diff --git a/.github/workflows/deploy-pages.yml b/.github/workflows/deploy-pages.yml
index 02abd54..80bb772 100644
--- a/.github/workflows/deploy-pages.yml
+++ b/.github/workflows/deploy-pages.yml
@@ -49,25 +49,33 @@ jobs:
       - name: Test parser
         run: npm run test:go
 
+      # The crawler is not part of the build — it only refreshes data/ by hand.
+      # It is still compiled and tested here so it cannot rot unnoticed, and
+      # because its filename test guards the parser: input filenames decide
+      # which row survives a duplicate exam number.
+      - name: Test crawler
+        run: npm run test:crawler
+
       # excelize carries an open advisory, and the 2017 refresh runbook feeds
       # network-downloaded spreadsheets straight into the parser.
       - name: Vulnerability scan
-        working-directory: go-parser
         run: |
           go install golang.org/x/vuln/cmd/govulncheck@latest
-          "$(go env GOPATH)/bin/govulncheck" ./...
+          GOVULNCHECK="$(go env GOPATH)/bin/govulncheck"
+          (cd go-parser && "$GOVULNCHECK" ./...)
+          (cd crawler && "$GOVULNCHECK" ./...)
 
       # One parser binary builds every dataset; build-db.js reads the dataset
-      # list from src/datasets.js, verifies each database against its known row
+      # list from web/src/datasets.js, verifies each database against its row
       # count, then gzips it in place leaving no uncompressed file behind.
       - name: Build databases
         run: |
           npm run build:go
           npm run build:db
 
-      # One Vite build produces every page. scripts/assemble-site.js copies the
-      # emitted index.html to each dataset path (and the legacy nested URLs),
-      # then fails the job if any uncompressed database reached the artifact.
+      # One Vite build produces every page. web/scripts/assemble-site.js copies
+      # the emitted index.html to each dataset path (and the legacy nested
+      # URLs), then fails the job if an uncompressed database reached _site.
       - name: Build and assemble site
         run: npm run build:site
 
diff --git a/README.md b/README.md
index 02c569d..192fdf8 100644
--- a/README.md
+++ b/README.md
@@ -2,7 +2,7 @@
 
 Tra cứu điểm thi THPT Quốc gia — exam-score lookup for Vietnam's national high
 school graduation exam. Client-side SQL (sql.js) over a SQLite database built
-from the ministry's raw `.xls` score files by the Rust `xlsxread` parser.
+from the ministry's raw `.xls` score files by the Go `xlsxread` parser.
 
 Live at **[tiennm99.github.io/thptqg](https://tiennm99.github.io/thptqg/)**.
 
@@ -10,43 +10,60 @@ Live at **[tiennm99.github.io/thptqg](https://tiennm99.github.io/thptqg/)**.
 | --- | --- | --- | --- |
 | `2016` | 2016 | 877,461 | [/2016/](https://tiennm99.github.io/thptqg/2016/) |
 | `2017` | 2017 | 861,068 | [/2017/](https://tiennm99.github.io/thptqg/2017/) |
-| `2017-old` | 2017 | 847,348 | [/2017-old/](https://tiennm99.github.io/thptqg/2017-old/) |
-| `2017-old2` | 2017 | 679,764 | [/2017-old2/](https://tiennm99.github.io/thptqg/2017-old2/) |
 
-The three 2017 datasets are successive publications of the same exam and they
-disagree; all three are kept so the differences stay inspectable.
+Two earlier 2017 publications (`2017-old`, `2017-old2`) were kept for a while
+because they disagreed with the current one. They have been removed; they remain
+in git history.
 
 ## Layout
 
 ```
-index.html + src/     the frontend — one app serving all four datasets and the hub
-  datasets.js           the four dataset ids and their per-dataset content
-  router.js             pathname → dataset
+web/                  the frontend — one Vite app serving both datasets and the hub
+  src/datasets.js       the dataset ids and their per-dataset content
+  src/router.js         pathname → dataset
+  scripts/              site assembly
+crawler/              Go — re-fetches the source spreadsheets
+  internal/sources/     one file per data source: its links and local filenames
+  internal/fetch/       concurrent, resumable downloading
+go-parser/            Go — Excel to SQLite
+  internal/schema/      canonical 22-column table: DDL, INSERT, subject regexes
+  configs/<id>.yml      per-dataset parse rules only, no SQL
+  scripts/              database build, parity verification
 data/<id>/            raw Excel files, one directory per dataset
-parser/               the Rust parser
-  src/schema.rs         canonical 22-column table: DDL, INSERT, subject regexes
-  configs/<id>.yml     per-dataset parse rules only, no SQL
-  scripts/              database build, crawler, parity verification
-scripts/              site assembly
 docs/                 architecture, data pipeline, deployment
 ```
 
+`web/` is the only npm workspace; `crawler/` and `go-parser/` are independent Go
+modules. The one cross-boundary import is `web/src/datasets.js`, which
+`go-parser/scripts/build-db.js` reads for the dataset list and expected sizes.
+
 The dataset id is one identifier end to end:
 
 ```
-data/2017-old/ → go-parser/configs/2017-old.yml → db/2017-old.db.gz → /thptqg/2017-old/
+data/2017/ → go-parser/configs/2017.yml → db/2017.db.gz → /thptqg/2017/
 ```
 
 ## Build
 
 ```bash
 npm ci
-npm run build:go     # compile the parser
-npm run build:db       # build + gzip all four databases (add an id for just one)
+npm run build:go       # compile the parser
+npm run build:db       # build + gzip both databases (add an id for just one)
 npm run build:site     # one Vite build, then assemble into _site/
 npx serve _site
 ```
 
+The source spreadsheets are committed, so a crawl is only needed to refresh
+them:
+
+```bash
+npm run crawl:2016     # re-fetch data/2016/
+npm run crawl:2017     # re-fetch data/2017/ from the baotintuc.vn CDN
+```
+
+Crawling is idempotent — files already present are skipped — and is never part
+of the build.
+
 Pushing to `main` runs the same steps in
 `.github/workflows/deploy-pages.yml` and publishes to GitHub Pages.
 
@@ -55,7 +72,7 @@ Pushing to `main` runs the same steps in
 1. Put the Excel files in `data/<id>/`
 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`
+3. Add an entry to `DATASETS` in `web/src/datasets.js`
 
 Everything else follows: the build script, the site assembly and the router all
 read that one list, and the UI adapts to whichever columns the dataset fills.
diff --git a/crawler/cmd/crawl/main.go b/crawler/cmd/crawl/main.go
new file mode 100644
index 0000000..ac95d6d
--- /dev/null
+++ b/crawler/cmd/crawl/main.go
@@ -0,0 +1,145 @@
+// Command crawl downloads a dataset's source spreadsheets into data/<id>/.
+//
+// The argument is the dataset id, the same one used by go-parser's configs and
+// the published site paths.
+//
+//	crawl 2017                      # from the baotintuc.vn CDN
+//	crawl 2016                      # source not yet configured
+//	crawl 2017 --list               # show what would be downloaded
+//
+// Runs are idempotent: a file already present and non-empty is skipped, so an
+// interrupted crawl can simply be re-run.
+package main
+
+import (
+	"context"
+	"flag"
+	"fmt"
+	"os"
+	"os/signal"
+	"path/filepath"
+	"syscall"
+	"time"
+
+	"github.com/tiennm99/thptqg/crawler/internal/fetch"
+	"github.com/tiennm99/thptqg/crawler/internal/sources"
+)
+
+func main() {
+	if err := run(os.Args[1:]); err != nil {
+		fmt.Fprintf(os.Stderr, "crawl: %v\n", err)
+		os.Exit(1)
+	}
+}
+
+func usage() {
+	fmt.Fprint(os.Stderr, "usage: crawl <dataset> [flags]\n\nDatasets:\n")
+	for _, s := range sources.All() {
+		fmt.Fprintf(os.Stderr, "  %-6s %s\n", s.ID, s.Summary)
+	}
+	fmt.Fprint(os.Stderr, "\nFlags:\n")
+	fmt.Fprint(os.Stderr, "  --out string          output directory (default ../data/<dataset>)\n")
+	fmt.Fprint(os.Stderr, "  --concurrency int     parallel downloads (default 6)\n")
+	fmt.Fprint(os.Stderr, "  --timeout duration    per-file timeout (default 2m)\n")
+	fmt.Fprint(os.Stderr, "  --list                print the file list and exit\n")
+}
+
+func run(args []string) error {
+	if len(args) == 0 || args[0] == "-h" || args[0] == "--help" {
+		usage()
+		if len(args) == 0 {
+			return fmt.Errorf("no dataset given")
+		}
+		return nil
+	}
+
+	src, err := sources.Lookup(args[0])
+	if err != nil {
+		usage()
+		return err
+	}
+
+	fs := flag.NewFlagSet(src.ID, flag.ContinueOnError)
+	// The default is relative to the crawler module directory, which is where
+	// both `go -C crawler run ./cmd/crawl` and a manual `cd crawler` land.
+	out := fs.String("out", filepath.Join("..", "data", src.ID), "output directory")
+	concurrency := fs.Int("concurrency", 6, "parallel downloads")
+	timeout := fs.Duration("timeout", 2*time.Minute, "per-file timeout")
+	list := fs.Bool("list", false, "print the file list and exit")
+	if err := fs.Parse(args[1:]); err != nil {
+		return err
+	}
+
+	files, err := src.Files()
+	if err != nil {
+		return err
+	}
+	if len(files) == 0 {
+		return fmt.Errorf("source %q produced no files", src.ID)
+	}
+
+	outDir, err := filepath.Abs(*out)
+	if err != nil {
+		return err
+	}
+
+	items := make([]fetch.Item, 0, len(files))
+	for _, f := range files {
+		items = append(items, fetch.Item{
+			Name: f.Name,
+			URL:  f.URL,
+			Path: filepath.Join(outDir, f.Dest),
+		})
+	}
+
+	if *list {
+		for _, it := range items {
+			fmt.Printf("%s\t%s\n", filepath.Base(it.Path), it.URL)
+		}
+		return nil
+	}
+
+	// Ctrl-C cancels in-flight requests; each worker deletes its partial file
+	// on the way out, so an interrupted run leaves nothing half-written.
+	ctx, stop := signal.NotifyContext(context.Background(), os.Interrupt, syscall.SIGTERM)
+	defer stop()
+
+	fmt.Printf("Downloading %d files to %s...\n", len(items), outDir)
+
+	results, runErr := fetch.Run(ctx, items, fetch.Options{
+		Concurrency: *concurrency,
+		Timeout:     *timeout,
+		Headers:     src.Headers,
+		OnResult:    printResult,
+	})
+	if runErr != nil {
+		return runErr
+	}
+
+	ok, skip, failed := fetch.Tally(results)
+	fmt.Printf("\nDone. ok=%d skip=%d fail=%d\n", ok, skip, len(failed))
+	if len(failed) > 0 {
+		for _, r := range failed {
+			fmt.Fprintf(os.Stderr, "  %s: %v\n", r.Item.Name, r.Err)
+		}
+		return fmt.Errorf("%d file(s) failed", len(failed))
+	}
+	return nil
+}
+
+func printResult(done, total int, r fetch.Result) {
+	tag := map[fetch.Status]string{
+		fetch.StatusOK:   "✓",
+		fetch.StatusSkip: "·",
+		fetch.StatusFail: "✗",
+	}[r.Status]
+
+	detail := ""
+	switch r.Status {
+	case fetch.StatusFail:
+		detail = r.Err.Error()
+	default:
+		detail = fmt.Sprintf("%.0f KB", float64(r.Size)/1024)
+	}
+	fmt.Printf("  %s [%d/%d] %-20s %s\n", tag, done, total, r.Item.Name, detail)
+}
diff --git a/crawler/go.mod b/crawler/go.mod
new file mode 100644
index 0000000..dd80379
--- /dev/null
+++ b/crawler/go.mod
@@ -0,0 +1,3 @@
+module github.com/tiennm99/thptqg/crawler
+
+go 1.26.5
diff --git a/crawler/internal/fetch/fetch.go b/crawler/internal/fetch/fetch.go
new file mode 100644
index 0000000..5afff19
--- /dev/null
+++ b/crawler/internal/fetch/fetch.go
@@ -0,0 +1,205 @@
+// Package fetch downloads a list of files concurrently, skipping any that are
+// already on disk.
+//
+// It is deliberately source-agnostic: it knows nothing about provinces, exam
+// years or spreadsheet formats. Each source in internal/sources produces a
+// []Item and this package moves the bytes.
+package fetch
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"net/http"
+	"os"
+	"path/filepath"
+	"sync"
+	"time"
+)
+
+// Item is one file to download.
+type Item struct {
+	// Name is the human label used in progress output, e.g. "An Giang".
+	Name string
+	URL  string
+	// Path is the absolute destination, including filename.
+	Path string
+}
+
+// Status is the outcome of one item.
+type Status int
+
+const (
+	// StatusOK means the file was downloaded during this run.
+	StatusOK Status = iota
+	// StatusSkip means a non-empty file was already present.
+	StatusSkip
+	// StatusFail means the download did not complete; Result.Err says why.
+	StatusFail
+)
+
+// Result records what happened to one item.
+type Result struct {
+	Item   Item
+	Status Status
+	Size   int64
+	Err    error
+}
+
+// Options configures a run. The zero value is usable: Concurrency and Timeout
+// fall back to defaults.
+type Options struct {
+	// Concurrency is the number of files downloaded at once.
+	Concurrency int
+	// Timeout bounds each individual request, headers and body together.
+	// The original JS crawler had no timeout, so one stalled connection could
+	// hang the whole run indefinitely.
+	Timeout time.Duration
+	// Headers are sent with every request. The CDN this was written against
+	// rejects requests without a browser User-Agent and a matching Referer.
+	Headers map[string]string
+	// OnResult, if set, is called once per completed item. Calls are
+	// serialised, so it does not need its own locking, but they arrive in
+	// completion order rather than list order.
+	OnResult func(done, total int, r Result)
+}
+
+const (
+	defaultConcurrency = 6
+	defaultTimeout     = 2 * time.Minute
+)
+
+// Run downloads every item, returning one Result each. It does not return an
+// error for a failed download — that is reported per item — only for a problem
+// that stops the run as a whole, such as an unusable output directory.
+//
+// Results come back in completion order, not input order.
+func Run(ctx context.Context, items []Item, opts Options) ([]Result, error) {
+	if opts.Concurrency <= 0 {
+		opts.Concurrency = defaultConcurrency
+	}
+	if opts.Timeout <= 0 {
+		opts.Timeout = defaultTimeout
+	}
+
+	// Every item's parent directory must exist before any worker starts, so a
+	// missing output directory fails once here rather than N times in parallel.
+	for _, it := range items {
+		if err := os.MkdirAll(filepath.Dir(it.Path), 0o755); err != nil {
+			return nil, fmt.Errorf("cannot create output directory: %w", err)
+		}
+	}
+
+	client := &http.Client{Timeout: opts.Timeout}
+
+	var (
+		mu      sync.Mutex
+		results = make([]Result, 0, len(items))
+		next    int
+	)
+
+	work := make(chan Item)
+	var wg sync.WaitGroup
+	for range opts.Concurrency {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			for it := range work {
+				r := download(ctx, client, it, opts.Headers)
+				mu.Lock()
+				results = append(results, r)
+				next++
+				if opts.OnResult != nil {
+					opts.OnResult(next, len(items), r)
+				}
+				mu.Unlock()
+			}
+		}()
+	}
+
+	for _, it := range items {
+		select {
+		case work <- it:
+		case <-ctx.Done():
+			close(work)
+			wg.Wait()
+			return results, ctx.Err()
+		}
+	}
+	close(work)
+	wg.Wait()
+
+	return results, nil
+}
+
+// download fetches one item, or reports that it was already present.
+//
+// The body is streamed to a .part file and renamed into place only once it is
+// complete. Without that, an interrupted run leaves a truncated file at the
+// final path — and because the skip check only tests for a non-empty file,
+// every later run would skip it and the corruption would persist silently.
+func download(ctx context.Context, client *http.Client, it Item, headers map[string]string) Result {
+	if st, err := os.Stat(it.Path); err == nil && st.Size() > 0 {
+		return Result{Item: it, Status: StatusSkip, Size: st.Size()}
+	}
+
+	fail := func(err error) Result {
+		return Result{Item: it, Status: StatusFail, Err: err}
+	}
+
+	req, err := http.NewRequestWithContext(ctx, http.MethodGet, it.URL, nil)
+	if err != nil {
+		return fail(err)
+	}
+	for k, v := range headers {
+		req.Header.Set(k, v)
+	}
+
+	resp, err := client.Do(req)
+	if err != nil {
+		return fail(err)
+	}
+	defer resp.Body.Close()
+	if resp.StatusCode != http.StatusOK {
+		return fail(fmt.Errorf("HTTP %d", resp.StatusCode))
+	}
+
+	part := it.Path + ".part"
+	f, err := os.Create(part)
+	if err != nil {
+		return fail(err)
+	}
+	n, err := io.Copy(f, resp.Body)
+	if cerr := f.Close(); err == nil {
+		err = cerr
+	}
+	if err != nil {
+		os.Remove(part)
+		return fail(err)
+	}
+	if n == 0 {
+		os.Remove(part)
+		return fail(fmt.Errorf("empty response body"))
+	}
+	if err := os.Rename(part, it.Path); err != nil {
+		os.Remove(part)
+		return fail(err)
+	}
+
+	return Result{Item: it, Status: StatusOK, Size: n}
+}
+
+// Tally counts results by status.
+func Tally(results []Result) (ok, skip int, failed []Result) {
+	for _, r := range results {
+		switch r.Status {
+		case StatusOK:
+			ok++
+		case StatusSkip:
+			skip++
+		case StatusFail:
+			failed = append(failed, r)
+		}
+	}
+	return ok, skip, failed
+}
diff --git a/crawler/internal/fetch/fetch_test.go b/crawler/internal/fetch/fetch_test.go
new file mode 100644
index 0000000..f4214a1
--- /dev/null
+++ b/crawler/internal/fetch/fetch_test.go
@@ -0,0 +1,184 @@
+package fetch
+
+import (
+	"context"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"path/filepath"
+	"testing"
+)
+
+func serve(t *testing.T, h http.HandlerFunc) string {
+	t.Helper()
+	s := httptest.NewServer(h)
+	t.Cleanup(s.Close)
+	return s.URL
+}
+
+func TestDownloadsAndSkips(t *testing.T) {
+	var hits int
+	url := serve(t, func(w http.ResponseWriter, _ *http.Request) {
+		hits++
+		w.Write([]byte("payload"))
+	})
+
+	dir := t.TempDir()
+	items := []Item{{Name: "one", URL: url, Path: filepath.Join(dir, "one.xls")}}
+
+	results, err := Run(context.Background(), items, Options{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if results[0].Status != StatusOK {
+		t.Fatalf("status = %v, err = %v", results[0].Status, results[0].Err)
+	}
+	if got, _ := os.ReadFile(items[0].Path); string(got) != "payload" {
+		t.Errorf("content = %q", got)
+	}
+
+	// Second run must not re-fetch.
+	results, err = Run(context.Background(), items, Options{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if results[0].Status != StatusSkip {
+		t.Errorf("status = %v, want StatusSkip", results[0].Status)
+	}
+	if hits != 1 {
+		t.Errorf("server hit %d times, want 1", hits)
+	}
+}
+
+func TestNon200Fails(t *testing.T) {
+	url := serve(t, func(w http.ResponseWriter, _ *http.Request) {
+		http.Error(w, "gone", http.StatusNotFound)
+	})
+
+	dir := t.TempDir()
+	path := filepath.Join(dir, "missing.xls")
+	results, err := Run(context.Background(), []Item{{Name: "x", URL: url, Path: path}}, Options{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if results[0].Status != StatusFail {
+		t.Fatalf("status = %v, want StatusFail", results[0].Status)
+	}
+	if _, err := os.Stat(path); !os.IsNotExist(err) {
+		t.Error("a failed download must leave no file at the destination")
+	}
+	assertNoPartFiles(t, dir)
+}
+
+// TestFailureLeavesNoPartial is the reason downloads land on a .part file
+// first. Writing straight to the destination would leave a truncated file
+// there, and the skip check — which only tests for a non-empty file — would
+// skip it on every later run, so the corruption would never be re-fetched.
+func TestFailureLeavesNoPartial(t *testing.T) {
+	url := serve(t, func(w http.ResponseWriter, _ *http.Request) {
+		w.Header().Set("Content-Length", "1000")
+		w.Write([]byte("short"))
+		// Closing early makes the body read fail mid-copy.
+		if f, ok := w.(http.Flusher); ok {
+			f.Flush()
+		}
+		panic(http.ErrAbortHandler)
+	})
+
+	dir := t.TempDir()
+	path := filepath.Join(dir, "truncated.xls")
+	results, err := Run(context.Background(), []Item{{Name: "x", URL: url, Path: path}}, Options{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if results[0].Status != StatusFail {
+		t.Fatalf("status = %v, want StatusFail", results[0].Status)
+	}
+	if _, err := os.Stat(path); !os.IsNotExist(err) {
+		t.Error("a truncated download must not be left at the destination")
+	}
+	assertNoPartFiles(t, dir)
+}
+
+func TestEmptyBodyFails(t *testing.T) {
+	url := serve(t, func(w http.ResponseWriter, _ *http.Request) {})
+
+	dir := t.TempDir()
+	path := filepath.Join(dir, "empty.xls")
+	results, err := Run(context.Background(), []Item{{Name: "x", URL: url, Path: path}}, Options{})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if results[0].Status != StatusFail {
+		t.Errorf("an empty body must fail, got %v", results[0].Status)
+	}
+	assertNoPartFiles(t, dir)
+}
+
+func TestHeadersAreSent(t *testing.T) {
+	var gotUA, gotRef string
+	url := serve(t, func(w http.ResponseWriter, r *http.Request) {
+		gotUA, gotRef = r.Header.Get("User-Agent"), r.Header.Get("Referer")
+		w.Write([]byte("ok"))
+	})
+
+	_, err := Run(context.Background(),
+		[]Item{{Name: "x", URL: url, Path: filepath.Join(t.TempDir(), "x.xls")}},
+		Options{Headers: map[string]string{"User-Agent": "test-agent", "Referer": "https://example.test/"}})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if gotUA != "test-agent" || gotRef != "https://example.test/" {
+		t.Errorf("headers not sent: ua=%q referer=%q", gotUA, gotRef)
+	}
+}
+
+func TestAllItemsRun(t *testing.T) {
+	url := serve(t, func(w http.ResponseWriter, _ *http.Request) {
+		w.Write([]byte("x"))
+	})
+
+	dir := t.TempDir()
+	var items []Item
+	for i := range 20 {
+		items = append(items, Item{
+			Name: "f",
+			URL:  url,
+			Path: filepath.Join(dir, "sub", string(rune('a'+i))+".xls"),
+		})
+	}
+
+	results, err := Run(context.Background(), items, Options{Concurrency: 4})
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(results) != len(items) {
+		t.Fatalf("got %d results, want %d", len(results), len(items))
+	}
+	ok, _, failed := Tally(results)
+	if ok != len(items) {
+		t.Errorf("ok = %d, want %d (failures: %v)", ok, len(items), failed)
+	}
+}
+
+func TestTally(t *testing.T) {
+	ok, skip, failed := Tally([]Result{
+		{Status: StatusOK}, {Status: StatusOK}, {Status: StatusSkip}, {Status: StatusFail},
+	})
+	if ok != 2 || skip != 1 || len(failed) != 1 {
+		t.Errorf("ok=%d skip=%d fail=%d", ok, skip, len(failed))
+	}
+}
+
+func assertNoPartFiles(t *testing.T, dir string) {
+	t.Helper()
+	entries, err := os.ReadDir(dir)
+	if err != nil {
+		t.Fatal(err)
+	}
+	for _, e := range entries {
+		if filepath.Ext(e.Name()) == ".part" {
+			t.Errorf("leftover partial file: %s", e.Name())
+		}
+	}
+}
diff --git a/crawler/internal/sources/source_2016.go b/crawler/internal/sources/source_2016.go
new file mode 100644
index 0000000..58f1764
--- /dev/null
+++ b/crawler/internal/sources/source_2016.go
@@ -0,0 +1,182 @@
+package sources
+
+// source2016 fetches the 2016 dataset: one spreadsheet per exam cluster
+// (cụm thi), 4 .xls and 115 .xlsx.
+//
+// # Provenance
+//
+// The list below was recovered from the aggregator article that originally
+// published these files:
+//
+//	cong-bo-diem-thi-thptqg-2016-toan-bo-120-cum-thi-da-co-diem.html
+//
+// The site that first carried it (dtntbacgiang.edu.vn) no longer resolves, so
+// the list was read from the Internet Archive's copy. It is verified, not
+// guessed: all 119 filenames match data/2016/ exactly in both directions, which
+// TestSource2016MatchesFilesOnDisk asserts.
+//
+// # Filenames are load-bearing — do not "tidy" them
+//
+// Dest is the server-assigned name, a 32-hex content hash followed by the
+// cluster slug and a millisecond timestamp. It is kept verbatim because
+// go-parser sorts input files bytewise and inserts last-wins, so filenames
+// decide which row survives a duplicate exam number. That is live for 2016, not
+// hypothetical: 877,464 source rows collapse to 877,461, so three rows' contents
+// depend on this sort order. The hashes in
+// go-parser/testdata/reader-fidelity-hashes.tsv are keyed by full path as well,
+// and they are frozen — they were produced by the Rust reader, which no longer
+// exists.
+//
+// # The host is not verified
+//
+// Only the paths are known to be right; the host that still serves them is not.
+// The original is gone and the Internet Archive captured the article but none of
+// the spreadsheets, so this could not be confirmed by fetching. baseURL points
+// at a mirror of the same article that is still online. If it serves these files
+// from a different directory, uploadPath is the single constant to change.
+const (
+	baseURL    = "https://dtnt.bacninh.edu.vn"
+	uploadPath = "/upload/s/20180102/"
+)
+
+var source2016 = Source{
+	ID:      "2016",
+	Summary: "119 exam-cluster files (4 .xls + 115 .xlsx)",
+	Files: func() ([]File, error) {
+		out := make([]File, 0, len(clusters2016))
+		for _, c := range clusters2016 {
+			out = append(out, File{
+				Name: c.name,
+				URL:  baseURL + uploadPath + c.file,
+				Dest: c.file,
+			})
+		}
+		return out, nil
+	},
+}
+
+type cluster struct{ name, file string }
+
+// clusters2016 is in the article's publication order. That order is incidental —
+// go-parser re-sorts by filename — but it is kept as published so the list can
+// be diffed against the source article.
+var clusters2016 = []cluster{
+	{"Trường Đại học Bách khoa Hà Nội", "b34c777942ca8a4de91f23a35cce1c6bdhbachkhoahn-1468901491486.xlsx"},
+	{"Trường Đại học Sư phạm Hà Nội", "76f19098881e9c79f66ac4f1b42a8a55dhsuphamhanoi-1468901144203.xlsx"},
+	{"Trường Đại học Thuỷ lợi * Cơ sở 1 ở phía Bắc", "4a3fa95a83fe9b3b2fafab309323fd22dhthuyloi-1468901840919.xlsx"},
+	{"Học viện Kỹ thuật Quân sự * Cơ sở 1 ở phía Bắc (Quân đội)", "4334ea476a2b8196d54ba40a341831echvkythuatquansu-1468920342994.xlsx"},
+	{"Trường Đại học Lâm nghiệp", "c5ecc58a731fd1a54b7e6cc3bc99449ddhlamnghiep-1468939300790.xlsx"},
+	{"Trường Đại học Bách khoa - Đại học Quốc gia Thành phố Hồ Chí Minh", "357c6aaf59aeaa6c11c3e83c595d38cfdhbachkhoa-tphcm-1468939678829.xlsx"},
+	{"Trường Đại học Khoa học tự nhiên - Đại học Quốc gia Thành phố Hồ Chí Minh", "64e44466ffeff9a587260795644805d0dhkhtunhien-dhqgtphcm-1468921015146.xlsx"},
+	{"Trường Đại học Khoa học xã hội và Nhân văn - Đại học Quốc gia Thành phố Hồ Chí Minh", "5fd13ae5287cbf0149bd733b507e0608dhkhxhnv-dhqgtphcm-1468920630193.xlsx"},
+	{"Trường Đại học Sư phạm Tp.HCM", "a995de3f7a7055cf10b2f002f0e194f7dhsptphcm-1468920792291.xlsx"},
+	{"Trường Đại học Hàng Hải", "c6ef16bc9f75ce0fd16c1229afad4e71dhhanghai-1468906457479.xlsx"},
+	{"Trường Đại học Sư phạm - Đại học Thái Nguyên", "ee09c723da4e86cdbe26213413450af2dhsupham-dhthainguyen-1468939061872.xlsx"},
+	{"Trường Đại học Kỹ thuật công nghiệp - Đại học Thái Nguyên", "8af8f1b29b6a40e26ccb4a2e19e8eededhkythuatcongnghiep-dhthainguyen-1468901919568.xlsx"},
+	{"Trường Đại học Nông lâm - Đại học Thái Nguyên", "53e5e20ebd74d84741513b703fbd13bfdhnonglam-dhthainguyen-1468939023637.xlsx"},
+	{"Học viện Ngân hàng", "f2846894bd5533e02b122a1e3b4333f6hvnganhang-1468939472850.xlsx"},
+	{"Trường Đại học Luật Hà Nội", "9e0e564ae801e347bd7649da8b89ad90dhluathn-1468920377978.xlsx"},
+	{"Trường Đại học Tân Trào", "20db5eaace927596f1806f151649ee6edhtantrao-1468901296040.xlsx"},
+	{"Trường Đại học Xây dựng Hà Nội", "7b5d74564ab9f6e0e137f6d161fab6f0dhxaydung-1468940561682.xlsx"},
+	{"Trường Đại học Khoa học - Đại học Thái Nguyên", "c2cfaaf287fb1e7120894e545771b027dhkhoahoc-dhthainguyen-1468900365545.xlsx"},
+	{"Đại học Thái Nguyên", "3daa6ee33fdafadc3c112c18bff18c2ddhthainguyen-1468940268065.xlsx"},
+	{"Học viện Tài chính", "96c95c85c5549a06f2e92ffcacf8ff21hvtaichinh-1468934470924.xlsx"},
+	{"Trường Đại học Tây Bắc", "74e6cd0dc78a3fb7dcf56ae57283b9b9dhtaybac-1468940302052.xlsx"},
+	{"Trường Đại học Hùng Vương", "778ab074ed6a95d209fa901fd5d68843dhhungvuong-1468901964127.xlsx"},
+	{"Trường Đại học Sư phạm Hà Nội 2", "a11e707e6ed5fe4d29a8e32380d7558cdhsphanoi-2-1468902006284.xlsx"},
+	{"Trường Đại học Ngoại thương * Cơ sở 1 ở phía Bắc", "ec84067268f47d523b1e80b493248377dhngoaithuong-1468920480753.xlsx"},
+	{"Trường Đại học Kinh tế Quốc dân", "8f948c2a5ab49a3c5fd03b6ea04887a7dhkinhtequocdanfix-1468902083782.xlsx"},
+	{"Trường Đại học Giao thông Vận tải", "51bb2f9b6e9f565c2d705606fd56958bdhgiaothongvantai-1468939107463.xlsx"},
+	{"Học viện Nông Nghiệp Việt Nam", "5aede4951e7af8c1e1056c747a553386hvnongnghiepvn-1468900610577.xlsx"},
+	{"Trường Đại học Sư phạm Kỹ thuật Hưng Yên", "f85f04bb2bbd84457b83ac846728a2a6dhspkythuathungyen-1468920667091.xlsx"},
+	{"Trường Đại học Hải Phòng", "deca197a916a8f633f6894b4e13c3ca4dhhaiphong-1468901178368.xlsx"},
+	{"Trường Đại học Thương mại", "b2552e20c9ff4b237244dfbf2e278a47dhthuongmai-1468902876264.xlsx"},
+	{"Trường Đại học Công nghiệp Hà Nội", "82338a4a93ea1e0cde26e0061066ca8adhcongnghiephanoi-1468920154757.xlsx"},
+	{"Y Dược Thái Bình", "22e8617886412abd90ae3d33cc6d7db4dhyduocthaibinh-1468941020145.xlsx"},
+	{"Trường Đại học Mỏ Địa chất", "c174631d90303276ac63ff38b31157bbdhmodiachat-1468900697349.xlsx"},
+	{"Trường Đại học Hồng Đức", "bd857add0cfe92c1c247469ff2104527dh-hong-duc-1468900427119.xlsx"},
+	{"Trường Đại học Vinh", "64d2635efd30a0d0db4f7c30b9735261dhvinh-1468940230027.xlsx"},
+	{"Trường ĐH Sư phạm - ĐH Huế", "ab2f259ef55cb08b3a436d3f15d42cc7dhsupham-dhhue-1468934431828.xlsx"},
+	{"Trường ĐH Khoa học - ĐH Huế", "87ffaf3197ba20f0919f98a0b3f33914dhkhoahoc-dhhue-1468938729472.xlsx"},
+	{"Trường ĐH Kinh tế - ĐH Huế", "8916da7b3c4d58cffc1dd29dc7a8aabfdhkinhte-dhhue-1468938560205.xlsx"},
+	{"Đại học Huế", "919276a0348634459f8937dfcd6c7129dhhue-1468938792115.xlsx"},
+	{"ĐH Đà Nẵng", "b42fd58ead4c54279d7df5f53160a851dhdanang-1468920187203.xlsx"},
+	{"Trường ĐH Bách khoa - ĐH Đà Nẵng", "43213f61074068d9177685eb6c3f2d3bdhbachkhoa-dhdanang-1468938494437.xlsx"},
+	{"Trường ĐH Sư phạm - ĐH Đà Nẵng", "54bbcf865c137d58cf62af15733bd376dhsupham-dhdanang-1468938526558.xlsx"},
+	{"Trường Đại học Quy Nhơn", "70bf8cd2b6e87523219e55149a8524c5dhquynhon-1468938878328.xlsx"},
+	{"Trường ĐH Xây dựng miền Trung", "4e9bb192d56c9fa92b3e0ec13d8236c8dhxaydungmientrung-1468940978346.xlsx"},
+	{"Trường ĐH Nông Lâm TPHCM", "b2a739584a8861502573d1c6372fd6e9dhnonglamtphcm-1468939555326.xlsx"},
+	{"Trường ĐH Ngoại ngữ - ĐH Đà Nẵng", "9ba098a7913ad837a66e173b3fc41fd7dhngoaingu-dhdanang-1468902750424.xlsx"},
+	{"Trường ĐH Tây Nguyên", "216ee4e385dbc236767be899b068bf98dhtaynguyen-1468940490240.xlsx"},
+	{"Trường ĐH Tài chính Marketing", "af6828f837a80e03d77a65f3d310ec60dh-tai-chinh-marketing-1468924711596.xlsx"},
+	{"Trường ĐH Nha Trang", "6375fd13557bee08d20f1f3e46dac130dhnhatrang-coso1-1468901350425.xlsx"},
+	{"Trường ĐH Giao thông vận tải TP.HCM", "c5c28959fcd0b4e39dd9fd538db5448adhgtvttphcm-1468939150821.xlsx"},
+	{"Trường ĐH Sư phạm Kĩ thuật TP.HCM", "514877c1cfebdc9c2251f651b5f02c7edhspkythuattphcm-1468906568655.xlsx"},
+	{"Trường ĐH Đà Lạt", "2b33640b0dd154490f2809d96be81797dhdalat-1468902130616.xlsx"},
+	{"Trường ĐH Kinh tế TP.HCM", "f0aa5a19c2c8c1fde8211341f0e98e26dhkinhtetphcm-1468939229043.xlsx"},
+	{"Trường ĐH Kinh tế - Luật - ĐHQG TP.HCM", "ff877788e43bd84f0119dae026996892dhkinhteluat-dhqgtphcm-1468939860894.xlsx"},
+	{"Trường ĐH Công nghiệp Thực phẩm TP.HCM", "167116af6c8a096c2a9307edfa115f77dhcnthucpham-tphcm-1468910578187.xls"},
+	{"Trường ĐH Công nghiệp TP.HCM", "487c7287c1c9735722b31d877942becddhcongnghieptphcm-1468983373303.xlsx"},
+	{"Trường ĐH Sài Gòn", "d75282bff0a4d8cd1421c3fb74f789c3dhsaigon-1468900867561.xlsx"},
+	{"Trường ĐH Đồng Tháp", "388fcdfdac18ef6eee6125a2a015714ddhdongthap-1468934545359.xlsx"},
+	{"Trường Đại học An Giang", "fc511852659150f6305f6f624445300ddhangiang-1468940189321.xlsx"},
+	{"Trường ĐH Tôn Đức Thắng", "9d78f6dce7965fb4b7a337279920d4b1dhtonducthang-1468902168623.xlsx"},
+	{"Trường ĐH Tiền Giang", "bacd1b2133b37d0ea9fbf46445265036dhtiengiang-1468940391597.xlsx"},
+	{"Trường ĐH Cần Thơ", "b35a3f26a162bec3740e2da5639327afdhcantho-1468907131146.xls"},
+	{"Trường ĐH Cần Thơ - Hậu Giang", "7f1cbe7f35d1f539ca817a0ab8bd00c2dhcantho-haugiang-1468907167921.xls"},
+	{"Trường ĐH Luật TP.HCM", "d652d24e8bd4fe6a48d634b116f37c27dhluattphcm-1468939398928.xlsx"},
+	{"Trường ĐH Sư phạm Kỹ thuật Vĩnh Long", "4c2fd846ed2a0670ba6d1200d1d68e1dspkythuatvinhlong-1468902640438.xlsx"},
+	{"Trường ĐH Trà Vinh", "7418e83ff9d07b146216a3beafa927bbdhtravinh-1468902716174.xlsx"},
+	{"Trường ĐH Ngân hàng TP.HCM", "86a692bebf9b0b7c58adb9876b9befd5dhnganhangtphcm-1468902244566.xlsx"},
+	{"Trường ĐH Cần Thơ - Bạc Liêu", "5ca241aa1de8d25a3e6ff6db75234025dhcantho-baclieu-1468907071460.xls"},
+	{"Trường ĐH Kiên Giang", "cdc912664d7217c75eb9b67c6ef3797adhkiengiang-1468901222446.xlsx"},
+	{"Trường ĐH Y Dược Cần Thơ", "023718c7d3cf7ace3a7116fabb12bd9cdhyduoccantho-1468920829104.xlsx"},
+	{"Sở GDĐT Hà Nội", "1c66db0265bf45df237b37a7bcaeaef3hanoi-1468932790746.xlsx"},
+	{"Sở GDĐT Hà Giang", "178814c33c8b85a77d0fc295fa6d2be1hagiang-1468932847230.xlsx"},
+	{"Sở GDĐT Cao Bằng", "ddb631f833af894200ef10cd408d2ef0caobang-1468932877769.xlsx"},
+	{"Sở GDĐT Lai Châu", "a3aa106104f114c6426deb2d5e5da5fdlaichau-1468932954079.xlsx"},
+	{"Sở GDĐT Lào Cai", "d035f470c13adbf5785797c6ccdcb204laocai-1468919805337.xlsx"},
+	{"xem TẠI ĐÂY ) 07 Sở GDĐT Tuyên Quang", "d5ecd2574b5ab36c62dffad435262cb8tuyenquang-1468902296807.xlsx"},
+	{"Sở GDĐT Lạng Sơn", "7afd5d7fdff7c170a6342cd8f9474b22langson-1468932984587.xlsx"},
+	{"Sở GDĐT Bắc Kạn", "17843bccde10f362c2b9600bc49cd3f8backan-1468933016076.xlsx"},
+	{"Sở GDĐT Thái Nguyên", "5fb3a1b66aff077ee3da482be08cd391thainguyen-1468902797020.xlsx"},
+	{"Sở GDĐT Yên Bái", "97aa741a4c2b1276559eb021779daa32yenbai-1468933116055.xlsx"},
+	{"Sở GDĐT Sơn La", "4b98fdd7300b86a94636e4d15394fc19sonla-1468899319124.xlsx"},
+	{"Sở GDĐT Phú Thọ", "5465cca59495dcd9a48f689e2d38b85ephutho-1468902409682.xlsx"},
+	{"xem TẠI ĐÂY ) 14 Sở GDĐT Vĩnh Phúc", "4e0ccee19eb64513d23a6dff96ff7f9cvinhphuc-1468919874814.xlsx"},
+	{"Sở GDĐT Quảng Ninh", "a31b3446bae5a17c517348d8c38c5b7equangninh-1468919920197.xlsx"},
+	{"Sở GDĐT Bắc Giang", "b7a0c1ecce7c8445b3a1021091a5e654bacgiang-1468983153932.xlsx"},
+	{"Sở GDĐT Bắc Ninh", "7e0d0b6fb981d407cffdb79d724e8aecbacninh-1468919984998.xlsx"},
+	{"Sở GDĐT Hải Dương", "7cfc91bd6a1f5ae9d7917e920f469b90haiduong-1468899640007.xlsx"},
+	{"Sở GDĐT Hưng Yên", "92045214159cb0335b2836aa1c4bb25ahungyen-1468899732002.xlsx"},
+	{"Sở GDĐT Hoà Bình", "505a876dfd1ea321291e29192aebe95ehoabinh-1468983220125.xlsx"},
+	{"Sở GDĐT Hà Nam", "e693bc9de67acc916032c6a3b2fc9eedhanam-1468933146526.xlsx"},
+	{"Sở GDĐT Nam Định", "dfa63c5b68a8f638379ca0e9c5a1c556namdinh-1468933234061.xlsx"},
+	{"Sở GDĐT Thái Bình", "6960b6fa493e1365a183af461e945389thaibinh-1468920019273.xlsx"},
+	{"xem TẠI ĐÂY ) 24 Sở GDĐT Ninh Bình", "0b583c9875443e65ffa449d5ca76fae8ninhbinh-1468899778557.xlsx"},
+	{"xem TẠI ĐÂY ) 25 Sở GDĐT Thanh Hoá", "238adc19e80c0daef089a307af23edcethanhhoa-1468933304999.xlsx"},
+	{"Sở GDĐT Nghệ An", "7f12b90225e91f9d00aac3bcdf9974e1nghean-1468920043121.xlsx"},
+	{"Sở GDĐT Quảng Bình", "9c730c43bf4309171528c3853ed53b19quangbinh-1468933376032.xlsx"},
+	{"Sở GDĐT Quảng Trị", "2208903e1fa12220d100f93e40d5ff3bquangtri-1468933440097.xlsx"},
+	{"Sở GDĐT Thừa Thiên -Huế", "6badc7cfa258b59ad67e2b64f713b95fthuathienhue-1468983258438.xlsx"},
+	{"Sở GDĐT Quảng Nam", "ec765c32190773f5a18cd8202a5eaf0dquangnam-1468933501467.xlsx"},
+	{"Sở GDĐT Quảng Ngãi", "e80dc028d53fa49ee247391a0653dee5quangngai-1468933592400.xlsx"},
+	{"Sở GDĐT Kon Tum", "cc57dc986acb525086b86e85df54949dkontum-1468933632113.xlsx"},
+	{"Sở GDĐT Bình Định", "412e62265f35ca0078a0d94af5d1ed8abinhdinh-1468899884192.xlsx"},
+	{"Sở GDĐT Gia Lai", "698b930ab7697a0672bbc39168024c9cgialai-1468933753098.xlsx"},
+	{"Sở GDĐT Đắk Lắk", "2004a8d225fd87524e324d2d519e87a4daclac-1468899960505.xlsx"},
+	{"Sở GDĐT Khánh Hoà", "4ad416ccf014bd5d01f415238566b210khanhhoa-1468933789968.xlsx"},
+	{"Sở GDĐT Lâm Đồng", "2fa6d9a308bd11bc09f02d787eef2c15lamdong-1468933821082.xlsx"},
+	{"Sở GDĐT Ninh Thuận", "15c85d901a73dd49f2bae71aadcdbb5cninhthuan-1468920071763.xlsx"},
+	{"Sở GDĐT Đồng Nai", "538c9676b70f2509573245d58fba93bcdongnai-1468938241935.xlsx"},
+	{"Sở GDĐT Đồng Tháp", "575df8a4df74e87abb8f55d807324374dongthap-1468933902699.xlsx"},
+	{"Sở GDĐT Kiên Giang", "7435bb054e8e4ec34c173af1b6ad8402kiengiang-1468900147395.xlsx"},
+	{"Sở GDĐT Cần Thơ", "5b16d7156a03bb7ed0618796bae87bc3cantho-1468933954714.xlsx"},
+	{"Sở GDĐT Bến Tre", "74145979ee30186bf4e40ee3e6a3ac74bentre-1468934010040.xlsx"},
+	{"Sở GDĐT Vĩnh Long", "7f4362ef567cdf74495368e0fc11072avinhlong-1468934067981.xlsx"},
+	{"Sở GDĐT Trà Vinh", "08cdeb3636dcb2adaef829d62968274atravinh-1468902443859.xlsx"},
+	{"Sở GDĐT Sóc Trăng", "a8ff80d9478747e36d21ec95cc9ecb8fsoctrang-1468902686409.xlsx"},
+	{"Sở GDĐT Bạc Liêu", "18055d17e46854ecc6c8be585bcb6e5cbaclieu-1468934144150.xlsx"},
+	{"Sở GDĐT Điện Biên", "3c0e56abd8c5334f4e1ff7df3ac23a15dienbien-1468920113820.xlsx"},
+	{"Sở GDĐT Đăk Nông", "bf259a908ff8a8dcc067a761204459f4daknong-1468934173803.xlsx"},
+	{"Sở GDĐT Hậu Giang", "964c4368131f6be6058fabb1caefba1chaugiang-1468934247324.xlsx"}}
diff --git a/crawler/internal/sources/source_2017.go b/crawler/internal/sources/source_2017.go
new file mode 100644
index 0000000..c3a8738
--- /dev/null
+++ b/crawler/internal/sources/source_2017.go
@@ -0,0 +1,121 @@
+package sources
+
+import (
+	"regexp"
+	"strings"
+)
+
+// article is the page the links were taken from. The CDN rejects requests that
+// do not carry it as a Referer.
+const article = "https://baotintuc.vn/tuyen-sinh/tra-cuu-diem-thi-thpt-2017-cua-63-tinh-thanh-pho-tren-baotintucvn-20170706073512672.htm"
+
+// browserUA is required too: the CDN 403s an unrecognised User-Agent.
+const browserUA = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Safari/537.36"
+
+// source2017 fetches the 2017 dataset from the baotintuc.vn CDN: one .xls per
+// province.
+//
+// This is the only dataset that can be re-fetched from its original source.
+// The CDN filenames are inconsistent (Angiang.xls, 1BaRiaVungTau.xls,
+// 23HaiPhong.xls, gia-lai.xls), so local names are derived from the province
+// name instead — which is what produced the data/2017/ files currently on disk,
+// and what any re-crawl must keep producing.
+var source2017 = Source{
+	ID:      "2017",
+	Summary: "63 province .xls files from the baotintuc.vn CDN",
+	Headers: map[string]string{
+		"User-Agent": browserUA,
+		"Referer":    article,
+	},
+	Files: func() ([]File, error) {
+		out := make([]File, 0, len(provinces2017))
+		for _, p := range provinces2017 {
+			out = append(out, File{
+				Name: p.name,
+				URL:  p.url,
+				Dest: slug(p.name) + ".xls",
+			})
+		}
+		return out, nil
+	},
+}
+
+var nonAlphanumeric = regexp.MustCompile(`[^a-z0-9]+`)
+
+// slug converts a province name to its local filename stem: lowercase, with
+// every run of non-alphanumeric characters collapsed to a single hyphen.
+//
+//	"Ba Ria - Vung Tau" -> "ba-ria-vung-tau"
+//
+// The names are already unaccented ASCII, so no transliteration is involved.
+func slug(name string) string {
+	return strings.Trim(nonAlphanumeric.ReplaceAllString(strings.ToLower(name), "-"), "-")
+}
+
+type province struct{ name, url string }
+
+var provinces2017 = []province{
+	{"An Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/17/57/Angiang.xls"},
+	{"Bac Lieu", "https://cdnmedia.baotintuc.vn/2017/07/06/17/59/Baclieu.xls"},
+	{"Ba Ria - Vung Tau", "https://cdnmedia.baotintuc.vn/2017/07/06/08/17/1BaRiaVungTau.xls"},
+	{"Bac Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/09/33/BacGiang.xls"},
+	{"Bac Kan", "https://cdnmedia.baotintuc.vn/2017/07/06/08/27/BacKan.xls"},
+	{"Bac Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/33/BacNinh.xls"},
+	{"Ben Tre", "https://cdnmedia.baotintuc.vn/2017/07/06/09/34/BenTre.xls"},
+	{"Binh Duong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/34/BinhDuong.xls"},
+	{"Binh Thuan", "https://cdnmedia.baotintuc.vn/2017/07/06/09/35/BinhThuan.xls"},
+	{"Binh Phuoc", "https://cdnmedia.baotintuc.vn/2017/07/06/09/09/BinhPhuoc.xls"},
+	{"Binh Dinh", "https://cdnmedia.baotintuc.vn/2017/07/06/13/05/Binhdinh.xls"},
+	{"Ca Mau", "https://cdnmedia.baotintuc.vn/2017/07/06/13/05/Camau.xls"},
+	{"Cao Bang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/02/Caobang.xls"},
+	{"Can Tho", "https://cdnmedia.baotintuc.vn/2017/07/06/13/07/Cantho.xls"},
+	{"Da Nang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/03/Danang.xls"},
+	{"Dak Nong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/35/DakNong.xls"},
+	{"Dak Lak", "https://cdnmedia.baotintuc.vn/2017/07/06/18/02/Daklak.xls"},
+	{"Dong Nai", "https://cdnmedia.baotintuc.vn/2017/07/06/13/08/dongnai.xls"},
+	{"Dong Thap", "https://cdnmedia.baotintuc.vn/2017/07/06/17/59/Dongthap.xls"},
+	{"Dien Bien", "https://cdnmedia.baotintuc.vn/2017/07/06/09/08/DienBien.xls"},
+	{"Gia Lai", "https://cdnmedia.baotintuc.vn/2017/07/06/13/09/Gia-Lai.xls"},
+	{"Ha Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/18/Hagiang.xls"},
+	{"Ha Noi", "https://cdnmedia.baotintuc.vn/2017/07/07/08/16/HaNoi.xls"},
+	{"Ha Nam", "https://cdnmedia.baotintuc.vn/2017/07/06/09/07/Hanam.xls"},
+	{"Ha Tinh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/36/HaTinh.xls"},
+	{"Hai Phong", "https://cdnmedia.baotintuc.vn/2017/07/06/08/18/23HaiPhong.xls"},
+	{"Hai Duong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/36/HaiDuong.xls"},
+	{"Hau Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/00/Haugiang.xls"},
+	{"Ho Chi Minh", "https://cdnmedia.baotintuc.vn/2017/07/06/08/26/HCM.xls"},
+	{"Hoa Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/06/HoaBinh.xls"},
+	{"Hung Yen", "https://cdnmedia.baotintuc.vn/2017/07/06/08/25/HungYen.xls"},
+	{"Khanh Hoa", "https://cdnmedia.baotintuc.vn/2017/07/06/18/04/Khanhhoa.xls"},
+	{"Kien Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/08/28/KienGiang.xls"},
+	{"Kon Tum", "https://cdnmedia.baotintuc.vn/2017/07/06/18/05/KonTum.xls"},
+	{"Nam Dinh", "https://cdnmedia.baotintuc.vn/2017/07/06/07/55/13NamDinh.xls"},
+	{"Nghe An", "https://cdnmedia.baotintuc.vn/2017/07/06/07/55/14NgheAn.xls"},
+	{"Ninh Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/37/NinhBinh.xls"},
+	{"Ninh Thuan", "https://cdnmedia.baotintuc.vn/2017/07/06/18/00/Ninhthuan.xls"},
+	{"Lao Cai", "https://cdnmedia.baotintuc.vn/2017/07/06/08/52/11LaoCai.xls"},
+	{"Lai Chau", "https://cdnmedia.baotintuc.vn/2017/07/06/13/33/LaiChau.xls"},
+	{"Lang Son", "https://cdnmedia.baotintuc.vn/2017/07/06/18/05/Langson.xls"},
+	{"Lam Dong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/00/LamDong.xls"},
+	{"Long An", "https://cdnmedia.baotintuc.vn/2017/07/06/08/53/12LongAn.xls"},
+	{"Quang Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/23/QuangBinh.xls"},
+	{"Quang Nam", "https://cdnmedia.baotintuc.vn/2017/07/06/09/24/QuangNam.xls"},
+	{"Quang Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/18/06/Quangninh.xls"},
+	{"Quang Ngai", "https://cdnmedia.baotintuc.vn/2017/07/06/09/24/QuangNgai.xls"},
+	{"Quang Tri", "https://cdnmedia.baotintuc.vn/2017/07/06/09/25/QuangTri.xls"},
+	{"Phu Tho", "https://cdnmedia.baotintuc.vn/2017/07/06/09/23/PhuTho.xls"},
+	{"Phu Yen", "https://cdnmedia.baotintuc.vn/2017/07/06/09/38/PhuYen.xls"},
+	{"Son La", "https://cdnmedia.baotintuc.vn/2017/07/06/18/01/Sonla.xls"},
+	{"Soc Trang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/06/Soctrang.xls"},
+	{"Tay Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/26/TayNinh.xls"},
+	{"Thai Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/09/ThaiBinh.xls"},
+	{"Thai Nguyen", "https://cdnmedia.baotintuc.vn/2017/07/06/13/35/ThaiNguyen.xls"},
+	{"Thanh Hoa", "https://cdnmedia.baotintuc.vn/2017/07/06/13/10/thanhoa.xls"},
+	{"Tra Vinh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/38/TraVinh.xls"},
+	{"Thua Thien Hue", "https://cdnmedia.baotintuc.vn/2017/07/06/13/10/thuathienhue.xls"},
+	{"Tien Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/13/11/tiengiang.xls"},
+	{"Tuyen Quang", "https://cdnmedia.baotintuc.vn/2017/07/06/09/39/TuyenQuang.xls"},
+	{"Vinh Phuc", "https://cdnmedia.baotintuc.vn/2017/07/06/09/40/VinhPhuc.xls"},
+	{"Vinh Long", "https://cdnmedia.baotintuc.vn/2017/07/06/13/13/vinhlong.xls"},
+	{"Yen Bai", "https://cdnmedia.baotintuc.vn/2017/07/06/18/07/Yenbai.xls"},
+}
diff --git a/crawler/internal/sources/sources.go b/crawler/internal/sources/sources.go
new file mode 100644
index 0000000..0efafc7
--- /dev/null
+++ b/crawler/internal/sources/sources.go
@@ -0,0 +1,53 @@
+// Package sources describes where each dataset's spreadsheets come from.
+//
+// A source is a dataset id plus the list of remote files that fill it.
+// Downloading is internal/fetch's job; deciding what to download and what to
+// call it locally is this package's.
+package sources
+
+import "fmt"
+
+// File is one remote spreadsheet.
+type File struct {
+	// Name is the human label used in progress output, e.g. "An Giang".
+	Name string
+	URL  string
+	// Dest is the filename within the dataset directory.
+	//
+	// This is not cosmetic. go-parser sorts its input files and inserts with
+	// INSERT OR REPLACE, which is last-wins, so filenames decide which row
+	// survives a duplicate exam number. A re-crawl that names files differently
+	// can produce a database with the same row count and different content.
+	Dest string
+}
+
+// Source is one crawlable dataset.
+type Source struct {
+	// ID is the dataset id: the subcommand, the directory under data/, and the
+	// go-parser config name, all at once. Keeping it single means a source
+	// cannot be pointed at the wrong dataset's directory.
+	ID      string
+	Summary string
+	// Headers are sent with every request for this source.
+	Headers map[string]string
+	// Files lists what to download. It returns an error rather than an empty
+	// list when a source is known but not yet configured, so "nothing to do"
+	// can never be mistaken for success.
+	Files func() ([]File, error)
+}
+
+// registry is ordered: it drives the help output.
+var registry = []Source{source2016, source2017}
+
+// All returns every known source, in help order.
+func All() []Source { return registry }
+
+// Lookup finds a source by ID.
+func Lookup(id string) (Source, error) {
+	for _, s := range registry {
+		if s.ID == id {
+			return s, nil
+		}
+	}
+	return Source{}, fmt.Errorf("unknown dataset %q", id)
+}
diff --git a/crawler/internal/sources/sources_test.go b/crawler/internal/sources/sources_test.go
new file mode 100644
index 0000000..dca1379
--- /dev/null
+++ b/crawler/internal/sources/sources_test.go
@@ -0,0 +1,156 @@
+package sources
+
+import (
+	"os"
+	"path/filepath"
+	"strings"
+	"testing"
+)
+
+func TestSlug(t *testing.T) {
+	for in, want := range map[string]string{
+		"An Giang":          "an-giang",
+		"Ba Ria - Vung Tau": "ba-ria-vung-tau",
+		"Ho Chi Minh":       "ho-chi-minh",
+		"Thua Thien Hue":    "thua-thien-hue",
+	} {
+		if got := slug(in); got != want {
+			t.Errorf("slug(%q) = %q, want %q", in, got, want)
+		}
+	}
+}
+
+func TestSource2017IsComplete(t *testing.T) {
+	files, err := source2017.Files()
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(files) != 63 {
+		t.Errorf("got %d provinces, want 63", len(files))
+	}
+
+	// A duplicate Dest would silently overwrite one province with another and
+	// leave the dataset one file short, with no error anywhere.
+	seenDest := map[string]string{}
+	seenURL := map[string]string{}
+	for _, f := range files {
+		if prev, dup := seenDest[f.Dest]; dup {
+			t.Errorf("%q and %q both write to %s", prev, f.Name, f.Dest)
+		}
+		seenDest[f.Dest] = f.Name
+		if prev, dup := seenURL[f.URL]; dup {
+			t.Errorf("%q and %q share a URL: %s", prev, f.Name, f.URL)
+		}
+		seenURL[f.URL] = f.Name
+
+		if !strings.HasPrefix(f.URL, "https://") {
+			t.Errorf("%s: URL is not https: %s", f.Name, f.URL)
+		}
+		if !strings.HasSuffix(f.Dest, ".xls") {
+			t.Errorf("%s: Dest %q does not end in .xls", f.Name, f.Dest)
+		}
+	}
+}
+
+// TestMatchesFilesOnDisk is the guard that matters.
+//
+// go-parser sorts its inputs and inserts last-wins, so filenames decide which
+// row survives a duplicate exam number. If a re-crawl started naming files
+// differently, the rebuilt database could differ in content while still passing
+// the row-count guard in build-db.js. The committed data/<id>/ is the oracle:
+// whatever a source would write must match it exactly, in both directions.
+//
+// For 2016 this is also the only evidence that the recovered link list is the
+// right one — the host that served those files is gone, so the list cannot be
+// checked by fetching it.
+func TestMatchesFilesOnDisk(t *testing.T) {
+	for _, src := range All() {
+		t.Run(src.ID, func(t *testing.T) {
+			dir := filepath.Join("..", "..", "..", "data", src.ID)
+			entries, err := os.ReadDir(dir)
+			if os.IsNotExist(err) {
+				t.Skipf("%s not present", dir)
+			}
+			if err != nil {
+				t.Fatal(err)
+			}
+
+			onDisk := map[string]bool{}
+			for _, e := range entries {
+				if !e.IsDir() {
+					onDisk[e.Name()] = true
+				}
+			}
+
+			files, err := src.Files()
+			if err != nil {
+				t.Fatal(err)
+			}
+			for _, f := range files {
+				if !onDisk[f.Dest] {
+					t.Errorf("crawler would write %q, which is not in %s — "+
+						"a renamed input can change which duplicate row survives", f.Dest, dir)
+				}
+				delete(onDisk, f.Dest)
+			}
+			for name := range onDisk {
+				t.Errorf("%s/%s exists but no source produces it", dir, name)
+			}
+		})
+	}
+}
+
+func TestLookup(t *testing.T) {
+	for _, s := range All() {
+		got, err := Lookup(s.ID)
+		if err != nil {
+			t.Errorf("Lookup(%q): %v", s.ID, err)
+		}
+		if got.ID != s.ID || got.Summary == "" {
+			t.Errorf("Lookup(%q) returned %+v", s.ID, got)
+		}
+	}
+	if _, err := Lookup("nope"); err == nil {
+		t.Error("Lookup of an unknown source must fail")
+	}
+}
+
+// TestIDsAreDatasetIDs: the ID doubles as the directory under data/, so a typo
+// would send a crawl into a new directory the parser never reads.
+func TestIDsAreDatasetIDs(t *testing.T) {
+	for _, s := range All() {
+		dir := filepath.Join("..", "..", "..", "data", s.ID)
+		if _, err := os.Stat(dir); os.IsNotExist(err) {
+			t.Errorf("source %q has no dataset directory at %s", s.ID, dir)
+		}
+	}
+}
+
+// TestSource2016IsComplete: 4 .xls + 115 .xlsx, matching what the source
+// article published and what data/2016/ holds.
+func TestSource2016IsComplete(t *testing.T) {
+	files, err := source2016.Files()
+	if err != nil {
+		t.Fatal(err)
+	}
+	if len(files) != 119 {
+		t.Errorf("got %d clusters, want 119", len(files))
+	}
+	var xls, xlsx int
+	for _, f := range files {
+		switch filepath.Ext(f.Dest) {
+		case ".xls":
+			xls++
+		case ".xlsx":
+			xlsx++
+		default:
+			t.Errorf("%s: unexpected extension in %q", f.Name, f.Dest)
+		}
+		if !strings.HasPrefix(f.URL, baseURL+uploadPath) {
+			t.Errorf("%s: URL does not use the configured base: %s", f.Name, f.URL)
+		}
+	}
+	if xls != 4 || xlsx != 115 {
+		t.Errorf("got %d .xls and %d .xlsx, want 4 and 115", xls, xlsx)
+	}
+}
diff --git a/data/2017-old/10_BinhThuan_RIIW.xls.xlsx b/data/2017-old/10_BinhThuan_RIIW.xls.xlsx
deleted file mode 100644
index 8ecaccd..0000000
Binary files a/data/2017-old/10_BinhThuan_RIIW.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/10_Ca_Mau_BKXT.xls.xlsx b/data/2017-old/10_Ca_Mau_BKXT.xls.xlsx
deleted file mode 100644
index 8fe2b4a..0000000
Binary files a/data/2017-old/10_Ca_Mau_BKXT.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/10_LamDong_GNFT.xls.xlsx b/data/2017-old/10_LamDong_GNFT.xls.xlsx
deleted file mode 100644
index 54b89a4..0000000
Binary files a/data/2017-old/10_LamDong_GNFT.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/10_Soc_Trang_XCGJ.xls.xlsx b/data/2017-old/10_Soc_Trang_XCGJ.xls.xlsx
deleted file mode 100644
index 70a0ff8..0000000
Binary files a/data/2017-old/10_Soc_Trang_XCGJ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/11_BinhDuong_RYQL.xls.xlsx b/data/2017-old/11_BinhDuong_RYQL.xls.xlsx
deleted file mode 100644
index ccb2951..0000000
Binary files a/data/2017-old/11_BinhDuong_RYQL.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/11_LaoCai_GMSU.xls.xlsx b/data/2017-old/11_LaoCai_GMSU.xls.xlsx
deleted file mode 100644
index 33ed6f0..0000000
Binary files a/data/2017-old/11_LaoCai_GMSU.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/12_BenTre_DKWF.xls.xlsx b/data/2017-old/12_BenTre_DKWF.xls.xlsx
deleted file mode 100644
index 8a7cddb..0000000
Binary files a/data/2017-old/12_BenTre_DKWF.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/12_LongAn_ZZUK.xls.xlsx b/data/2017-old/12_LongAn_ZZUK.xls.xlsx
deleted file mode 100644
index 8f11d78..0000000
Binary files a/data/2017-old/12_LongAn_ZZUK.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/13_NamDinh_ESEL.xls.xlsx b/data/2017-old/13_NamDinh_ESEL.xls.xlsx
deleted file mode 100644
index b296577..0000000
Binary files a/data/2017-old/13_NamDinh_ESEL.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/13_TraVinh_LKUJ.xls.xlsx b/data/2017-old/13_TraVinh_LKUJ.xls.xlsx
deleted file mode 100644
index 340c20e..0000000
Binary files a/data/2017-old/13_TraVinh_LKUJ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/14_NgheAn_BSLY.xls.xlsx b/data/2017-old/14_NgheAn_BSLY.xls.xlsx
deleted file mode 100644
index b1f2891..0000000
Binary files a/data/2017-old/14_NgheAn_BSLY.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/15_PhuTho_ABWQ.xls.xlsx b/data/2017-old/15_PhuTho_ABWQ.xls.xlsx
deleted file mode 100644
index 5135dac..0000000
Binary files a/data/2017-old/15_PhuTho_ABWQ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/16_QuangBinh_KGEU.xls.xlsx b/data/2017-old/16_QuangBinh_KGEU.xls.xlsx
deleted file mode 100644
index c12f598..0000000
Binary files a/data/2017-old/16_QuangBinh_KGEU.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/17_QuangNam_AMTK.xls.xlsx b/data/2017-old/17_QuangNam_AMTK.xls.xlsx
deleted file mode 100644
index 761a8df..0000000
Binary files a/data/2017-old/17_QuangNam_AMTK.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/18_QuangNgai_KOFP.xls.xlsx b/data/2017-old/18_QuangNgai_KOFP.xls.xlsx
deleted file mode 100644
index 349bede..0000000
Binary files a/data/2017-old/18_QuangNgai_KOFP.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/19_QuangTri_OMZF.xls.xlsx b/data/2017-old/19_QuangTri_OMZF.xls.xlsx
deleted file mode 100644
index f66a8bb..0000000
Binary files a/data/2017-old/19_QuangTri_OMZF.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/1_BaRia_VungTau_HJKG.xls.xlsx b/data/2017-old/1_BaRia_VungTau_HJKG.xls.xlsx
deleted file mode 100644
index 35be1a9..0000000
Binary files a/data/2017-old/1_BaRia_VungTau_HJKG.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/1_Da_Nang_AHWJ.xls.xlsx b/data/2017-old/1_Da_Nang_AHWJ.xls.xlsx
deleted file mode 100644
index 4a63447..0000000
Binary files a/data/2017-old/1_Da_Nang_AHWJ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/1_Ha_Noi_CVXG.xls.xlsx b/data/2017-old/1_Ha_Noi_CVXG.xls.xlsx
deleted file mode 100644
index e55102c..0000000
Binary files a/data/2017-old/1_Ha_Noi_CVXG.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/1_Son_La_JIDP.xls.xlsx b/data/2017-old/1_Son_La_JIDP.xls.xlsx
deleted file mode 100644
index 62d7b18..0000000
Binary files a/data/2017-old/1_Son_La_JIDP.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/1_TuyenQuang_JBYF.xls.xlsx b/data/2017-old/1_TuyenQuang_JBYF.xls.xlsx
deleted file mode 100644
index 601bb9e..0000000
Binary files a/data/2017-old/1_TuyenQuang_JBYF.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/20_TayNinh_ILFA.xls.xlsx b/data/2017-old/20_TayNinh_ILFA.xls.xlsx
deleted file mode 100644
index 2b636fc..0000000
Binary files a/data/2017-old/20_TayNinh_ILFA.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/21_ThaiBinh_FTVG.xls.xlsx b/data/2017-old/21_ThaiBinh_FTVG.xls.xlsx
deleted file mode 100644
index a9f5bae..0000000
Binary files a/data/2017-old/21_ThaiBinh_FTVG.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/22_ThaiNguyen_TLTW.xls.xlsx b/data/2017-old/22_ThaiNguyen_TLTW.xls.xlsx
deleted file mode 100644
index 8f9c2bb..0000000
Binary files a/data/2017-old/22_ThaiNguyen_TLTW.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/23_HaiPhong_HXBV.xls.xlsx b/data/2017-old/23_HaiPhong_HXBV.xls.xlsx
deleted file mode 100644
index c219f1f..0000000
Binary files a/data/2017-old/23_HaiPhong_HXBV.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/24_HCM_XULN.xls.xlsx b/data/2017-old/24_HCM_XULN.xls.xlsx
deleted file mode 100644
index a526aac..0000000
Binary files a/data/2017-old/24_HCM_XULN.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/2_BacKan_GFVQ.xls.xlsx b/data/2017-old/2_BacKan_GFVQ.xls.xlsx
deleted file mode 100644
index 115d551..0000000
Binary files a/data/2017-old/2_BacKan_GFVQ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/2_Ha_Giang_QNCM.xls.xlsx b/data/2017-old/2_Ha_Giang_QNCM.xls.xlsx
deleted file mode 100644
index 2750bed..0000000
Binary files a/data/2017-old/2_Ha_Giang_QNCM.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/2_Ninh_Thuan_VHLY.xls.xlsx b/data/2017-old/2_Ninh_Thuan_VHLY.xls.xlsx
deleted file mode 100644
index 042f58d..0000000
Binary files a/data/2017-old/2_Ninh_Thuan_VHLY.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/2_Thanh_Hoa_AUFV.xls.xlsx b/data/2017-old/2_Thanh_Hoa_AUFV.xls.xlsx
deleted file mode 100644
index 351d53e..0000000
Binary files a/data/2017-old/2_Thanh_Hoa_AUFV.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/2_VinhPhuc_GUDK.xls.xlsx b/data/2017-old/2_VinhPhuc_GUDK.xls.xlsx
deleted file mode 100644
index cfb98f9..0000000
Binary files a/data/2017-old/2_VinhPhuc_GUDK.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/3_BacGiang_TOIF.xls.xlsx b/data/2017-old/3_BacGiang_TOIF.xls.xlsx
deleted file mode 100644
index 1e9661b..0000000
Binary files a/data/2017-old/3_BacGiang_TOIF.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/3_BinhPhuoc_YFMU.xls.xlsx b/data/2017-old/3_BinhPhuoc_YFMU.xls.xlsx
deleted file mode 100644
index 592d676..0000000
Binary files a/data/2017-old/3_BinhPhuoc_YFMU.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/3_Cao_Bang_CIEY.xls.xlsx b/data/2017-old/3_Cao_Bang_CIEY.xls.xlsx
deleted file mode 100644
index f68c4e7..0000000
Binary files a/data/2017-old/3_Cao_Bang_CIEY.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/3_Dong_Thap_GSXQ.xls.xlsx b/data/2017-old/3_Dong_Thap_GSXQ.xls.xlsx
deleted file mode 100644
index 5afdc2d..0000000
Binary files a/data/2017-old/3_Dong_Thap_GSXQ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/3_Thua_Thien_Hue_HMDB.xls.xlsx b/data/2017-old/3_Thua_Thien_Hue_HMDB.xls.xlsx
deleted file mode 100644
index 34d591f..0000000
Binary files a/data/2017-old/3_Thua_Thien_Hue_HMDB.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/4_An_Giang_JNOS.xls.xlsx b/data/2017-old/4_An_Giang_JNOS.xls.xlsx
deleted file mode 100644
index b4032d3..0000000
Binary files a/data/2017-old/4_An_Giang_JNOS.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/4_BacNinh_STLR.xls.xlsx b/data/2017-old/4_BacNinh_STLR.xls.xlsx
deleted file mode 100644
index cc08868..0000000
Binary files a/data/2017-old/4_BacNinh_STLR.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/4_Binh_Dinh_WWZW.xls.xlsx b/data/2017-old/4_Binh_Dinh_WWZW.xls.xlsx
deleted file mode 100644
index ddb0fc5..0000000
Binary files a/data/2017-old/4_Binh_Dinh_WWZW.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/4_DienBien_SOJG.xls.xlsx b/data/2017-old/4_DienBien_SOJG.xls.xlsx
deleted file mode 100644
index 0dddaac..0000000
Binary files a/data/2017-old/4_DienBien_SOJG.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/4_Lang_Son_NYQL.xls.xlsx b/data/2017-old/4_Lang_Son_NYQL.xls.xlsx
deleted file mode 100644
index a845623..0000000
Binary files a/data/2017-old/4_Lang_Son_NYQL.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/5_Bac_Lieu_XEKH.xls.xlsx b/data/2017-old/5_Bac_Lieu_XEKH.xls.xlsx
deleted file mode 100644
index 73c1bef..0000000
Binary files a/data/2017-old/5_Bac_Lieu_XEKH.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/5_Gia_Lai_ABZI.xls.xlsx b/data/2017-old/5_Gia_Lai_ABZI.xls.xlsx
deleted file mode 100644
index c0f9294..0000000
Binary files a/data/2017-old/5_Gia_Lai_ABZI.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/5_HaiDuong_WWWG.xls.xlsx b/data/2017-old/5_HaiDuong_WWWG.xls.xlsx
deleted file mode 100644
index 190b19a..0000000
Binary files a/data/2017-old/5_HaiDuong_WWWG.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/5_Hanam_QGJS.xls.xlsx b/data/2017-old/5_Hanam_QGJS.xls.xlsx
deleted file mode 100644
index 1740b8b..0000000
Binary files a/data/2017-old/5_Hanam_QGJS.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/5_Yen_Bai_FAQR.xls.xlsx b/data/2017-old/5_Yen_Bai_FAQR.xls.xlsx
deleted file mode 100644
index 09a6e5f..0000000
Binary files a/data/2017-old/5_Yen_Bai_FAQR.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/6_Dong_Nai_WOTM.xls.xlsx b/data/2017-old/6_Dong_Nai_WOTM.xls.xlsx
deleted file mode 100644
index 35fbd77..0000000
Binary files a/data/2017-old/6_Dong_Nai_WOTM.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/6_Hau_Giang_KWDM.xls.xlsx b/data/2017-old/6_Hau_Giang_KWDM.xls.xlsx
deleted file mode 100644
index 7543315..0000000
Binary files a/data/2017-old/6_Hau_Giang_KWDM.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/6_HoaBinh_TPZY.xls.xlsx b/data/2017-old/6_HoaBinh_TPZY.xls.xlsx
deleted file mode 100644
index 932db3e..0000000
Binary files a/data/2017-old/6_HoaBinh_TPZY.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/6_NinhBinh_IGFT.xls.xlsx b/data/2017-old/6_NinhBinh_IGFT.xls.xlsx
deleted file mode 100644
index 2553671..0000000
Binary files a/data/2017-old/6_NinhBinh_IGFT.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/6_Quang_Ninh_DQCJ.xls.xlsx b/data/2017-old/6_Quang_Ninh_DQCJ.xls.xlsx
deleted file mode 100644
index c04950f..0000000
Binary files a/data/2017-old/6_Quang_Ninh_DQCJ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/7_HaTinh_DDHD.xls.xlsx b/data/2017-old/7_HaTinh_DDHD.xls.xlsx
deleted file mode 100644
index 9327104..0000000
Binary files a/data/2017-old/7_HaTinh_DDHD.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/7_HungYen_LTIK.xls.xlsx b/data/2017-old/7_HungYen_LTIK.xls.xlsx
deleted file mode 100644
index bfff635..0000000
Binary files a/data/2017-old/7_HungYen_LTIK.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/7_Kon_Tum_RSLR.xls.xlsx b/data/2017-old/7_Kon_Tum_RSLR.xls.xlsx
deleted file mode 100644
index 3563aab..0000000
Binary files a/data/2017-old/7_Kon_Tum_RSLR.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/7_Tien_Giang_EFHX.xls.xlsx b/data/2017-old/7_Tien_Giang_EFHX.xls.xlsx
deleted file mode 100644
index 8e51dda..0000000
Binary files a/data/2017-old/7_Tien_Giang_EFHX.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/8_Can_Tho_RQZM.xls.xlsx b/data/2017-old/8_Can_Tho_RQZM.xls.xlsx
deleted file mode 100644
index 172a18d..0000000
Binary files a/data/2017-old/8_Can_Tho_RQZM.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/8_Dak_Lak_YKPR.xls.xlsx b/data/2017-old/8_Dak_Lak_YKPR.xls.xlsx
deleted file mode 100644
index 98c80a6..0000000
Binary files a/data/2017-old/8_Dak_Lak_YKPR.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/8_KienGiang_OTOB.xls.xlsx b/data/2017-old/8_KienGiang_OTOB.xls.xlsx
deleted file mode 100644
index fa07efa..0000000
Binary files a/data/2017-old/8_KienGiang_OTOB.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/8_PhuYen_OGIM.xls.xlsx b/data/2017-old/8_PhuYen_OGIM.xls.xlsx
deleted file mode 100644
index 02e391c..0000000
Binary files a/data/2017-old/8_PhuYen_OGIM.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/9_DakNong_FBOP.xls.xlsx b/data/2017-old/9_DakNong_FBOP.xls.xlsx
deleted file mode 100644
index 6ae062c..0000000
Binary files a/data/2017-old/9_DakNong_FBOP.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/9_Khanh_Hoa_KPKQ.xls.xlsx b/data/2017-old/9_Khanh_Hoa_KPKQ.xls.xlsx
deleted file mode 100644
index d30f3f7..0000000
Binary files a/data/2017-old/9_Khanh_Hoa_KPKQ.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/9_LaiChau_ALKN.xls.xlsx b/data/2017-old/9_LaiChau_ALKN.xls.xlsx
deleted file mode 100644
index 9806b8e..0000000
Binary files a/data/2017-old/9_LaiChau_ALKN.xls.xlsx and /dev/null differ
diff --git a/data/2017-old/9_Vinh_Long_OLWA.xls.xlsx b/data/2017-old/9_Vinh_Long_OLWA.xls.xlsx
deleted file mode 100644
index 7ba2885..0000000
Binary files a/data/2017-old/9_Vinh_Long_OLWA.xls.xlsx and /dev/null differ
diff --git a/data/2017-old2/1.BaRia-VungTau_PGZT.xlsx b/data/2017-old2/1.BaRia-VungTau_PGZT.xlsx
deleted file mode 100644
index 3a26735..0000000
Binary files a/data/2017-old2/1.BaRia-VungTau_PGZT.xlsx and /dev/null differ
diff --git a/data/2017-old2/1.Da Nang_ABWU.xlsx b/data/2017-old2/1.Da Nang_ABWU.xlsx
deleted file mode 100644
index 5396aef..0000000
Binary files a/data/2017-old2/1.Da Nang_ABWU.xlsx and /dev/null differ
diff --git a/data/2017-old2/1.Son La_XLFN.xlsx b/data/2017-old2/1.Son La_XLFN.xlsx
deleted file mode 100644
index 7eafde3..0000000
Binary files a/data/2017-old2/1.Son La_XLFN.xlsx and /dev/null differ
diff --git a/data/2017-old2/1.TuyenQuang_PTMR.xlsx b/data/2017-old2/1.TuyenQuang_PTMR.xlsx
deleted file mode 100644
index 06f7c27..0000000
Binary files a/data/2017-old2/1.TuyenQuang_PTMR.xlsx and /dev/null differ
diff --git a/data/2017-old2/10.BinhThuan_MVVG.xlsx b/data/2017-old2/10.BinhThuan_MVVG.xlsx
deleted file mode 100644
index 8cabd6d..0000000
Binary files a/data/2017-old2/10.BinhThuan_MVVG.xlsx and /dev/null differ
diff --git a/data/2017-old2/10.LamDong_YUQA.xlsx b/data/2017-old2/10.LamDong_YUQA.xlsx
deleted file mode 100644
index e4debd8..0000000
Binary files a/data/2017-old2/10.LamDong_YUQA.xlsx and /dev/null differ
diff --git a/data/2017-old2/10.Soc Trang_LQWU.xlsx b/data/2017-old2/10.Soc Trang_LQWU.xlsx
deleted file mode 100644
index c8cd707..0000000
Binary files a/data/2017-old2/10.Soc Trang_LQWU.xlsx and /dev/null differ
diff --git a/data/2017-old2/11.BinhDuong_HVAH.xlsx b/data/2017-old2/11.BinhDuong_HVAH.xlsx
deleted file mode 100644
index 8b3f7ad..0000000
Binary files a/data/2017-old2/11.BinhDuong_HVAH.xlsx and /dev/null differ
diff --git a/data/2017-old2/11.LaoCai_ZTBP.xlsx b/data/2017-old2/11.LaoCai_ZTBP.xlsx
deleted file mode 100644
index 63870d1..0000000
Binary files a/data/2017-old2/11.LaoCai_ZTBP.xlsx and /dev/null differ
diff --git a/data/2017-old2/12.BenTre_NQTU.xlsx b/data/2017-old2/12.BenTre_NQTU.xlsx
deleted file mode 100644
index ae9edb4..0000000
Binary files a/data/2017-old2/12.BenTre_NQTU.xlsx and /dev/null differ
diff --git a/data/2017-old2/12.LongAn_PDRH.xlsx b/data/2017-old2/12.LongAn_PDRH.xlsx
deleted file mode 100644
index e0b66c8..0000000
Binary files a/data/2017-old2/12.LongAn_PDRH.xlsx and /dev/null differ
diff --git a/data/2017-old2/13.NamDinh_NAYR.xlsx b/data/2017-old2/13.NamDinh_NAYR.xlsx
deleted file mode 100644
index b16ead2..0000000
Binary files a/data/2017-old2/13.NamDinh_NAYR.xlsx and /dev/null differ
diff --git a/data/2017-old2/13.TraVinh_FODZ.xlsx b/data/2017-old2/13.TraVinh_FODZ.xlsx
deleted file mode 100644
index eb2c6ce..0000000
Binary files a/data/2017-old2/13.TraVinh_FODZ.xlsx and /dev/null differ
diff --git a/data/2017-old2/14.NgheAn_HTKD.xlsx b/data/2017-old2/14.NgheAn_HTKD.xlsx
deleted file mode 100644
index f872da1..0000000
Binary files a/data/2017-old2/14.NgheAn_HTKD.xlsx and /dev/null differ
diff --git a/data/2017-old2/15.PhuTho_IJZW.xlsx b/data/2017-old2/15.PhuTho_IJZW.xlsx
deleted file mode 100644
index e075d3a..0000000
Binary files a/data/2017-old2/15.PhuTho_IJZW.xlsx and /dev/null differ
diff --git a/data/2017-old2/17.QuangNam_NQMG.xlsx b/data/2017-old2/17.QuangNam_NQMG.xlsx
deleted file mode 100644
index 3e75262..0000000
Binary files a/data/2017-old2/17.QuangNam_NQMG.xlsx and /dev/null differ
diff --git a/data/2017-old2/18.QuangNgai_IUPY.xlsx b/data/2017-old2/18.QuangNgai_IUPY.xlsx
deleted file mode 100644
index 5e37ea2..0000000
Binary files a/data/2017-old2/18.QuangNgai_IUPY.xlsx and /dev/null differ
diff --git a/data/2017-old2/19.QuangTri_MKNN.xlsx b/data/2017-old2/19.QuangTri_MKNN.xlsx
deleted file mode 100644
index 454256c..0000000
Binary files a/data/2017-old2/19.QuangTri_MKNN.xlsx and /dev/null differ
diff --git a/data/2017-old2/2.BacKan_YQNX.xlsx b/data/2017-old2/2.BacKan_YQNX.xlsx
deleted file mode 100644
index 6fa41c9..0000000
Binary files a/data/2017-old2/2.BacKan_YQNX.xlsx and /dev/null differ
diff --git a/data/2017-old2/2.Ha Giang_PIYK.xlsx b/data/2017-old2/2.Ha Giang_PIYK.xlsx
deleted file mode 100644
index d65b471..0000000
Binary files a/data/2017-old2/2.Ha Giang_PIYK.xlsx and /dev/null differ
diff --git a/data/2017-old2/2.Ninh Thuan_BAGG.xlsx b/data/2017-old2/2.Ninh Thuan_BAGG.xlsx
deleted file mode 100644
index a77706b..0000000
Binary files a/data/2017-old2/2.Ninh Thuan_BAGG.xlsx and /dev/null differ
diff --git a/data/2017-old2/2.Thanh Hoa_UOPE.xlsx b/data/2017-old2/2.Thanh Hoa_UOPE.xlsx
deleted file mode 100644
index d40be28..0000000
Binary files a/data/2017-old2/2.Thanh Hoa_UOPE.xlsx and /dev/null differ
diff --git a/data/2017-old2/2.VinhPhuc_QZJK.xlsx b/data/2017-old2/2.VinhPhuc_QZJK.xlsx
deleted file mode 100644
index be86a48..0000000
Binary files a/data/2017-old2/2.VinhPhuc_QZJK.xlsx and /dev/null differ
diff --git a/data/2017-old2/20.TayNinh_KJAQ.xlsx b/data/2017-old2/20.TayNinh_KJAQ.xlsx
deleted file mode 100644
index 65eedff..0000000
Binary files a/data/2017-old2/20.TayNinh_KJAQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/21.ThaiBinh_KTQN.xlsx b/data/2017-old2/21.ThaiBinh_KTQN.xlsx
deleted file mode 100644
index 02a3895..0000000
Binary files a/data/2017-old2/21.ThaiBinh_KTQN.xlsx and /dev/null differ
diff --git a/data/2017-old2/22.ThaiNguyen_BKIF.xlsx b/data/2017-old2/22.ThaiNguyen_BKIF.xlsx
deleted file mode 100644
index a144e9c..0000000
Binary files a/data/2017-old2/22.ThaiNguyen_BKIF.xlsx and /dev/null differ
diff --git a/data/2017-old2/24.HCM_UTLQ.xlsx b/data/2017-old2/24.HCM_UTLQ.xlsx
deleted file mode 100644
index cc42e51..0000000
Binary files a/data/2017-old2/24.HCM_UTLQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/3.BacGiang_SAVS.xlsx b/data/2017-old2/3.BacGiang_SAVS.xlsx
deleted file mode 100644
index bc9b5f6..0000000
Binary files a/data/2017-old2/3.BacGiang_SAVS.xlsx and /dev/null differ
diff --git a/data/2017-old2/3.BinhPhuoc_IPHL.xlsx b/data/2017-old2/3.BinhPhuoc_IPHL.xlsx
deleted file mode 100644
index 8c5871d..0000000
Binary files a/data/2017-old2/3.BinhPhuoc_IPHL.xlsx and /dev/null differ
diff --git a/data/2017-old2/3.Cao Bang_WMUU.xlsx b/data/2017-old2/3.Cao Bang_WMUU.xlsx
deleted file mode 100644
index 36200fb..0000000
Binary files a/data/2017-old2/3.Cao Bang_WMUU.xlsx and /dev/null differ
diff --git a/data/2017-old2/3.Dong Thap_HKJX.xlsx b/data/2017-old2/3.Dong Thap_HKJX.xlsx
deleted file mode 100644
index ae37166..0000000
Binary files a/data/2017-old2/3.Dong Thap_HKJX.xlsx and /dev/null differ
diff --git a/data/2017-old2/3.Thua Thien -Hue_MAET.xlsx b/data/2017-old2/3.Thua Thien -Hue_MAET.xlsx
deleted file mode 100644
index 587af78..0000000
Binary files a/data/2017-old2/3.Thua Thien -Hue_MAET.xlsx and /dev/null differ
diff --git a/data/2017-old2/4.An Giang_PMJD.xlsx b/data/2017-old2/4.An Giang_PMJD.xlsx
deleted file mode 100644
index b651ffb..0000000
Binary files a/data/2017-old2/4.An Giang_PMJD.xlsx and /dev/null differ
diff --git a/data/2017-old2/4.BacNinh_NNIS.xlsx b/data/2017-old2/4.BacNinh_NNIS.xlsx
deleted file mode 100644
index 04de119..0000000
Binary files a/data/2017-old2/4.BacNinh_NNIS.xlsx and /dev/null differ
diff --git a/data/2017-old2/4.Binh Dinh_VOMJ.xlsx b/data/2017-old2/4.Binh Dinh_VOMJ.xlsx
deleted file mode 100644
index 704c4ae..0000000
Binary files a/data/2017-old2/4.Binh Dinh_VOMJ.xlsx and /dev/null differ
diff --git a/data/2017-old2/4.DienBien_FYGN.xlsx b/data/2017-old2/4.DienBien_FYGN.xlsx
deleted file mode 100644
index aae73eb..0000000
Binary files a/data/2017-old2/4.DienBien_FYGN.xlsx and /dev/null differ
diff --git a/data/2017-old2/4.Lang Son_QWOG.xlsx b/data/2017-old2/4.Lang Son_QWOG.xlsx
deleted file mode 100644
index e7e9534..0000000
Binary files a/data/2017-old2/4.Lang Son_QWOG.xlsx and /dev/null differ
diff --git a/data/2017-old2/5.Bac Lieu_VIVY.xlsx b/data/2017-old2/5.Bac Lieu_VIVY.xlsx
deleted file mode 100644
index 575cac0..0000000
Binary files a/data/2017-old2/5.Bac Lieu_VIVY.xlsx and /dev/null differ
diff --git a/data/2017-old2/5.Gia Lai_TAAS.xlsx b/data/2017-old2/5.Gia Lai_TAAS.xlsx
deleted file mode 100644
index f358c30..0000000
Binary files a/data/2017-old2/5.Gia Lai_TAAS.xlsx and /dev/null differ
diff --git a/data/2017-old2/5.HaiDuong_WNHD.xlsx b/data/2017-old2/5.HaiDuong_WNHD.xlsx
deleted file mode 100644
index 7b1169f..0000000
Binary files a/data/2017-old2/5.HaiDuong_WNHD.xlsx and /dev/null differ
diff --git a/data/2017-old2/5.Hanam_SDKN.xlsx b/data/2017-old2/5.Hanam_SDKN.xlsx
deleted file mode 100644
index 5b9d113..0000000
Binary files a/data/2017-old2/5.Hanam_SDKN.xlsx and /dev/null differ
diff --git a/data/2017-old2/5.Yen Bai_BSLV.xlsx b/data/2017-old2/5.Yen Bai_BSLV.xlsx
deleted file mode 100644
index 915dc38..0000000
Binary files a/data/2017-old2/5.Yen Bai_BSLV.xlsx and /dev/null differ
diff --git a/data/2017-old2/6.Hau Giang_SIAJ.xlsx b/data/2017-old2/6.Hau Giang_SIAJ.xlsx
deleted file mode 100644
index f4a67e4..0000000
Binary files a/data/2017-old2/6.Hau Giang_SIAJ.xlsx and /dev/null differ
diff --git a/data/2017-old2/6.HoaBinh_HLYQ.xlsx b/data/2017-old2/6.HoaBinh_HLYQ.xlsx
deleted file mode 100644
index 768a885..0000000
Binary files a/data/2017-old2/6.HoaBinh_HLYQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/6.NinhBinh_PKMQ.xlsx b/data/2017-old2/6.NinhBinh_PKMQ.xlsx
deleted file mode 100644
index 7fab324..0000000
Binary files a/data/2017-old2/6.NinhBinh_PKMQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/6.Quang Ninh_YAKJ.xlsx b/data/2017-old2/6.Quang Ninh_YAKJ.xlsx
deleted file mode 100644
index 51ddb21..0000000
Binary files a/data/2017-old2/6.Quang Ninh_YAKJ.xlsx and /dev/null differ
diff --git a/data/2017-old2/7.HaTinh_XMFQ.xlsx b/data/2017-old2/7.HaTinh_XMFQ.xlsx
deleted file mode 100644
index 0aa1629..0000000
Binary files a/data/2017-old2/7.HaTinh_XMFQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/7.HungYen_TCBE.xlsx b/data/2017-old2/7.HungYen_TCBE.xlsx
deleted file mode 100644
index be95a39..0000000
Binary files a/data/2017-old2/7.HungYen_TCBE.xlsx and /dev/null differ
diff --git a/data/2017-old2/7.Kon Tum_CAQU.xlsx b/data/2017-old2/7.Kon Tum_CAQU.xlsx
deleted file mode 100644
index aa2d88a..0000000
Binary files a/data/2017-old2/7.Kon Tum_CAQU.xlsx and /dev/null differ
diff --git a/data/2017-old2/7.Tien Giang_AOWZ.xlsx b/data/2017-old2/7.Tien Giang_AOWZ.xlsx
deleted file mode 100644
index 43206a0..0000000
Binary files a/data/2017-old2/7.Tien Giang_AOWZ.xlsx and /dev/null differ
diff --git a/data/2017-old2/8.PhuYen_MLTQ.xlsx b/data/2017-old2/8.PhuYen_MLTQ.xlsx
deleted file mode 100644
index 7af0186..0000000
Binary files a/data/2017-old2/8.PhuYen_MLTQ.xlsx and /dev/null differ
diff --git a/data/2017-old2/9.DakNong_ZWJT.xlsx b/data/2017-old2/9.DakNong_ZWJT.xlsx
deleted file mode 100644
index cffd8a6..0000000
Binary files a/data/2017-old2/9.DakNong_ZWJT.xlsx and /dev/null differ
diff --git a/data/2017-old2/9.Khanh Hoa_IBTX.xlsx b/data/2017-old2/9.Khanh Hoa_IBTX.xlsx
deleted file mode 100644
index 058419d..0000000
Binary files a/data/2017-old2/9.Khanh Hoa_IBTX.xlsx and /dev/null differ
diff --git a/data/2017-old2/9.Vinh Long_WVKI.xlsx b/data/2017-old2/9.Vinh Long_WVKI.xlsx
deleted file mode 100644
index febfad4..0000000
Binary files a/data/2017-old2/9.Vinh Long_WVKI.xlsx and /dev/null differ
diff --git a/docs/data-pipeline.md b/docs/data-pipeline.md
index bbfe48d..b88c711 100644
--- a/docs/data-pipeline.md
+++ b/docs/data-pipeline.md
@@ -2,32 +2,63 @@
 
 From raw Excel files to a compressed SQLite file the browser can load.
 
-One Rust binary (`go-parser/`) builds every dataset. What differs per dataset is
+One Go binary (`go-parser/`) builds every dataset. What differs per dataset is
 parse rules only — sheet strategy, column layout, validation guards — declared
 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
 
-| id | Files | Reproducible | Origin |
+| id | Files | Origin | Host live? |
 | --- | --- | --- | --- |
-| `2016` | 4 `.xls` + 115 `.xlsx` | no | Bộ GD&ĐT, collected 2016 |
-| `2017` | 63 `.xls` | **yes** | baotintuc.vn CDN |
-| `2017-old` | 63 `.xlsx` | no | pre-refresh archive |
-| `2017-old2` | 54 `.xlsx` | no | corrected re-export |
+| `2016` | 4 `.xls` + 115 `.xlsx` | aggregator article, 119 exam clusters | unconfirmed |
+| `2017` | 63 `.xls` | baotintuc.vn CDN | **yes** |
 
-Only `2017` can be re-fetched:
+Crawling lives in `crawler/`, a separate Go module. It is never part of the
+build — the source files are committed, so a crawl only refreshes them. Both
+runs are idempotent: files already present are skipped.
 
 ```bash
-node go-parser/scripts/crawl-baotintuc.js
+npm run crawl:2016              # crawler/internal/sources/source_2016.go
+npm run crawl:2017              # crawler/internal/sources/source_2017.go
+npm run crawl:2017 -- --list    # show what would be downloaded, fetch nothing
 ```
 
-Idempotent — skips files already present, saves to `data/2017/`. Source article:
+**2017** comes from the article
 `https://baotintuc.vn/tuyen-sinh/tra-cuu-diem-thi-thpt-2017-cua-63-tinh-thanh-pho-tren-baotintucvn-20170706073512672.htm`
+and its CDN is still serving the files.
 
-The other three were collected from Vietnamese news sites at the time and the
-publisher URLs were not recorded. **The files committed in git are the only
-copy.**
+**2016** comes from the aggregator article
+`cong-bo-diem-thi-thptqg-2016-toan-bo-120-cum-thi-da-co-diem.html`. The site
+that originally published it (`dtntbacgiang.edu.vn`) no longer resolves, so the
+link list was read from the Internet Archive's copy of the page. Two caveats
+that matter:
+
+- **The list is verified; the host is not.** All 119 filenames match `data/2016/`
+  exactly in both directions, which is asserted by a test. But the Archive
+  captured only the article, not the spreadsheets, so the URLs themselves could
+  not be confirmed by fetching. `baseURL` in `source_2016.go` points at a mirror
+  of the same article that is still online; if it serves the files from a
+  different directory, `uploadPath` is the one constant to change.
+- **`data/2016/` is still the only confirmed copy.** Do not delete it on the
+  assumption that a crawl can restore it.
+
+### Local filenames are load-bearing
+
+`go-parser` sorts its input files and inserts with `INSERT OR REPLACE`, which is
+last-wins, so **filenames decide which row survives a duplicate exam number**. A
+re-crawl that names files differently can produce a database with the same row
+count and different content, which the row-count guard in `build-db.js` would
+not catch.
+
+Each source therefore pins its local names, and
+`crawler/internal/sources/sources_test.go` checks every source against its
+committed `data/<id>/` in both directions — every name it would write exists,
+and every file present is accounted for.
+
+2017 derives its names from the province (the CDN's own names are inconsistent:
+`Angiang.xls`, `1BaRiaVungTau.xls`, `23HaiPhong.xls`). 2016 keeps the
+server-assigned names verbatim, since it did not choose them.
 
 ## Source Excel shapes
 
@@ -41,7 +72,7 @@ The 2017 datasets share one layout:
 | 3 | DIEM_THI | concatenated per-subject scores, e.g. `"Toán: 6.80 Ngữ văn: 5.25 …"` |
 
 2016 has **three** layouts across its 119 files, chosen per file at runtime via
-`format_detection = "thptqg2016"` in its config:
+`format_detection: thptqg2016` in its config:
 
 | Format | Detected by | Notes |
 | --- | --- | --- |
@@ -95,8 +126,6 @@ scores rather than false matches.
 | --- | --- | --- |
 | `2016` | all sheets | per-file format detection; header-token rows rejected |
 | `2017` | all sheets — Hà Nội and HCM overflow | none |
-| `2017-old` | first sheet only | reject non-numeric SBD (rejects a header leak in this export) |
-| `2017-old2` | all sheets — HCM overflows | reject non-numeric SBD; skip blank rows before counting |
 
 ### Overflow-sheet gotcha
 
@@ -111,8 +140,6 @@ silently drops 13,720 students** (Hanoi +7,275, HCM +6,445). That is what
 | --- | --- | --- | --- |
 | `2016` | 877,464 | 3 duplicate SBDs collapsed | **877,461** |
 | `2017` | 861,068 | 0 | **861,068** |
-| `2017-old` | 847,349 | 1 header leak | **847,348** |
-| `2017-old2` | 679,764 | 0 | **679,764** |
 
 ## Verifying a rebuild
 
@@ -143,11 +170,14 @@ cell of any input file reads differently.
 
 ```bash
 rm data/2017/*.xls
-node go-parser/scripts/crawl-baotintuc.js
-node go-parser/scripts/build-db.js 2017
+npm run crawl:2017
+npm run build:db 2017
 ```
 
-The row-count guard in `build:db` confirms the rebuild matches the expected total.
+The row-count guard in `build:db` confirms the rebuild matches the expected
+total. That guard checks the count only, so if the crawl was expected to change
+the data, compare content with `differential-parity.mjs` against a copy of the
+previous database rather than trusting the count.
 
 ## Removed scripts
 
@@ -157,3 +187,9 @@ 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.
+
+`crawl-baotintuc.js` was not dropped but rewritten in Go as
+`crawler/internal/sources/source_2017.go`, keeping the same link list and the same
+local filenames. The rewrite added a per-file timeout and made downloads land on
+a `.part` file first — writing straight to the destination meant an interrupted
+run left a truncated file that the skip check would then skip forever.
diff --git a/docs/deployment-guide.md b/docs/deployment-guide.md
index 80710bc..88b8342 100644
--- a/docs/deployment-guide.md
+++ b/docs/deployment-guide.md
@@ -1,7 +1,7 @@
 # Deployment Guide
 
 Deploys to GitHub Pages via `.github/workflows/deploy-pages.yml`. Every push to
-`main` rebuilds all four datasets and redeploys the whole site.
+`main` rebuilds both datasets and redeploys the whole site.
 
 One-time setup: **Settings → Pages → Source: GitHub Actions**.
 
@@ -9,9 +9,9 @@ One-time setup: **Settings → Pages → Source: GitHub Actions**.
 
 1. Checkout, Go toolchain, Node 24, `npm ci`
 2. `npm run build:go` — one parser binary
-3. `npm run build:db` — builds and gzips all four databases into
+3. `npm run build:db` — builds and gzips both databases into
    `.build/public/db/`
-4. `npm run build:site` — one Vite build, then `scripts/assemble-site.js`
+4. `npm run build:site` — one Vite build, then `web/scripts/assemble-site.js`
 5. `actions/upload-pages-artifact` + `actions/deploy-pages`
 
 The database build dominates the runtime: roughly 419 MB of Excel is parsed on
@@ -23,20 +23,18 @@ every deploy.
 https://<user>.github.io/thptqg/
 https://<user>.github.io/thptqg/2016/
 https://<user>.github.io/thptqg/2017/
-https://<user>.github.io/thptqg/2017-old/
-https://<user>.github.io/thptqg/2017-old2/
 ```
 
-`/thptqg/2017/old/` and `/thptqg/2017/old2/` were the pre-flattening URLs. They
-are still served, and the router rewrites them to the flat form with the query
-string intact.
+`/thptqg/2017/old/` and `/thptqg/2017/old2/` were the pre-flattening URLs for
+the two removed 2017 archives. They are no longer served; like any unknown path
+they now render the hub via `404.html`.
 
 ## Local reproduction
 
 ```bash
 npm ci
 npm run build:go
-npm run build:db      # all four; pass an id to build just one
+npm run build:db      # both; pass an id to build just one
 npm run build:site    # vite build + assemble into _site/
 npx serve _site
 ```
@@ -44,7 +42,7 @@ npx serve _site
 To rebuild a single dataset:
 
 ```bash
-node go-parser/scripts/build-db.js 2017-old
+npm run build:db 2017
 ```
 
 ## Base path
@@ -59,7 +57,7 @@ up as a blank page with 404s on `/assets/...`.
 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`
+3. Add an entry to `DATASETS` in `web/src/datasets.js`
 
 Nothing else. The build script, the site assembly and the router all read that
 one list, and the frontend adapts to whichever columns the dataset populates.
@@ -98,7 +96,7 @@ run rebuilds the older state. There is no data to migrate.
 | Symptom | Typical cause |
 | --- | --- |
 | Blank page, 404 on assets | `base` in `vite.config.js` does not match the repo name |
-| `Failed to fetch database: 404` | Dataset id in `src/datasets.js` does not match the file in `db/` |
+| `Failed to fetch database: 404` | Dataset id in `web/src/datasets.js` does not match the file in `db/` |
 | A route 404s | `assemble-site.js` did not run, or the id is missing from `DATASETS` |
 | WASM fails to load | `sql.js.org` unreachable — self-host `sql-wasm.wasm` and update `SQL_WASM_URL` in `use-sqlite.js` |
 | Deploy fails on assembly | An uncompressed database artefact reached the output; the error names the files |
diff --git a/docs/project-overview.md b/docs/project-overview.md
index af8dfdd..7b1cba5 100644
--- a/docs/project-overview.md
+++ b/docs/project-overview.md
@@ -4,7 +4,7 @@
 
 A public lookup tool for Vietnam's National High School Graduation Exam scores,
 running entirely in the browser and hosted for free on GitHub Pages. Covers the
-2016 and 2017 exams — 1.7 million candidates across four datasets.
+2016 and 2017 exams — 1.7 million candidates across two datasets.
 
 ## Scope
 
@@ -21,7 +21,7 @@ running entirely in the browser and hosted for free on GitHub Pages. Covers the
 
 ## Constraints
 
-- **Zero backend.** The full database (38–48 MB gzipped per dataset) is
+- **Zero backend.** The full database (44–48 MB gzipped per dataset) is
   downloaded to the browser and queried in-process.
 - **Read-only.** `INSERT`/`UPDATE`/`DELETE` are rejected, so nobody is misled
   into thinking edits persist. `sql.js` is in-memory anyway.
@@ -36,16 +36,17 @@ running entirely in the browser and hosted for free on GitHub Pages. Covers the
 | --- | --- | --- | --- |
 | `2016` | 2016 | 877,461 | 119 files, three column layouts |
 | `2017` | 2017 | 861,068 | current generation, reproducible from source |
-| `2017-old` | 2017 | 847,348 | pre-refresh archive |
-| `2017-old2` | 2017 | 679,764 | corrected re-export |
 
-The three 2017 datasets are kept side by side because they disagree, and the
-disagreement is itself informative. Only `2017` is re-fetchable; the rest exist
-solely as the copies committed here.
+Two further 2017 datasets (`2017-old`, `2017-old2`) were kept alongside these
+because the three publications disagreed. They have been removed; git history
+still has them.
 
-Original 2016 aggregator link
-(<https://dtntbacgiang.edu.vn/tin-tuc/tin-tuc-su-kien/cong-bo-diem-thi-thptqg-2016-toan-bo-120-cum-thi-da-co-diem.html>)
-is no longer accessible, which is why the raw files are mirrored in `data/2016/`.
+Both datasets now have a crawler source. 2017 comes from the baotintuc.vn CDN,
+which is still live. 2016 comes from an aggregator article whose original host
+(`dtntbacgiang.edu.vn`) no longer resolves — its link list was recovered from
+the Internet Archive and is pointed at a mirror that is still online. See
+[data-pipeline](./data-pipeline.md#sources) for what that does and does not
+guarantee.
 
 ## History
 
@@ -53,7 +54,7 @@ Each year began as a standalone repository (`thptqg2016`, `thptqg2017`), merged
 here with full history. They initially kept separate frontends and separate
 copies of the same Rust parser, synchronised by hand. That duplication was
 removed: there is now one frontend, one parser, and one canonical schema, with
-per-dataset differences confined to four small config files and one registry
+per-dataset differences confined to one small config file and one registry
 entry each.
 
 The unification also fixed a latent data-loss bug — neither year's parser
diff --git a/docs/system-architecture.md b/docs/system-architecture.md
index ab5040d..803b756 100644
--- a/docs/system-architecture.md
+++ b/docs/system-architecture.md
@@ -3,23 +3,24 @@
 Static site, no backend. The browser downloads a compressed SQLite file at boot
 and every query runs locally via `sql.js` (SQLite compiled to WebAssembly).
 
-One frontend, one parser, one schema, four datasets.
+One frontend, one parser, one schema, two datasets.
 
 ## Data flow
 
 ```
+        ▲  crawler/  (Go — manual refresh only, never part of the build)
 data/<id>/*.xls(x)
         │
-        ▼  parser/  (Rust, one binary, one config per dataset)
+        ▼  go-parser/  (Go, one binary, one config per dataset)
    .build/public/db/<id>.db
         │
         ▼  gzip -9   (no -k: the raw file does not survive)
    .build/public/db/<id>.db.gz
         │
-        ▼  vite build   (publicDir = .build/public)
-   dist/
+        ▼  vite build   (root = web/, publicDir = .build/public)
+   web/dist/
         │
-        ▼  scripts/assemble-site.js
+        ▼  web/scripts/assemble-site.js
    _site/   →  GitHub Pages
         │
         ▼  browser
@@ -31,26 +32,28 @@ data/<id>/*.xls(x)
 One identifier ties the whole pipeline together:
 
 ```
-data/2017-old/  →  go-parser/configs/2017-old.yml  →  db/2017-old.db.gz  →  /thptqg/2017-old/
+data/2017/  →  go-parser/configs/2017.yml  →  db/2017.db.gz  →  /thptqg/2017/
 ```
 
-`src/datasets.js` declares the four ids once. The frontend, the database build
-(`go-parser/scripts/build-db.js`) and the site assembly all import that list, so
-adding a dataset means adding one entry and one config file.
+`web/src/datasets.js` declares the two ids once. The frontend, the database
+build (`go-parser/scripts/build-db.js`) and the site assembly all import that
+list, so adding a dataset means adding one entry and one config file.
+
+That import crosses a package boundary — go-parser reaches into the web
+workspace for it. It stays there because the same entries also carry the UI's
+labels and SQL presets, and splitting them would mean two lists to keep in step.
 
 | id | Exam | Rows | Source |
 | --- | --- | --- | --- |
 | `2016` | 2016 | 877,461 | Bộ GD&ĐT |
 | `2017` | 2017 | 861,068 | baotintuc.vn |
-| `2017-old` | 2017 | 847,348 | pre-refresh archive |
-| `2017-old2` | 2017 | 679,764 | corrected re-export |
 
 ## Canonical schema
 
 Defined once in `go-parser/internal/schema/schema.go` — DDL, INSERT, column order and the 16
-subject regexes. The four TOML configs carry no SQL at all, only per-dataset
-parse rules. Config parsing uses `deny_unknown_fields`, so a leftover `[schema]`
-block fails loudly instead of looking effective while `schema.rs` drives the
+subject regexes. The two YAML configs carry no SQL at all, only per-dataset
+parse rules. Config parsing sets `KnownFields(true)`, so a leftover `schema:`
+block fails loudly instead of looking effective while `schema.go` drives the
 build.
 
 ```sql
@@ -73,7 +76,7 @@ CREATE INDEX idx_ten_cum_thi  ON student(ten_cum_thi) WHERE ten_cum_thi IS NOT N
 
 Every dataset gets all 22 columns; ones it has no data for are NULL, costing
 about a byte per row. `khtn`, `khxh` and `gdcd` are empty on 2016;
-`ten_cum_thi` and `gioi_tinh` are empty on the 2017 datasets.
+`ten_cum_thi` and `gioi_tinh` are empty on 2017.
 
 `idx_ten_cum_thi` is partial, so it holds zero entries where the column is
 always NULL.
@@ -86,15 +89,13 @@ URLs are flat, one segment per dataset, and the segment is the id:
 /thptqg/            hub
 /thptqg/2016/
 /thptqg/2017/
-/thptqg/2017-old/
-/thptqg/2017-old2/
 ```
 
-`src/router.js` is an exact match on that segment. The nested form used before
-(`/thptqg/2017/old/`) would have needed longest-prefix matching, since it also
-starts with `/thptqg/2017/`. Those two legacy URLs still resolve: the router
-rewrites them to the flat equivalent with `history.replaceState`, preserving the
-query string so `?q=` deep links survive.
+`web/src/router.js` is an exact match on that segment. The nested form used
+before (`/thptqg/2017/old/`) would have needed longest-prefix matching, since it
+also starts with `/thptqg/2017/`. Both of those URLs addressed the two removed
+2017 archives, so the rewrite that kept them working has gone with them; they
+fall through to the hub like any other unknown path.
 
 A single Vite build emits one `index.html`. Because `base` is absolute
 (`/thptqg/`), that file references `/thptqg/assets/...` regardless of the
@@ -115,11 +116,11 @@ No component contains a per-dataset conditional. Two mechanisms do the work:
   self-exclude wherever those languages were not sat.
 
 Anything genuinely per-dataset — title, source, database size, search examples,
-SQL presets — lives in `src/datasets.js`.
+SQL presets — lives in `web/src/datasets.js`.
 
 ## Exam ID formats
 
-`src/lib/query-mode.js` decides whether a query is an exam ID or a name, and is
+`web/src/lib/query-mode.js` decides whether a query is an exam ID or a name, and is
 shared by `App.jsx` and `search-form.jsx` (they previously held separate copies
 and had drifted apart on exactly this rule).
 
@@ -135,7 +136,7 @@ candidates (70.3%). Letter prefixes are upper-cased before lookup, so
 
 ## Score tiers
 
-Six-level ladder in `scoreTier()` (`src/lib/admission-blocks.js`), paired with a
+Six-level ladder in `scoreTier()` (`web/src/lib/admission-blocks.js`), paired with a
 symbol so meaning is never colour-only.
 
 | Tier | Range | Vietnamese |
@@ -150,7 +151,7 @@ symbol so meaning is never colour-only.
 ## Admission blocks
 
 Vietnamese universities admit on three-subject combinations (khối thi).
-`src/lib/admission-blocks.js` lists the blocks computable from this schema
+`web/src/lib/admission-blocks.js` lists the blocks computable from this schema
 (A00–A11, B00–B08, C00–C20, D01–D15, plus D05/D06 for German and Japanese).
 `computeBlocks(student)` returns those where all three scores exist, sorted by
 total descending.
@@ -165,11 +166,11 @@ total descending.
 | Diacritics search | Pre-computed `ho_ten_ascii` | `LOWER(REPLACE(...))` at query time defeats the index |
 | SQL safety | Leading-keyword allowlist | `sql.js` is in-memory so writes cannot persist; the allowlist prevents confusion |
 | Row caps | 100 (lookup), 1000 (SQL) | Keeps DOM render sizes reasonable |
-| Routing | Hand-rolled, ~20 lines | Five static routes do not justify a router dependency |
+| Routing | Hand-rolled, ~15 lines | Three static routes do not justify a router dependency |
 
 ## Risks and limitations
 
-- **Database size.** 38–48 MB gzipped per dataset; slow links wait, mitigated by
+- **Database size.** 44–48 MB gzipped per dataset; slow links wait, mitigated by
   a progress bar.
 - **Browser memory.** The full database lives in RAM; older mobile devices may
   run out.
@@ -177,4 +178,4 @@ total descending.
   load. Self-hosting `sql-wasm.wasm` and updating `SQL_WASM_URL` in
   `use-sqlite.js` is the fix.
 - **Excel format drift.** A new source file with an unseen header layout needs a
-  new branch in `format_detect_2016.rs` or a new config.
+  new branch in `go-parser/internal/ingest/detect2016.go` or a new config.
diff --git a/eslint.config.js b/eslint.config.js
index 7ce9419..bbaa88e 100644
--- a/eslint.config.js
+++ b/eslint.config.js
@@ -5,7 +5,9 @@ import reactRefresh from 'eslint-plugin-react-refresh'
 import { defineConfig, globalIgnores } from 'eslint/config'
 
 export default defineConfig([
-  globalIgnores(['dist', '.build', '_site']),
+  // '**/dist', not 'dist': the build output moved to web/dist when the app
+  // became a workspace, and a root-anchored pattern would stop matching it.
+  globalIgnores(['**/dist', '.build', '_site']),
   {
     files: ['**/*.{js,jsx}'],
     extends: [
@@ -27,8 +29,9 @@ export default defineConfig([
     },
   },
   {
-    // Node-executed files (Vite config, parser tooling) run with Node globals.
-    files: ['vite.config.js', 'scripts/**/*.js', 'go-parser/scripts/**/*.js'],
+    // Node-executed files (Vite config, site assembly, parser tooling) run with
+    // Node globals. The crawler is Go, so it has nothing here.
+    files: ['web/vite.config.js', 'web/scripts/**/*.js', 'go-parser/scripts/**/*.js'],
     languageOptions: {
       globals: { ...globals.node },
     },
diff --git a/go-parser/configs/2016.yml b/go-parser/configs/2016.yml
index 412adcf..f823d44 100644
--- a/go-parser/configs/2016.yml
+++ b/go-parser/configs/2016.yml
@@ -1,7 +1,7 @@
 # thptqg2016 data/ — 4 .xls + 115 .xlsx mixed files.
 #
 # Three column layouts exist across the 119 files; the binary selects the
-# right one per-file at runtime via format_detection = "thptqg2016":
+# right one per-file at runtime via format_detection: thptqg2016:
 #
 #   separate-scores  header SBD(0)/HOTEN(1)/TOAN(2)...  — dhhanghai files
 #   mapped           header SOBAODANH|SBD + DIEM_THI     — most provinces
@@ -13,7 +13,7 @@
 # sheet_mode = "all": several provinces overflow into Sheet2 (65k Excel row cap).
 # strip_blank_rows = false: no blank-row anomaly observed in this dataset.
 #
-# Table shape, INSERT and subject regexes are canonical — see src/schema.rs.
+# Table shape, INSERT and subject regexes are canonical — see go-parser/internal/schema/schema.go.
 
 format_detection: thptqg2016
 
diff --git a/go-parser/configs/2017-old.yml b/go-parser/configs/2017-old.yml
deleted file mode 100644
index 4d10c04..0000000
--- a/go-parser/configs/2017-old.yml
+++ /dev/null
@@ -1,25 +0,0 @@
-# thptqg2017 data-old/ — 63 .xlsx files (pre-baotintuc refresh).
-#
-# sheet_mode = "first": single-sheet workbooks, never hit the 65k row cap.
-# SBD validation: require ^\d+$ (build-database-old.js:55 guard).
-# strip_blank_rows = false: no explicit blank-skip in build-database-old.js.
-#
-# Table shape, INSERT and subject regexes are canonical — see src/schema.rs.
-
-reader:
-  sheet_mode: first
-  strip_blank_rows: false
-
-columns:
-  ho_ten: 0
-  ngay_sinh: 1
-  so_bao_danh: 2
-  diem_thi: 3
-
-validation:
-  require_numeric_sbd: true
-  require_nonempty_name: true
-  require_nonempty_sbd: true
-
-header:
-  tokens: ["HO_TEN", "HỌ TÊN", "STT"]
diff --git a/go-parser/configs/2017-old2.yml b/go-parser/configs/2017-old2.yml
deleted file mode 100644
index dc9de9a..0000000
--- a/go-parser/configs/2017-old2.yml
+++ /dev/null
@@ -1,26 +0,0 @@
-# thptqg2017 data-old2/ — 54 .xlsx files (corrected-export set).
-#
-# sheet_mode = "all": 24.HCM_UTLQ.xlsx overflows into Sheet2 (+6,446 rows).
-# SBD validation: require ^\d+$ (build-database-old2.js:57 guard).
-# strip_blank_rows = true: skip fully blank rows BEFORE counting sourceRows
-#   (build-database-old2.js:50-51 checks blank before sourceRows++).
-#
-# Table shape, INSERT and subject regexes are canonical — see src/schema.rs.
-
-reader:
-  sheet_mode: all
-  strip_blank_rows: true
-
-columns:
-  ho_ten: 0
-  ngay_sinh: 1
-  so_bao_danh: 2
-  diem_thi: 3
-
-validation:
-  require_numeric_sbd: true
-  require_nonempty_name: true
-  require_nonempty_sbd: true
-
-header:
-  tokens: ["HO_TEN", "HỌ TÊN", "STT"]
diff --git a/go-parser/configs/2017.yml b/go-parser/configs/2017.yml
index 6f28b7b..49dcff6 100644
--- a/go-parser/configs/2017.yml
+++ b/go-parser/configs/2017.yml
@@ -4,7 +4,7 @@
 # SBD validation: no numeric guard — build-database.js did not apply ^\d+$.
 # strip_blank_rows = false: no blank-row anomaly in this dataset.
 #
-# Table shape, INSERT and subject regexes are canonical — see src/schema.rs.
+# Table shape, INSERT and subject regexes are canonical — see go-parser/internal/schema/schema.go.
 
 reader:
   sheet_mode: all
diff --git a/go-parser/internal/config/config_test.go b/go-parser/internal/config/config_test.go
index c3a1a1c..28f91fd 100644
--- a/go-parser/internal/config/config_test.go
+++ b/go-parser/internal/config/config_test.go
@@ -135,9 +135,8 @@ header:
 	}
 }
 
-// TestLoadRealConfigs loads the four shipped configs — the Go binary reads the
-// same files as Rust, never a fork — and asserts the per-dataset differences
-// recorded during scouting.
+// TestLoadRealConfigs loads the shipped configs and asserts the per-dataset
+// differences recorded during scouting.
 func TestLoadRealConfigs(t *testing.T) {
 	root := repoRoot(t)
 	want := map[string]struct {
@@ -148,10 +147,8 @@ func TestLoadRealConfigs(t *testing.T) {
 		formatDet  string
 		tokenCount int
 	}{
-		"2016":      {SheetModeAll, false, false, false, "thptqg2016", 6},
-		"2017":      {SheetModeAll, false, false, true, "", 3},
-		"2017-old":  {SheetModeFirst, false, true, true, "", 3},
-		"2017-old2": {SheetModeAll, true, true, true, "", 3},
+		"2016": {SheetModeAll, false, false, false, "thptqg2016", 6},
+		"2017": {SheetModeAll, false, false, true, "", 3},
 	}
 	for id, w := range want {
 		t.Run(id, func(t *testing.T) {
diff --git a/go-parser/scripts/build-db.js b/go-parser/scripts/build-db.js
index f3ac9da..51bca9d 100644
--- a/go-parser/scripts/build-db.js
+++ b/go-parser/scripts/build-db.js
@@ -2,8 +2,8 @@
 /**
  * Build the SQLite database for one or all datasets, verify it, then gzip it.
  *
- * The dataset list comes from src/datasets.js so it is written in exactly one
- * place.
+ * The dataset list comes from web/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
@@ -14,7 +14,7 @@
  * Until now nothing between the parser and the public site asserted that a
  * database actually had data in it. The parser logs a file-level failure and
  * continues, returns success regardless, and finishes cleanly even at zero rows;
- * this script gzipped whatever it got; and scripts/assemble-site.js only greps
+ * this script gzipped whatever it got; and web/scripts/assemble-site.js greps
  * *filenames* for stray .db files. So a reader that silently under-produced
  * would publish a truncated dataset with green CI and no red signal anywhere.
  *
@@ -32,7 +32,7 @@ import { dirname, resolve } from "node:path";
 import { fileURLToPath } from "node:url";
 import { DatabaseSync } from "node:sqlite";
 
-import { DATASET_IDS, DATASETS } from "../../src/datasets.js";
+import { DATASET_IDS, DATASETS } from "../../web/src/datasets.js";
 
 const ROOT = resolve(dirname(fileURLToPath(import.meta.url)), "../..");
 const BIN = resolve(ROOT, "go-parser/bin/xlsxread");
@@ -48,8 +48,6 @@ const OUT_DIR = resolve(ROOT, ".build/public/db");
 const EXPECTED_ROWS = {
   "2016": 877461,
   "2017": 861068,
-  "2017-old": 847348,
-  "2017-old2": 679764,
 };
 
 /** A gzipped database far below its usual size means a truncated build. */
diff --git a/go-parser/scripts/crawl-baotintuc.js b/go-parser/scripts/crawl-baotintuc.js
deleted file mode 100644
index 8f4c20c..0000000
--- a/go-parser/scripts/crawl-baotintuc.js
+++ /dev/null
@@ -1,137 +0,0 @@
-// Crawl Excel score files for all 63 provinces from baotintuc.vn article:
-// https://baotintuc.vn/tuyen-sinh/tra-cuu-diem-thi-thpt-2017-cua-63-tinh-thanh-pho-tren-baotintucvn-20170706073512672.htm
-// Saves to data/2017/<ascii-kebab-name>.xls (63 files). Idempotent: skips already-downloaded files.
-import fs from "fs";
-import path from "path";
-import { fileURLToPath } from "url";
-
-const __dirname = path.dirname(fileURLToPath(import.meta.url));
-const OUT_DIR = path.join(__dirname, "..", "..", "data", "2017");
-
-const LINKS = [
-  ["An Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/17/57/Angiang.xls"],
-  ["Bac Lieu", "https://cdnmedia.baotintuc.vn/2017/07/06/17/59/Baclieu.xls"],
-  ["Ba Ria - Vung Tau", "https://cdnmedia.baotintuc.vn/2017/07/06/08/17/1BaRiaVungTau.xls"],
-  ["Bac Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/09/33/BacGiang.xls"],
-  ["Bac Kan", "https://cdnmedia.baotintuc.vn/2017/07/06/08/27/BacKan.xls"],
-  ["Bac Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/33/BacNinh.xls"],
-  ["Ben Tre", "https://cdnmedia.baotintuc.vn/2017/07/06/09/34/BenTre.xls"],
-  ["Binh Duong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/34/BinhDuong.xls"],
-  ["Binh Thuan", "https://cdnmedia.baotintuc.vn/2017/07/06/09/35/BinhThuan.xls"],
-  ["Binh Phuoc", "https://cdnmedia.baotintuc.vn/2017/07/06/09/09/BinhPhuoc.xls"],
-  ["Binh Dinh", "https://cdnmedia.baotintuc.vn/2017/07/06/13/05/Binhdinh.xls"],
-  ["Ca Mau", "https://cdnmedia.baotintuc.vn/2017/07/06/13/05/Camau.xls"],
-  ["Cao Bang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/02/Caobang.xls"],
-  ["Can Tho", "https://cdnmedia.baotintuc.vn/2017/07/06/13/07/Cantho.xls"],
-  ["Da Nang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/03/Danang.xls"],
-  ["Dak Nong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/35/DakNong.xls"],
-  ["Dak Lak", "https://cdnmedia.baotintuc.vn/2017/07/06/18/02/Daklak.xls"],
-  ["Dong Nai", "https://cdnmedia.baotintuc.vn/2017/07/06/13/08/dongnai.xls"],
-  ["Dong Thap", "https://cdnmedia.baotintuc.vn/2017/07/06/17/59/Dongthap.xls"],
-  ["Dien Bien", "https://cdnmedia.baotintuc.vn/2017/07/06/09/08/DienBien.xls"],
-  ["Gia Lai", "https://cdnmedia.baotintuc.vn/2017/07/06/13/09/Gia-Lai.xls"],
-  ["Ha Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/18/Hagiang.xls"],
-  ["Ha Noi", "https://cdnmedia.baotintuc.vn/2017/07/07/08/16/HaNoi.xls"],
-  ["Ha Nam", "https://cdnmedia.baotintuc.vn/2017/07/06/09/07/Hanam.xls"],
-  ["Ha Tinh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/36/HaTinh.xls"],
-  ["Hai Phong", "https://cdnmedia.baotintuc.vn/2017/07/06/08/18/23HaiPhong.xls"],
-  ["Hai Duong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/36/HaiDuong.xls"],
-  ["Hau Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/00/Haugiang.xls"],
-  ["Ho Chi Minh", "https://cdnmedia.baotintuc.vn/2017/07/06/08/26/HCM.xls"],
-  ["Hoa Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/06/HoaBinh.xls"],
-  ["Hung Yen", "https://cdnmedia.baotintuc.vn/2017/07/06/08/25/HungYen.xls"],
-  ["Khanh Hoa", "https://cdnmedia.baotintuc.vn/2017/07/06/18/04/Khanhhoa.xls"],
-  ["Kien Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/08/28/KienGiang.xls"],
-  ["Kon Tum", "https://cdnmedia.baotintuc.vn/2017/07/06/18/05/KonTum.xls"],
-  ["Nam Dinh", "https://cdnmedia.baotintuc.vn/2017/07/06/07/55/13NamDinh.xls"],
-  ["Nghe An", "https://cdnmedia.baotintuc.vn/2017/07/06/07/55/14NgheAn.xls"],
-  ["Ninh Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/37/NinhBinh.xls"],
-  ["Ninh Thuan", "https://cdnmedia.baotintuc.vn/2017/07/06/18/00/Ninhthuan.xls"],
-  ["Lao Cai", "https://cdnmedia.baotintuc.vn/2017/07/06/08/52/11LaoCai.xls"],
-  ["Lai Chau", "https://cdnmedia.baotintuc.vn/2017/07/06/13/33/LaiChau.xls"],
-  ["Lang Son", "https://cdnmedia.baotintuc.vn/2017/07/06/18/05/Langson.xls"],
-  ["Lam Dong", "https://cdnmedia.baotintuc.vn/2017/07/06/09/00/LamDong.xls"],
-  ["Long An", "https://cdnmedia.baotintuc.vn/2017/07/06/08/53/12LongAn.xls"],
-  ["Quang Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/23/QuangBinh.xls"],
-  ["Quang Nam", "https://cdnmedia.baotintuc.vn/2017/07/06/09/24/QuangNam.xls"],
-  ["Quang Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/18/06/Quangninh.xls"],
-  ["Quang Ngai", "https://cdnmedia.baotintuc.vn/2017/07/06/09/24/QuangNgai.xls"],
-  ["Quang Tri", "https://cdnmedia.baotintuc.vn/2017/07/06/09/25/QuangTri.xls"],
-  ["Phu Tho", "https://cdnmedia.baotintuc.vn/2017/07/06/09/23/PhuTho.xls"],
-  ["Phu Yen", "https://cdnmedia.baotintuc.vn/2017/07/06/09/38/PhuYen.xls"],
-  ["Son La", "https://cdnmedia.baotintuc.vn/2017/07/06/18/01/Sonla.xls"],
-  ["Soc Trang", "https://cdnmedia.baotintuc.vn/2017/07/06/18/06/Soctrang.xls"],
-  ["Tay Ninh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/26/TayNinh.xls"],
-  ["Thai Binh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/09/ThaiBinh.xls"],
-  ["Thai Nguyen", "https://cdnmedia.baotintuc.vn/2017/07/06/13/35/ThaiNguyen.xls"],
-  ["Thanh Hoa", "https://cdnmedia.baotintuc.vn/2017/07/06/13/10/thanhoa.xls"],
-  ["Tra Vinh", "https://cdnmedia.baotintuc.vn/2017/07/06/09/38/TraVinh.xls"],
-  ["Thua Thien Hue", "https://cdnmedia.baotintuc.vn/2017/07/06/13/10/thuathienhue.xls"],
-  ["Tien Giang", "https://cdnmedia.baotintuc.vn/2017/07/06/13/11/tiengiang.xls"],
-  ["Tuyen Quang", "https://cdnmedia.baotintuc.vn/2017/07/06/09/39/TuyenQuang.xls"],
-  ["Vinh Phuc", "https://cdnmedia.baotintuc.vn/2017/07/06/09/40/VinhPhuc.xls"],
-  ["Vinh Long", "https://cdnmedia.baotintuc.vn/2017/07/06/13/13/vinhlong.xls"],
-  ["Yen Bai", "https://cdnmedia.baotintuc.vn/2017/07/06/18/07/Yenbai.xls"],
-];
-
-function slug(name) {
-  return name.toLowerCase().replace(/[^a-z0-9]+/g, "-").replace(/^-|-$/g, "");
-}
-
-async function downloadOne(name, url) {
-  const outPath = path.join(OUT_DIR, `${slug(name)}.xls`);
-  if (fs.existsSync(outPath) && fs.statSync(outPath).size > 0) {
-    return { name, url, outPath, status: "skip", size: fs.statSync(outPath).size };
-  }
-  const res = await fetch(url, {
-    headers: {
-      "User-Agent":
-        "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0 Safari/537.36",
-      Referer:
-        "https://baotintuc.vn/tuyen-sinh/tra-cuu-diem-thi-thpt-2017-cua-63-tinh-thanh-pho-tren-baotintucvn-20170706073512672.htm",
-    },
-  });
-  if (!res.ok) {
-    return { name, url, outPath, status: "fail", httpStatus: res.status };
-  }
-  const buf = Buffer.from(await res.arrayBuffer());
-  fs.writeFileSync(outPath, buf);
-  return { name, url, outPath, status: "ok", size: buf.length };
-}
-
-async function main() {
-  fs.mkdirSync(OUT_DIR, { recursive: true });
-  console.log(`Downloading ${LINKS.length} files to ${OUT_DIR}...`);
-
-  const CONCURRENCY = 6;
-  const results = [];
-  let i = 0;
-  async function worker() {
-    while (i < LINKS.length) {
-      const idx = i++;
-      const [name, url] = LINKS[idx];
-      try {
-        const r = await downloadOne(name, url);
-        results.push(r);
-        const tag = r.status === "ok" ? "✓" : r.status === "skip" ? "·" : "✗";
-        const sz = r.size ? `${(r.size / 1024).toFixed(0)} KB` : "";
-        console.log(`  ${tag} [${idx + 1}/${LINKS.length}] ${name.padEnd(20)} ${sz} ${r.status === "fail" ? "HTTP " + r.httpStatus : ""}`);
-      } catch (err) {
-        console.log(`  ✗ [${idx + 1}/${LINKS.length}] ${name}: ${err.message}`);
-        results.push({ name, url, status: "error", message: err.message });
-      }
-    }
-  }
-  await Promise.all(Array.from({ length: CONCURRENCY }, worker));
-
-  const ok = results.filter((r) => r.status === "ok").length;
-  const skip = results.filter((r) => r.status === "skip").length;
-  const fail = results.filter((r) => r.status === "fail" || r.status === "error");
-  console.log(`\nDone. ok=${ok} skip=${skip} fail=${fail.length}`);
-  if (fail.length) {
-    console.log("Failed:", fail.map((f) => f.name).join(", "));
-    process.exit(1);
-  }
-}
-
-main();
diff --git a/go-parser/testdata/reader-fidelity-hashes.tsv b/go-parser/testdata/reader-fidelity-hashes.tsv
index 8c9ce2f..76bb266 100644
--- a/go-parser/testdata/reader-fidelity-hashes.tsv
+++ b/go-parser/testdata/reader-fidelity-hashes.tsv
@@ -190,120 +190,3 @@ data/2017/tuyen-quang.xls	053d293e4acd9dfa6ce9532a047cc64b4d2edee1ff13b387723586
 data/2017/vinh-long.xls	6e7584a3047b0ba7487dbb8c71f5b594db220f11b747149d41d31ea97e1e4eab
 data/2017/vinh-phuc.xls	9753e2d864584e76aef1e690e4e2de98b28240fbba810e614eab9926668d66d2
 data/2017/yen-bai.xls	fe6824a1254116bc6ed372e0a1fd635a816a09e010b7efde9659baca2e143138
-data/2017-old/10_BinhThuan_RIIW.xls.xlsx	2833e7366ea1a21ea869978a6bb920b033160bb8042fe0bfe9b7573dcfe85409
-data/2017-old/10_Ca_Mau_BKXT.xls.xlsx	53cb69b13325cb9cc9004b8c89f0cd1c85b5f943e7c78c8b010ddabad3f816c9
-data/2017-old/10_LamDong_GNFT.xls.xlsx	f05b4cdc8a9a8413b29801fbddeb48a7c60d3c6bb2228bbe238113495e89bc98
-data/2017-old/10_Soc_Trang_XCGJ.xls.xlsx	8e540cef6f2ff387f3e825a5aa3190af394c6f9c7e9dfc66d5f7032ee47f7f14
-data/2017-old/11_BinhDuong_RYQL.xls.xlsx	11953116054a2889929c3fe05b68e8932c7b79ffd1b85d697afaac63fcdd9adf
-data/2017-old/11_LaoCai_GMSU.xls.xlsx	5a73becf1bdb8027102faff74035d1df71e13c1cac39cc38ec6efaab6699ca52
-data/2017-old/12_BenTre_DKWF.xls.xlsx	579c0d94b20440f1c60a94fb0c762b71dd1c7719c8aa4db0b9418ecaba9b605e
-data/2017-old/12_LongAn_ZZUK.xls.xlsx	c4cf569706bfc30582d774ba9e699dcbc4bdc734f46bf6493df58eaecc37a755
-data/2017-old/13_NamDinh_ESEL.xls.xlsx	7e5256cb095262fef5399092f8f843c1ee82e117bca8e6571e363646bb1170a6
-data/2017-old/13_TraVinh_LKUJ.xls.xlsx	2b7321d68a4a910c4a506d7ca207531cfbdd9619f1d8ba0ce475baf1b91043aa
-data/2017-old/14_NgheAn_BSLY.xls.xlsx	034234a8a5461910f8beff3977aafa29a6541c8a56c007816118e5c4eb2b7218
-data/2017-old/15_PhuTho_ABWQ.xls.xlsx	9891c17001cfcc49ccd587419339698ef59b2dd43b5ca7f2765b09a17c26659d
-data/2017-old/16_QuangBinh_KGEU.xls.xlsx	01e249d45c1269d62ae1977bef0bb192f833941510471b723e57c32adb0eadaa
-data/2017-old/17_QuangNam_AMTK.xls.xlsx	4c9f05d24d861ec3c82a1c564e4baa1541bd4882785f143459a1a81a6b8359c7
-data/2017-old/18_QuangNgai_KOFP.xls.xlsx	ee9c26f2c16d36dabfde79901a054d2266c28c4acbfd19b628cfe50f8eb03957
-data/2017-old/19_QuangTri_OMZF.xls.xlsx	6c961bb461274c73b4589de4087624a9ef5edd43c04ea54675148df00f6420e5
-data/2017-old/1_BaRia_VungTau_HJKG.xls.xlsx	cb03f933442a8cea6b1622da4e5f804ab476dff9f3462b404e957b6a2e87b2cd
-data/2017-old/1_Da_Nang_AHWJ.xls.xlsx	59892fc01b6cb10497679f2bc69f59023215755d2f00fd764b23b895c98be2be
-data/2017-old/1_Ha_Noi_CVXG.xls.xlsx	2718fe9c2b58843621c0082da5880a4b957266525b83123af02c78fd51d4f4a5
-data/2017-old/1_Son_La_JIDP.xls.xlsx	0a0e273c5def980406b6316cceb63e5cee53aae56e112dfae7e795261eb1cd06
-data/2017-old/1_TuyenQuang_JBYF.xls.xlsx	45826ed10248268ac3344d0b592cf2dba6f86278b7402dc60626a76b18fa4d6b
-data/2017-old/20_TayNinh_ILFA.xls.xlsx	684a1820d02f658904c87c26aa195faa6aceb3d7ca00c674df11a63f9274ab14
-data/2017-old/21_ThaiBinh_FTVG.xls.xlsx	1a12e43d4ea368b8e3a2e4d881b5d66214a16140219e80e2e72fcb269899b285
-data/2017-old/22_ThaiNguyen_TLTW.xls.xlsx	29227bdc3204591327e803680525d8621228fc371ce7f802dc667ec70eb8d7b3
-data/2017-old/23_HaiPhong_HXBV.xls.xlsx	e450e2344fbd724555de9892412f521414fc1fcc53e69887c6381dd266b4e92c
-data/2017-old/24_HCM_XULN.xls.xlsx	49ed22e823607b78920a24f5999eaf1ef8634a141108dfd4bf970ee000aecbf9
-data/2017-old/2_BacKan_GFVQ.xls.xlsx	77a08fc4edb41d639fcdd798980903519181639db00a360ef0c6c6c2bd88f529
-data/2017-old/2_Ha_Giang_QNCM.xls.xlsx	6bf4d62413c1b92f88f980e5e3e414a7e25971c7546c6390c7e67a6b0f742b68
-data/2017-old/2_Ninh_Thuan_VHLY.xls.xlsx	6dfb6517f9ccc3420c222b6a88aabac54f6f25f6b743d993d6a5f28b4fa78ff2
-data/2017-old/2_Thanh_Hoa_AUFV.xls.xlsx	2f2ce90425c19a9089024061f328f37b9ef580ddadc076ddcf452f549690f4b3
-data/2017-old/2_VinhPhuc_GUDK.xls.xlsx	fc0e4952445ebda1af28f7330aa808c032f2e3a7a3c03dfbb9692bf6a248867a
-data/2017-old/3_BacGiang_TOIF.xls.xlsx	8043510ff89f11ba063f898dc1d2c8879a8da9063a420308647e20a2f39d3092
-data/2017-old/3_BinhPhuoc_YFMU.xls.xlsx	8869970f4e6005381264f3f4bb508c5181f15ae990a75fe904a65b305c0a123a
-data/2017-old/3_Cao_Bang_CIEY.xls.xlsx	94a06bdb2333a9f44216001ba8d8463700b8cf3e51944b94db11eef7eac8dbd3
-data/2017-old/3_Dong_Thap_GSXQ.xls.xlsx	0cdd0763fc4f20f1d990621ca9e8dab107a3d4eb7c0ec3e53a4d8c240fbc569f
-data/2017-old/3_Thua_Thien_Hue_HMDB.xls.xlsx	62e54eebda7bbd542ca63443bb923a81b472e7892e55dce39fd0611fc0f00abd
-data/2017-old/4_An_Giang_JNOS.xls.xlsx	778ad205bbabb7c4aea3df34325770aa4a1daec0a1683f73f0405ac1cc5aaf9a
-data/2017-old/4_BacNinh_STLR.xls.xlsx	009c3b7821f55210a86f9ace74c31eb0d5c16aafdb0cac8f2b71cac616abdfec
-data/2017-old/4_Binh_Dinh_WWZW.xls.xlsx	fbacd256d58627377d698462d69aac0e83d124c8c8ff1f5b63ba0bfb5442dd39
-data/2017-old/4_DienBien_SOJG.xls.xlsx	4931e8d6604013df48b2b469fda3eb05fdf1203b67bbc2ae4ff6d7949a115a24
-data/2017-old/4_Lang_Son_NYQL.xls.xlsx	cc7d2454caadb18740ed6baf6edc62baa47e98b3a907b3ce06f72b9a16b7838a
-data/2017-old/5_Bac_Lieu_XEKH.xls.xlsx	dc1dca45fd5357b45abc930b0ebb95ab589ef80a09bef8b5ad4dab9e00be87e5
-data/2017-old/5_Gia_Lai_ABZI.xls.xlsx	5b7be25274d628c284c37efb3df0587171cc9273dc3bd8dc824c7bd1b8bccf1a
-data/2017-old/5_HaiDuong_WWWG.xls.xlsx	1d800d0aa7338fb507c4eeb3e752620d2608c30ac820895ca086325dbaf3643d
-data/2017-old/5_Hanam_QGJS.xls.xlsx	cfc233ac37ab8e263b6cc569815ea2c687da30364f139109032c8f2cac1ad98e
-data/2017-old/5_Yen_Bai_FAQR.xls.xlsx	4c8915417d8747c291f643ea4b378c87982c397460e0e9e3cc6df19670d6720f
-data/2017-old/6_Dong_Nai_WOTM.xls.xlsx	9b582a8421c419156c2eab6e532d854736385735f8208f9d390c251f22ac764e
-data/2017-old/6_Hau_Giang_KWDM.xls.xlsx	c11473e3e1f964302c5378c90dd6b9d478eb11397351b7ef74df9ae709c9e908
-data/2017-old/6_HoaBinh_TPZY.xls.xlsx	e1210961dc821642f6951cd6e4a71b5fbc81d5070ee3816b7c8b5dc56bf2b234
-data/2017-old/6_NinhBinh_IGFT.xls.xlsx	9399b8ab58e2a7a1c0ebf5425359319e567f8a28602349aa587efced72142fc8
-data/2017-old/6_Quang_Ninh_DQCJ.xls.xlsx	8cf95259b11d48267f710f151c3f1419ccac49d823b78b70a2708c59ab25a4f7
-data/2017-old/7_HaTinh_DDHD.xls.xlsx	d96df9f2c8365dfc73be3e10c268367b93fd00fe92bb39c745d8e353ba75c05b
-data/2017-old/7_HungYen_LTIK.xls.xlsx	e80eb38af8307c8e193a91bdda5368c437693d7669e281b543bbe891a46e6f36
-data/2017-old/7_Kon_Tum_RSLR.xls.xlsx	41d2adea059449b6e214241414748f10776861d261cbaa1fd213629757deb846
-data/2017-old/7_Tien_Giang_EFHX.xls.xlsx	00d9bbc70be89f8f7dad6aff990ad36236c6b7244ddeb2e3d9b3d8ed1a6e21d0
-data/2017-old/8_Can_Tho_RQZM.xls.xlsx	3a1182b26d34fbdfabf4326098ff265609d8f1fe2b766b6eef897db9685d4cd4
-data/2017-old/8_Dak_Lak_YKPR.xls.xlsx	379b45f8a347bf4842daf9402d84be47426ed7107a64b2e7710bdc8e71eb57be
-data/2017-old/8_KienGiang_OTOB.xls.xlsx	3e1789b8565a5ae7c187598cb7e9c3261029c06bc7f0cccc4eb503e8e302cb18
-data/2017-old/8_PhuYen_OGIM.xls.xlsx	91d1f202cba3bf4c57a7f26e7611d3d29faf8dd13de30552cbfc96eab7a451e9
-data/2017-old/9_DakNong_FBOP.xls.xlsx	c611bf7f1d1c95cbfc78bdab3ea99bf8278e94c8db09a9fd3ccac55a538fe0b2
-data/2017-old/9_Khanh_Hoa_KPKQ.xls.xlsx	0e90c87dd74a5eb293a0a71b93da45653a19ed2e3a2600d3b880a867a695f31d
-data/2017-old/9_LaiChau_ALKN.xls.xlsx	3b7a022f7727134620c6756925bd5adb041e2d80cabd4aab13a4063501739514
-data/2017-old/9_Vinh_Long_OLWA.xls.xlsx	43d7abb9c6b8d3205ce90f9cceaf9623875c87998d9e030c6dc8da9bc18bfa09
-data/2017-old2/10.BinhThuan_MVVG.xlsx	2833e7366ea1a21ea869978a6bb920b033160bb8042fe0bfe9b7573dcfe85409
-data/2017-old2/10.LamDong_YUQA.xlsx	f05b4cdc8a9a8413b29801fbddeb48a7c60d3c6bb2228bbe238113495e89bc98
-data/2017-old2/10.Soc Trang_LQWU.xlsx	8e540cef6f2ff387f3e825a5aa3190af394c6f9c7e9dfc66d5f7032ee47f7f14
-data/2017-old2/11.BinhDuong_HVAH.xlsx	11953116054a2889929c3fe05b68e8932c7b79ffd1b85d697afaac63fcdd9adf
-data/2017-old2/11.LaoCai_ZTBP.xlsx	5a73becf1bdb8027102faff74035d1df71e13c1cac39cc38ec6efaab6699ca52
-data/2017-old2/12.BenTre_NQTU.xlsx	579c0d94b20440f1c60a94fb0c762b71dd1c7719c8aa4db0b9418ecaba9b605e
-data/2017-old2/12.LongAn_PDRH.xlsx	d1c65184b475fc29ae273cc89b8574b7e273bc3dab83cd39489d72a5de1bb1e4
-data/2017-old2/13.NamDinh_NAYR.xlsx	7e5256cb095262fef5399092f8f843c1ee82e117bca8e6571e363646bb1170a6
-data/2017-old2/13.TraVinh_FODZ.xlsx	2b7321d68a4a910c4a506d7ca207531cfbdd9619f1d8ba0ce475baf1b91043aa
-data/2017-old2/14.NgheAn_HTKD.xlsx	034234a8a5461910f8beff3977aafa29a6541c8a56c007816118e5c4eb2b7218
-data/2017-old2/15.PhuTho_IJZW.xlsx	9891c17001cfcc49ccd587419339698ef59b2dd43b5ca7f2765b09a17c26659d
-data/2017-old2/17.QuangNam_NQMG.xlsx	4c9f05d24d861ec3c82a1c564e4baa1541bd4882785f143459a1a81a6b8359c7
-data/2017-old2/18.QuangNgai_IUPY.xlsx	ee9c26f2c16d36dabfde79901a054d2266c28c4acbfd19b628cfe50f8eb03957
-data/2017-old2/19.QuangTri_MKNN.xlsx	6c961bb461274c73b4589de4087624a9ef5edd43c04ea54675148df00f6420e5
-data/2017-old2/1.BaRia-VungTau_PGZT.xlsx	cb03f933442a8cea6b1622da4e5f804ab476dff9f3462b404e957b6a2e87b2cd
-data/2017-old2/1.Da Nang_ABWU.xlsx	59892fc01b6cb10497679f2bc69f59023215755d2f00fd764b23b895c98be2be
-data/2017-old2/1.Son La_XLFN.xlsx	0a0e273c5def980406b6316cceb63e5cee53aae56e112dfae7e795261eb1cd06
-data/2017-old2/1.TuyenQuang_PTMR.xlsx	45826ed10248268ac3344d0b592cf2dba6f86278b7402dc60626a76b18fa4d6b
-data/2017-old2/20.TayNinh_KJAQ.xlsx	684a1820d02f658904c87c26aa195faa6aceb3d7ca00c674df11a63f9274ab14
-data/2017-old2/21.ThaiBinh_KTQN.xlsx	1a12e43d4ea368b8e3a2e4d881b5d66214a16140219e80e2e72fcb269899b285
-data/2017-old2/22.ThaiNguyen_BKIF.xlsx	29227bdc3204591327e803680525d8621228fc371ce7f802dc667ec70eb8d7b3
-data/2017-old2/24.HCM_UTLQ.xlsx	49ed22e823607b78920a24f5999eaf1ef8634a141108dfd4bf970ee000aecbf9
-data/2017-old2/2.BacKan_YQNX.xlsx	77a08fc4edb41d639fcdd798980903519181639db00a360ef0c6c6c2bd88f529
-data/2017-old2/2.Ha Giang_PIYK.xlsx	6bf4d62413c1b92f88f980e5e3e414a7e25971c7546c6390c7e67a6b0f742b68
-data/2017-old2/2.Ninh Thuan_BAGG.xlsx	6dfb6517f9ccc3420c222b6a88aabac54f6f25f6b743d993d6a5f28b4fa78ff2
-data/2017-old2/2.Thanh Hoa_UOPE.xlsx	2f2ce90425c19a9089024061f328f37b9ef580ddadc076ddcf452f549690f4b3
-data/2017-old2/2.VinhPhuc_QZJK.xlsx	fc0e4952445ebda1af28f7330aa808c032f2e3a7a3c03dfbb9692bf6a248867a
-data/2017-old2/3.BacGiang_SAVS.xlsx	8043510ff89f11ba063f898dc1d2c8879a8da9063a420308647e20a2f39d3092
-data/2017-old2/3.BinhPhuoc_IPHL.xlsx	8869970f4e6005381264f3f4bb508c5181f15ae990a75fe904a65b305c0a123a
-data/2017-old2/3.Cao Bang_WMUU.xlsx	94a06bdb2333a9f44216001ba8d8463700b8cf3e51944b94db11eef7eac8dbd3
-data/2017-old2/3.Dong Thap_HKJX.xlsx	0cdd0763fc4f20f1d990621ca9e8dab107a3d4eb7c0ec3e53a4d8c240fbc569f
-data/2017-old2/3.Thua Thien -Hue_MAET.xlsx	62e54eebda7bbd542ca63443bb923a81b472e7892e55dce39fd0611fc0f00abd
-data/2017-old2/4.An Giang_PMJD.xlsx	778ad205bbabb7c4aea3df34325770aa4a1daec0a1683f73f0405ac1cc5aaf9a
-data/2017-old2/4.BacNinh_NNIS.xlsx	009c3b7821f55210a86f9ace74c31eb0d5c16aafdb0cac8f2b71cac616abdfec
-data/2017-old2/4.Binh Dinh_VOMJ.xlsx	fbacd256d58627377d698462d69aac0e83d124c8c8ff1f5b63ba0bfb5442dd39
-data/2017-old2/4.DienBien_FYGN.xlsx	4931e8d6604013df48b2b469fda3eb05fdf1203b67bbc2ae4ff6d7949a115a24
-data/2017-old2/4.Lang Son_QWOG.xlsx	cc7d2454caadb18740ed6baf6edc62baa47e98b3a907b3ce06f72b9a16b7838a
-data/2017-old2/5.Bac Lieu_VIVY.xlsx	dc1dca45fd5357b45abc930b0ebb95ab589ef80a09bef8b5ad4dab9e00be87e5
-data/2017-old2/5.Gia Lai_TAAS.xlsx	5b7be25274d628c284c37efb3df0587171cc9273dc3bd8dc824c7bd1b8bccf1a
-data/2017-old2/5.HaiDuong_WNHD.xlsx	1d800d0aa7338fb507c4eeb3e752620d2608c30ac820895ca086325dbaf3643d
-data/2017-old2/5.Hanam_SDKN.xlsx	cfc233ac37ab8e263b6cc569815ea2c687da30364f139109032c8f2cac1ad98e
-data/2017-old2/5.Yen Bai_BSLV.xlsx	4c8915417d8747c291f643ea4b378c87982c397460e0e9e3cc6df19670d6720f
-data/2017-old2/6.Hau Giang_SIAJ.xlsx	c11473e3e1f964302c5378c90dd6b9d478eb11397351b7ef74df9ae709c9e908
-data/2017-old2/6.HoaBinh_HLYQ.xlsx	e1210961dc821642f6951cd6e4a71b5fbc81d5070ee3816b7c8b5dc56bf2b234
-data/2017-old2/6.NinhBinh_PKMQ.xlsx	9399b8ab58e2a7a1c0ebf5425359319e567f8a28602349aa587efced72142fc8
-data/2017-old2/6.Quang Ninh_YAKJ.xlsx	8cf95259b11d48267f710f151c3f1419ccac49d823b78b70a2708c59ab25a4f7
-data/2017-old2/7.HaTinh_XMFQ.xlsx	d96df9f2c8365dfc73be3e10c268367b93fd00fe92bb39c745d8e353ba75c05b
-data/2017-old2/7.HungYen_TCBE.xlsx	e80eb38af8307c8e193a91bdda5368c437693d7669e281b543bbe891a46e6f36
-data/2017-old2/7.Kon Tum_CAQU.xlsx	41d2adea059449b6e214241414748f10776861d261cbaa1fd213629757deb846
-data/2017-old2/7.Tien Giang_AOWZ.xlsx	00d9bbc70be89f8f7dad6aff990ad36236c6b7244ddeb2e3d9b3d8ed1a6e21d0
-data/2017-old2/8.PhuYen_MLTQ.xlsx	9dd53293f99d86801692a1995a5cf7a08af8394b820c76e68c88c2b7644b9b67
-data/2017-old2/9.DakNong_ZWJT.xlsx	c611bf7f1d1c95cbfc78bdab3ea99bf8278e94c8db09a9fd3ccac55a538fe0b2
-data/2017-old2/9.Khanh Hoa_IBTX.xlsx	0e90c87dd74a5eb293a0a71b93da45653a19ed2e3a2600d3b880a867a695f31d
-data/2017-old2/9.Vinh Long_WVKI.xlsx	43d7abb9c6b8d3205ce90f9cceaf9623875c87998d9e030c6dc8da9bc18bfa09
diff --git a/package-lock.json b/package-lock.json
index cc1c89f..e755a6a 100644
--- a/package-lock.json
+++ b/package-lock.json
@@ -8,21 +8,15 @@
       "name": "thptqg",
       "version": "1.0.0",
       "license": "ISC",
-      "dependencies": {
-        "react": "^19.2.4",
-        "react-dom": "^19.2.4",
-        "sql.js": "^1.14.1"
-      },
+      "workspaces": [
+        "web"
+      ],
       "devDependencies": {
         "@eslint/js": "^9.39.4",
-        "@types/react": "^19.2.14",
-        "@types/react-dom": "^19.2.3",
-        "@vitejs/plugin-react": "^6.0.1",
         "eslint": "^9.39.4",
         "eslint-plugin-react-hooks": "^7.0.1",
         "eslint-plugin-react-refresh": "^0.5.2",
-        "globals": "^17.4.0",
-        "vite": "^8.0.16"
+        "globals": "^17.4.0"
       }
     },
     "node_modules/@babel/code-frame": {
@@ -2347,6 +2341,10 @@
         "node": ">=8"
       }
     },
+    "node_modules/thptqg-web": {
+      "resolved": "web",
+      "link": true
+    },
     "node_modules/tinyglobby": {
       "version": "0.2.17",
       "resolved": "https://registry.npmjs.org/tinyglobby/-/tinyglobby-0.2.17.tgz",
@@ -2564,6 +2562,21 @@
       "peerDependencies": {
         "zod": "^3.25.0 || ^4.0.0"
       }
+    },
+    "web": {
+      "name": "thptqg-web",
+      "version": "1.0.0",
+      "dependencies": {
+        "react": "^19.2.4",
+        "react-dom": "^19.2.4",
+        "sql.js": "^1.14.1"
+      },
+      "devDependencies": {
+        "@types/react": "^19.2.14",
+        "@types/react-dom": "^19.2.3",
+        "@vitejs/plugin-react": "^6.0.1",
+        "vite": "^8.0.16"
+      }
     }
   }
 }
diff --git a/package.json b/package.json
index 3ba77e3..55325af 100644
--- a/package.json
+++ b/package.json
@@ -4,36 +4,33 @@
   "version": "1.0.0",
   "type": "module",
   "description": "Tra cứu điểm thi THPT Quốc gia — 2016 và 2017",
+  "workspaces": [
+    "web"
+  ],
   "scripts": {
+    "crawl:2016": "go -C crawler run ./cmd/crawl 2016",
+    "crawl:2017": "go -C crawler run ./cmd/crawl 2017",
+    "test:crawler": "go -C crawler test ./...",
     "build:go": "go -C go-parser build -o bin/xlsxread ./cmd/xlsxread",
     "build:db": "node go-parser/scripts/build-db.js",
-    "dev": "vite",
-    "build": "vite build",
-    "assemble": "node scripts/assemble-site.js",
+    "test:go": "go -C go-parser test ./...",
+    "dev": "npm run dev -w web",
+    "build": "npm run build -w web",
+    "preview": "npm run preview -w web",
+    "assemble": "node web/scripts/assemble-site.js",
     "build:site": "npm run build && npm run assemble",
-    "preview": "vite preview",
-    "lint": "eslint .",
-    "test:go": "go -C go-parser test ./..."
+    "lint": "eslint ."
   },
   "repository": {
     "type": "git",
     "url": "git+https://github.com/tiennm99/thptqg.git"
   },
   "license": "ISC",
-  "dependencies": {
-    "react": "^19.2.4",
-    "react-dom": "^19.2.4",
-    "sql.js": "^1.14.1"
-  },
   "devDependencies": {
     "@eslint/js": "^9.39.4",
-    "@types/react": "^19.2.14",
-    "@types/react-dom": "^19.2.3",
-    "@vitejs/plugin-react": "^6.0.1",
     "eslint": "^9.39.4",
     "eslint-plugin-react-hooks": "^7.0.1",
     "eslint-plugin-react-refresh": "^0.5.2",
-    "globals": "^17.4.0",
-    "vite": "^8.0.16"
+    "globals": "^17.4.0"
   }
 }
diff --git a/index.html b/web/index.html
similarity index 100%
rename from index.html
rename to web/index.html
diff --git a/web/package.json b/web/package.json
new file mode 100644
index 0000000..3459e34
--- /dev/null
+++ b/web/package.json
@@ -0,0 +1,23 @@
+{
+  "name": "thptqg-web",
+  "private": true,
+  "version": "1.0.0",
+  "type": "module",
+  "description": "Frontend — one Vite app serving all four datasets and the hub",
+  "scripts": {
+    "dev": "vite",
+    "build": "vite build",
+    "preview": "vite preview"
+  },
+  "dependencies": {
+    "react": "^19.2.4",
+    "react-dom": "^19.2.4",
+    "sql.js": "^1.14.1"
+  },
+  "devDependencies": {
+    "@types/react": "^19.2.14",
+    "@types/react-dom": "^19.2.3",
+    "@vitejs/plugin-react": "^6.0.1",
+    "vite": "^8.0.16"
+  }
+}
diff --git a/scripts/assemble-site.js b/web/scripts/assemble-site.js
similarity index 78%
rename from scripts/assemble-site.js
rename to web/scripts/assemble-site.js
index 919aa0b..6907d30 100644
--- a/scripts/assemble-site.js
+++ b/web/scripts/assemble-site.js
@@ -12,7 +12,7 @@
  * rewrites the URL and would interfere with the ?q= deep links the app relies
  * on.
  *
- * Usage: node scripts/assemble-site.js
+ * Usage: node web/scripts/assemble-site.js
  */
 
 import { cpSync, mkdirSync, rmSync, existsSync, readdirSync, statSync } from "node:fs";
@@ -21,12 +21,15 @@ import { fileURLToPath } from "node:url";
 
 import { DATASET_IDS } from "../src/datasets.js";
 
-const ROOT = resolve(dirname(fileURLToPath(import.meta.url)), "..");
-const DIST = join(ROOT, "dist");
-const SITE = join(ROOT, "_site");
+// The build output belongs to this workspace; the Pages artifact does not. The
+// deploy action uploads _site from the repository root, so that one climbs out.
+const WEB = resolve(dirname(fileURLToPath(import.meta.url)), "..");
+const DIST = join(WEB, "dist");
+const SITE = join(WEB, "..", "_site");
 
-/** URLs published before the switch to flat paths; the router rewrites these. */
-const LEGACY_PATHS = ["2017/old", "2017/old2"];
+// The two legacy nested URLs (2017/old, 2017/old2) are gone with the datasets
+// they addressed. Unknown paths render the hub via 404.html, so old bookmarks
+// land somewhere useful rather than on a Pages error page.
 
 if (!existsSync(join(DIST, "index.html"))) {
   console.error(`no build found at ${DIST} — run: npm run build`);
@@ -42,8 +45,8 @@ cpSync(DIST, SITE, { recursive: true });
 // Unknown paths render the hub rather than the default Pages 404.
 cpSync(join(DIST, "index.html"), join(SITE, "404.html"));
 
-// One entry point per dataset, plus the legacy nested URLs.
-for (const path of [...DATASET_IDS, ...LEGACY_PATHS]) {
+// One entry point per dataset.
+for (const path of DATASET_IDS) {
   mkdirSync(join(SITE, path), { recursive: true });
   cpSync(join(DIST, "index.html"), join(SITE, path, "index.html"));
 }
@@ -72,6 +75,6 @@ if (stray.length) {
 }
 
 console.log(`assembled ${SITE}`);
-for (const path of ["", "404.html", ...DATASET_IDS, ...LEGACY_PATHS]) {
+for (const path of ["", "404.html", ...DATASET_IDS]) {
   console.log(`  /thptqg/${path}`);
 }
diff --git a/src/App.css b/web/src/App.css
similarity index 100%
rename from src/App.css
rename to web/src/App.css
diff --git a/src/App.jsx b/web/src/App.jsx
similarity index 100%
rename from src/App.jsx
rename to web/src/App.jsx
diff --git a/src/assets/vite.svg b/web/src/assets/vite.svg
similarity index 100%
rename from src/assets/vite.svg
rename to web/src/assets/vite.svg
diff --git a/src/components/custom-query.jsx b/web/src/components/custom-query.jsx
similarity index 100%
rename from src/components/custom-query.jsx
rename to web/src/components/custom-query.jsx
diff --git a/src/components/hub.jsx b/web/src/components/hub.jsx
similarity index 100%
rename from src/components/hub.jsx
rename to web/src/components/hub.jsx
diff --git a/src/components/score-table.jsx b/web/src/components/score-table.jsx
similarity index 100%
rename from src/components/score-table.jsx
rename to web/src/components/score-table.jsx
diff --git a/src/components/search-form.jsx b/web/src/components/search-form.jsx
similarity index 100%
rename from src/components/search-form.jsx
rename to web/src/components/search-form.jsx
diff --git a/src/components/student-detail.jsx b/web/src/components/student-detail.jsx
similarity index 100%
rename from src/components/student-detail.jsx
rename to web/src/components/student-detail.jsx
diff --git a/src/datasets.js b/web/src/datasets.js
similarity index 71%
rename from src/datasets.js
rename to web/src/datasets.js
index 6f5509e..d370b23 100644
--- a/src/datasets.js
+++ b/web/src/datasets.js
@@ -11,6 +11,10 @@
  * Imported by the Vite app *and* by go-parser/scripts/build-db.js under plain
  * Node, so this module must stay free of `import.meta.env` and any Vite-only
  * syntax. Callers pass the base URL in explicitly for that reason.
+ *
+ * That cross-package import is why this file cannot move into a Vite-only
+ * corner of the app: go-parser reads DATASET_IDS and dbSizeMb straight from
+ * here, and reaches across the workspace boundary to do it.
  */
 
 // Extension is required: this module is also imported by plain Node
@@ -44,30 +48,6 @@ export const DATASETS = [
     examples: ["49008235", "Nguyễn Minh Tiến"],
     presets: PRESETS_2017,
   },
-  {
-    id: "2017-old",
-    year: 2017,
-    label: "2017 — bản cũ",
-    blurb: "847.348 thí sinh · bản trước khi cập nhật",
-    title: "Tra cứu điểm thi THPT Quốc gia 2017 (bản cũ)",
-    subtitle: SUBTITLE,
-    source: "bản lưu trước đợt cập nhật baotintuc.vn",
-    dbSizeMb: 47,
-    examples: ["49008235", "Nguyễn Minh Tiến"],
-    presets: PRESETS_2017,
-  },
-  {
-    id: "2017-old2",
-    year: 2017,
-    label: "2017 — bản cũ 2",
-    blurb: "679.764 thí sinh · bản xuất lại",
-    title: "Tra cứu điểm thi THPT Quốc gia 2017 (bản cũ 2)",
-    subtitle: SUBTITLE,
-    source: "bản xuất lại đã hiệu chỉnh",
-    dbSizeMb: 38,
-    examples: ["49008235", "Nguyễn Minh Tiến"],
-    presets: PRESETS_2017,
-  },
 ];
 
 /** Dataset IDs in build order. */
diff --git a/src/hooks/use-sqlite.js b/web/src/hooks/use-sqlite.js
similarity index 100%
rename from src/hooks/use-sqlite.js
rename to web/src/hooks/use-sqlite.js
diff --git a/src/index.css b/web/src/index.css
similarity index 100%
rename from src/index.css
rename to web/src/index.css
diff --git a/src/lib/admission-blocks.js b/web/src/lib/admission-blocks.js
similarity index 100%
rename from src/lib/admission-blocks.js
rename to web/src/lib/admission-blocks.js
diff --git a/src/lib/query-mode.js b/web/src/lib/query-mode.js
similarity index 100%
rename from src/lib/query-mode.js
rename to web/src/lib/query-mode.js
diff --git a/src/lib/sql-presets.js b/web/src/lib/sql-presets.js
similarity index 100%
rename from src/lib/sql-presets.js
rename to web/src/lib/sql-presets.js
diff --git a/src/lib/subjects.js b/web/src/lib/subjects.js
similarity index 100%
rename from src/lib/subjects.js
rename to web/src/lib/subjects.js
diff --git a/src/main.jsx b/web/src/main.jsx
similarity index 100%
rename from src/main.jsx
rename to web/src/main.jsx
diff --git a/src/router.js b/web/src/router.js
similarity index 50%
rename from src/router.js
rename to web/src/router.js
index d278465..f17117b 100644
--- a/src/router.js
+++ b/web/src/router.js
@@ -3,27 +3,23 @@
  *
  * URLs are flat, one segment per dataset, and the segment IS the dataset id:
  *
- *   /thptqg/            → null (hub)
- *   /thptqg/2016/       → 2016
- *   /thptqg/2017-old/   → 2017-old
+ *   /thptqg/        → null (hub)
+ *   /thptqg/2016/   → 2016
+ *   /thptqg/2017/   → 2017
  *
  * That makes this an exact match on a single segment. The nested form this
  * replaced (/thptqg/2017/old/) needed longest-prefix matching, because
  * "/thptqg/2017/old/" also starts with "/thptqg/2017/" — an entire class of
  * routing bug the flat scheme designs out.
+ *
+ * Those nested URLs used to be rewritten here so old bookmarks kept working.
+ * Both of them addressed the 2017-old and 2017-old2 datasets, which no longer
+ * exist, so there is nothing left to rewrite them to: they now fall through to
+ * the hub like any other unknown path.
  */
 
 import { DATASETS } from "./datasets";
 
-/**
- * URLs published before the switch to flat paths. Kept working so existing
- * links and bookmarks resolve instead of 404ing.
- */
-const LEGACY_PATHS = {
-  "2017/old": "2017-old",
-  "2017/old2": "2017-old2",
-};
-
 const BASE = import.meta.env.BASE_URL;
 
 /** Strip the site base and any trailing slash, leaving a bare route segment. */
@@ -32,25 +28,8 @@ function toSegment(pathname) {
   return path.replace(/^\/+/, "").replace(/\/+$/, "");
 }
 
-/**
- * Resolve the current location to a dataset, or null for the hub.
- *
- * Rewrites legacy nested URLs to their flat equivalent, preserving the query
- * string — a `?q=` deep link must survive the redirect, which is exactly what
- * the usual SPA 404-fallback hack would break.
- */
+/** Resolve the current location to a dataset, or null for the hub. */
 export function resolveRoute(pathname = window.location.pathname) {
-  let segment = toSegment(pathname);
-
-  const flat = LEGACY_PATHS[segment];
-  if (flat) {
-    window.history.replaceState(
-      {},
-      "",
-      `${BASE}${flat}/${window.location.search}${window.location.hash}`,
-    );
-    segment = flat;
-  }
-
+  const segment = toSegment(pathname);
   return DATASETS.find((d) => d.id === segment) ?? null;
 }
diff --git a/vite.config.js b/web/vite.config.js
similarity index 62%
rename from vite.config.js
rename to web/vite.config.js
index e047942..10b02f6 100644
--- a/vite.config.js
+++ b/web/vite.config.js
@@ -12,8 +12,17 @@ import react from "@vitejs/plugin-react";
 //
 // publicDir holds only the gzipped databases, staged there by
 // go-parser/scripts/build-db.js. Nothing uncompressed is ever placed in it.
+//
+// It sits at the repository root rather than inside this workspace because
+// go-parser writes it, so the path has to climb out of web/. Vite resolves
+// publicDir against the project root, which is this directory.
+//
+// outDir is left at its default, so the build lands in web/dist and this
+// workspace stays self-contained. The Pages artifact (_site) is assembled at
+// the repository root by web/scripts/assemble-site.js, since that is where the
+// deploy action uploads from.
 export default defineConfig({
   plugins: [react()],
   base: "/thptqg/",
-  publicDir: ".build/public",
+  publicDir: "../.build/public",
 });