takaebato commented on code in PR #497:
URL: https://github.com/apache/echarts-doc/pull/497#discussion_r3035177127
##########
build/build-llms.js:
##########
@@ -0,0 +1,350 @@
+/**
+ * Converts built part JSONs (HTML desc) to Markdown using turndown,
+ * and generates llms.txt + individual .md files.
+ *
+ * Mechanically converts documents/*-parts/*.json to llms-documents/ (.md
files).
+ * Root files (e.g. option.md) are placed at llms-documents/, while part files
+ * (e.g. option.title.md) are placed at llms-documents/*-parts/.
+ * Type information is extracted from documents/*.json (full schema) via
traverse.
+ *
+ * Prerequisites: JSON must be built first (node build.js --env dev)
+ * Usage: node build/build-llms.js --env dev
+ */
+const fs = require('fs');
+const fse = require('fs-extra');
+const path = require('path');
+const globby = require('globby');
+const TurndownService = require('turndown');
+const {gfm} = require('turndown-plugin-gfm');
+const {traverse} = require('../tool/schemaHelper');
+const {readConfigEnvFile} = require('./helper');
+
+// --- Constants ---
+
+const LANGUAGES = ['en', 'zh'];
+const OUTPUT_DIR_NAME = 'llms-documents';
+const MAX_HEADING_DEPTH = 6;
+
+const SECTION_LABELS = {
+ en: {'option-parts': 'Option', 'option-gl-parts': 'Option GL',
'api-parts': 'API', 'tutorial-parts': 'Tutorial'},
+ zh: {'option-parts': '配置项 (Option)', 'option-gl-parts': 'Option GL',
'api-parts': 'API', 'tutorial-parts': '教程 (Tutorial)'}
+};
+
+const LLMS_TXT_HEADER = [
+ '# Apache ECharts Documentation',
+ '',
+ '> Apache ECharts is a free, powerful charting and visualization library
offering easy ways to add intuitive, interactive, and highly customizable
charts to your commercial products.',
+ ''
+].join('\n');
+
+// --- Config ---
+
+const argv = require('yargs').argv;
+const envType = (argv.dev != null || argv.debug != null || argv.env === 'dev')
? 'dev' : argv.env;
+if (!envType) throw new Error('--env MUST be specified');
+const config = readConfigEnvFile(envType);
+
+// --- Turndown ---
+
+const td = new TurndownService({headingStyle: 'atx', codeBlockStyle:
'fenced'});
+td.use(gfm);
+td.addRule('iframe', {filter: 'iframe', replacement: () => ''});
+
+function htmlToMd(html) {
+ return html ? td.turndown(html).replace(/\n{3,}/g, '\n\n').trim() : '';
+}
+
+// --- Extract type info from full schema JSON ---
+
+/**
+ * Extract type and default value info from a full schema JSON by traversing
+ * the nested schema tree.
+ *
+ * @param {string} schemaJsonPath - path to schema JSON (e.g.
"documents/option.json")
+ * @param {string} docName - e.g. "option", "api"
+ * @returns {Object<string, {type: string|null, default: string|null}>}
+ * e.g. { "option.title.show": {type: "boolean", default: "true"} }
+ */
+function buildTypeMap(schemaJsonPath, docName) {
+ if (!fs.existsSync(schemaJsonPath)) return {};
+ const schema = JSON.parse(fs.readFileSync(schemaJsonPath, 'utf-8'));
+ const typeMap = {};
+ traverse(schema, docName, (schemaPath, node) => {
+ if (node.type || node.default != null) {
+ typeMap[schemaPath] = {
+ type: node.type ? (Array.isArray(node.type) ?
node.type.join('|') : node.type) : null,
+ default: node.default != null ? String(node.default) : null
+ };
+ }
Review Comment:
Fixed. Non-primitive defaults do not exist now, but fixed for in case.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
