fix(i18n): repair broken front matter in 302 translated pages

[bloxster]

fix(i18n): strip a stray leading --- from 260 translated pages

pages build and deployment has been failing since 2026-08-14, and the published
GitHub Pages site (https://zechub.github.io/zechub/, status errored) has been
frozen on that day's build ever since.

The LLM sync prepends a --- that the English source does not have. Jekyll reads
a leading --- as a YAML front-matter opener, runs to the next --- far down the
page, and fails to parse. The build dies on
translations/ja/site/Zcash_Tech/What_a_Block_Explorer_Can_See.md.

Emulating Jekyll's front-matter parse across the whole tree: 251 translated pages
break, 0 English pages do. Every breaking file is a translation, and in all 251
the English source has no leading ---, so there is no judgment call — the
dashes are simply not supposed to be there.

Traced by counting leading---- files at each merge:

commit	files	pages build
d951213 (#1948)	96	last success
c301ff2 (#1960)	127	first failure
636861a (#1964)	356	failing

96 such files predate this and the build was green, so a leading --- is not
automatically fatal — it breaks when the resulting pseudo-front-matter is invalid
YAML, which prose in these locales reliably produces. #1960 tipped it over and
#1964 made it much worse.

This strips the stray line from all 260 pages where the English source lacks it,
not only the 251 that currently break, since the other 9 are the same defect and
a latent break.

Reader-facing impact was limited: the frontend calls next-mdx-remote's
serialize(..., {}), whose parseFrontmatter defaults to false, and gray-matter
— though in package.json — is imported nowhere in src/, so the dashes rendered as
a stray horizontal rule rather than swallowing content. Worth noting gray-matter
DOES throw on these files, so anything that starts using it would break hard.

Two sidebar labels were wrong, though, and this fixes them: Using_Zcash/Zecmap.md
read "Kurzfassung" in de and "Resumo" in pt — both meaning "summary" — because the
stray --- made the title extractor skip # ZecMap and pick up ## TL;DR. Both
now read "ZecMap". Menu-title manifests regenerated accordingly.

No prose changed: the only content line touched anywhere is the bare ---.
src is untouched because it hashes the ENGLISH source, which has not moved.
edited stays false throughout — this is a mechanical pass, and flipping it
would remove these pages from automated sync permanently.

Verification, all on this branch:

check	result
Jekyll front-matter emulation	0 breaking pages, was 251
protected-terms (scoped)	passed, 3,654 pages
manifest-invariants	holds, 203 x 18
translation lib tests	51 pass / 0 fail
menu-titles-fresh	up to date, 19 files
non---- content lines changed	0

Incidentally a first real exercise of #1973: with the identical content repair but
the manifest left untagged, manifest-invariants now fails with 260 violations.
Before #1973 it passed green, so this repair could have landed with no provenance
trace at all.

Not done here, having checked and found nothing wrong: the 108 translated pages
using the markdown badge form [![Edit](...)](...) rather than the HTML
<a><img/></a> form. All 108 match their English source exactly.

---

Second commit: two residual front-matter defects (42 more pages)

A sweep of all 4800 .md files — rather than only the pages the last resync
touched — turned up two more defects of the same family. Neither breaks the
Jekyll build on its own; both are invisible until someone reads the page.

24 pages carry a duplicated front-matter opener. On disk:

---
---
published: 2025-08-19
---

Jekyll consumes the first, empty block as the front matter, so the date leaks
into the body — in 22 of them the following --- turns it into a setext H2
heading, and in the 2 with a blank line between, a paragraph followed by a
rule. The extra delimiter is dropped; the value is untouched.
Locales: ru 4, ja 3, ko 2, hi 4, fr 1, uk 5, tr 5.

18 pages had the front-matter key itself machine-translated — ee wotae:,
sw iliyochapishwa:, ak wotintimii:, six Research pages each — so no
consumer could read the publication date at all. The published: key is
restored, value untouched. (published is a reserved Jekyll key, but every
value is a date and therefore truthy, so no page's visibility changes.)

Provenance for all 42 is recorded in tool
(+collapse-empty-frontmatter / +restore-frontmatter-key), the same
convention this branch already uses for +strip-stray-frontmatter. src is
deliberately not bumped — these pages are no fresher against their English
source than before — and edited stays false, since flipping it for a
mechanical pass would hold the page out of automated re-translation forever.

Gates run locally against upstream/main as the base, so change-tracking is
live rather than skipped:

gate	result
check-invariants --base upstream/main	invariants hold: 203 curated pages, 18 locales (302 changed files, all with provenance)
gen-menu-titles --check	up to date (19 files)
node --test translation/lib/*.test.mjs	51 pass / 0 fail
check-protected-terms --base	passed for 3654 pages (pre-existing drift reported, none in the 42)

After this, no .md in the tree opens with an empty front-matter block, an
unterminated block, or a front-matter key other than published.

Follow-up, not in this PR: the generator's own guard cannot catch either
defect. It compares only the first line of the translation against the first
line of the source, so a doubled opener on a page whose source legitimately
has front matter passes, and a translated key is never checked. Without
tightening that gate to compare the parsed block (delimiter count + key set),
a future resync can reintroduce both.

Third commit: a third shape, found by the tightened generator guard (14 pages)

The guard in the generator was rewritten to compare the parsed front-matter
BLOCK against the source's instead of only the first line. Pointed at the whole
corpus it found 14 pages that the scans above could not see, because those only
examined files that HAVE front matter:

12 Research pages (ig 6, yo 6) lost the opening delimiter while the key was
translated and the closing --- survived — bipụtara: 2025-08-02,
tí a tẹ̀ jáde: 2026-04-14. The page therefore has no front matter at all and
the line renders as a setext H2 heading. The source block is restored and the
mangled line dropped.

2 ja pages kept a blank line inside the block after the second commit
collapsed their duplicated opener. Valid YAML with the right value either way;
normalised so every translated block is byte-identical to its source.

Replaying the new guard over all 54 pre-fix files from this branch's history
reproduces the fixes in these three commits byte-for-byte, and the repaired
corpus is a fixed point: 3654 pages, 0 repairs, 0 failures.

One deliberate non-change: six of the restored values keep a trailing space
(published: 2023-10-23 ) because the English source has it, and one source is
CRLF. Stripping it in the translations alone would make them the only pages that
deviate from their source, which the new guard would then repair straight back.
Whitespace in the English sources is a separate cleanup.