test: add tests for parsing phase

Author: chrispcampbell

packages/docs-builder/src/parse.spec.ts

@@ -0,0 +1,278 @@
+// Copyright (c) 2025 Climate Interactive / New Venture Fund. All rights reserved.
+
+import { describe, expect, it } from 'vitest'
+
+import type { Config } from './config'
+import { Context } from './context'
+import { parseMarkdownPageContent } from './parse'
+
+const config: Config = {
+  mode: 'development',
+  baseProjDir: 'xxx',
+  sourceDir: 'xxx',
+  outDir: 'xxx',
+  version: '25.1.0',
+  langs: [],
+  formats: [],
+  template: 'default',
+  author: 'Climate Interactive',
+  logoPath: 'xxx',
+  defs: [],
+  pages: ['page_1.md'],
+  untranslated: [],
+  options: {}
+}
+
+describe('parseMarkdownPageContent', () => {
+  it('should parse valid Markdown content', () => {
+    const mdCommon = `\
+<!-- Common strings -->
+
+<!-- def:content_placeholder -->
+This text was defined in \`common.md\`.
+
+<!-- def:github_project -->
+This page was generated by [\`docs-builder\`](https://github.com/climateinteractive/docs-builder).
+
+<!-- Titles for untranslated pages -->
+
+<!-- def:appendix__title -->
+Appendix
+
+<!-- Common section headers -->
+
+<!-- def:section_examples -->
+Examples
+
+<!-- def:section_footnotes -->
+Footnotes
+`
+
+    const mdGlossary = `\
+# <!-- section:glossary --><!-- def:title -->Glossary
+
+## <!-- section:page -->
+<!-- def:def -->
+**page**: one side of a sheet of paper in a collection of sheets bound together
+`
+
+    const mdPage = `\
+---
+fragments:
+  head: ['example']
+---
+
+# <!-- section:page_1 --><!-- def:title -->Page 1
+
+<!-- def:intro -->
+This is the first [page][glossary_page].
+
+## <!-- section:examples -->:section_examples:
+
+<!-- begin-def:example_1 -->
+
+This is a sentence with __bold__ and _italic_ text.
+
+This block has two paragraphs that are captured using a \`begin-def\` / \`end-def\` pair.
+
+<!-- end-def -->
+
+<!-- def[hidden]:hidden_text -->
+this text was defined using "hidden" flag
+
+<!-- def:example_2 -->
+Use \`def[hidden]\` to define some text without making it appear, then "use" it later on the page (like this: _:page_1__examples__hidden_text:_).
+
+:content_placeholder:
+
+<!-- def:example_3 -->
+This sentence refers to a footnote. footnote-ref:fn_example
+
+| <!-- def:header_person-->person | <!-- def:header_age -->age |
+|--|--|
+| Alice | 42 |
+| Bob | 99 |
+
+
+\`\`\`js
+// This is a code block
+const one = 1
+const two = 2
+const answer = one + two
+\`\`\`
+
+## <!-- section:footnotes -->:section_footnotes:
+
+footnote:fn_example This is a footnote (only in English for now).
+
+_:github_project:_
+`
+
+    // Parse the common and glossary strings so that the strings are part of the context
+    const enContext = new Context(config, 'en')
+    parseMarkdownPageContent(enContext, 'common.md', mdCommon)
+    parseMarkdownPageContent(enContext, 'glossary.md', mdGlossary)
+
+    // Verify that an error is thrown if the English content contains invalid link syntax
+    const parsed = parseMarkdownPageContent(enContext, 'page_1.md', mdPage)
+    expect(parsed.frontmatter).toEqual({
+      fragments: {
+        head: ['example']
+      }
+    })
+    expect(parsed.raw).toEqual(`
+# <a name="page_1"></a>Page 1<a class="heading-link" href="#page_1">&#128279;</a>
+
+This is the first [page](glossary:page).
+
+## <a name="page_1__examples"></a>Examples<a class="heading-link" href="#page_1__examples">&#128279;</a>
+
+This is a sentence with **bold** and _italic_ text.
+
+This block has two paragraphs that are captured using a \`begin-def\` / \`end-def\` pair.
+
+
+
+Use \`def[hidden]\` to define some text without making it appear, then &quot;use&quot; it later on the page (like this: _this text was defined using "hidden" flag_).
+
+This text was defined in \`common.md\`.
+
+This sentence refers to a footnote. footnote-ref:fn_example
+
+| person | age |
+|--|--|
+| Alice | 42 |
+| Bob | 99 |
+
+\`\`\`js
+// This is a code block
+const one = 1
+const two = 2
+const answer = one + two
+\`\`\`
+
+## <a name="page_1__footnotes"></a>Footnotes<a class="heading-link" href="#page_1__footnotes">&#128279;</a>
+
+footnote:fn_example This is a footnote (only in English for now).
+
+_This page was generated by [\`docs-builder\`](https://github.com/climateinteractive/docs-builder)._
+
+
+`)
+  })
+
+  it('should throw an error if an unknown command is used', () => {
+    const md = `\
+<!-- somecommand:key -->
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Unknown command 'somecommand' (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if a command is used with an invalid identifier', () => {
+    const md = `\
+<!-- def:key_with_INVALID_chars -->
+Hello
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Identifier (key_with_INVALID_chars) must contain only lowercase letters, digits, and underscores (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if a begin-def command is not closed', () => {
+    const md = `\
+<!-- begin-def:example_1 -->
+
+Hello
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Command 'begin-def' was not resolved or closed (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if an unexpected command is used before a begin-def is closed', () => {
+    const md = `\
+<!-- begin-def:example_1 -->
+
+Hello
+
+<!-- def:example_2 -->
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Unexpected command 'def' while current command 'begin-def' (id=example_1) is in effect (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if a begin-def / end-def pair does not contain any text', () => {
+    const md = `\
+<!-- begin-def:example_1 -->
+
+<!-- end-def -->
+`
+
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Saw 'end-def' but no tokens were included (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if an end-def command is used without a corresponding begin-def command', () => {
+    const md = `\
+Hello
+
+<!-- end-def -->
+`
+
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      `Saw 'end-def' without corresponding 'begin-def' (page=page_1.md)`
+    )
+  })
+
+  it('should throw an error if text replacement syntax is used with an unknown string identifier', () => {
+    const md = `\
+Hello
+
+:some_id:
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      'Unknown replacement for id=some_id (page=page_1.md)'
+    )
+  })
+
+  it('should throw an error if two spaces are detected at the end of a line', () => {
+    const spaces = ' '.repeat(2)
+    const md = `\
+# <!-- section:section_1 -->Section 1
+
+Hello${spaces}
+there
+`
+
+    let expectedMsg = ''
+    expectedMsg += 'Detected two or more spaces at the end of a text line. '
+    expectedMsg += 'Markdown interprets this as a line break, which can be surprising. '
+    expectedMsg += 'If the spaces were added unintentionally, remove the extra spaces. '
+    expectedMsg += `If you do want a line break, use an explicit HTML 'br' tag instead. `
+    expectedMsg += '(page=page_1.md scope=section_1)'
+
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(expectedMsg)
+  })
+
+  it('should throw an error if a reference-style link cannot be resolved', () => {
+    const md = `\
+[This is a link][unknown_ref]
+`
+    const enContext = new Context(config, 'en')
+    expect(() => parseMarkdownPageContent(enContext, 'page_1.md', md)).toThrow(
+      'Unresolved reference-style link found for lang=en link=[This is a link][unknown_ref]'
+    )
+  })
+})

packages/docs-builder/src/parse.ts

@@ -38,23 +38,41 @@ type Links = {
  * @return The translated Markdown content.
  */
 export function parseMarkdownPage(context: Context, relPath: string): MarkdownPage {
+  // Read the Markdown file
+  const filePath = resolve(context.config.baseProjDir, relPath)
+  const origMarkdownWithFrontmatter = readTextFile(filePath)
+
+  // Parse the Markdown content
+  return parseMarkdownPageContent(context, relPath, origMarkdownWithFrontmatter)
+}
+
+/**
+ * Parse the given base (English) Markdown page and translate it into the language
+ * associated with the context.
+ *
+ * @param context The language-specific context.
+ * @param relPath The path to the Markdown file, relative to the base project directory.
+ * @param origMarkdownWithFrontmatter The Markdown content with optional frontmatter.
+ * @return The translated Markdown content.
+ */
+export function parseMarkdownPageContent(
+  context: Context,
+  relPath: string,
+  origMarkdownWithFrontmatter: string
+): MarkdownPage {
   // Configure marked.js
   marked.setOptions({
     headerIds: false
   })
 
-  // Set the current page (for error reporting)
-  context.setCurrentPage(relPath)
-
-  // Read the Markdown file
-  const filePath = resolve(context.config.baseProjDir, relPath)
-  const origMarkdownWithFrontmatter = readTextFile(filePath)
-
   // Separate frontmatter from the content
   const origMarkdownSeparated = matter(origMarkdownWithFrontmatter)
   const origMarkdown = origMarkdownSeparated.content
   const frontmatter = origMarkdownSeparated.data
 
+  // Set the current page (for error reporting)
+  context.setCurrentPage(relPath)
+
   // Append synthesized link info for glossary references so that source files can
   // use `[link text][glossary_term]` without manually defining a reference for
   // "glossary_term"; it will be converted to `[...](glossary:term)` automatically
@@ -445,7 +463,7 @@ function processToken(context: Context, state: ProcessState, token: marked.Token
         let msg = 'Detected two or more spaces at the end of a text line.'
         msg += ' Markdown interprets this as a line break, which can be surprising.'
         msg += ' If the spaces were added unintentionally, remove the extra spaces.'
-        msg += ' If you do want a line break, use an explicit HTML `<br/>` tag instead.'
+        msg += ` If you do want a line break, use an explicit HTML 'br' tag instead.`
         throw new Error(context.getScopedMessage(msg))
       }
       default: