter().skip(start_idx) {
-            if row.len() < 2 {
-                continue;
-            }
-
-            *total_source_rows += 1;
-
-            match process_row_2016(row, &fmt, patterns) {
-                None => {
-                    // Row was empty/invalid — skipped (mirrors JS `if (!record) continue`)
-                }
-                Some(parsed) => {
-                    match insert_row(conn, &parsed) {
-                        Ok(()) => file_rows += 1,
-                        Err(e) => {
-                            *total_errors += 1;
-                            if *total_errors <= 5 {
-                                eprintln!("  [warn] {base}: {e}");
-                            }
-                        }
-                    }
-                }
-            }
-        }
-    }
-
-    Ok(file_rows)
-}
diff --git a/parser/src/reader.rs b/parser/src/reader.rs
deleted file mode 100644
index ac71da1..0000000
--- a/parser/src/reader.rs
+++ /dev/null
@@ -1,197 +0,0 @@
-/// Spreadsheet reader: wraps calamine to iterate rows across sheets.
-///
-/// Sheet selection mirrors the JS scripts:
-///   - sheet_mode = "all"   → iterate every sheet (handles HCM/HN 65k overflow in data/)
-///   - sheet_mode = "first" → sheet 0 only (data-old/)
-///
-/// Header detection mirrors build-lib.js isHeaderRow:
-///   row[0].toUpperCase() in {"HO_TEN", "HỌ TÊN", "STT"}
-use std::path::Path;
-
-use calamine::{open_workbook_auto, Data, Reader, Sheets};
-
-use crate::config::{DatasetConfig, HeaderCfg, SheetMode};
-use crate::error::BuildError;
-
-// ---------------------------------------------------------------------------
-// Public row representation from calamine
-// ---------------------------------------------------------------------------
-
-pub type RawRow = Vec<Data>;
-
-// ---------------------------------------------------------------------------
-// Header detection — mirrors build-lib.js isHeaderRow
-// ---------------------------------------------------------------------------
-
-/// Returns true when the first cell (uppercased) matches one of the configured
-/// header tokens. Used to skip the header row on the first row of each sheet.
-pub fn is_header_row(row: &[Data], header_cfg: &HeaderCfg) -> bool {
-    if row.len() < 3 {
-        return false;
-    }
-    let first = row[0].to_string().trim().to_uppercase();
-    header_cfg.tokens.iter().any(|t| t.to_uppercase() == first)
-}
-
-// ---------------------------------------------------------------------------
-// All-blank row check (data-old2: strip_blank_rows)
-// ---------------------------------------------------------------------------
-
-pub fn is_all_blank(row: &[Data]) -> bool {
-    row.iter()
-        .all(|c| matches!(c, Data::Empty) || c.to_string().trim().is_empty())
-}
-
-// ---------------------------------------------------------------------------
-// File processor — yields all data rows from the file
-// ---------------------------------------------------------------------------
-
-/// Process one spreadsheet file, calling `on_row` for each data row.
-///
-/// `on_row` receives `(sheet_index, row_index_in_sheet, raw_row)` where
-/// `row_index_in_sheet` is 0-based AFTER the header has been consumed.
-/// Returns `(sheets_seen, total_rows_yielded)`.
-pub fn process_file<F>(
-    path: &Path,
-    cfg: &DatasetConfig,
-    mut on_row: F,
-) -> Result<(usize, usize), BuildError>
-where
-    F: FnMut(usize, &RawRow),
-{
-    let path_str = path.display().to_string();
-
-    // calamine::open_workbook_auto dispatches on file extension
-    let mut workbook: Sheets<_> = open_workbook_auto(path).map_err(|e| BuildError::Calamine {
-        path: path_str.clone(),
-        source: e,
-    })?;
-
-    let sheet_names: Vec<String> = workbook.sheet_names().to_vec();
-    if sheet_names.is_empty() {
-        return Err(BuildError::NoSheets(path_str.clone()));
-    }
-
-    // Sheet selection per config
-    let sheets_to_read: Vec<String> = match cfg.reader.sheet_mode {
-        SheetMode::All => sheet_names.clone(),
-        SheetMode::First => vec![sheet_names[0].clone()],
-    };
-
-    let mut total_rows = 0usize;
-
-    for (sheet_idx, sheet_name) in sheets_to_read.iter().enumerate() {
-        let range = workbook
-            .worksheet_range(sheet_name)
-            .map_err(|e| BuildError::Calamine {
-                path: path_str.clone(),
-                source: e,
-            })?;
-
-        let mut first_row = true;
-
-        for raw in range.rows() {
-            let row: RawRow = raw.to_vec();
-
-            // Skip header row on first row of each sheet (matches JS: `if (i === 0 && isHeaderRow(...))`)
-            if first_row {
-                first_row = false;
-                if is_header_row(&row, &cfg.header) {
-                    continue;
-                }
-            }
-
-            on_row(sheet_idx, &row);
-            total_rows += 1;
-        }
-    }
-
-    Ok((sheets_to_read.len(), total_rows))
-}
-
-// ---------------------------------------------------------------------------
-// Unit tests
-// ---------------------------------------------------------------------------
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-    use crate::config::HeaderCfg;
-
-    fn hdr(tokens: &[&str]) -> HeaderCfg {
-        HeaderCfg {
-            tokens: tokens.iter().map(|s| s.to_string()).collect(),
-        }
-    }
-
-    #[test]
-    fn header_detects_ho_ten() {
-        let row = vec![
-            Data::String("HO_TEN".into()),
-            Data::String("NGAY_SINH".into()),
-            Data::String("SBD".into()),
-        ];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn header_detects_stt() {
-        let row = vec![
-            Data::String("STT".into()),
-            Data::String("B".into()),
-            Data::String("C".into()),
-        ];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn header_detects_ho_ten_unicode() {
-        let row = vec![
-            Data::String("HỌ TÊN".into()),
-            Data::String("B".into()),
-            Data::String("C".into()),
-        ];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn header_rejects_data_row() {
-        let row = vec![
-            Data::String("Nguyen Van A".into()),
-            Data::String("01/01/2000".into()),
-            Data::String("12345678".into()),
-        ];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(!is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn header_rejects_short_row() {
-        let row = vec![Data::String("HO_TEN".into()), Data::Empty];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(!is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn header_case_insensitive() {
-        let row = vec![
-            Data::String("ho_ten".into()),
-            Data::String("B".into()),
-            Data::String("C".into()),
-        ];
-        let cfg = hdr(&["HO_TEN", "HỌ TÊN", "STT"]);
-        assert!(is_header_row(&row, &cfg));
-    }
-
-    #[test]
-    fn blank_row_detection() {
-        let row = vec![Data::Empty, Data::Empty, Data::String("".into())];
-        assert!(is_all_blank(&row));
-
-        let row2 = vec![Data::String("Nguyen".into()), Data::Empty, Data::Empty];
-        assert!(!is_all_blank(&row2));
-    }
-}
diff --git a/parser/src/schema.rs b/parser/src/schema.rs
deleted file mode 100644
index 6eaf739..0000000
--- a/parser/src/schema.rs
+++ /dev/null
@@ -1,213 +0,0 @@
-//! Canonical `student` table definition — the single source of truth for the
-//! SQL shape of every dataset.
-//!
-//! Every dataset (2016, 2017, 2017-old, 2017-old2) is written into this same
-//! 22-column table. Columns a dataset has no data for bind NULL, which costs
-//! ~1 byte per row in SQLite's record header.
-//!
-//! Column provenance:
-//!   ten_cum_thi, gioi_tinh, tieng_duc, tieng_nhat  → 2016 only
-//!   khtn, khxh, gdcd, tieng_nga                    → 2017 datasets only
-//!   everything else                                → both
-//!
-//! Before this module existed the DDL, the INSERT statement and the subject
-//! regex table were duplicated across four TOML configs, which is how the 2016
-//! and 2017 schemas drifted apart in the first place. The configs now carry only
-//! per-dataset parse rules.
-// ---------------------------------------------------------------------------
-// DDL
-// ---------------------------------------------------------------------------
-
-/// Executed verbatim after the output database is (re)created.
-///
-/// `idx_ten_cum_thi` is partial so it holds zero entries on the three 2017
-/// datasets — where the column is always NULL — while staying fully useful for
-/// the 2016 cluster-grouping queries.
-pub const DDL: &str = "
-CREATE TABLE student (
-  so_bao_danh   TEXT PRIMARY KEY,
-  ho_ten        TEXT NOT NULL,
-  ho_ten_ascii  TEXT NOT NULL,
-  ngay_sinh     TEXT,
-  ten_cum_thi   TEXT,
-  gioi_tinh     TEXT,
-  toan          REAL,
-  ngu_van       REAL,
-  vat_ly        REAL,
-  hoa_hoc       REAL,
-  sinh_hoc      REAL,
-  khtn          REAL,
-  lich_su       REAL,
-  dia_ly        REAL,
-  gdcd          REAL,
-  khxh          REAL,
-  tieng_anh     REAL,
-  tieng_phap    REAL,
-  tieng_nga     REAL,
-  tieng_duc     REAL,
-  tieng_nhat    REAL,
-  tieng_trung   REAL
-);
-CREATE INDEX idx_ho_ten       ON student(ho_ten);
-CREATE INDEX idx_ho_ten_ascii ON student(ho_ten_ascii);
-CREATE INDEX idx_ten_cum_thi  ON student(ten_cum_thi) WHERE ten_cum_thi IS NOT NULL;
-";
-
-// ---------------------------------------------------------------------------
-// Column order
-// ---------------------------------------------------------------------------
-
-/// Identity columns, in INSERT parameter order.
-pub const IDENTITY_FIELDS: &[&str] = &[
-    "so_bao_danh",
-    "ho_ten",
-    "ho_ten_ascii",
-    "ngay_sinh",
-    "ten_cum_thi",
-    "gioi_tinh",
-];
-
-/// Subject columns, in INSERT parameter order. Bound as NULL when a row has no
-/// score for that subject.
-pub const SCORE_FIELDS: &[&str] = &[
-    "toan",
-    "ngu_van",
-    "vat_ly",
-    "hoa_hoc",
-    "sinh_hoc",
-    "khtn",
-    "lich_su",
-    "dia_ly",
-    "gdcd",
-    "khxh",
-    "tieng_anh",
-    "tieng_phap",
-    "tieng_nga",
-    "tieng_duc",
-    "tieng_nhat",
-    "tieng_trung",
-];
-
-/// Total bound parameters per row.
-pub const PARAM_COUNT: usize = IDENTITY_FIELDS.len() + SCORE_FIELDS.len();
-
-// ---------------------------------------------------------------------------
-// INSERT
-// ---------------------------------------------------------------------------
-
-/// Positional INSERT matching IDENTITY_FIELDS followed by SCORE_FIELDS.
-///
-/// `OR REPLACE` preserves the pre-existing behaviour where a repeated SBD
-/// overwrites the earlier row rather than aborting the transaction.
-pub const INSERT_SQL: &str = "
-INSERT OR REPLACE INTO student
-  (so_bao_danh, ho_ten, ho_ten_ascii, ngay_sinh, ten_cum_thi, gioi_tinh,
-   toan, ngu_van, vat_ly, hoa_hoc, sinh_hoc, khtn,
-   lich_su, dia_ly, gdcd, khxh,
-   tieng_anh, tieng_phap, tieng_nga, tieng_duc, tieng_nhat, tieng_trung)
-VALUES
-  (?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?, ?)
-";
-
-// ---------------------------------------------------------------------------
-// Subject score patterns
-// ---------------------------------------------------------------------------
-
-/// Regex per subject, applied to the DIEM_THI cell text.
-///
-/// Every pattern runs against every dataset. A subject absent from a given exam
-/// year simply never matches and stays NULL — 2016 source files contain no
-/// "KHTN:" or "Tiếng Nga:" tokens, and 2017 files contain no "Tiếng Đức:" or
-/// "Tiếng Nhật:". The parity check asserts those counts are exactly zero rather
-/// than assuming it.
-///
-/// Order here is irrelevant (matching is by name); SCORE_FIELDS fixes the
-/// INSERT order.
-pub const SCORE_PATTERNS: &[(&str, &str)] = &[
-    ("toan", r"Toán:\s*(\d+(?:\.\d+)?)"),
-    ("ngu_van", r"Ngữ văn:\s*(\d+(?:\.\d+)?)"),
-    ("vat_ly", r"Vật lí:\s*(\d+(?:\.\d+)?)"),
-    ("hoa_hoc", r"Hóa học:\s*(\d+(?:\.\d+)?)"),
-    ("sinh_hoc", r"Sinh học:\s*(\d+(?:\.\d+)?)"),
-    ("khtn", r"KHTN:\s*(\d+(?:\.\d+)?)"),
-    ("lich_su", r"Lịch sử:\s*(\d+(?:\.\d+)?)"),
-    ("dia_ly", r"Địa lí:\s*(\d+(?:\.\d+)?)"),
-    ("gdcd", r"GDCD:\s*(\d+(?:\.\d+)?)"),
-    ("khxh", r"KHXH:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_anh", r"Tiếng Anh:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_phap", r"Tiếng Pháp:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_nga", r"Tiếng Nga:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_duc", r"Tiếng Đức:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_nhat", r"Tiếng Nhật:\s*(\d+(?:\.\d+)?)"),
-    ("tieng_trung", r"Tiếng Trung:\s*(\d+(?:\.\d+)?)"),
-];
-
-// ---------------------------------------------------------------------------
-// Tests
-// ---------------------------------------------------------------------------
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    /// The INSERT placeholder count, the named column list and the field
-    /// constants must agree, or rows silently land in the wrong columns.
-    #[test]
-    fn insert_matches_field_order() {
-        assert_eq!(PARAM_COUNT, 22);
-        assert_eq!(INSERT_SQL.matches('?').count(), PARAM_COUNT);
-
-        let named = INSERT_SQL
-            .split_once('(')
-            .and_then(|(_, rest)| rest.split_once(')'))
-            .map(|(cols, _)| cols)
-            .expect("INSERT must contain a column list");
-
-        let listed: Vec<&str> = named
-            .split(',')
-            .map(str::trim)
-            .filter(|s| !s.is_empty())
-            .collect();
-
-        let expected: Vec<&str> = IDENTITY_FIELDS
-            .iter()
-            .chain(SCORE_FIELDS.iter())
-            .copied()
-            .collect();
-
-        assert_eq!(listed, expected);
-    }
-
-    /// Every subject column must have a pattern and vice versa.
-    #[test]
-    fn score_patterns_cover_score_fields() {
-        assert_eq!(SCORE_PATTERNS.len(), SCORE_FIELDS.len());
-        for (field, _) in SCORE_PATTERNS {
-            assert!(
-                SCORE_FIELDS.contains(field),
-                "pattern {field} has no column"
-            );
-        }
-        for field in SCORE_FIELDS {
-            assert!(
-                SCORE_PATTERNS.iter().any(|(f, _)| f == field),
-                "column {field} has no pattern"
-            );
-        }
-    }
-
-    /// Every column named in the DDL must be bound by the INSERT.
-    #[test]
-    fn ddl_columns_match_insert() {
-        for field in IDENTITY_FIELDS.iter().chain(SCORE_FIELDS.iter()) {
-            assert!(DDL.contains(field), "DDL missing column {field}");
-        }
-    }
-
-    #[test]
-    fn score_patterns_compile() {
-        for (field, src) in SCORE_PATTERNS {
-            regex::Regex::new(src).unwrap_or_else(|e| panic!("{field}: {e}"));
-        }
-    }
-}
diff --git a/parser/src/transform.rs b/parser/src/transform.rs
deleted file mode 100644
index c228f41..0000000
--- a/parser/src/transform.rs
+++ /dev/null
@@ -1,409 +0,0 @@
-/// Row transformation: ascii normalisation, score regex parsing, validation.
-///
-/// `to_ascii` replicates build-lib.js `toAscii` exactly:
-///   str.normalize("NFD").replace(/[̀-ͯ]/g,"").replace(/đ/gi,"d").toLowerCase()
-use std::collections::HashMap;
-
-use regex::Regex;
-use unicode_normalization::UnicodeNormalization;
-
-use crate::config::{DatasetConfig, ValidationCfg};
-use crate::error::BuildError;
-
-// ---------------------------------------------------------------------------
-// Compiled score patterns (built once at startup from config)
-// ---------------------------------------------------------------------------
-
-pub struct CompiledPatterns {
-    /// One compiled regex per subject column in `schema::SCORE_PATTERNS`.
-    pub patterns: Vec<(String, Regex)>,
-}
-
-impl CompiledPatterns {
-    /// Compile the canonical subject patterns once at startup.
-    ///
-    /// All 16 patterns run against every dataset. A subject that did not exist
-    /// in a given exam year simply never matches and stays NULL — the parity
-    /// check asserts those counts are exactly zero rather than assuming it.
-    pub fn new() -> Result<Self, BuildError> {
-        let mut patterns = Vec::with_capacity(crate::schema::SCORE_PATTERNS.len());
-        for (field, src) in crate::schema::SCORE_PATTERNS {
-            let re = Regex::new(src).map_err(|e| BuildError::Regex {
-                pattern: (*src).to_string(),
-                source: e,
-            })?;
-            patterns.push(((*field).to_string(), re));
-        }
-        Ok(Self { patterns })
-    }
-}
-
-// ---------------------------------------------------------------------------
-// to_ascii — must be byte-for-byte equivalent to build-lib.js toAscii
-// ---------------------------------------------------------------------------
-
-/// Normalise a Vietnamese name to an ASCII slug.
-///
-/// Algorithm mirrors the JavaScript `toAscii` in build-lib.js:
-///   1. NFD decompose (splits base + combining diacritics)
-///   2. Drop all Unicode combining marks (U+0300–U+036F)
-///   3. Replace đ/Đ with d (NFD does not decompose đ)
-///   4. Lowercase
-pub fn to_ascii(s: &str) -> String {
-    // Step 1 + 2: NFD then filter out combining marks (Unicode category M)
-    let decomposed: String = s
-        .nfd()
-        .filter(|c| !('\u{0300}'..='\u{036f}').contains(c))
-        .collect();
-
-    // Step 3: đ/Đ are not decomposed by NFD — replace explicitly
-    let replaced = decomposed.replace(['đ', 'Đ'], "d");
-
-    // Step 4: lowercase
-    replaced.to_lowercase()
-}
-
-// ---------------------------------------------------------------------------
-// Parsed row ready for DB insert
-// ---------------------------------------------------------------------------
-
-pub struct ParsedRow {
-    pub so_bao_danh: String,
-    pub ho_ten: String,
-    pub ho_ten_ascii: String,
-    pub ngay_sinh: Option<String>,
-    /// thptqg2016 only: examination cluster name (TEN_CUMTHI column)
-    pub ten_cum_thi: Option<String>,
-    /// thptqg2016 only: gender (GIOI_TINH column), normalised to "Nam"/"Nữ" or None
-    pub gioi_tinh: Option<String>,
-    /// Subject field → float value; absent subjects not in map → NULL
-    pub scores: HashMap<String, f64>,
-}
-
-// ---------------------------------------------------------------------------
-// Row validation — mirrors the per-script skip logic
-// ---------------------------------------------------------------------------
-
-/// Returns `None` when the row should be skipped entirely (before sourceRows counter).
-/// Returns `Some(reason)` when the row should be counted as sourceRows but skipped.
-#[derive(Debug, PartialEq, Eq)]
-pub enum SkipReason {
-    /// Row is fully blank (data-old2 only, before sourceRows counter)
-    BlankRow,
-    /// soBaoDanh or hoTen empty/missing
-    EmptyField,
-    /// soBaoDanh contains non-digit characters (data-old / data-old2 guard)
-    NonNumericSbd,
-}
-
-/// Validates a raw cell slice against the dataset's `ValidationCfg`.
-/// Returns `Ok(())` on pass, `Err(SkipReason)` on fail.
-pub fn validate_row(
-    ho_ten: &str,
-    so_bao_danh: &str,
-    cfg: &ValidationCfg,
-    strip_blank_rows: bool,
-    all_blank: bool,
-) -> Result<(), SkipReason> {
-    // data-old2: skip fully blank rows BEFORE counting sourceRows
-    if strip_blank_rows && all_blank {
-        return Err(SkipReason::BlankRow);
-    }
-
-    if cfg.require_nonempty_sbd && so_bao_danh.is_empty() {
-        return Err(SkipReason::EmptyField);
-    }
-    if cfg.require_nonempty_name && ho_ten.is_empty() {
-        return Err(SkipReason::EmptyField);
-    }
-    if cfg.require_numeric_sbd && !so_bao_danh.chars().all(|c| c.is_ascii_digit()) {
-        return Err(SkipReason::NonNumericSbd);
-    }
-    Ok(())
-}
-
-// ---------------------------------------------------------------------------
-// Score parsing — mirrors build-lib.js parseScores
-// ---------------------------------------------------------------------------
-
-/// Parse a DIEM_THI cell string and extract matching subject scores.
-pub fn parse_scores(diem_thi: &str, patterns: &CompiledPatterns) -> HashMap<String, f64> {
-    let mut out = HashMap::new();
-    for (field, re) in &patterns.patterns {
-        if let Some(caps) = re.captures(diem_thi) {
-            if let Some(m) = caps.get(1) {
-                if let Ok(v) = m.as_str().parse::<f64>() {
-                    if v.is_finite() {
-                        out.insert(field.clone(), v);
-                    }
-                }
-            }
-        }
-    }
-    out
-}
-
-// ---------------------------------------------------------------------------
-// Full row transform (thptqg2017 fixed-column path)
-// ---------------------------------------------------------------------------
-
-/// Extract and transform one spreadsheet row into a `ParsedRow` using fixed column indices.
-/// `raw` is the full cell slice; column indices come from `cfg.columns`.
-/// Used for thptqg2017 configs that have a static [columns] table.
-pub fn transform_row(
-    raw: &[calamine::Data],
-    cfg: &DatasetConfig,
-    patterns: &CompiledPatterns,
-) -> ParsedRow {
-    let cols = cfg
-        .columns
-        .as_ref()
-        .expect("transform_row requires [columns] section in config");
-
-    let get = |idx: usize| -> String {
-        raw.get(idx)
-            .map(|cell| cell.to_string().trim().to_owned())
-            .unwrap_or_default()
-    };
-
-    let ho_ten = get(cols.ho_ten);
-    let ngay_sinh = get(cols.ngay_sinh);
-    let so_bao_danh = get(cols.so_bao_danh);
-    let diem_thi = raw
-        .get(cols.diem_thi)
-        .map(|c| c.to_string())
-        .unwrap_or_default();
-
-    let ho_ten_ascii = to_ascii(&ho_ten);
-    let scores = parse_scores(&diem_thi, patterns);
-    let ngay_sinh_opt = if ngay_sinh.is_empty() {
-        None
-    } else {
-        Some(ngay_sinh)
-    };
-
-    ParsedRow {
-        so_bao_danh,
-        ho_ten,
-        ho_ten_ascii,
-        ngay_sinh: ngay_sinh_opt,
-        ten_cum_thi: None,
-        gioi_tinh: None,
-        scores,
-    }
-}
-
-// ---------------------------------------------------------------------------
-// Unit tests — 20 cases for to_ascii (real Vietnamese names)
-// ---------------------------------------------------------------------------
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    // Helper: assert to_ascii(input) == expected
-    fn check(input: &str, expected: &str) {
-        assert_eq!(
-            to_ascii(input),
-            expected,
-            "to_ascii({input:?}) expected {expected:?}"
-        );
-    }
-
-    #[test]
-    fn ascii_plain_latin() {
-        check("Nguyen Van A", "nguyen van a");
-    }
-
-    #[test]
-    fn ascii_nguyen_thi_hoa() {
-        check("Nguyễn Thị Hoa", "nguyen thi hoa");
-    }
-
-    #[test]
-    fn ascii_tran_van_duc() {
-        // đ/Đ replacement
-        check("Trần Văn Đức", "tran van duc");
-    }
-
-    #[test]
-    fn ascii_le_thi_my_duyen() {
-        check("Lê Thị Mỹ Duyên", "le thi my duyen");
-    }
-
-    #[test]
-    fn ascii_pham_thi_lan() {
-        check("Phạm Thị Lan", "pham thi lan");
-    }
-
-    #[test]
-    fn ascii_bui_thi_thu() {
-        check("Bùi Thị Thu", "bui thi thu");
-    }
-
-    #[test]
-    fn ascii_hoang_van_truong() {
-        check("Hoàng Văn Trường", "hoang van truong");
-    }
-
-    #[test]
-    fn ascii_do_thi_ngan() {
-        // Đ uppercase at start
-        check("Đỗ Thị Ngân", "do thi ngan");
-    }
-
-    #[test]
-    fn ascii_nguyen_van_khanh() {
-        check("Nguyễn Văn Khánh", "nguyen van khanh");
-    }
-
-    #[test]
-    fn ascii_trinh_thi_bich_ngoc() {
-        check("Trịnh Thị Bích Ngọc", "trinh thi bich ngoc");
-    }
-
-    #[test]
-    fn ascii_vu_thi_dieu() {
-        // ề = e + combining grave + combining circumflex (after NFD)
-        check("Vũ Thị Diệu", "vu thi dieu");
-    }
-
-    #[test]
-    fn ascii_nguyen_thi_tuong_vi() {
-        check("Nguyễn Thị Tường Vi", "nguyen thi tuong vi");
-    }
-
-    #[test]
-    fn ascii_lowercase_d_stroke() {
-        // Lowercase đ → d
-        check("đặng thị hằng", "dang thi hang");
-    }
-
-    #[test]
-    fn ascii_uppercase_d_stroke() {
-        check("ĐẶNG THỊ HẰNG", "dang thi hang");
-    }
-
-    #[test]
-    fn ascii_mixed_case() {
-        check("NGUYỄN VĂN AN", "nguyen van an");
-    }
-
-    #[test]
-    fn ascii_tran_thi_kim_anh() {
-        check("Trần Thị Kim Anh", "tran thi kim anh");
-    }
-
-    #[test]
-    fn ascii_nguyen_thi_phuong_thao() {
-        check("Nguyễn Thị Phương Thảo", "nguyen thi phuong thao");
-    }
-
-    #[test]
-    fn ascii_le_van_long() {
-        check("Lê Văn Long", "le van long");
-    }
-
-    #[test]
-    fn ascii_vo_thi_xuan_mai() {
-        check("Võ Thị Xuân Mai", "vo thi xuan mai");
-    }
-
-    #[test]
-    fn ascii_empty_string() {
-        check("", "");
-    }
-
-    // --- Score parsing tests ---
-
-    fn make_patterns() -> CompiledPatterns {
-        CompiledPatterns::new().unwrap()
-    }
-
-    #[test]
-    fn parse_scores_single() {
-        let p = make_patterns();
-        let s = "Toán: 8.5";
-        let scores = parse_scores(s, &p);
-        assert_eq!(scores.get("toan"), Some(&8.5));
-        assert!(!scores.contains_key("ngu_van"));
-    }
-
-    #[test]
-    fn parse_scores_multiple() {
-        let p = make_patterns();
-        let s = "Toán: 7.25  Ngữ văn: 6.0  Vật lí: 9";
-        let scores = parse_scores(s, &p);
-        assert_eq!(scores.get("toan"), Some(&7.25));
-        assert_eq!(scores.get("ngu_van"), Some(&6.0));
-        assert_eq!(scores.get("vat_ly"), Some(&9.0));
-    }
-
-    #[test]
-    fn parse_scores_empty_cell() {
-        let p = make_patterns();
-        let scores = parse_scores("", &p);
-        assert!(scores.is_empty());
-    }
-
-    // --- Validation tests ---
-
-    fn default_validation() -> ValidationCfg {
-        ValidationCfg {
-            require_numeric_sbd: false,
-            require_nonempty_name: true,
-            require_nonempty_sbd: true,
-        }
-    }
-
-    #[test]
-    fn validate_ok() {
-        let v = default_validation();
-        assert!(validate_row("Nguyen Van A", "12345678", &v, false, false).is_ok());
-    }
-
-    #[test]
-    fn validate_empty_sbd() {
-        let v = default_validation();
-        assert_eq!(
-            validate_row("Nguyen Van A", "", &v, false, false),
-            Err(SkipReason::EmptyField)
-        );
-    }
-
-    #[test]
-    fn validate_empty_name() {
-        let v = default_validation();
-        assert_eq!(
-            validate_row("", "12345678", &v, false, false),
-            Err(SkipReason::EmptyField)
-        );
-    }
-
-    #[test]
-    fn validate_non_numeric_sbd_rejected() {
-        let mut v = default_validation();
-        v.require_numeric_sbd = true;
-        assert_eq!(
-            validate_row("Nguyen Van A", "12AB5678", &v, false, false),
-            Err(SkipReason::NonNumericSbd)
-        );
-    }
-
-    #[test]
-    fn validate_numeric_sbd_accepted() {
-        let mut v = default_validation();
-        v.require_numeric_sbd = true;
-        assert!(validate_row("Nguyen Van A", "12345678", &v, false, false).is_ok());
-    }
-
-    #[test]
-    fn validate_blank_row_skipped() {
-        let v = default_validation();
-        // strip_blank_rows=true AND all_blank=true → BlankRow
-        assert_eq!(
-            validate_row("", "", &v, true, true),
-            Err(SkipReason::BlankRow)
-        );
-    }
-}
diff --git a/parser/src/writer.rs b/parser/src/writer.rs
deleted file mode 100644
index e8fc17a..0000000
--- a/parser/src/writer.rs
+++ /dev/null
@@ -1,133 +0,0 @@
-/// SQLite writer: DDL setup, batched INSERT OR REPLACE, VACUUM, stats output.
-///
-/// Mirrors build-lib.js createDb + the transaction loop in each build-database*.js.
-/// Stats output lines match the JS stdout exactly so existing CI log-greps still work.
-///
-/// Every dataset writes the same canonical table (see `crate::schema`), so there
-/// is exactly one insert path. Columns a dataset carries no data for bind NULL.
-use std::fs;
-use std::path::Path;
-
-use rusqlite::{params_from_iter, Connection, ToSql};
-
-use crate::error::BuildError;
-use crate::schema;
-use crate::transform::ParsedRow;
-
-// ---------------------------------------------------------------------------
-// DB initialisation — mirrors build-lib.js createDb (delete + recreate)
-// ---------------------------------------------------------------------------
-
-/// Open (or recreate) the output SQLite database, execute the canonical DDL,
-/// and return the open connection ready for inserts.
-pub fn open_db(db_path: &Path) -> Result<Connection, BuildError> {
-    // Mirror Node behaviour: delete existing file before creating (build-lib.js:54)
-    if db_path.exists() {
-        fs::remove_file(db_path).map_err(|e| BuildError::Io {
-            path: db_path.display().to_string(),
-            source: e,
-        })?;
-    }
-
-    // Ensure parent directory exists
-    if let Some(parent) = db_path.parent() {
-        if !parent.as_os_str().is_empty() {
-            fs::create_dir_all(parent).map_err(|e| BuildError::Io {
-                path: parent.display().to_string(),
-                source: e,
-            })?;
-        }
-    }
-
-    let conn = Connection::open(db_path)?;
-    conn.execute_batch(schema::DDL)?;
-    Ok(conn)
-}
-
-// ---------------------------------------------------------------------------
-// Insert a single parsed row
-// ---------------------------------------------------------------------------
-
-/// Bind every field from `row` into the canonical INSERT and execute it.
-///
-/// Parameter order is `schema::IDENTITY_FIELDS` followed by
-/// `schema::SCORE_FIELDS`. Subjects absent from `row.scores` — and the two
-/// identity columns only the thptqg2016 layouts populate — bind NULL.
-pub fn insert_row(conn: &Connection, row: &ParsedRow) -> Result<(), BuildError> {
-    let mut params: Vec<Box<dyn ToSql>> = Vec::with_capacity(schema::PARAM_COUNT);
-
-    params.push(Box::new(row.so_bao_danh.clone()));
-    params.push(Box::new(row.ho_ten.clone()));
-    params.push(Box::new(row.ho_ten_ascii.clone()));
-    params.push(Box::new(row.ngay_sinh.clone()));
-    params.push(Box::new(row.ten_cum_thi.clone()));
-    params.push(Box::new(row.gioi_tinh.clone()));
-
-    for field in schema::SCORE_FIELDS {
-        let val: Option<f64> = row.scores.get(*field).copied();
-        params.push(Box::new(val));
-    }
-
-    debug_assert_eq!(params.len(), schema::PARAM_COUNT);
-
-    conn.execute(
-        schema::INSERT_SQL,
-        params_from_iter(params.iter().map(|p| p.as_ref())),
-    )?;
-    Ok(())
-}
-
-// ---------------------------------------------------------------------------
-// Post-build: VACUUM + stats output
-// ---------------------------------------------------------------------------
-
-/// Run VACUUM and print statistics lines that mirror the Node scripts' stdout.
-/// The exact prefix tokens ("Source data rows", "DB rows", "Size:") are preserved
-/// so any log-grep in the deploy pipeline keeps working.
-#[allow(clippy::too_many_arguments)]
-pub fn finish_db(
-    conn: &Connection,
-    db_path: &Path,
-    source_rows: u64,
-    skipped: u64,
-    errors: u64,
-    dataset_label: &str,
-    _file_count: usize,
-    is_old2: bool,
-) -> Result<(), BuildError> {
-    conn.execute_batch("VACUUM")?;
-
-    let db_count: i64 = conn.query_row("SELECT COUNT(*) FROM student", [], |row| row.get(0))?;
-
-    let insertable = source_rows - skipped;
-
-    println!();
-    if is_old2 {
-        println!("Source non-blank data rows:      {source_rows}");
-        println!("  skipped (empty/non-numeric SBD): {skipped}");
-    } else {
-        println!("Source data rows (post-header):  {source_rows}");
-        if dataset_label.contains("old") {
-            println!("  skipped (empty/non-numeric SBD): {skipped}");
-        } else {
-            println!("  skipped (empty/invalid):        {skipped}");
-        }
-    }
-    println!("  insertable:                     {insertable}");
-    println!("  insert errors:                  {errors}");
-    println!("DB rows (distinct SBD):           {db_count}");
-
-    if !dataset_label.contains("old") && errors == 0 {
-        let gap = insertable as i64 - db_count;
-        if gap == 0 {
-            println!("Audit: OK — every source row made it in.");
-        } else {
-            println!("Audit: {gap} row(s) collapsed (duplicate SBDs overwriting).");
-        }
-    }
-
-    let sz = fs::metadata(db_path).map(|m| m.len()).unwrap_or(0);
-    println!("Size: {:.1} MB", sz as f64 / 1024.0 / 1024.0);
-
-    Ok(())
-}
diff --git a/parser/tests/fixtures/README.md b/parser/tests/fixtures/README.md
deleted file mode 100644
index c0885f3..0000000
--- a/parser/tests/fixtures/README.md
+++ /dev/null
@@ -1,18 +0,0 @@
-# Test Fixtures
-
-Anonymised `.xlsx` files for integration testing. All student PII has been replaced:
-
-- `ho_ten` replaced with `Nguyen Van Test NNN` / `Tran Thi Test NNN` patterns
-- `so_bao_danh` replaced with sequential synthetic numbers (e.g. `10000001`)
-- `ngay_sinh` replaced with fixed synthetic dates
-- Scores are realistic random values in the 0–10 range
-
-Files:
-
-- `province-100.xlsx` — 100-row single-sheet file (simulates a normal province)
-- `hcm-overflow.xlsx` — 2-sheet file (200 rows Sheet1 + 200 rows Sheet2, simulating HCM overflow)
-- `province-numeric-sbd.xlsx` — 100 rows with strictly numeric SBDs (for data-old variant)
-
-These files are generated by `tests/golden.rs` `generate_fixtures()` if they do not already exist
-on disk. The generator is pure Rust (uses the `zip` crate already pulled in via calamine).
-No external Python or Node tooling required for unit/integration tests.
diff --git a/parser/tests/fixtures/hcm-overflow.xlsx b/parser/tests/fixtures/hcm-overflow.xlsx
deleted file mode 100644
index c92e498..0000000
Binary files a/parser/tests/fixtures/hcm-overflow.xlsx and /dev/null differ
diff --git a/parser/tests/fixtures/province-100.xlsx b/parser/tests/fixtures/province-100.xlsx
deleted file mode 100644
index 5d1a109..0000000
Binary files a/parser/tests/fixtures/province-100.xlsx and /dev/null differ
diff --git a/parser/tests/fixtures/province-numeric-sbd.xlsx b/parser/tests/fixtures/province-numeric-sbd.xlsx
deleted file mode 100644
index 5eb518c..0000000
Binary files a/parser/tests/fixtures/province-numeric-sbd.xlsx and /dev/null differ
diff --git a/parser/tests/golden.rs b/parser/tests/golden.rs
deleted file mode 100644
index 3a290d6..0000000
--- a/parser/tests/golden.rs
+++ /dev/null
@@ -1,591 +0,0 @@
-/// Golden tests — integration tests using anonymised fixture files.
-///
-/// Fixture files are generated in-process via raw OOXML + zip if they do not
-/// already exist on disk. No external tooling is required.
-use std::io::Write as IoWrite;
-use std::path::{Path, PathBuf};
-
-// ---------------------------------------------------------------------------
-// Minimal OOXML xlsx generator
-//
-// Produces a valid .xlsx that calamine can read. Only uses the `zip` crate
-// which is already pulled in as a transitive dependency of calamine.
-// ---------------------------------------------------------------------------
-
-/// One row of cell data for a fixture sheet.
-struct XlsxRow {
-    values: Vec<String>,
-}
-
-/// Write a minimal .xlsx to `path` with the given sheets.
-/// `sheets`: Vec<(sheet_name, rows)> where rows[0] is the header.
-fn write_xlsx(path: &Path, sheets: &[(String, Vec<XlsxRow>)]) {
-    use zip::{write::SimpleFileOptions, ZipWriter};
-
-    let file = std::fs::File::create(path).expect("create fixture xlsx");
-    let mut zip = ZipWriter::new(file);
-    let opts = SimpleFileOptions::default();
-
-    // [Content_Types].xml
-    let mut content_types = String::from(
-        r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<Types xmlns="http://schemas.openxmlformats.org/package/2006/content-types">
-  <Default Extension="rels" ContentType="application/vnd.openxmlformats-package.relationships+xml"/>
-  <Default Extension="xml" ContentType="application/xml"/>
-  <Override PartName="/xl/workbook.xml" ContentType="application/vnd.openxmlformats-officedocument.spreadsheetml.sheet.main+xml"/>
-"#,
-    );
-    for (i, _) in sheets.iter().enumerate() {
-        content_types.push_str(&format!(
-            r#"  <Override PartName="/xl/worksheets/sheet{}.xml" ContentType="application/vnd.openxmlformats-officedocument.spreadsheetml.worksheet+xml"/>
-"#,
-            i + 1
-        ));
-    }
-    content_types.push_str("</Types>");
-    zip.start_file("[Content_Types].xml", opts).unwrap();
-    zip.write_all(content_types.as_bytes()).unwrap();
-
-    // _rels/.rels
-    zip.start_file("_rels/.rels", opts).unwrap();
-    zip.write_all(
-        br#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
-  <Relationship Id="rId1" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument" Target="xl/workbook.xml"/>
-</Relationships>"#,
-    )
-    .unwrap();
-
-    // xl/_rels/workbook.xml.rels
-    let mut wb_rels = String::from(
-        r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<Relationships xmlns="http://schemas.openxmlformats.org/package/2006/relationships">
-"#,
-    );
-    for (i, _) in sheets.iter().enumerate() {
-        wb_rels.push_str(&format!(
-            r#"  <Relationship Id="rId{}" Type="http://schemas.openxmlformats.org/officeDocument/2006/relationships/worksheet" Target="worksheets/sheet{}.xml"/>
-"#,
-            i + 1,
-            i + 1
-        ));
-    }
-    wb_rels.push_str("</Relationships>");
-    zip.start_file("xl/_rels/workbook.xml.rels", opts).unwrap();
-    zip.write_all(wb_rels.as_bytes()).unwrap();
-
-    // xl/workbook.xml
-    let mut wb = String::from(
-        r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<workbook xmlns="http://schemas.openxmlformats.org/spreadsheetml/2006/main"
-          xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships">
-  <sheets>
-"#,
-    );
-    for (i, (name, _)) in sheets.iter().enumerate() {
-        let escaped = xml_escape(name);
-        wb.push_str(&format!(
-            r#"    <sheet name="{}" sheetId="{}" r:id="rId{}"/>
-"#,
-            escaped,
-            i + 1,
-            i + 1
-        ));
-    }
-    wb.push_str("  </sheets>\n</workbook>");
-    zip.start_file("xl/workbook.xml", opts).unwrap();
-    zip.write_all(wb.as_bytes()).unwrap();
-
-    // xl/worksheets/sheetN.xml
-    for (i, (_, rows)) in sheets.iter().enumerate() {
-        let mut ws = String::from(
-            r#"<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<worksheet xmlns="http://schemas.openxmlformats.org/spreadsheetml/2006/main">
-  <sheetData>
-"#,
-        );
-        for (row_idx, row) in rows.iter().enumerate() {
-            ws.push_str(&format!(
-                r#"    <row r="{}">
-"#,
-                row_idx + 1
-            ));
-            for (col_idx, val) in row.values.iter().enumerate() {
-                let col_letter = col_letter(col_idx);
-                let cell_ref = format!("{}{}", col_letter, row_idx + 1);
-                let escaped = xml_escape(val);
-                ws.push_str(&format!(
-                    r#"      <c r="{}" t="inlineStr"><is><t>{}</t></is></c>
-"#,
-                    cell_ref, escaped
-                ));
-            }
-            ws.push_str("    </row>\n");
-        }
-        ws.push_str("  </sheetData>\n</worksheet>");
-        zip.start_file(format!("xl/worksheets/sheet{}.xml", i + 1), opts)
-            .unwrap();
-        zip.write_all(ws.as_bytes()).unwrap();
-    }
-
-    zip.finish().unwrap();
-}
-
-fn col_letter(idx: usize) -> &'static str {
-    const LETTERS: &[&str] = &[
-        "A", "B", "C", "D", "E", "F", "G", "H", "I", "J", "K", "L", "M", "N", "O", "P", "Q", "R",
-        "S", "T", "U", "V", "W", "X", "Y", "Z",
-    ];
-    LETTERS[idx % 26]
-}
-
-fn xml_escape(s: &str) -> String {
-    s.replace('&', "&amp;")
-        .replace('<', "&lt;")
-        .replace('>', "&gt;")
-        .replace('"', "&quot;")
-}
-
-// ---------------------------------------------------------------------------
-// Fixture data builders
-// ---------------------------------------------------------------------------
-
-fn header_row() -> XlsxRow {
-    XlsxRow {
-        values: vec![
-            "HO_TEN".into(),
-            "NGAY_SINH".into(),
-            "SO_BAO_DANH".into(),
-            "DIEM_THI".into(),
-        ],
-    }
-}
-
-fn data_row(idx: usize, scores: &str) -> XlsxRow {
-    // Anonymised: name uses sequential pattern, SBD is purely synthetic
-    let name = if idx.is_multiple_of(2) {
-        format!("Nguyen Van Test {:03}", idx)
-    } else {
-        format!("Tran Thi Test {:03}", idx)
-    };
-    XlsxRow {
-        values: vec![
-            name,
-            format!("15/0{}/{}", (idx % 9) + 1, 1999 + (idx % 5)),
-            format!("1000{:04}", idx),
-            scores.to_owned(),
-        ],
-    }
-}
-
-fn sample_scores(idx: usize) -> String {
-    // Realistic scores in 0–10 range, varies by idx
-    let toan = 4.0 + (idx % 60) as f64 / 10.0;
-    let van = 3.5 + (idx % 65) as f64 / 10.0;
-    format!("Toán: {toan:.1}  Ngữ văn: {van:.1}  Tiếng Anh: 7.5")
-}
-
-// ---------------------------------------------------------------------------
-// Fixture file paths
-// ---------------------------------------------------------------------------
-
-fn fixtures_dir() -> PathBuf {
-    // tests/fixtures/ relative to the crate root
-    let mut p = PathBuf::from(env!("CARGO_MANIFEST_DIR"));
-    p.push("tests");
-    p.push("fixtures");
-    p
-}
-
-fn province_fixture_path() -> PathBuf {
-    fixtures_dir().join("province-100.xlsx")
-}
-
-fn hcm_overflow_fixture_path() -> PathBuf {
-    fixtures_dir().join("hcm-overflow.xlsx")
-}
-
-fn numeric_sbd_fixture_path() -> PathBuf {
-    fixtures_dir().join("province-numeric-sbd.xlsx")
-}
-
-// ---------------------------------------------------------------------------
-// Fixture generation — called once per test run if files missing
-// ---------------------------------------------------------------------------
-
-fn ensure_fixtures() {
-    let dir = fixtures_dir();
-    std::fs::create_dir_all(&dir).expect("create fixtures dir");
-
-    // province-100.xlsx — 100 data rows, single sheet, with header
-    if !province_fixture_path().exists() {
-        let mut rows = vec![header_row()];
-        for i in 0..100 {
-            rows.push(data_row(i, &sample_scores(i)));
-        }
-        write_xlsx(&province_fixture_path(), &[("Sheet1".to_owned(), rows)]);
-    }
-
-    // hcm-overflow.xlsx — 2 sheets × 200 rows each (no header on sheet 2)
-    if !hcm_overflow_fixture_path().exists() {
-        let mut sheet1 = vec![header_row()];
-        for i in 0..200 {
-            sheet1.push(data_row(i, &sample_scores(i)));
-        }
-        // Sheet2: continuation rows, no header row (as in real HCM overflow)
-        let mut sheet2 = Vec::new();
-        for i in 200..400 {
-            sheet2.push(data_row(i, &sample_scores(i)));
-        }
-        write_xlsx(
-            &hcm_overflow_fixture_path(),
-            &[("Sheet1".to_owned(), sheet1), ("Sheet2".to_owned(), sheet2)],
-        );
-    }
-
-    // province-numeric-sbd.xlsx — strictly numeric SBDs for data-old config
-    if !numeric_sbd_fixture_path().exists() {
-        let mut rows = vec![header_row()];
-        for i in 0..100 {
-            rows.push(XlsxRow {
-                values: vec![
-                    format!("Nguyen Van Test {:03}", i),
-                    "01/01/2000".to_owned(),
-                    format!("{:08}", 20000000 + i), // pure digits
-                    sample_scores(i),
-                ],
-            });
-        }
-        write_xlsx(&numeric_sbd_fixture_path(), &[("Sheet1".to_owned(), rows)]);
-    }
-}
-
-// ---------------------------------------------------------------------------
-// Config helpers
-// ---------------------------------------------------------------------------
-
-fn make_data_config() -> xlsxread::config::DatasetConfig {
-    let cfg_path = PathBuf::from(env!("CARGO_MANIFEST_DIR"))
-        .join("configs")
-        .join("2017.yml");
-    xlsxread::config::load_config(&cfg_path).expect("load data config")
-}
-
-// ---------------------------------------------------------------------------
-// Integration tests — pure Rust, no Node dependency
-// ---------------------------------------------------------------------------
-
-#[test]
-fn province_100_builds_100_rows() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-    std::fs::copy(province_fixture_path(), fixture_dir.join("province.xlsx")).unwrap();
-
-    run_build_cmd(&fixture_dir, &db_path, "2017.yml");
-
-    let count = query_count(&db_path);
-    assert_eq!(count, 100, "expected 100 rows from province-100 fixture");
-}
-
-#[test]
-fn hcm_overflow_builds_400_rows() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-    std::fs::copy(hcm_overflow_fixture_path(), fixture_dir.join("hcm.xlsx")).unwrap();
-
-    run_build_cmd(&fixture_dir, &db_path, "2017.yml");
-
-    let count = query_count(&db_path);
-    assert_eq!(
-        count, 400,
-        "expected 400 rows (200 × 2 sheets) from hcm-overflow fixture"
-    );
-}
-
-#[test]
-fn data_old_first_sheet_only_100_rows() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-    // Use the overflow file but with data-old config (first sheet only → 200 rows)
-    std::fs::copy(hcm_overflow_fixture_path(), fixture_dir.join("hcm.xlsx")).unwrap();
-
-    run_build_cmd(&fixture_dir, &db_path, "2017-old.yml");
-
-    // data-old: sheet_mode=first → only 200 rows from sheet1; but SBDs "1000NNNN" are
-    // all digits so all pass the numeric guard
-    let count = query_count(&db_path);
-    assert_eq!(
-        count, 200,
-        "data-old config should read only first sheet (200 rows)"
-    );
-}
-
-#[test]
-fn numeric_sbd_guard_rejects_non_numeric() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-
-    // Write a fixture with one non-numeric SBD mixed in
-    let mut rows = vec![header_row()];
-    for i in 0..10 {
-        rows.push(XlsxRow {
-            values: vec![
-                format!("Test {:03}", i),
-                "01/01/2000".to_owned(),
-                if i == 5 {
-                    "ABC123".to_owned()
-                } else {
-                    format!("{:08}", 20000000 + i)
-                },
-                sample_scores(i),
-            ],
-        });
-    }
-    let mixed_path = fixture_dir.join("mixed.xlsx");
-    write_xlsx(&mixed_path, &[("Sheet1".to_owned(), rows)]);
-
-    run_build_cmd(&fixture_dir, &db_path, "2017-old.yml");
-
-    // Row i=5 has non-numeric SBD → rejected by data-old config
-    let count = query_count(&db_path);
-    assert_eq!(
-        count, 9,
-        "non-numeric SBD row should be skipped by data-old config"
-    );
-}
-
-#[test]
-fn scores_parsed_correctly_into_db() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-
-    let rows = vec![
-        header_row(),
-        XlsxRow {
-            values: vec![
-                "Nguyen Van Test 001".to_owned(),
-                "01/01/2000".to_owned(),
-                "10000001".to_owned(),
-                "Toán: 8.5  Ngữ văn: 7.0  Tiếng Anh: 9.25".to_owned(),
-            ],
-        },
-    ];
-    write_xlsx(
-        &fixture_dir.join("one.xlsx"),
-        &[("Sheet1".to_owned(), rows)],
-    );
-
-    run_build_cmd(&fixture_dir, &db_path, "2017.yml");
-
-    let conn = rusqlite::Connection::open(&db_path).unwrap();
-    let (toan, van, anh): (f64, f64, f64) = conn
-        .query_row(
-            "SELECT toan, ngu_van, tieng_anh FROM student WHERE so_bao_danh = '10000001'",
-            [],
-            |r| Ok((r.get(0)?, r.get(1)?, r.get(2)?)),
-        )
-        .expect("row not found");
-    assert!((toan - 8.5).abs() < 1e-9);
-    assert!((van - 7.0).abs() < 1e-9);
-    assert!((anh - 9.25).abs() < 1e-9);
-}
-
-#[test]
-fn to_ascii_stored_correctly() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-
-    let rows = vec![
-        header_row(),
-        XlsxRow {
-            values: vec![
-                "Nguyễn Văn Đức".to_owned(),
-                "".to_owned(),
-                "20000001".to_owned(),
-                "".to_owned(),
-            ],
-        },
-    ];
-    write_xlsx(
-        &fixture_dir.join("one.xlsx"),
-        &[("Sheet1".to_owned(), rows)],
-    );
-
-    run_build_cmd(&fixture_dir, &db_path, "2017.yml");
-
-    let conn = rusqlite::Connection::open(&db_path).unwrap();
-    let ascii: String = conn
-        .query_row(
-            "SELECT ho_ten_ascii FROM student WHERE so_bao_danh = '20000001'",
-            [],
-            |r| r.get(0),
-        )
-        .expect("row not found");
-    assert_eq!(ascii, "nguyen van duc");
-}
-
-#[test]
-fn audit_subcommand_matches_after_build() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-    std::fs::copy(province_fixture_path(), fixture_dir.join("province.xlsx")).unwrap();
-
-    run_build_cmd(&fixture_dir, &db_path, "2017.yml");
-
-    // audit should match (100 distinct SBDs in xlsx == 100 rows in DB)
-    let cfg = make_data_config();
-    let result = xlsxread::audit::run_audit(&fixture_dir, &db_path, &cfg).expect("audit failed");
-    assert!(result.matched, "audit should match after build");
-    assert_eq!(result.distinct_sbds, 100);
-    assert_eq!(result.db_count, 100);
-}
-
-#[test]
-fn audit_subcommand_mismatch_detected() {
-    ensure_fixtures();
-    let dir = tempdir();
-    let db_path = dir.join("test.db");
-    let fixture_dir = dir.join("input");
-    std::fs::create_dir_all(&fixture_dir).unwrap();
-
-    // Write 10 rows to xlsx but build DB from only 5 rows
-    let mut all_rows = vec![header_row()];
-    for i in 0..10 {
-        all_rows.push(data_row(i, &sample_scores(i)));
-    }
-    write_xlsx(
-        &fixture_dir.join("all.xlsx"),
-        &[("Sheet1".to_owned(), all_rows)],
-    );
-
-    // Build DB with only first 5 rows in a different file
-    let build_dir = dir.join("build_input");
-    std::fs::create_dir_all(&build_dir).unwrap();
-    let mut five_rows = vec![header_row()];
-    for i in 0..5 {
-        five_rows.push(data_row(i, &sample_scores(i)));
-    }
-    write_xlsx(
-        &build_dir.join("five.xlsx"),
-        &[("Sheet1".to_owned(), five_rows)],
-    );
-
-    run_build_cmd(&build_dir, &db_path, "2017.yml");
-
-    // audit against fixture_dir (10 xlsx rows) but DB has 5 rows → mismatch
-    let cfg = make_data_config();
-    let result = xlsxread::audit::run_audit(&fixture_dir, &db_path, &cfg).expect("audit failed");
-    assert!(!result.matched, "audit should not match (10 xlsx vs 5 db)");
-    assert_eq!(result.distinct_sbds, 10);
-    assert_eq!(result.db_count, 5);
-}
-
-
-// ---------------------------------------------------------------------------
-// Helpers
-// ---------------------------------------------------------------------------
-
-fn tempdir() -> PathBuf {
-    let base = std::env::temp_dir().join(format!(
-        "xlsxread-test-{}",
-        std::time::SystemTime::now()
-            .duration_since(std::time::UNIX_EPOCH)
-            .unwrap()
-            .subsec_nanos()
-    ));
-    std::fs::create_dir_all(&base).unwrap();
-    base
-}
-
-fn run_build_cmd(input_dir: &Path, db_path: &Path, config_name: &str) {
-    let cfg_path = PathBuf::from(env!("CARGO_MANIFEST_DIR"))
-        .join("configs")
-        .join(config_name);
-
-    let cfg = xlsxread::config::load_config(&cfg_path)
-        .unwrap_or_else(|e| panic!("load config {config_name}: {e}"));
-    let patterns = xlsxread::transform::CompiledPatterns::new().expect("compile patterns");
-
-    // Collect files
-    let mut files: Vec<PathBuf> = std::fs::read_dir(input_dir)
-        .unwrap()
-        .filter_map(|e| e.ok())
-        .map(|e| e.path())
-        .filter(|p| {
-            p.is_file()
-                && p.extension()
-                    .and_then(|e| e.to_str())
-                    .map(|e| {
-                        let l = e.to_lowercase();
-                        l == "xls" || l == "xlsx"
-                    })
-                    .unwrap_or(false)
-        })
-        .collect();
-    files.sort();
-
-    let conn = xlsxread::writer::open_db(db_path).expect("open db");
-    conn.execute_batch("BEGIN").unwrap();
-
-    for file in &files {
-        xlsxread::reader::process_file(file, &cfg, |_, raw| {
-            let all_blank = xlsxread::reader::is_all_blank(raw);
-            if cfg.reader.strip_blank_rows && all_blank {
-                return;
-            }
-            let cols = cfg.columns.as_ref().expect("golden test requires [columns]");
-            let ho_ten = raw
-                .get(cols.ho_ten)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-            let so_bao_danh = raw
-                .get(cols.so_bao_danh)
-                .map(|c| c.to_string().trim().to_owned())
-                .unwrap_or_default();
-            if xlsxread::transform::validate_row(
-                &ho_ten,
-                &so_bao_danh,
-                &cfg.validation,
-                cfg.reader.strip_blank_rows,
-                all_blank,
-            )
-            .is_err()
-            {
-                return;
-            }
-            let row = xlsxread::transform::transform_row(raw, &cfg, &patterns);
-            let _ = xlsxread::writer::insert_row(&conn, &row);
-        })
-        .expect("process file");
-    }
-
-    conn.execute_batch("COMMIT").unwrap();
-    conn.execute_batch("VACUUM").unwrap();
-}
-
-fn query_count(db_path: &Path) -> i64 {
-    let conn = rusqlite::Connection::open(db_path).expect("open db for count");
-    conn.query_row("SELECT COUNT(*) FROM student", [], |r| r.get(0))
-        .expect("count query")
-}
diff --git a/src/datasets.js b/src/datasets.js
index 0100672..6f5509e 100644
--- a/src/datasets.js
+++ b/src/datasets.js
@@ -3,7 +3,7 @@
  *
  * `id` is the single identifier used end to end:
  *
- *   data/<id>/  →  parser/configs/<id>.yml  →  db/<id>.db.gz  →  /thptqg/<id>/
+ *   data/<id>/  →  go-parser/configs/<id>.yml  →  db/<id>.db.gz  →  /thptqg/<id>/
  *
  * Site path and database URL are derived from `id` rather than stored, so a
  * dataset cannot be misconfigured into pointing at the wrong database.