@promptbook/website-crawler 0.84.0-12 → 0.84.0-14

package/esm/index.es.js

@@ -9,7 +9,7 @@ import { format } from 'prettier';
 import parserHtml from 'prettier/parser-html';
 import { forTime } from 'waitasecond';
 import sha256 from 'crypto-js/sha256';
-import { lookup } from 'mime-types';
+import { lookup, extension } from 'mime-types';
 import { unparse, parse } from 'papaparse';
 import { Converter } from 'showdown';
 
@@ -27,7 +27,7 @@ var BOOK_LANGUAGE_VERSION = '1.0.0';
  * @generated
  * @see https://github.com/webgptorg/promptbook
  */
-var PROMPTBOOK_ENGINE_VERSION = '0.84.0-11';
+var PROMPTBOOK_ENGINE_VERSION = '0.84.0-13';
 /**
  * TODO: string_promptbook_version should be constrained to the all versions of Promptbook engine
  * Note: [💞] Ignore a discrepancy between file name and entity name
@@ -3590,6 +3590,17 @@ function isFileExisting(filename, fs) {
  * TODO: [🖇] What about symlinks?
  */
 
+/**
+ * Convert mime type to file extension
+ *
+ * Note: If the mime type is invalid, `null` is returned
+ *
+ * @private within the repository
+ */
+function mimeTypeToExtension(value) {
+    return extension(value) || null;
+}
+
 /**
  * The built-in `fetch' function with a lightweight error handling wrapper as default fetch function used in Promptbook scrapers
  *
@@ -3625,7 +3636,7 @@ var scraperFetch = function (url, init) { return __awaiter(void 0, void 0, void
 function makeKnowledgeSourceHandler(knowledgeSource, tools, options) {
     var _a;
     return __awaiter(this, void 0, void 0, function () {
-        var _b, fetch, knowledgeSourceContent, name, _c, _d, rootDirname, url, response, mimeType, filename, hash, rootDirname_1, filepath, _f, _g, _h, _j, _k, filename_1, fileExtension, mimeType;
+        var _b, fetch, knowledgeSourceContent, name, _c, _d, rootDirname, url, response_1, mimeType, basename, hash, rootDirname_1, filepath, _f, _g, _h, _j, _k, filename_1, fileExtension, mimeType;
         return __generator(this, function (_l) {
             switch (_l.label) {
                 case 0:
@@ -3641,25 +3652,67 @@ function makeKnowledgeSourceHandler(knowledgeSource, tools, options) {
                     url = knowledgeSourceContent;
                     return [4 /*yield*/, fetch(url)];
                 case 1:
-                    response = _l.sent();
-                    mimeType = ((_a = response.headers.get('content-type')) === null || _a === void 0 ? void 0 : _a.split(';')[0]) || 'text/html';
-                    filename = url.split('/').pop() || titleToName(url);
+                    response_1 = _l.sent();
+                    mimeType = ((_a = response_1.headers.get('content-type')) === null || _a === void 0 ? void 0 : _a.split(';')[0]) || 'text/html';
+                    if (tools.fs === undefined || !url.endsWith('.pdf' /* <- TODO: [💵] */)) {
+                        return [2 /*return*/, {
+                                source: name,
+                                filename: null,
+                                url: url,
+                                mimeType: mimeType,
+                                /*
+                              TODO: [🥽]
+                                  > async asBlob() {
+                                  >     // TODO: [👨🏻‍🤝‍👨🏻] This can be called multiple times BUT when called second time, response in already consumed
+                                  >     const content = await response.blob();
+                                  >     return content;
+                                  > },
+                              */
+                                asJson: function () {
+                                    return __awaiter(this, void 0, void 0, function () {
+                                        var content;
+                                        return __generator(this, function (_a) {
+                                            switch (_a.label) {
+                                                case 0: return [4 /*yield*/, response_1.json()];
+                                                case 1:
+                                                    content = _a.sent();
+                                                    return [2 /*return*/, content];
+                                            }
+                                        });
+                                    });
+                                },
+                                asText: function () {
+                                    return __awaiter(this, void 0, void 0, function () {
+                                        var content;
+                                        return __generator(this, function (_a) {
+                                            switch (_a.label) {
+                                                case 0: return [4 /*yield*/, response_1.text()];
+                                                case 1:
+                                                    content = _a.sent();
+                                                    return [2 /*return*/, content];
+                                            }
+                                        });
+                                    });
+                                },
+                            }];
+                    }
+                    basename = url.split('/').pop() || titleToName(url);
                     hash = sha256(hexEncoder.parse(url)).toString( /* hex */);
                     rootDirname_1 = join(process.cwd(), DEFAULT_DOWNLOAD_CACHE_DIRNAME);
-                    filepath = join.apply(void 0, __spreadArray(__spreadArray([], __read(nameToSubfolderPath(hash /* <- TODO: [🎎] Maybe add some SHA256 prefix */)), false), ["".concat(filename.substring(0, MAX_FILENAME_LENGTH), ".pdf")], false));
+                    filepath = join.apply(void 0, __spreadArray(__spreadArray([], __read(nameToSubfolderPath(hash /* <- TODO: [🎎] Maybe add some SHA256 prefix */)), false), ["".concat(basename.substring(0, MAX_FILENAME_LENGTH), ".").concat(mimeTypeToExtension(mimeType))], false));
                     return [4 /*yield*/, tools.fs.mkdir(dirname(join(rootDirname_1, filepath)), { recursive: true })];
                 case 2:
                     _l.sent();
                     _g = (_f = tools.fs).writeFile;
                     _h = [join(rootDirname_1, filepath)];
                     _k = (_j = Buffer).from;
-                    return [4 /*yield*/, response.arrayBuffer()];
+                    return [4 /*yield*/, response_1.arrayBuffer()];
                 case 3: return [4 /*yield*/, _g.apply(_f, _h.concat([_k.apply(_j, [_l.sent()])]))];
                 case 4:
                     _l.sent();
-                    // TODO: !!!!!!!! Check the file security
+                    // TODO: [💵] Check the file security
                     // TODO: !!!!!!!! Check the file size (if it is not too big)
-                    // TODO: !!!!!!!! Delete the file
+                    // TODO: !!!!!!!! Delete the file after the scraping is done
                     return [2 /*return*/, makeKnowledgeSourceHandler({ name: name, knowledgeSourceContent: filepath }, tools, __assign(__assign({}, options), { rootDirname: rootDirname_1 }))];
                 case 5:
                     if (!isValidFilePath(knowledgeSourceContent)) return [3 /*break*/, 7];