Allow reading from files (#4)

* Add tests

* Read markdown file

* Fix tests

* Read notebooks too

* Include file path in error message

* Make file reading async

* Fix for windows?

* Actually fix for windows

Should have read the errors properly

* Update src/lib.rs

Co-authored-by: Eric Arellano <14852634+Eric-Arellano@users.noreply.github.com>

---------

Co-authored-by: Eric Arellano <14852634+Eric-Arellano@users.noreply.github.com>

Author: frankharkins

Cargo.toml

@@ -10,8 +10,11 @@ crate-type = ["cdylib"]
 
 [dependencies]
 markdown    = "1.0.0"
-napi        = "3.0.0"
+napi        = { version = "3.0.0", features = ["async"] }
 napi-derive = "3.0.0"
+serde       = { version = "1.0.228", features = ["derive"] }
+serde_json  = "1.0.149"
+tokio       = { version = "1.49.0", features = ["fs"] }
 
 [build-dependencies]
 napi-build = "2"

__test__/fixtures/invalid-notebook-json.ipynb

@@ -0,0 +1,10 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "source": [
+        "This is not valid JSON due to the following trailing comma:",
+      ]
+    }
+  ]
+}

__test__/fixtures/invalid-notebook-schema.ipynb

@@ -0,0 +1,12 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "sauce": [
+        "# Example notebook\n",
+        "\n",
+        "With a [link](/path)!"
+      ]
+    }
+  ]
+}

__test__/fixtures/markdown.ipynb

@@ -0,0 +1,73 @@
+{
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "id": "48bb7c88-5f83-4f6d-9fe8-aa20f188c4ff",
+      "metadata": {},
+      "source": [
+        "# Example notebook\n",
+        "\n",
+        "With a [link](/path)!"
+      ]
+    },
+    {
+      "cell_type": "code",
+      "execution_count": 1,
+      "id": "bcaf0da4-226f-46f7-b575-8a68880acdfd",
+      "metadata": {},
+      "outputs": [
+        {
+          "data": {
+            "text/plain": [
+              "Some output"
+            ]
+          },
+          "execution_count": 1,
+          "metadata": {},
+          "output_type": "execute_result"
+        }
+      ],
+      "source": [
+        "print(\"Some output\")"
+      ]
+    },
+    {
+      "cell_type": "raw",
+      "id": "e30adffd-fb91-4711-abfa-79ebe7eeee46",
+      "metadata": {},
+      "source": [
+        "???"
+      ]
+    },
+    {
+      "cell_type": "markdown",
+      "id": "48bb7c88-5f83-4f6d-9fe8-aa20f188c4ff",
+      "metadata": {},
+      "source": [
+        "And a second [link](/path2)"
+      ]
+    }
+  ],
+  "metadata": {
+    "kernelspec": {
+      "display_name": "Python 3",
+      "language": "python",
+      "name": "python3"
+    },
+    "language_info": {
+      "codemirror_mode": {
+        "name": "ipython",
+        "version": 3
+      },
+      "file_extension": ".py",
+      "mimetype": "text/x-python",
+      "name": "python",
+      "nbconvert_exporter": "python",
+      "pygments_lexer": "ipython3",
+      "version": "3"
+    },
+    "title": "Example notebook"
+  },
+  "nbformat": 4,
+  "nbformat_minor": 4
+}

__test__/fixtures/markdown.mdx

@@ -0,0 +1,3 @@
+# Example document
+
+Here's a simple page with a [link](/path)!

__test__/index.spec.ts

@@ -1,6 +1,6 @@
 import test from 'ava'
 
-import { extractLinks } from '../index'
+import { extractLinks, extractLinksFromFile } from '../index'
 
 const dedent = (s: string) => s.replace('\n    ', '')
 
@@ -108,3 +108,49 @@ test('extractLinks: appropriate jsx error message', (t) => {
   t.is(error.name, 'Error')
   t.is(error.message, '1:13: Expected a closing tag for `<Admonition>` (1:1) (markdown-rs:end-tag-mismatch)')
 })
+
+test('extractLinksFromFile: mdx file', async (t) => {
+  const links = await extractLinksFromFile('__test__/fixtures/markdown.mdx')
+  t.deepEqual(links, ['/path'])
+})
+
+test('extractLinksFromFile: notebook', async (t) => {
+  const links = (await extractLinksFromFile('__test__/fixtures/markdown.ipynb')).sort()
+  t.deepEqual(links, ['/path', '/path2'].sort())
+})
+
+test('extractLinksFromFile: markdown file not found', async (t) => {
+  const error = await t.throwsAsync(
+    async () => await extractLinksFromFile('__test__/fixtures/file_that_does_not_exist.md'),
+  )
+  t.is(error.name, 'Error')
+
+  // The error message changes depending on OS, but both are acceptable
+  const acceptableMessages = [
+    'Could not read "__test__/fixtures/file_that_does_not_exist.md": No such file or directory (os error 2)',
+    'Could not read "__test__/fixtures/file_that_does_not_exist.md": The system cannot find the file specified. (os error 2)',
+  ]
+  t.assert(acceptableMessages.includes(error.message))
+})
+
+test('extractLinksFromFile: invalid notebook (not JSON)', async (t) => {
+  const error = await t.throwsAsync(
+    async () => await extractLinksFromFile('__test__/fixtures/invalid-notebook-json.ipynb'),
+  )
+  t.is(error.name, 'Error')
+  t.is(
+    error.message,
+    'Could not read "__test__/fixtures/invalid-notebook-json.ipynb": trailing comma at line 7 column 7',
+  )
+})
+
+test('extractLinksFromFile: invalid notebook (bad schema)', async (t) => {
+  const error = await t.throwsAsync(
+    async () => await extractLinksFromFile('__test__/fixtures/invalid-notebook-schema.ipynb'),
+  )
+  t.is(error.name, 'Error')
+  t.is(
+    error.message,
+    'Could not read "__test__/fixtures/invalid-notebook-schema.ipynb": missing field `source` at line 10 column 5',
+  )
+})

index.d.ts

@@ -5,3 +5,5 @@
  * (gfm), math, and JSX.
  */
 export declare function extractLinks(markdown: string): Array<string>
+
+export declare function extractLinksFromFile(filePath: string): Promise<Array<string>>

index.js

@@ -573,3 +573,4 @@ if (!nativeBinding) {
 
 module.exports = nativeBinding
 module.exports.extractLinks = nativeBinding.extractLinks
+module.exports.extractLinksFromFile = nativeBinding.extractLinksFromFile

src/lib.rs

@@ -5,6 +5,36 @@ use markdown::{mdast::Node, to_mdast, Constructs, ParseOptions};
 use napi::Error;
 use napi_derive::napi;
 use std::collections::HashSet;
+use tokio::fs;
+
+use crate::notebook::extract_markdown_from_notebook_source;
+
+mod notebook;
+
+fn file_read_error(path: String, reason: String) -> Result<Vec<String>, Error> {
+  let message = format!("Could not read \"{path}\": {reason}");
+  Err(Error::from_reason(message))
+}
+
+#[napi]
+pub async fn extract_links_from_file(file_path: String) -> Result<Vec<String>, Error> {
+  let is_notebook = file_path.ends_with(".ipynb");
+  let source = match fs::read_to_string(&file_path).await {
+    Ok(s) => s,
+    Err(e) => return file_read_error(file_path, e.to_string()),
+  };
+
+  let markdown = if is_notebook {
+    match extract_markdown_from_notebook_source(source) {
+      Ok(md) => md,
+      Err(e) => return file_read_error(file_path, e.to_string()),
+    }
+  } else {
+    source
+  };
+
+  extract_links(markdown)
+}
 
 /// Extract links from a markdown string. Supports GitHub-flavored markdown
 /// (gfm), math, and JSX.

src/notebook/mod.rs

@@ -0,0 +1,29 @@
+use serde::Deserialize;
+
+pub fn extract_markdown_from_notebook_source(source: String) -> Result<String, String> {
+  let notebook: Notebook = match serde_json::from_str(&source) {
+    Ok(s) => s,
+    Err(e) => return Err(e.to_string()),
+  };
+
+  let markdown: String = notebook
+    .cells
+    .into_iter()
+    .filter(|cell| cell.cell_type == "markdown")
+    .map(|cell| cell.source.join(""))
+    .collect::<Vec<String>>()
+    .join("\n\n");
+
+  Ok(markdown)
+}
+
+#[derive(Deserialize)]
+struct Notebook {
+  cells: Vec<NotebookCell>,
+}
+
+#[derive(Deserialize)]
+struct NotebookCell {
+  cell_type: String,
+  source: Vec<String>,
+}