Translations update from Hosted Weblate

[weblate]

Translations update from Hosted Weblate for OpenWrt/LuCI/modules/luci-base.

Current translation status:

[openwrt-ai]

Reviewed 1 new commit (a52b92d) covering 65 .po files across es/tr/ta/pl/sv.

Mechanical validation of all 309 changed entries came back clean: no printf format-specifier mismatches between msgid and msgstr, no %{...} LuCI template-placeholder count drift, no HTML tag or &entity; imbalance, no leading/trailing whitespace or trailing-newline drift, no duplicate msgid/msgctxt pairs, and every file's Language: header matches its directory with msgstr[] counts consistent with its declared nplurals. Three inline notes below.

Commit checks

• a52b92d "Added translation using Weblate (Polish)" — the subject names only Polish, but the diff spans five languages and the bulk of it is Spanish, Turkish and Tamil (Polish is 4 of 65 files). The squashed body does enumerate everything, and this is inherent to how Weblate picks the first commit of a squash as the subject, so flagging for awareness only.

---

Generated by Claude Code

[openwrt-ai]

ENCODING is an unreplaced .pot template placeholder, not a valid MIME Content-Transfer-Encoding value. This header block is newly generated by Weblate for a file that previously had no header at all, and it is the only .po file in the tree with this value — all 3822 others use 8bit.

Suggested change

	"Content-Transfer-Encoding: ENCODING\n"
	"Content-Transfer-Encoding: 8bit\n"

---

Generated by Claude Code

[openwrt-ai]

nit: this reads as the imperative "update the URL", but the source string is a noun label. It is the title of a read-only form.DummyValue that displays the provider's update URL — see _("Update URL") at overview.js:682.``

The same file already renders the same concept as a noun for Custom update-URL — "Özel güncelleme URL'si" at tr/ddns.po:129 — and es/pt_BR both use the noun form ("URL de actualización" / "URL de atualização"), so the two entries are now inconsistent with each other. Is the verb form intended here?

---

Generated by Claude Code

[openwrt-ai]

nit: this new file contains 0 translated strings out of 34 — every msgstr is empty and the header is Last-Translator: Automatically generated / Language-Team: none. The commit subject is "Added translation using Weblate (Polish)", but unlike the other three new po/pl/ files in this PR (ledtrig-rssi, ledtrig-usbport, tailscale-community — all 100%), nothing is actually translated here, so it adds 166 lines with no user-visible effect.

Empty auto-generated .po files do already exist in the tree, so this may just be Weblate provisioning the component for pl — worth a look before merge if you'd rather not carry an empty file.

---

Generated by Claude Code

[openwrt-ai]

Re-reviewed at e2a78dc (previous review was at a52b92d, which was replaced by a Weblate force-push). The incremental range a52b92d..e2a78dc touches 66 .po files across es, pl, lt, ko and tr, and is dominated by a large Spanish terminology pass ("Ajustes" → "Configuración", "UPS" → "SAI", tú → usted).

Mechanical validation of the newly changed entries came back clean: no printf format-specifier drift, no %{...} placeholder count mismatches, no HTML tag or &entity; imbalance, no leading/trailing whitespace or trailing-newline drift, no duplicate msgid/msgctxt pairs, and every file's Language: header still matches its directory with msgstr[] counts consistent with its declared nplurals. The luci-app-acl/po/lt Plural-Forms rewrite is a Weblate normalisation and is functionally equivalent to the previous expression. CI on e2a78dc is green (eslint success, build/formality skipped).

Seven inline notes below — one is a real regression (duplicated qosify string), two are concrete Spanish errors, four are nits. All of them ultimately need fixing in Weblate rather than in this branch; the suggestion blocks are only there to show the intended text.

Commit checks

• e2a78dc "Translated using Weblate (Polish)" — the subject names only Polish, but the squashed commit spans five languages in this range and the overwhelming majority of the changes are Spanish (Polish is roughly a dozen of the 66 files). The body does enumerate every component, and this is inherent to Weblate taking the first commit of a squash as the subject, so this is flagged for awareness only — same observation as the previous review.

---

Generated by Claude Code

[openwrt-ai]

Clear rules editor? now carries the exact same translation as Clear config editor? two entries above (lines 98-101) — both read "el editor de configuración". The confirmation dialog for clearing the rules editor will therefore tell the user it is about to clear the config editor. The string this replaced ("¿Editor de reglas vacío? …") did name the right editor, so this is a regression introduced by the rewording.

Suggested change

	"¿Vaciar el editor de configuración? El contenido no se guardará hasta que "
	"haga clic en Guardar."
	"¿Vaciar el editor de reglas? El contenido no se guardará hasta que haga "
	"clic en Guardar."

(The authoritative fix belongs in Weblate — this block just shows the intended text.)

---

Generated by Claude Code

[openwrt-ai]

syn-flood is banIP's label for the TCP SYN flag counter, not the adjective "synchronous" — "sincrónica" reads as the latter and loses the protocol meaning. The two sibling entries added in this same commit keep the token as an uppercase acronym: icmp-flood → "inundación ICMP" and udp-flood → "inundación UDP" (lines 1487 and 1495).

Suggested change

	msgstr "inundación sincrónica"
	msgstr "inundación SYN"

(Fix belongs in Weblate; block shown for clarity.)

---

Generated by Claude Code

[openwrt-ai]

Typo: "intefaz" is missing the r — should be "interfaz". Every other occurrence in this file spells it correctly (e.g. msgstr "Interfaces").

Suggested change

	msgstr "intefaz wan"
	msgstr "interfaz wan"

(Fix belongs in Weblate; block shown for clarity.)

---

Generated by Claude Code

[openwrt-ai]

nit: this entry renders "invalid" as "no válido" while the immediately preceding entry invalid ct uses "inválido" (line 1415). Both were added in this commit and they render side by side in the same set-report legend, so picking one form would read better. The rest of the file consistently uses "no válido" (e.g. "Formato de URL no válido").

---

Generated by Claude Code

[openwrt-ai]

nit: this commit rewrites the whole "Network UPS Tools" family to a single pattern — NUT CGI → "Herramientas del SAI de red (CGI)", NUT Server → "Herramientas del SAI de red (servidor)", Network UPS Tools → "Herramientas del SAI de red" — but NUT Monitor breaks out of it and becomes "Monitor de NUT". These three strings are the sibling menu entries under the same NUT menu, so the odd one out is visible in the UI. Was the divergence intentional, or should it be "Herramientas del SAI de red (Monitor)" to match?

---

Generated by Claude Code

[openwrt-ai]

nit: "Basicstation" here is the Semtech LoRa Basics™ Station product name, and this commit translates it in the menu entry ("Estación básica LoRaWAN") while leaving it untranslated everywhere else in the same file — see "Radio para proporcionar reloj a Basicstation".` The previous value was the untranslated "LoRaWAN Basicstation". Intentional, or should the product name stay as-is in the menu too?

---

Generated by Claude Code

[openwrt-ai]

nit: number disagreement — the msgid is singular ("CA certificate", labelling one general.js:92 file-upload field) but the translation is plural. The neighbouring certificate/key labels in this same commit all stay singular ("Certificado de estación privada", "Clave de estación privada").

Suggested change

	msgstr "Certificados CA"
	msgstr "Certificado CA"

(Fix belongs in Weblate; block shown for clarity.)

---

Generated by Claude Code