@sillsdev/docu-notion 1.0.0-alpha.2 → 1.0.0-alpha.5

package/dist/pull.js

@@ -66,16 +66,19 @@ function getOptionsForLogging(options) {
 const kNotionApiVersion = "2026-03-11";
 let layoutStrategy;
 let notionToMarkdown;
-const pages = new Array();
-const counts = {
-    output_normally: 0,
-    skipped_because_empty: 0,
-    skipped_because_status: 0,
-    skipped_because_level_cannot_have_content: 0,
-    error_because_no_slug: 0,
-};
 function notionPull(options) {
     return __awaiter(this, void 0, void 0, function* () {
+        // These are local to each pull so that repeated calls (e.g. in tests or
+        // programmatic multi-run scenarios) don't accumulate pages or counts from
+        // previous runs.
+        const pages = new Array();
+        const counts = {
+            output_normally: 0,
+            skipped_because_empty: 0,
+            skipped_because_status: 0,
+            skipped_because_level_cannot_have_content: 0,
+            error_because_no_slug: 0,
+        };
         // It's helpful when troubleshooting CI secrets and environment variables to see what options actually made it to docu-notion.
         const optionsForLogging = getOptionsForLogging(options);
         const config = yield (0, configuration_1.loadConfigAsync)();
@@ -97,12 +100,12 @@ function notionPull(options) {
             (0, process_1.exit)(1);
         }
         (0, log_1.group)("Stage 1: walk children of the page named 'Outline', looking for pages...");
-        yield getPagesRecursively(options, "", options.rootPage, 0, true);
+        yield getPagesRecursively(options, "", options.rootPage, 0, true, pages, counts);
         (0, log_1.logDebug)("getPagesRecursively", JSON.stringify(pages, null, 2));
         (0, log_1.info)(`Found ${pages.length} pages`);
         (0, log_1.endGroup)();
         (0, log_1.group)(`Stage 2: convert ${pages.length} Notion pages to markdown and save locally...`);
-        yield outputPages(options, config, pages);
+        yield outputPages(options, config, pages, counts);
         (0, log_1.endGroup)();
         (0, log_1.group)("Stage 3: clean up old files & images...");
         yield layoutStrategy.cleanupOldFiles();
@@ -110,7 +113,7 @@ function notionPull(options) {
         (0, log_1.endGroup)();
     });
 }
-function outputPages(options, config, pages) {
+function outputPages(options, config, pages, counts) {
     return __awaiter(this, void 0, void 0, function* () {
         const context = {
             getBlockChildren: getBlockChildren,
@@ -124,7 +127,7 @@ function outputPages(options, config, pages) {
             notionToMarkdown: notionToMarkdown,
             options: options,
             pages: pages,
-            counts: counts, // review will this get copied or pointed to?
+            counts: counts,
             imports: [],
             convertNotionLinkToLocalDocusaurusLink: (url) => (0, internalLinks_1.convertInternalUrl)(context, url),
         };
@@ -141,7 +144,7 @@ function outputPages(options, config, pages) {
                 context.options.statusTag != "*" &&
                 page.status !== context.options.statusTag) {
                 (0, log_1.verbose)(`Skipping page because status is not '${context.options.statusTag}': ${page.nameOrTitle}`);
-                ++context.counts.skipped_because_status;
+                ++counts.skipped_because_status;
             }
             else {
                 if (options.requireSlugs && !page.hasExplicitSlug) {
@@ -149,7 +152,7 @@ function outputPages(options, config, pages) {
                     ++counts.error_because_no_slug;
                 }
                 const markdown = yield (0, transform_1.getMarkdownForPage)(config, context, page);
-                writePage(page, markdown);
+                writePage(page, markdown, counts);
             }
         }
         if (counts.error_because_no_slug > 0)
@@ -164,7 +167,7 @@ function outputPages(options, config, pages) {
 // outline that contain content (which we call "Simple" pages). Later, we can
 // then step through this list creating the files we need, and, crucially, be
 // able to figure out what the url will be for any links between content pages.
-function getPagesRecursively(options, incomingContext, pageIdOfThisParent, orderOfThisParent, rootLevel) {
+function getPagesRecursively(options, incomingContext, pageIdOfThisParent, orderOfThisParent, rootLevel, pages, counts) {
     return __awaiter(this, void 0, void 0, function* () {
         var _a;
         const pageInTheOutline = yield fromPageId(incomingContext, pageIdOfThisParent, orderOfThisParent, true);
@@ -197,7 +200,7 @@ function getPagesRecursively(options, incomingContext, pageIdOfThisParent, order
                 layoutContext = layoutStrategy.newLevel(options.markdownOutputPath, pageInTheOutline.order, incomingContext, pageInTheOutline.nameOrTitle);
             }
             for (const childPageInfo of pageInfo.childPageIdsAndOrder) {
-                yield getPagesRecursively(options, layoutContext, childPageInfo.id, childPageInfo.order, false);
+                yield getPagesRecursively(options, layoutContext, childPageInfo.id, childPageInfo.order, false, pages, counts);
             }
             for (const linkPageInfo of pageInfo.linksPageIdsAndOrder) {
                 pages.push(yield fromPageId(layoutContext, linkPageInfo.id, linkPageInfo.order, false));
@@ -209,7 +212,7 @@ function getPagesRecursively(options, incomingContext, pageIdOfThisParent, order
         }
     });
 }
-function writePage(page, finalMarkdown) {
+function writePage(page, finalMarkdown, counts) {
     const mdPath = layoutStrategy.getPathForPage(page, ".md");
     (0, log_1.verbose)(`writing ${mdPath}`);
     fs.writeFileSync(mdPath, finalMarkdown, {});

package/dist/transform.js

@@ -113,13 +113,19 @@ function doTransformsOnMarkdown(context, config, input) {
         let body = input;
         //console.log("body before regex: " + body);
         let match;
-        // eslint-disable-next-line @typescript-eslint/no-unused-vars
         for (const mod of regexMods) {
-            let replacement = undefined;
             // regex.exec is stateful, so we don't want to mess up the plugin's use of its own regex, so we clone it.
             // we also add the "g" flag to make sure we get all matches
             const regex = new RegExp(`${codeBlocks.source}|(${mod.regex.source})`, "g");
-            while ((match = regex.exec(input)) !== null) {
+            // We match against `source` (a snapshot of the current body) and build up a
+            // fresh `result`. Mutating the string we're iterating over would invalidate
+            // `match.index` for every subsequent match once a replacement changes the
+            // length, so instead we copy the spans between matches using indices that
+            // always refer to the unchanged `source`.
+            const source = body;
+            let result = "";
+            let lastIndex = 0;
+            while ((match = regex.exec(source)) !== null) {
                 if (match[0]) {
                     const original = match[0];
                     if (original.startsWith("```") &&
@@ -127,6 +133,7 @@ function doTransformsOnMarkdown(context, config, input) {
                         !mod.includeCodeBlocks) {
                         continue; // code block, and they didn't say to include them
                     }
+                    let replacement = undefined;
                     if (mod.getReplacement) {
                         // our match here has an extra group, which is an implementation detail
                         // that shouldn't be made visible to the plugin
@@ -138,11 +145,11 @@ function doTransformsOnMarkdown(context, config, input) {
                     }
                     if (replacement !== undefined) {
                         (0, log_1.verbose)(`[${mod.name}] ${original} --> ${replacement}`);
-                        const precedingPart = body.substring(0, match.index); // ?
-                        const partStartingFromThisMatch = body.substring(match.index); // ?
-                        body =
-                            precedingPart +
-                                partStartingFromThisMatch.replace(original, replacement);
+                        // copy everything since the last match, then the replacement (the
+                        // replacement is appended literally, so `$` sequences in it are not
+                        // given any special meaning)
+                        result += source.substring(lastIndex, match.index) + replacement;
+                        lastIndex = match.index + original.length;
                         // add any library imports
                         if (!context.imports)
                             context.imports = [];
@@ -150,6 +157,9 @@ function doTransformsOnMarkdown(context, config, input) {
                     }
                 }
             }
+            // copy whatever follows the last replacement (including any skipped code blocks)
+            result += source.substring(lastIndex);
+            body = result;
         }
         (0, log_1.logDebug)("doTransformsOnMarkdown", "body after regex: " + body);
         return body;

package/dist/transform.spec.d.ts

@@ -0,0 +1 @@
+export {};

package/dist/transform.spec.js

@@ -0,0 +1,63 @@
+"use strict";
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+const log_1 = require("./log");
+const pluginTestRun_1 = require("./plugins/pluginTestRun");
+function paragraph(text) {
+    return {
+        type: "paragraph",
+        paragraph: {
+            rich_text: [
+                {
+                    type: "text",
+                    text: { content: text, link: null },
+                    annotations: {
+                        bold: false,
+                        italic: false,
+                        strikethrough: false,
+                        underline: false,
+                        code: false,
+                        color: "default",
+                    },
+                    plain_text: text,
+                    href: null,
+                },
+            ],
+        },
+    };
+}
+// Regression test for the bug where doTransformsOnMarkdown matched against the
+// original `input` but sliced/replaced the mutating `body`. Once the first
+// replacement changes the string length, every subsequent match.index is wrong.
+// We make the first replacement SHRINK the text so the second match.index lands
+// past the real match, and the inner .replace silently drops the transformation.
+test("applies every match even after a length-changing replacement", () => __awaiter(void 0, void 0, void 0, function* () {
+    (0, log_1.setLogLevel)("verbose");
+    const p = {
+        name: "test",
+        regexMarkdownModifications: [
+            {
+                regex: /REMOVE/,
+                replacementPattern: `X`, // shorter than the match, so the body shrinks
+            },
+        ],
+    };
+    const config = { plugins: [p] };
+    // Produces markdown: "REMOVE\n\nREMOVE keep"
+    const result = yield (0, pluginTestRun_1.blocksToMarkdown)(config, [
+        paragraph("REMOVE"),
+        paragraph("REMOVE keep"),
+    ]);
+    // Both occurrences must be transformed; none should survive.
+    expect(result).not.toContain("REMOVE");
+    expect((result.match(/X/g) || []).length).toBe(2);
+    expect(result).toContain("keep");
+}));

package/package.json

@@ -54,6 +54,7 @@
     "lint-staged": "^10.5.4",
     "prettier": "^2.2.1",
     "rimraf": "^4.1.2",
+    "semantic-release": "25.0.3",
     "typescript": "^5.9.3",
     "vite": "^4.2.1",
     "vitest": "^0.30.1"
@@ -87,8 +88,15 @@
   "publishConfig": {
     "access": "public"
   },
+  "packageManager": "npm@11.16.0",
   "volta": {
-    "node": "22.21.0"
+    "node": "24.16.0",
+    "npm": "11.16.0"
   },
-  "version": "1.0.0-alpha.2"
+  "allowScripts": {
+    "core-js@2.6.12": true,
+    "esbuild@0.17.17": true,
+    "fsevents": false
+  },
+  "version": "1.0.0-alpha.5"
 }