@promptbook/markitdown 0.100.0-2 → 0.100.0-4
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/esm/index.es.js +39 -5
- package/esm/index.es.js.map +1 -1
- package/esm/typings/src/version.d.ts +1 -1
- package/esm/typings/src/wizard/wizard.d.ts +14 -4
- package/package.json +2 -2
- package/umd/index.umd.js +39 -5
- package/umd/index.umd.js.map +1 -1
package/esm/index.es.js
CHANGED
|
@@ -26,7 +26,7 @@ const BOOK_LANGUAGE_VERSION = '1.0.0';
|
|
|
26
26
|
* @generated
|
|
27
27
|
* @see https://github.com/webgptorg/promptbook
|
|
28
28
|
*/
|
|
29
|
-
const PROMPTBOOK_ENGINE_VERSION = '0.100.0-
|
|
29
|
+
const PROMPTBOOK_ENGINE_VERSION = '0.100.0-4';
|
|
30
30
|
/**
|
|
31
31
|
* TODO: string_promptbook_version should be constrained to the all versions of Promptbook engine
|
|
32
32
|
* Note: [💞] Ignore a discrepancy between file name and entity name
|
|
@@ -834,7 +834,23 @@ async function getScraperIntermediateSource(source, options) {
|
|
|
834
834
|
.join('/') +
|
|
835
835
|
'.' +
|
|
836
836
|
extension;
|
|
837
|
-
|
|
837
|
+
// Note: Try to create cache directory, but don't fail if filesystem has issues
|
|
838
|
+
try {
|
|
839
|
+
await mkdir(dirname(cacheFilename), { recursive: true });
|
|
840
|
+
}
|
|
841
|
+
catch (error) {
|
|
842
|
+
// Note: If we can't create cache directory, continue without it
|
|
843
|
+
// This handles read-only filesystems, permission issues, and missing parent directories
|
|
844
|
+
if (error instanceof Error && (error.message.includes('EROFS') ||
|
|
845
|
+
error.message.includes('read-only') ||
|
|
846
|
+
error.message.includes('EACCES') ||
|
|
847
|
+
error.message.includes('EPERM') ||
|
|
848
|
+
error.message.includes('ENOENT'))) ;
|
|
849
|
+
else {
|
|
850
|
+
// Re-throw other unexpected errors
|
|
851
|
+
throw error;
|
|
852
|
+
}
|
|
853
|
+
}
|
|
838
854
|
let isDestroyed = true;
|
|
839
855
|
const fileHandler = {
|
|
840
856
|
filename: cacheFilename,
|
|
@@ -3484,7 +3500,23 @@ async function makeKnowledgeSourceHandler(knowledgeSource, tools, options) {
|
|
|
3484
3500
|
// <- TODO: [🥬] Encapsulate sha256 to some private utility function
|
|
3485
3501
|
const rootDirname = join(process.cwd(), DEFAULT_DOWNLOAD_CACHE_DIRNAME);
|
|
3486
3502
|
const filepath = join(...nameToSubfolderPath(hash /* <- TODO: [🎎] Maybe add some SHA256 prefix */), `${basename.substring(0, MAX_FILENAME_LENGTH)}.${mimeTypeToExtension(mimeType)}`);
|
|
3487
|
-
|
|
3503
|
+
// Note: Try to create cache directory, but don't fail if filesystem has issues
|
|
3504
|
+
try {
|
|
3505
|
+
await tools.fs.mkdir(dirname(join(rootDirname, filepath)), { recursive: true });
|
|
3506
|
+
}
|
|
3507
|
+
catch (error) {
|
|
3508
|
+
// Note: If we can't create cache directory, we'll handle it when trying to write the file
|
|
3509
|
+
// This handles read-only filesystems, permission issues, and missing parent directories
|
|
3510
|
+
if (error instanceof Error && (error.message.includes('EROFS') ||
|
|
3511
|
+
error.message.includes('read-only') ||
|
|
3512
|
+
error.message.includes('EACCES') ||
|
|
3513
|
+
error.message.includes('EPERM') ||
|
|
3514
|
+
error.message.includes('ENOENT'))) ;
|
|
3515
|
+
else {
|
|
3516
|
+
// Re-throw other unexpected errors
|
|
3517
|
+
throw error;
|
|
3518
|
+
}
|
|
3519
|
+
}
|
|
3488
3520
|
const fileContent = Buffer.from(await response.arrayBuffer());
|
|
3489
3521
|
if (fileContent.length > DEFAULT_MAX_FILE_SIZE /* <- TODO: Allow to pass different value to remote server */) {
|
|
3490
3522
|
throw new LimitReachedError(`File is too large (${Math.round(fileContent.length / 1024 / 1024)}MB). Maximum allowed size is ${Math.round(DEFAULT_MAX_FILE_SIZE / 1024 / 1024)}MB.`);
|
|
@@ -3499,7 +3531,8 @@ async function makeKnowledgeSourceHandler(knowledgeSource, tools, options) {
|
|
|
3499
3531
|
if (error instanceof Error && (error.message.includes('EROFS') ||
|
|
3500
3532
|
error.message.includes('read-only') ||
|
|
3501
3533
|
error.message.includes('EACCES') ||
|
|
3502
|
-
error.message.includes('EPERM')
|
|
3534
|
+
error.message.includes('EPERM') ||
|
|
3535
|
+
error.message.includes('ENOENT'))) {
|
|
3503
3536
|
// Return a handler that works directly with the downloaded content
|
|
3504
3537
|
return {
|
|
3505
3538
|
source: name,
|
|
@@ -6314,7 +6347,8 @@ class MarkitdownScraper {
|
|
|
6314
6347
|
if (error instanceof Error && (error.message.includes('EROFS') ||
|
|
6315
6348
|
error.message.includes('read-only') ||
|
|
6316
6349
|
error.message.includes('EACCES') ||
|
|
6317
|
-
error.message.includes('EPERM')
|
|
6350
|
+
error.message.includes('EPERM') ||
|
|
6351
|
+
error.message.includes('ENOENT'))) ;
|
|
6318
6352
|
else {
|
|
6319
6353
|
// Re-throw other unexpected errors
|
|
6320
6354
|
throw error;
|