@promptbook/website-crawler 0.78.0-0 → 0.78.2

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -23,10 +23,6 @@
23
23
 
24
24
 
25
25
 
26
- <blockquote style="color: #ff8811">
27
- <b>⚠ Warning:</b> This is a pre-release version of the library. It is not yet ready for production use. Please look at <a href="https://www.npmjs.com/package/@promptbook/core?activeTab=versions">latest stable release</a>.
28
- </blockquote>
29
-
30
26
  ## 📦 Package `@promptbook/website-crawler`
31
27
 
32
28
  - Promptbooks are [divided into several](#-packages) packages, all are published from [single monorepo](https://github.com/webgptorg/promptbook).
package/esm/index.es.js CHANGED
@@ -24,7 +24,7 @@ var BOOK_LANGUAGE_VERSION = '1.0.0';
24
24
  *
25
25
  * @see https://github.com/webgptorg/promptbook
26
26
  */
27
- var PROMPTBOOK_ENGINE_VERSION = '0.77.1';
27
+ var PROMPTBOOK_ENGINE_VERSION = '0.78.1';
28
28
  /**
29
29
  * TODO: string_promptbook_version should be constrained to the all versions of Promptbook engine
30
30
  * Note: [💞] Ignore a discrepancy between file name and entity name
@@ -204,164 +204,25 @@ function just(value) {
204
204
  }
205
205
 
206
206
  /**
207
- * This error type indicates that the error should not happen and its last check before crashing with some other error
207
+ * Name for the Promptbook
208
+ *
209
+ * TODO: [🗽] Unite branding and make single place for it
208
210
  *
209
211
  * @public exported from `@promptbook/core`
210
212
  */
211
- var UnexpectedError = /** @class */ (function (_super) {
212
- __extends(UnexpectedError, _super);
213
- function UnexpectedError(message) {
214
- var _this = _super.call(this, spaceTrim(function (block) { return "\n ".concat(block(message), "\n\n Note: This error should not happen.\n It's probbably a bug in the pipeline collection\n\n Please report issue:\n https://github.com/webgptorg/promptbook/issues\n\n Or contact us on me@pavolhejny.com\n\n "); })) || this;
215
- _this.name = 'UnexpectedError';
216
- Object.setPrototypeOf(_this, UnexpectedError.prototype);
217
- return _this;
218
- }
219
- return UnexpectedError;
220
- }(Error));
221
-
213
+ var NAME = "Promptbook";
222
214
  /**
223
- * Checks if the value is [🚉] serializable as JSON
224
- * If not, throws an UnexpectedError with a rich error message and tracking
225
- *
226
- * - Almost all primitives are serializable BUT:
227
- * - `undefined` is not serializable
228
- * - `NaN` is not serializable
229
- * - Objects and arrays are serializable if all their properties are serializable
230
- * - Functions are not serializable
231
- * - Circular references are not serializable
232
- * - `Date` objects are not serializable
233
- * - `Map` and `Set` objects are not serializable
234
- * - `RegExp` objects are not serializable
235
- * - `Error` objects are not serializable
236
- * - `Symbol` objects are not serializable
237
- * - And much more...
215
+ * Email of the responsible person
238
216
  *
239
- * @throws UnexpectedError if the value is not serializable as JSON
240
- * @public exported from `@promptbook/utils`
241
- */
242
- function checkSerializableAsJson(name, value) {
243
- var e_1, _a;
244
- if (value === undefined) {
245
- throw new UnexpectedError("".concat(name, " is undefined"));
246
- }
247
- else if (value === null) {
248
- return;
249
- }
250
- else if (typeof value === 'boolean') {
251
- return;
252
- }
253
- else if (typeof value === 'number' && !isNaN(value)) {
254
- return;
255
- }
256
- else if (typeof value === 'string') {
257
- return;
258
- }
259
- else if (typeof value === 'symbol') {
260
- throw new UnexpectedError("".concat(name, " is symbol"));
261
- }
262
- else if (typeof value === 'function') {
263
- throw new UnexpectedError("".concat(name, " is function"));
264
- }
265
- else if (typeof value === 'object' && Array.isArray(value)) {
266
- for (var i = 0; i < value.length; i++) {
267
- checkSerializableAsJson("".concat(name, "[").concat(i, "]"), value[i]);
268
- }
269
- }
270
- else if (typeof value === 'object') {
271
- if (value instanceof Date) {
272
- throw new UnexpectedError(spaceTrim$1("\n ".concat(name, " is Date\n\n Use `string_date_iso8601` instead\n ")));
273
- }
274
- else if (value instanceof Map) {
275
- throw new UnexpectedError("".concat(name, " is Map"));
276
- }
277
- else if (value instanceof Set) {
278
- throw new UnexpectedError("".concat(name, " is Set"));
279
- }
280
- else if (value instanceof RegExp) {
281
- throw new UnexpectedError("".concat(name, " is RegExp"));
282
- }
283
- else if (value instanceof Error) {
284
- throw new UnexpectedError(spaceTrim$1("\n ".concat(name, " is unserialized Error\n\n Use function `serializeError`\n ")));
285
- }
286
- else {
287
- try {
288
- for (var _b = __values(Object.entries(value)), _c = _b.next(); !_c.done; _c = _b.next()) {
289
- var _d = __read(_c.value, 2), subName = _d[0], subValue = _d[1];
290
- if (subValue === undefined) {
291
- // Note: undefined in object is serializable - it is just omited
292
- continue;
293
- }
294
- checkSerializableAsJson("".concat(name, ".").concat(subName), subValue);
295
- }
296
- }
297
- catch (e_1_1) { e_1 = { error: e_1_1 }; }
298
- finally {
299
- try {
300
- if (_c && !_c.done && (_a = _b.return)) _a.call(_b);
301
- }
302
- finally { if (e_1) throw e_1.error; }
303
- }
304
- try {
305
- JSON.stringify(value); // <- TODO: [0]
306
- }
307
- catch (error) {
308
- if (!(error instanceof Error)) {
309
- throw error;
310
- }
311
- throw new UnexpectedError(spaceTrim$1(function (block) { return "\n ".concat(name, " is not serializable\n\n ").concat(block(error.toString()), "\n "); }));
312
- }
313
- /*
314
- TODO: [0] Is there some more elegant way to check circular references?
315
- const seen = new Set();
316
- const stack = [{ value }];
317
- while (stack.length > 0) {
318
- const { value } = stack.pop()!;
319
- if (typeof value === 'object' && value !== null) {
320
- if (seen.has(value)) {
321
- throw new UnexpectedError(`${name} has circular reference`);
322
- }
323
- seen.add(value);
324
- if (Array.isArray(value)) {
325
- stack.push(...value.map((value) => ({ value })));
326
- } else {
327
- stack.push(...Object.values(value).map((value) => ({ value })));
328
- }
329
- }
330
- }
331
- */
332
- return;
333
- }
334
- }
335
- else {
336
- throw new UnexpectedError("".concat(name, " is unknown"));
337
- }
338
- }
339
- /**
340
- * TODO: [🧠][🛣] More elegant way to tracking than passing `name`
341
- * TODO: [🧠][main] !!! In-memory cache of same values to prevent multiple checks
342
- * Note: [🐠] This is how `checkSerializableAsJson` + `isSerializableAsJson` together can just retun true/false or rich error message
217
+ * @public exported from `@promptbook/core`
343
218
  */
344
-
219
+ var ADMIN_EMAIL = 'me@pavolhejny.com';
345
220
  /**
346
- * @@@
347
- * @@@
348
- *
349
- * Note: This function mutates the object and returns the original (but mutated-deep-freezed) object
221
+ * Name of the responsible person for the Promptbook on GitHub
350
222
  *
351
- * @param name - Name of the object for debugging purposes
352
- * @param objectValue - Object to be deeply frozen
353
- * @returns The same object as the input, but deeply frozen
354
- * @private this is in comparison to `deepFreeze` a more specific utility and maybe not very good practice to use without specific reason and considerations
355
- */
356
- function $asDeeplyFrozenSerializableJson(name, objectValue) {
357
- checkSerializableAsJson(name, objectValue);
358
- return $deepFreeze(objectValue);
359
- }
360
- /**
361
- * TODO: [🧠][🛣] More elegant way to tracking than passing `name`
362
- * TODO: [🧠] Is there a way how to meaningfully test this utility
223
+ * @public exported from `@promptbook/core`
363
224
  */
364
-
225
+ var ADMIN_GITHUB_NAME = 'hejny';
365
226
  // <- TODO: [🧠] Better system for generator warnings - not always "code" and "by `@promptbook/cli`"
366
227
  /**
367
228
  * The maximum number of iterations for a loops
@@ -413,7 +274,8 @@ var REPLACING_NONCE = 'u$k42k%!V2zo34w7Fu#@QUHYPW';
413
274
  *
414
275
  * @public exported from `@promptbook/core`
415
276
  */
416
- var RESERVED_PARAMETER_NAMES = $asDeeplyFrozenSerializableJson('RESERVED_PARAMETER_NAMES', [
277
+ var RESERVED_PARAMETER_NAMES =
278
+ /* !!!!!! $asDeeplyFrozenSerializableJson('RESERVED_PARAMETER_NAMES', _____ as const); */ [
417
279
  'content',
418
280
  'context',
419
281
  'knowledge',
@@ -423,7 +285,7 @@ var RESERVED_PARAMETER_NAMES = $asDeeplyFrozenSerializableJson('RESERVED_PARAMET
423
285
  // <- TODO: list here all command names
424
286
  // <- TODO: Add more like 'date', 'modelName',...
425
287
  // <- TODO: Add [emoji] + instructions ACRY when adding new reserved parameter
426
- ]);
288
+ ];
427
289
  /**
428
290
  * @@@
429
291
  *
@@ -491,6 +353,40 @@ var NotYetImplementedError = /** @class */ (function (_super) {
491
353
  return NotYetImplementedError;
492
354
  }(Error));
493
355
 
356
+ /**
357
+ * Make error report URL for the given error
358
+ *
359
+ * @private !!!!!!
360
+ */
361
+ function getErrorReportUrl(error) {
362
+ var report = {
363
+ title: "\uD83D\uDC1C Error report from ".concat(NAME),
364
+ body: spaceTrim$1(function (block) { return "\n\n\n `".concat(error.name || 'Error', "` has occurred in the [").concat(NAME, "], please look into it @").concat(ADMIN_GITHUB_NAME, ".\n\n ```\n ").concat(block(error.message || '(no error message)'), "\n ```\n\n\n ## More info:\n\n - **Promptbook engine version:** ").concat(PROMPTBOOK_ENGINE_VERSION, "\n - **Book language version:** ").concat(BOOK_LANGUAGE_VERSION, "\n - **Time:** ").concat(new Date().toISOString(), "\n\n <details>\n <summary>Stack trace:</summary>\n\n ## Stack trace:\n\n ```stacktrace\n ").concat(block(error.stack || '(empty)'), "\n ```\n </details>\n\n "); }),
365
+ };
366
+ var reportUrl = new URL("https://github.com/webgptorg/promptbook/issues/new");
367
+ reportUrl.searchParams.set('labels', 'bug');
368
+ reportUrl.searchParams.set('assignees', ADMIN_GITHUB_NAME);
369
+ reportUrl.searchParams.set('title', report.title);
370
+ reportUrl.searchParams.set('body', report.body);
371
+ return reportUrl;
372
+ }
373
+
374
+ /**
375
+ * This error type indicates that the error should not happen and its last check before crashing with some other error
376
+ *
377
+ * @public exported from `@promptbook/core`
378
+ */
379
+ var UnexpectedError = /** @class */ (function (_super) {
380
+ __extends(UnexpectedError, _super);
381
+ function UnexpectedError(message) {
382
+ var _this = _super.call(this, spaceTrim(function (block) { return "\n ".concat(block(message), "\n\n Note: This error should not happen.\n It's probbably a bug in the pipeline collection\n\n Please report issue:\n ").concat(block(getErrorReportUrl(new Error(message)).href), "\n\n Or contact us on ").concat(ADMIN_EMAIL, "\n\n "); })) || this;
383
+ _this.name = 'UnexpectedError';
384
+ Object.setPrototypeOf(_this, UnexpectedError.prototype);
385
+ return _this;
386
+ }
387
+ return UnexpectedError;
388
+ }(Error));
389
+
494
390
  /**
495
391
  * @@@
496
392
  *
@@ -1936,6 +1832,149 @@ function extractParameterNames(template) {
1936
1832
  return parameterNames;
1937
1833
  }
1938
1834
 
1835
+ /**
1836
+ * Checks if the value is [🚉] serializable as JSON
1837
+ * If not, throws an UnexpectedError with a rich error message and tracking
1838
+ *
1839
+ * - Almost all primitives are serializable BUT:
1840
+ * - `undefined` is not serializable
1841
+ * - `NaN` is not serializable
1842
+ * - Objects and arrays are serializable if all their properties are serializable
1843
+ * - Functions are not serializable
1844
+ * - Circular references are not serializable
1845
+ * - `Date` objects are not serializable
1846
+ * - `Map` and `Set` objects are not serializable
1847
+ * - `RegExp` objects are not serializable
1848
+ * - `Error` objects are not serializable
1849
+ * - `Symbol` objects are not serializable
1850
+ * - And much more...
1851
+ *
1852
+ * @throws UnexpectedError if the value is not serializable as JSON
1853
+ * @public exported from `@promptbook/utils`
1854
+ */
1855
+ function checkSerializableAsJson(name, value) {
1856
+ var e_1, _a;
1857
+ if (value === undefined) {
1858
+ throw new UnexpectedError("".concat(name, " is undefined"));
1859
+ }
1860
+ else if (value === null) {
1861
+ return;
1862
+ }
1863
+ else if (typeof value === 'boolean') {
1864
+ return;
1865
+ }
1866
+ else if (typeof value === 'number' && !isNaN(value)) {
1867
+ return;
1868
+ }
1869
+ else if (typeof value === 'string') {
1870
+ return;
1871
+ }
1872
+ else if (typeof value === 'symbol') {
1873
+ throw new UnexpectedError("".concat(name, " is symbol"));
1874
+ }
1875
+ else if (typeof value === 'function') {
1876
+ throw new UnexpectedError("".concat(name, " is function"));
1877
+ }
1878
+ else if (typeof value === 'object' && Array.isArray(value)) {
1879
+ for (var i = 0; i < value.length; i++) {
1880
+ checkSerializableAsJson("".concat(name, "[").concat(i, "]"), value[i]);
1881
+ }
1882
+ }
1883
+ else if (typeof value === 'object') {
1884
+ if (value instanceof Date) {
1885
+ throw new UnexpectedError(spaceTrim$1("\n ".concat(name, " is Date\n\n Use `string_date_iso8601` instead\n ")));
1886
+ }
1887
+ else if (value instanceof Map) {
1888
+ throw new UnexpectedError("".concat(name, " is Map"));
1889
+ }
1890
+ else if (value instanceof Set) {
1891
+ throw new UnexpectedError("".concat(name, " is Set"));
1892
+ }
1893
+ else if (value instanceof RegExp) {
1894
+ throw new UnexpectedError("".concat(name, " is RegExp"));
1895
+ }
1896
+ else if (value instanceof Error) {
1897
+ throw new UnexpectedError(spaceTrim$1("\n ".concat(name, " is unserialized Error\n\n Use function `serializeError`\n ")));
1898
+ }
1899
+ else {
1900
+ try {
1901
+ for (var _b = __values(Object.entries(value)), _c = _b.next(); !_c.done; _c = _b.next()) {
1902
+ var _d = __read(_c.value, 2), subName = _d[0], subValue = _d[1];
1903
+ if (subValue === undefined) {
1904
+ // Note: undefined in object is serializable - it is just omited
1905
+ continue;
1906
+ }
1907
+ checkSerializableAsJson("".concat(name, ".").concat(subName), subValue);
1908
+ }
1909
+ }
1910
+ catch (e_1_1) { e_1 = { error: e_1_1 }; }
1911
+ finally {
1912
+ try {
1913
+ if (_c && !_c.done && (_a = _b.return)) _a.call(_b);
1914
+ }
1915
+ finally { if (e_1) throw e_1.error; }
1916
+ }
1917
+ try {
1918
+ JSON.stringify(value); // <- TODO: [0]
1919
+ }
1920
+ catch (error) {
1921
+ if (!(error instanceof Error)) {
1922
+ throw error;
1923
+ }
1924
+ throw new UnexpectedError(spaceTrim$1(function (block) { return "\n ".concat(name, " is not serializable\n\n ").concat(block(error.toString()), "\n "); }));
1925
+ }
1926
+ /*
1927
+ TODO: [0] Is there some more elegant way to check circular references?
1928
+ const seen = new Set();
1929
+ const stack = [{ value }];
1930
+ while (stack.length > 0) {
1931
+ const { value } = stack.pop()!;
1932
+ if (typeof value === 'object' && value !== null) {
1933
+ if (seen.has(value)) {
1934
+ throw new UnexpectedError(`${name} has circular reference`);
1935
+ }
1936
+ seen.add(value);
1937
+ if (Array.isArray(value)) {
1938
+ stack.push(...value.map((value) => ({ value })));
1939
+ } else {
1940
+ stack.push(...Object.values(value).map((value) => ({ value })));
1941
+ }
1942
+ }
1943
+ }
1944
+ */
1945
+ return;
1946
+ }
1947
+ }
1948
+ else {
1949
+ throw new UnexpectedError("".concat(name, " is unknown"));
1950
+ }
1951
+ }
1952
+ /**
1953
+ * TODO: [🧠][🛣] More elegant way to tracking than passing `name`
1954
+ * TODO: [🧠][main] !!! In-memory cache of same values to prevent multiple checks
1955
+ * Note: [🐠] This is how `checkSerializableAsJson` + `isSerializableAsJson` together can just retun true/false or rich error message
1956
+ */
1957
+
1958
+ /**
1959
+ * @@@
1960
+ * @@@
1961
+ *
1962
+ * Note: This function mutates the object and returns the original (but mutated-deep-freezed) object
1963
+ *
1964
+ * @param name - Name of the object for debugging purposes
1965
+ * @param objectValue - Object to be deeply frozen
1966
+ * @returns The same object as the input, but deeply frozen
1967
+ * @private this is in comparison to `deepFreeze` a more specific utility and maybe not very good practice to use without specific reason and considerations
1968
+ */
1969
+ function $asDeeplyFrozenSerializableJson(name, objectValue) {
1970
+ checkSerializableAsJson(name, objectValue);
1971
+ return $deepFreeze(objectValue);
1972
+ }
1973
+ /**
1974
+ * TODO: [🧠][🛣] More elegant way to tracking than passing `name`
1975
+ * TODO: [🧠] Is there a way how to meaningfully test this utility
1976
+ */
1977
+
1939
1978
  /**
1940
1979
  * Unprepare just strips the preparation data of the pipeline
1941
1980
  *
@@ -3701,6 +3740,7 @@ function preparePipeline(pipeline, tools, options) {
3701
3740
  */
3702
3741
  function extractVariablesFromScript(script) {
3703
3742
  var variables = new Set();
3743
+ var originalScript = script;
3704
3744
  script = "(()=>{".concat(script, "})()");
3705
3745
  try {
3706
3746
  for (var i = 0; i < 100 /* <- TODO: This limit to configuration */; i++)
@@ -3732,7 +3772,9 @@ function extractVariablesFromScript(script) {
3732
3772
  if (!(error instanceof Error)) {
3733
3773
  throw error;
3734
3774
  }
3735
- throw new ParseError(spaceTrim(function (block) { return "\n Can not extract variables from the script\n\n ".concat(block(error.toString()), "}\n "); }));
3775
+ throw new ParseError(spaceTrim(function (block) { return "\n Can not extract variables from the script\n\n ".concat(block(error.toString()), "}\n\n\n Found variables:\n\n ").concat(Array.from(variables)
3776
+ .map(function (variableName, i) { return "".concat(i + 1, ") ").concat(variableName); })
3777
+ .join('\n'), "\n\n\n The script:\n\n ```javascript\n ").concat(block(originalScript), "\n ```\n "); }));
3736
3778
  }
3737
3779
  return variables;
3738
3780
  }