Merge pull request #13643 from quarto-dev/fix/tinytex/babel-error

Author: cderv

news/changelog-1.9.md

@@ -6,6 +6,7 @@ All changes included in 1.9:
 - ([#13441](https://github.com/quarto-dev/quarto-cli/pull/13441)): Catch `undefined` exceptions in Pandoc failure to avoid spurious error message.
 - ([#13046](https://github.com/quarto-dev/quarto-cli/issues/13046)): Use new url for multiplex socket.io server <https://multiplex.up.railway.app/> as default for `format: revealjs` and `revealjs.multiplex: true`.
 - ([#13506](https://github.com/quarto-dev/quarto-cli/issues/13506)): Fix navbar active state detection when sidebar has no logo configured. Prevents empty logo links from interfering with navigation highlighting.
+- ([#13633](https://github.com/quarto-dev/quarto-cli/issues/13633)): Fix detection and auto-installation of babel language packages from newer error format that doesn't explicitly mention `.ldf` filename.
 
 ## Dependencies
 
@@ -37,6 +38,7 @@ All changes included in 1.9:
 
 ### `pdf`
 
+- ([#10291](https://github.com/quarto-dev/quarto-cli/issues/10291)): Fix detection of babel hyphenation warnings with straight-quote format instead of backtick-quote format.
 - ([rstudio/tinytex-releases#49](https://github.com/rstudio/tinytex-releases/issues/49)): Fix detection of LuaTeX-ja missing file errors by matching both "File" and "file" in error messages.
 
 ## Projects

src/command/render/latexmk/parse-error.ts

@@ -133,7 +133,7 @@ export function findMissingHyphenationFiles(logText: string) {
   const babelWarningRegex = /^Package babel Warning:/m;
   const hasWarning = logText.match(babelWarningRegex);
   if (hasWarning) {
-    const languageRegex = /^\(babel\).* language `(\S+)'.*$/m;
+    const languageRegex = /^\(babel\).* language [`'](\S+)[`'].*$/m;
     const languageMatch = logText.match(languageRegex);
     if (languageMatch) {
       return filterLang(languageMatch[1]);
@@ -283,6 +283,12 @@ const packageMatchers = [
   { regex: /.*! LaTeX Error: File [`']([^']+)' not found.*/g },
   { regex: /.* [fF]ile ['`]?([^' ]+)'? not found.*/g },
   { regex: /.*the language definition file ([^\s]*).*/g },
+  {
+    regex: /.*! Package babel Error: Unknown option [`']([^'`]+)'[.].*/g,
+    filter: (match: string, _text: string) => {
+      return `${match}.ldf`;
+    },
+  },
   { regex: /.* \\(file ([^)]+)\\): cannot open .*/g },
   { regex: /.*file [`']([^']+)' .*is missing.*/g },
   { regex: /.*! CTeX fontset [`']([^']+)' is unavailable.*/g },

tests/docs/smoke-all/2024/08/30/10291/latex-hyphen-lang-es-no-install.qmd

@@ -1,16 +1,22 @@
 ---
 format: pdf
-# new babel support in Pandoc 3.6.3 solves missing package when specific lang is used.
 lang: es
 latex-auto-install: false
 _quarto:
+  tests-on-ci: false
   tests:
-    pdf: default
+    pdf:
+      noErrors: true
+      printsMessage:
+        level: WARN
+        regex: "Possibly missing hyphenation"
 ---
 
+** This test is to be run manually as we don't want to uninstall and reinstall package in CI**
+
 ```{r}
 #| include: false
-
+#| messages: 'NA'
 # Remove the hyphen package for spanish so that the test is meaningful
 if (tinytex::check_installed("hyphen-spanish")) {
   message("Removing 'hyphen-spanish' package for the render")

tests/docs/smoke-all/2024/08/30/10291/latex-hyphen-lang-es.qmd

@@ -6,14 +6,21 @@ _quarto:
     pdf: null
 ---
 
+**Only remove package on CI**
 ```{r}
+#| eval: !expr isFALSE(as.logical(Sys.getenv("CI", "false")))
 #| include: false
+#| messages: "NA"
 
 # Remove the hyphen package for spanish
 if (tinytex::check_installed("hyphen-spanish")) {
   message("Removing 'hyphen-spanish' package for the render")
   tinytex::tlmgr_remove("hyphen-spanish")
 }
+if (tinytex::check_installed("babel-spanish")) {
+  message("Removing 'babel-spanish' package for the render")
+  tinytex::tlmgr_remove("babel-spanish")
+}
 ```
 
 # Hola !

tests/docs/smoke-all/2025/11/04/13633.qmd

@@ -0,0 +1,22 @@
+---
+title: "Babel English Auto-Install (#13633)"
+format: pdf
+lang: en
+_quarto:
+  tests:
+    pdf: null
+---
+
+```{r}
+#| eval: !expr isFALSE(as.logical(Sys.getenv("CI", "false")))
+#| include: false
+#| messages: "NA"
+
+# Remove babel-english package to test auto-installation
+if (tinytex::check_installed("babel-english")) {
+  message("Removing 'babel-english' package for the render")
+  tinytex::tlmgr_remove("babel-english")
+}
+```
+
+This document tests that Quarto correctly detects and auto-installs babel-english when rendering with `lang: en`.

tests/unit/latexmk/parse-error.test.ts

@@ -5,7 +5,7 @@
 *
 */
 
-import { findMissingFontsAndPackages } from "../../../src/command/render/latexmk/parse-error.ts"
+import { findMissingFontsAndPackages, findMissingHyphenationFiles } from "../../../src/command/render/latexmk/parse-error.ts"
 import { unitTest } from "../../test.ts";
 import { assert } from "testing/asserts";
 
@@ -42,6 +42,10 @@ unitTest("Detect missing files with `findMissingFontsAndPackages`", async () =>
     (babel)                There is a locale ini file for this language.
     (babel)                If it’s the main language, try adding \`provide=*'
     (babel)                to the babel package options.`, "ngerman.ldf")
+  assertFound("! Package babel Error: Unknown option 'english'.", "english.ldf");
+  assertFound(`! Package babel Error: Unknown option 'ngerman'.
+(babel)                Suggested actions:
+(babel)                * Make sure you haven't misspelled it`, "ngerman.ldf");
   assertFound("!pdfTeX error: pdflatex (file 8r.enc): cannot open encoding file for reading", "8r.enc");
   assertFound("! CTeX fontset `fandol' is unavailable in current mode", "fandol");
   assertFound("! CTeX fontset 'fandol' is unavailable in current mode", "fandol");
@@ -70,4 +74,61 @@ unitTest("Detect missing files with `findMissingFontsAndPackages`", async () =>
   assertFound("No file LGRcmr.fd. ! LaTeX Error: This NFSS system isn't set up properly.", "lgrcmr.fd");
 },{
   cwd: () => "unit/latexmk/"
+})
+
+unitTest("Detect missing hyphenation with babel warnings", async () => {
+  // Test backtick-quote format (old format)
+  const logWithBacktick = `Package babel Warning: No hyphenation patterns were preloaded for
+(babel)                the language \`Spanish' into the format.
+(babel)                Please, configure your TeX system to add them and
+(babel)                rebuild the format. Now I will use the patterns
+(babel)                preloaded for \\language=0 instead on input line 51.`;
+  assert(
+    findMissingHyphenationFiles(logWithBacktick) === "hyphen-spanish",
+    "Should detect hyphen-spanish from backtick-quote format"
+  );
+
+  // Test straight-quote format (new format - the bug we're fixing)
+  const logWithStraightQuotes = `Package babel Warning: No hyphenation patterns were preloaded for
+(babel)                the language 'Spanish' into the format.
+(babel)                Please, configure your TeX system to add them and
+(babel)                rebuild the format. Now I will use the patterns
+(babel)                preloaded for \\language=0 instead on input line 51.`;
+  assert(
+    findMissingHyphenationFiles(logWithStraightQuotes) === "hyphen-spanish",
+    "Should detect hyphen-spanish from straight-quote format"
+  );
+
+  // Test ngerman special case (should return hyphen-german, not hyphen-ngerman)
+  const logGerman = `Package babel Warning: No hyphenation patterns were preloaded for
+(babel)                the language 'ngerman' into the format.`;
+  assert(
+    findMissingHyphenationFiles(logGerman) === "hyphen-german",
+    "Should map ngerman to hyphen-german"
+  );
+
+  // Test Chinese - no hyphen package exists
+  const logChinese = `Package babel Warning: No hyphenation patterns were preloaded for
+(babel)                the language 'chinese' into the format.`;
+  assert(
+    findMissingHyphenationFiles(logChinese) === undefined,
+    "Should return undefined for Chinese (no hyphen package)"
+  );
+
+  // Test alternative Info pattern (issue #10291)
+  const logInfoChinese = `Package babel Info: Hyphen rules for 'chinese-hans' set to \\l@nil
+(babel)             (\\language10). Reported on input line 143.`;
+  assert(
+    findMissingHyphenationFiles(logInfoChinese) === undefined,
+    "Should return undefined for chinese-hans via Info pattern"
+  );
+
+  // Test no warning present
+  const logNoWarning = "Some other log text without babel warnings";
+  assert(
+    findMissingHyphenationFiles(logNoWarning) === undefined,
+    "Should return undefined when no babel warning present"
+  );
+}, {
+  cwd: () => "unit/latexmk/"
 })