Part 2 - applying File URL or Path improvements #17091
Conversation
|
GTFalcao
moved this from Changes Required
to Ready for QA
in Component (Source and Action) Backlog
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 30
🔭 Outside diff range comments (6)
components/microsoft_onedrive/actions/upload-file/upload-file.mjs (1)
43-53: Ensure filename fallback & propagate content-typeIf neither the optional
filenameprop normetadata.nameis set (e.g. when the URL lacks a terminating filename), OneDrive will receive an empty name and will reject the request.
Additionally, you’re not forwardingmetadata.contentType, so uploads may be stored with a generic MIME type.-const name = filename || metadata.name; +const name = filename || metadata.name; +if (!name) { + throw new ConfigurationError("Unable to determine a filename – provide the *Name* prop or use a URL/path that contains one."); +}and pass the MIME type:
- const response = await this.onedrive.uploadFile({ + const response = await this.onedrive.uploadFile({ uploadFolderId, name, data: stream, + contentType: metadata.contentType, $, });This keeps failures explicit and preserves the original file’s MIME information.
components/docsbot_ai/actions/upload-source-file/upload-source-file.mjs (1)
35-57:$summarymay print “undefined” when metadata lacksnameYou already fallback to
filePath.split("/").pop()when building the upload URL, but the success summary always usesmetadata.name, which can beundefined(e.g. when uploading from a public URL that lacks a filename header).- $.export("$summary", `Successfully uploaded "${metadata.name}"`); + const fileName = metadata.name || filePath.split("/").pop(); + $.export("$summary", `Successfully uploaded "${fileName}"`);This keeps the summary meaningful in all scenarios.
components/onlyoffice_docspace/onlyoffice_docspace.app.mjs (1)
71-75:getHeadersis executed twice – drop the inner call to prevent redundant work
_makeRequest()already expandsheaderswiththis.getHeaders(headers)before passing it togetConfig().
InsidegetConfig(),this.getHeaders(headers)is invoked again, duplicating default header enrichment and risking key overwrites.Remove one layer:
-const config = await this.getConfig({ - url: this.getUrl(path), - headers: this.getHeaders(headers), - ...args, -}); +const config = await this.getConfig({ + url: this.getUrl(path), + headers, // already normalized + ...args, +});Less work, same result.
components/pandadoc/actions/create-document-attachment/create-document-attachment.mjs (1)
55-62: Leverage FormData’s own headers instead of private_boundaryUsing the non-public property
data._boundarymay break with futureform-dataversions.
formData.getHeaders()already populates the correctContent-Typewith the boundary.- headers: { - "Content-Type": `multipart/form-data; boundary=${data._boundary}`, - }, + headers: data.getHeaders(),components/boldsign/actions/send-document-template/send-document-template.mjs (1)
260-273:JSON.parse(message)inside catch block can throw and mask root errorsIf BoldSign returns plain-text, the secondary exception hides the original. Wrap the parse in a try/catch.
- const parsedMessage = JSON.parse(message); + let parsedMessage = {}; + try { + parsedMessage = JSON.parse(message); + } catch { /* non-JSON message */ }components/onlyoffice_docspace/actions/upload-file/upload-file.mjs (1)
12-19:folderIdprop incorrectly re-uses the “file” prop definition
propDefinition: [ app, "file" ]will inject the file property schema where a folder ID is expected, breaking the UI and runtime validation. This looks like a copy/paste slip.- folderId: { - label: "Folder ID", - description: "The ID of the folder where you want the file to be uploaded.", - propDefinition: [ - app, - "file", - ], - }, + folderId: { + propDefinition: [ + app, + "folderId", // assuming the app already exposes this + ], + },
🧹 Nitpick comments (39)
components/mistral_ai/package.json (1)
3-3: Confirm semver compliance for version bump.
You've bumped@pipedream/mistral_aifrom0.1.0to0.2.0, which indicates minor, backward-compatible changes. Ensure the CHANGELOG and release notes are updated to document the new file URL/path support and any behavioral changes.components/google_drive/package.json (1)
3-3: Major version bump—review for breaking changes.
Releasing@pipedream/google_driveas1.0.0signals potential breaking changes. Ensure downstream workflows, documentation, and CHANGELOG entries reflect any API alterations introduced by the refactor.components/cradl_ai/package.json (1)
3-3: Update minor version for new features.
Bumping@pipedream/cradl_aifrom0.1.0to0.2.0covers the new file handling improvements. Don’t forget to update your CHANGELOG or release notes with details of the async file stream support.components/alttext_ai/package.json (1)
3-3: Document version bump in CHANGELOG.
You’ve increased@pipedream/alttext_aifrom0.1.0to0.2.0. Make sure to log the new file URL/path handling feature in the component’s release notes.components/ocrspace/actions/process-pdf/process-pdf.mjs (1)
8-8: Increment action version.
The action version was bumped from0.0.1to0.1.0. Ensure that this change is mirrored in your CHANGELOG and that any consumers reference the correct version.components/imagga/package.json (1)
3-3: Confirm semantic versioning for the major release
Bumpingversionto1.0.0signals a major release with potential breaking changes. Ensure CHANGELOG, migration guides, and documentation clearly outline any API or behavior changes for consumers.components/box/package.json (1)
3-3: Confirm semantic versioning for the minor bump
Updatingversionfrom0.3.0to0.4.0indicates non-breaking enhancements. Ensure CHANGELOG reflects updated file-stream handling workflows.components/openai/package.json (1)
3-3: Confirm major version bump to 1.0.0
The jump from 0.x to 1.0.0 indicates breaking changes across OpenAI components. Ensure this is documented in CHANGELOG/release notes and that downstream consumers are aware of any migration steps.components/onlinecheckwriter/package.json (1)
3-3: Document version bump to 0.2.0
The minor version increase reflects added file-URL/path support. Please update CHANGELOG and any release notes to highlight the new capability.components/askyourpdf/package.json (1)
3-3: Update package version to 0.2.0
Reflects async refactors and file-input enhancements. Ensure release notes capture the switch to asynchronous utilities and any required user-facing changes.components/autodesk/package.json (1)
3-3: Bump version to 0.2.0
Indicates the upload-file action now supports file URLs and uses the new streaming utility. Add a CHANGELOG entry to note the enhanced file-input flexibility.components/ilovepdf/package.json (1)
3-3: Major version bump to 1.0.0
Switching from 0.x to 1.0.0 signals significant changes (URL/path handling, drop offileUrls). Make sure upgrade instructions are clearly communicated since this is likely a breaking release.components/claid_ai/claid_ai.app.mjs (1)
9-11: Description now mirrors new behaviour but drops the original docs linkThe clearer label/description is great, but removing the link to the ClaiD API docs slightly reduces discoverability for users. Consider re-adding the link at the end of the description for quick reference.
components/imagga/imagga.app.mjs (1)
34-36: Example extension doesn’t match the “image” contextThe description example points to
/tmp/myFile.txt, which suggests a text file.
Using an image-specific extension (e.g.,.jpgor.png) avoids confusing users about the expected MIME type.- (for example, `/tmp/myFile.txt`) + (for example, `/tmp/myImage.jpg`)components/flippingbook/flippingbook.app.mjs (1)
37-41: Property name no longer reflects behaviour
filePathnow accepts URLs as well, but the identifier still contains “Path”.
Consider renaming tofile(mirroring other components) to keep the code self-documenting, or at minimum add an inline comment acknowledging the broadened scope.components/claid_ai/actions/remove-background/remove-background.mjs (1)
99-108: RedundanturlExistsround-trip
getFileStreamAndMetadataalready handles both local paths and HTTP(S) URLs.
You could remove the extra HEAD request and simply attemptgetFileStreamAndMetadata; fall back to theeditImagebranch only when the call throws for non-local URLs.This avoids an additional network hop and simplifies control flow.
components/canva/actions/upload-asset/upload-asset.mjs (1)
32-46: Prefer actual MIME & omit unknown length
metadata.sizemay be absent for remote URLs, and hard-settingContent-Lengthwithundefinedproduces an invalid header. Also, using the real MIME type helps Canva classify the asset.-const { stream, metadata } = await getFileStreamAndMetadata(this.filePath); +const { stream, metadata } = await getFileStreamAndMetadata(this.filePath); ... -"Content-Length": metadata.size, -"Content-Type": "application/octet-stream", +...(Number.isFinite(metadata.size) && { "Content-Length": metadata.size }), +"Content-Type": metadata.contentType || "application/octet-stream", ... -data: stream, +data: stream,This keeps the request compliant and improves downstream processing accuracy.
components/ocrspace/actions/common/process-base.mjs (1)
48-51: Variable naming now misleading after util refactor
getUrlOrFile()now returns a base-64 string, not a{ url, file }object.
Storing it in a variable calledfilecan confuse future readers and static-analysis tools.- const file = await getUrlOrFile(this.file); + const base64Image = await getUrlOrFile(this.file); + const file = base64Image; // preserve old semantic if callers rely on `file`Renaming clarifies intent with zero runtime impact.
components/box/actions/upload-file-version/upload-file-version.mjs (1)
56-65: MissingparentIdprop – may always beundefined
parentIdis destructured fromthisand forwarded togetFileUploadBody, yet noparentIdprop exists.
IfgetFileUploadBodydoesn’t require it, drop the argument to avoid confusion; otherwise expose the prop.- file, fileId, createdAt, modifiedAt, fileName, parentId, + file, fileId, createdAt, modifiedAt, fileName,(or add a
parentIdprop mirroringfolderId).Tiny cleanup but prevents dead-code drift.
components/_2markdown/actions/pdf-to-markdown/pdf-to-markdown.mjs (1)
30-37: Wrap stream acquisition in atry / catchto surface user-friendly errorsA missing or inaccessible file/URL will currently bubble an unhandled rejection and surface as “Internal error”. Catch it early and throw a meaningful
ConfigurationError.-const { - stream, metadata, -} = await getFileStreamAndMetadata(this.filePath); +let stream, metadata; +try { + ({ stream, metadata } = await getFileStreamAndMetadata(this.filePath)); +} catch (err) { + throw new Error(`Unable to fetch file "${ this.filePath }": ${ err.message }`); +}components/_2markdown/actions/html-file-to-markdown/html-file-to-markdown.mjs (1)
23-30: Add defensive error handling aroundgetFileStreamAndMetadataSame rationale as in the PDF action – shield users from opaque stack traces.
-const { - stream, metadata, -} = await getFileStreamAndMetadata(this.filePath); +let stream, metadata; +try { + ({ stream, metadata } = await getFileStreamAndMetadata(this.filePath)); +} catch (err) { + throw new Error(`Unable to fetch file "${ this.filePath }": ${ err.message }`); +}components/onlyoffice_docspace/onlyoffice_docspace.app.mjs (1)
49-55: Avoid the terse boolean-chain for clarity & potential precedence pitfallsThe
dataassignment combines&&and||in a single expression. While functionally correct, the mixed short-circuit logic is harder to scan and easy to mis-read, especially onceawaitis added.-const data = hasMultipartHeader && await utils.getFormData(preData) || preData; +const data = hasMultipartHeader + ? await utils.getFormData(preData) + : preData;The explicit ternary improves readability and eliminates precedence ambiguity.
components/akeneo/actions/create-a-new-product-media-file/create-a-new-product-media-file.mjs (1)
34-37: Prop name now mismatches behaviourThe prop is still called
filenamebut it expects a path or URL. To avoid confusion and to stay in line with other components, rename the prop (and references) tofilePathorfile.components/mistral_ai/actions/upload-file/upload-file.mjs (1)
15-17: Clarify example extension in descriptionThe property description now says “for example,
/tmp/myFile.txt”.
Mistral only accepts specific formats (JSON, CSV, etc.). Consider replacing the.txtexample with an actually supported extension to prevent user confusion.components/autodesk/actions/upload-file/upload-file.mjs (1)
3-4: Import grouping nit
axiosandgetFileStreamAndMetadatacome from different logical domains (HTTP vs. file util). Consider separate import lines for readability.components/openai/actions/create-batch/create-batch.mjs (1)
33-36: Example file extension is misleadingThe description references
/tmp/myFile.jpg, but the OpenAI batch endpoint requires a.jsonlfile. Recommend updating the example to avoid accidental misuse.components/ocrspace/common/utils.mjs (1)
16-21:checkTmp()appears to be dead code
checkTmp()is no longer referenced after the refactor. Keeping unused helpers adds noise and can confuse future maintainers.-export const checkTmp = (filename) => { - if (filename.indexOf("/tmp") === -1) { - return `/tmp/${filename}`; - } - return filename; -};If no other module imports this function, please remove it (and any tests) to keep the surface area minimal.
components/airparser/actions/upload-document-parse/upload-document-parse.mjs (1)
20-23: Prop name & description drift
filePathstill implies “path”, but the label/description now mentions URLs as well. Consider renaming the prop tofileorfilePathOrUrlto avoid misleading users.components/onlyoffice_docspace/common/utils.mjs (1)
5-9: Serialawaitin deep recursion could be slow
for…of / awaitprocesses nested objects sequentially. When many files are nested, building the form may be noticeably slower. Consider collecting promises andawait Promise.all(...)where order is irrelevant.components/askyourpdf/askyourpdf.app.mjs (1)
35-36: Use a ternary for clarityThe short-circuit expression mixes
await,&&, and||, which is easy to mis-read.-const data = hasMultipartHeader && await utils.getFormData(preData) || preData; +const data = hasMultipartHeader + ? await utils.getFormData(preData) + : preData;components/boldsign/actions/send-document-template/send-document-template.mjs (2)
191-204: Inefficient in-memory buffering & missing content-type fallbackReading each file into memory can exhaust RAM for large PDFs. Consider:
- Streaming directly to a base64 encoder (or chunked upload if BoldSign API supports multipart).
- Defaulting
contentTypeto"application/octet-stream"whenmetadata.contentTypeis falsy.At minimum, document the memory limitation.
208-237: DuplicatereminderSettingscauses needless overwrites
reminderSettingsis built twice – once inadditionalDataand once inline – then the spread operator overwrites the first instance. Drop the inline copy to avoid confusion.- reminderSettings: { - enableAutoReminder: this.enableAutoReminder, - reminderDays: this.reminderDays, - reminderCount: this.reminderCount, - }, ... - ...additionalData, + ...additionalData,components/ilovepdf/actions/process-files/process-files.mjs (1)
64-71: SameknownLengthconcern as other componentsSee earlier PandaDoc comment – omit
knownLengthwhen it isundefined.components/nyckel/common/common-image.mjs (1)
15-24: Behaviour change: URLs are now downloaded instead of passed as JSONDownloading remote images server-side may increase latency and bandwidth costs but yields a consistent interface – good trade-off.
Consider retaining the original “pass-through URL” optimisation for large public images via a prop toggle.components/openai/actions/chat/chat.mjs (1)
49-54: Removal of model/audio compatibility check may surface late-stage API errorsPreviously the action failed fast when
audiowas supplied with a non-audio model. Now the request will reach OpenAI and return a 400, which is a poorer DX.Re-introduce a lightweight guard:
+if (this.audio && !this.modelId.includes("gpt-4o-audio-preview")) { + throw new Error("Audio input is only supported with the gpt-4o-audio-preview model"); +}This keeps users from incurring an unnecessary API call.
components/flippingbook/actions/update-flipbook/update-flipbook.mjs (1)
50-57: Minor duplication: consider extractinggetFileDatainto a shared utilBoth create- and update-flipbook actions expose identical
getFileDatalogic throughcreateFlipbook.methods. A small shared helper undercomponents/flippingbook/common/file-utils.mjswould keep cross-action behaviour in sync and simplify future fixes.components/imagga/common/utils.mjs (1)
42-49: Optional: dropknownLengthwhen unavailable
metadata.sizemay beundefinedfor remote URLs, which causes some servers to reject the upload. Consider omittingknownLengthwhen the value is falsy.- formData.append("image", stream, { - contentType: metadata.contentType, - knownLength: metadata.size, - filename: metadata.name, - }); + formData.append("image", stream, { + contentType: metadata.contentType, + ...(metadata.size && { knownLength: metadata.size }), + filename: metadata.name, + });components/openai/actions/analyze-image-content/analyze-image-content.mjs (1)
62-69: Dead code —imageUrlprop was removed but the branch remains
this.imageUrlis alwaysundefined, so this block will never run and only confuses future readers.- if (this.imageUrl) { - data.thread.messages[0].content.push({ - type: "image_url", - image_url: { - url: this.imageUrl, - }, - }); - }components/alttext_ai/actions/generate-alt-text/generate-alt-text.mjs (1)
69-82: Reading entire file into memory can blow up large workflowsFor very large images the
chunks -> Buffer.concatpattern will hold the whole file in memory.
Consider streaming to base64 on-the-fly or enforcing a reasonable size limit to avoid OOM crashes.If the API actually accepts multipart files, you could avoid base64 altogether and pipe the stream.
📜 Review details
Configuration used: CodeRabbit UI
Review profile: CHILL
Plan: Pro
⛔ Files ignored due to path filters (1)
pnpm-lock.yamlis excluded by!**/pnpm-lock.yaml
📒 Files selected for processing (107)
components/_2markdown/_2markdown.app.mjs(1 hunks)components/_2markdown/actions/html-file-to-markdown/html-file-to-markdown.mjs(2 hunks)components/_2markdown/actions/pdf-to-markdown/pdf-to-markdown.mjs(3 hunks)components/_2markdown/package.json(2 hunks)components/airparser/actions/upload-document-parse/upload-document-parse.mjs(3 hunks)components/airparser/package.json(2 hunks)components/akeneo/actions/create-a-new-product-media-file/create-a-new-product-media-file.mjs(3 hunks)components/akeneo/package.json(2 hunks)components/alttext_ai/actions/generate-alt-text/generate-alt-text.mjs(2 hunks)components/alttext_ai/package.json(2 hunks)components/askyourpdf/actions/add-document-via-file-upload/add-document-via-file-upload.mjs(1 hunks)components/askyourpdf/askyourpdf.app.mjs(2 hunks)components/askyourpdf/common/utils.mjs(1 hunks)components/askyourpdf/package.json(2 hunks)components/autodesk/actions/upload-file/upload-file.mjs(3 hunks)components/autodesk/package.json(2 hunks)components/azure_storage/actions/upload-blob/upload-blob.mjs(4 hunks)components/azure_storage/common/utils.mjs(0 hunks)components/azure_storage/package.json(2 hunks)components/boldsign/actions/send-document-template/send-document-template.mjs(3 hunks)components/boldsign/package.json(2 hunks)components/box/actions/upload-file-version/upload-file-version.mjs(2 hunks)components/box/actions/upload-file/upload-file.mjs(2 hunks)components/box/box.app.mjs(1 hunks)components/box/common/common-file-upload.mjs(2 hunks)components/box/common/utils.mjs(1 hunks)components/box/package.json(2 hunks)components/canva/actions/create-design-import-job/create-design-import-job.mjs(2 hunks)components/canva/actions/upload-asset/upload-asset.mjs(2 hunks)components/canva/canva.app.mjs(1 hunks)components/canva/package.json(2 hunks)components/claid_ai/actions/correct-color-lighting/correct-color-lighting.mjs(2 hunks)components/claid_ai/actions/remove-background/remove-background.mjs(2 hunks)components/claid_ai/actions/upscale-image/upscale-image.mjs(2 hunks)components/claid_ai/claid_ai.app.mjs(1 hunks)components/claid_ai/package.json(2 hunks)components/cradl_ai/actions/common/common.mjs(2 hunks)components/cradl_ai/actions/parse-document-human-in-loop/parse-document-human-in-loop.mjs(1 hunks)components/cradl_ai/actions/parse-document/parse-document.mjs(1 hunks)components/cradl_ai/package.json(2 hunks)components/docsbot_ai/actions/upload-source-file/upload-source-file.mjs(3 hunks)components/docsbot_ai/docsbot_ai.app.mjs(1 hunks)components/docsbot_ai/package.json(2 hunks)components/egnyte/actions/upload-file/upload-file.mjs(2 hunks)components/egnyte/package.json(2 hunks)components/fileforge/actions/generate-pdf/generate-pdf.mjs(3 hunks)components/fileforge/common/utils.mjs(0 hunks)components/fileforge/package.json(2 hunks)components/flippingbook/actions/create-flipbook/create-flipbook.mjs(2 hunks)components/flippingbook/actions/update-flipbook/update-flipbook.mjs(2 hunks)components/flippingbook/flippingbook.app.mjs(1 hunks)components/flippingbook/package.json(2 hunks)components/google_drive/actions/update-file/update-file.mjs(3 hunks)components/google_drive/actions/upload-file/upload-file.mjs(3 hunks)components/google_drive/common/filePathOrUrl.mjs(0 hunks)components/google_drive/common/utils.mjs(1 hunks)components/google_drive/google_drive.app.mjs(1 hunks)components/google_drive/package.json(1 hunks)components/happy_scribe/actions/submit-file/submit-file.mjs(2 hunks)components/happy_scribe/package.json(2 hunks)components/ilovepdf/actions/process-files/process-files.mjs(3 hunks)components/ilovepdf/ilovepdf.app.mjs(1 hunks)components/ilovepdf/package.json(2 hunks)components/imagga/actions/analyze-image/analyze-image.mjs(1 hunks)components/imagga/actions/process-batch/process-batch.mjs(1 hunks)components/imagga/common/utils.mjs(1 hunks)components/imagga/imagga.app.mjs(1 hunks)components/imagga/package.json(2 hunks)components/leaddyno/leaddyno.app.mjs(1 hunks)components/llmwhisperer/actions/extract-text/extract-text.mjs(4 hunks)components/llmwhisperer/package.json(2 hunks)components/lokalise/actions/upload-file/upload-file.mjs(3 hunks)components/lokalise/package.json(2 hunks)components/mapbox/actions/create-tileset/create-tileset.mjs(4 hunks)components/mapbox/package.json(2 hunks)components/microsoft_onedrive/actions/upload-file/upload-file.mjs(3 hunks)components/microsoft_onedrive/package.json(2 hunks)components/mistral_ai/actions/upload-file/upload-file.mjs(2 hunks)components/mistral_ai/package.json(2 hunks)components/monday/actions/update-column-values/update-column-values.mjs(4 hunks)components/monday/package.json(2 hunks)components/nyckel/actions/classify-image/classify-image.mjs(2 hunks)components/nyckel/actions/extract-text-from-image/extract-text-from-image.mjs(2 hunks)components/nyckel/common/common-image.mjs(2 hunks)components/nyckel/nyckel.app.mjs(1 hunks)components/nyckel/package.json(2 hunks)components/ocrspace/actions/common/process-base.mjs(1 hunks)components/ocrspace/actions/process-image/process-image.mjs(1 hunks)components/ocrspace/actions/process-pdf/process-pdf.mjs(1 hunks)components/ocrspace/common/utils.mjs(2 hunks)components/ocrspace/ocrspace.app.mjs(1 hunks)components/ocrspace/package.json(2 hunks)components/onlinecheckwriter/actions/mail-pdf-document/mail-pdf-document.mjs(3 hunks)components/onlinecheckwriter/package.json(2 hunks)components/onlyoffice_docspace/actions/upload-file/upload-file.mjs(2 hunks)components/onlyoffice_docspace/common/utils.mjs(1 hunks)components/onlyoffice_docspace/onlyoffice_docspace.app.mjs(2 hunks)components/onlyoffice_docspace/package.json(2 hunks)components/openai/actions/analyze-image-content/analyze-image-content.mjs(3 hunks)components/openai/actions/chat/chat.mjs(3 hunks)components/openai/actions/common/common.mjs(5 hunks)components/openai/actions/create-batch/create-batch.mjs(3 hunks)components/openai/actions/create-transcription/create-transcription.mjs(2 hunks)components/openai/actions/upload-file/upload-file.mjs(2 hunks)components/openai/openai.app.mjs(1 hunks)components/openai/package.json(2 hunks)components/pandadoc/actions/create-document-attachment/create-document-attachment.mjs(3 hunks)
⛔ Files not processed due to max files limit (27)
- components/pandadoc/actions/create-document-from-file/create-document-from-file.mjs
- components/pandadoc/package.json
- components/pandadoc/pandadoc.app.mjs
- components/pdf4me/actions/compress-pdf/compress-pdf.mjs
- components/pdf4me/actions/convert-to-pdf/convert-to-pdf.mjs
- components/pdf4me/actions/merge-pdfs/merge-pdfs.mjs
- components/pdf4me/common/utils.mjs
- components/pdf4me/package.json
- components/pdf4me/pdf4me.app.mjs
- components/podio/actions/upload-file/upload-file.mjs
- components/podio/package.json
- components/slack/actions/upload-file/upload-file.mjs
- components/slack/package.json
- components/slack/slack.app.mjs
- components/stannp/actions/create-campaign/create-campaign.mjs
- components/stannp/common/utils.mjs
- components/stannp/package.json
- components/zamzar/actions/start-job-from-file/start-job-from-file.mjs
- components/zamzar/common/utils.mjs
- components/zamzar/package.json
- components/zamzar/zamzar.app.mjs
- components/zoho_crm/actions/upload-attachment/upload-attachment.mjs
- components/zoho_crm/package.json
- components/zoho_desk/actions/add-ticket-attachment/add-ticket-attachment.mjs
- components/zoho_desk/common/utils.mjs
- components/zoho_desk/package.json
- components/zoho_desk/zoho_desk.app.mjs
💤 Files with no reviewable changes (3)
- components/fileforge/common/utils.mjs
- components/azure_storage/common/utils.mjs
- components/google_drive/common/filePathOrUrl.mjs
⏰ Context from checks skipped due to timeout of 90000ms (4)
- GitHub Check: pnpm publish
- GitHub Check: Publish TypeScript components
- GitHub Check: Lint Code Base
- GitHub Check: Verify TypeScript components
🔇 Additional comments (90)
components/leaddyno/leaddyno.app.mjs (1)
11-11: Trivial change
Added a trailing newline at end of file; no functional impact.components/mistral_ai/package.json (1)
16-16: Verify platform dependency update compatibility.
Upgrading@pipedream/platformto^3.1.0is required forgetFileStreamAndMetadata. Please confirm there are no unintended breaking changes in v3.1.0 that impact this component’s functionality.components/cradl_ai/package.json (1)
16-16: Confirm platform SDK upgrade.
Moving to@pipedream/platform@^3.1.0is necessary for the new file utilities. Verify that all existing actions in this component remain compatible with the updated SDK.components/alttext_ai/package.json (1)
16-16: Validate platform dependency update.
Upgrading to@pipedream/platform@^3.1.0unlocks the new streaming API. Confirm there are no regressions with other actions in this component.components/claid_ai/package.json (2)
3-3: Bump package version to 0.2.0
Aligns with the coordinated versioning across components in this PR.
16-16: ```shell
#!/bin/bashRe-verify that no direct fs imports or requires remain in the claid_ai component
rg -n "require(['"]fs['"])" components/claid_ai
rg -n "import fs" components/claid_ai</details> <details> <summary>components/azure_storage/package.json (2)</summary> `3-3`: **Bump package version to 0.2.0** Consistent with the Azure Storage component’s update cycle. --- `16-16`: Let’s re-run the search without the invalid `-R` flag to verify there are truly no lingering imports: ```shell #!/bin/bash # Correctly check for any references to common/utils.mjs in the azure_storage component rg -n "common/utils\.mjs" components/azure_storageThis will confirm whether any stray imports remain.
components/microsoft_onedrive/package.json (2)
3-3: Bump package version to 1.7.0
Keeps parity with other components handling file URLs/paths improvements.
13-13: ```shell
#!/bin/bashCheck for any remaining file-type imports in the OneDrive component
rg -n "require(['"]file-type['"]]|import .*\bfile-type\b" components/microsoft_onedrive
Verify file-type is not listed as a dependency in package.json
grep -R '"file-type"' components/microsoft_onedrive/package.json || true
</details> <details> <summary>components/mapbox/package.json (2)</summary> `3-3`: **Bump package version to 0.2.0** Aligns with the standard release cadence for Mapbox components. --- `16-16`: **Upgrade @pipedream/platform to ^3.1.0** Enables use of the latest file‐stream utilities; no additional changes needed here. </details> <details> <summary>components/cradl_ai/actions/parse-document-human-in-loop/parse-document-human-in-loop.mjs (1)</summary> `8-8`: **Bump action version to 0.1.0** Matches the new minor release for the human-in-loop parsing action; functionality remains unchanged. </details> <details> <summary>components/imagga/package.json (1)</summary> `16-16`: **Validate compatibility with @pipedream/platform v3.1.0** Upgrading the dependency to `^3.1.0` should support all new file utilities (e.g., `getFileStreamAndMetadata`). Run integration tests to confirm no missing exports or breaking changes in the platform SDK. </details> <details> <summary>components/llmwhisperer/package.json (2)</summary> `3-3`: **Approve minor version bump to 0.2.0** Incrementing from `0.1.0` to `0.2.0` is a backward-compatible change. Versioning aligns with non-breaking enhancements in file handling. --- `16-16`: **Verify removal of direct fs dependency and upgrade to @pipedream/platform v3.1.0** Ensure no residual `fs` imports remain and that all file operations now use platform utilities. ```shell #!/bin/bash # Check for lingering fs imports in llmwhisperer rg -n "require\(['\"]fs['\"]\)|import .*fs" -l components/llmwhisperercomponents/box/package.json (1)
13-13: Validate @pipedream/platform v3.1.0 upgrade
The dependency has been updated to^3.1.0. Verify that Box actions usinggetFileUploadBodyand related async utilities work correctly with this SDK version.components/boldsign/package.json (2)
3-3: Approve version bump to 0.2.0
Moving from0.1.0to0.2.0is aligned with the non-breaking refactor to platform file utilities.
16-16: Validate compatibility with @pipedream/platform v3.1.0
Confirm thatgetFileStreamAndMetadatais fully supported in v3.1.0 and that no legacy fs calls remain in the BoldSign actions.components/fileforge/package.json (2)
3-3: Approve version bump to 0.2.0
Increment from0.1.0to0.2.0indicates backward-compatible enhancements for fileforge’s PDF generation workflows.
16-16: Validate @pipedream/platform v3.1.0 upgrade
Ensure thatgetFileStreamAndMetadataproperly replaces priorfs-based logic in bothgenerate-pdfandutils.mjs. Run end-to-end tests in a sandbox.components/ocrspace/actions/process-image/process-image.mjs (1)
8-8: Bump action version
Updated the action version from0.0.1to0.1.0to reflect the standardized file input handling changes.components/canva/package.json (2)
3-3: Bump package version
Incremented version to0.3.0following internal file handling refactor.
16-16: Upgrade platform dependency
Updated@pipedream/platformto^3.1.0to leverage the new file streaming utilities.components/flippingbook/package.json (2)
3-3: Major version bump
Raised version to1.0.0—ensure the breaking change (removal offileUrlproperty and consolidation tofilePath) is documented in the CHANGELOG and upgrade guides.
16-16: Upgrade platform dependency
Bumped@pipedream/platformto^3.1.0to align with standardized async file handling utilities.components/onlyoffice_docspace/package.json (2)
3-3: Bump package version
Version updated to0.2.0to reflect async form data and file handling enhancements.
16-16: Upgrade platform dependency
Updated@pipedream/platformto^3.1.0forgetFileStreamAndMetadatasupport.components/lokalise/package.json (2)
3-3: Bump package version
Incremented to0.2.0following migration to async streaming withgetFileStream.
16-16: Upgrade platform dependency
Updated@pipedream/platformto^3.1.0to standardize file URL and path handling.components/akeneo/package.json (2)
3-3: Bump component version to 0.1.0
Aligns with the new action version for file URL/path improvements.
13-13: Upgrade @pipedream/platform dependency to ^3.1.0
Matches the PR-wide migration to v3 utilities for async file handling.components/monday/package.json (2)
3-3: Bump component version to 0.8.0
Reflects the new release containing file URL/path enhancements.
17-17: Upgrade @pipedream/platform dependency to ^3.1.0
Ensures access to the updated getFileStreamAndMetadata utilities.components/ocrspace/package.json (2)
3-3: Bump component version to 0.2.0
Prepares OCRSpace component for the file handling improvements release.
16-16: Upgrade @pipedream/platform dependency to ^3.1.0
Brings in the latest async file utilities across components.components/airparser/package.json (2)
3-3: Bump component version to 0.2.0
Marks Airparser release with platform v3 compatibility.
16-16: Upgrade @pipedream/platform dependency to ^3.1.0
Aligns Airparser with the new file-stream utilities.components/happy_scribe/package.json (2)
3-3: Bump component version to 0.2.0
Signals Happy Scribe component update for async file input support.
16-16: Upgrade @pipedream/platform dependency to ^3.1.0
Ensures availability of getFileStream for base64, URL, and path inputs.components/docsbot_ai/package.json (2)
3-3: Semantic version bump
Version updated from0.1.0to0.2.0to reflect the file URL/path enhancements.
16-16: Upgrade@pipedream/platformto ^3.1.0
This aligns with the platform utility refactor (getFileStreamAndMetadata). Ensure no breaking changes in the new version.components/cradl_ai/actions/parse-document/parse-document.mjs (1)
8-8: Bump action version
Updatedversionto"0.1.0"consistent with other component revisions. Confirm corresponding release notes or changelog updates.components/nyckel/package.json (2)
3-3: Semantic version bump
Version bumped from0.1.0to0.2.0to capture the file handling improvements.
16-16: Upgrade platform dependency
Updated@pipedream/platformto^3.1.0, matching the standardized file input utilities.components/_2markdown/package.json (2)
3-3: Version increment
Updated component version to0.2.0following the bulk refactor.
16-16: Platform dependency update
Upgraded@pipedream/platformto^3.1.0for consistent use of file streaming utilities.components/egnyte/package.json (2)
3-3: Package version bump
Incremented version to0.2.0in alignment with other packages.
16-16: Upgrade@pipedream/platform
Upgraded to^3.1.0to leverage the new file input utilities.components/openai/package.json (1)
17-17: Verify platform dependency upgrade to ^3.1.0
Upgrading@pipedream/platformto ^3.1.0 aligns with the newgetFileStreamAndMetadataAPI, but please confirm there are no unresolved peer dependency conflicts and all imports remain compatible.components/onlinecheckwriter/package.json (1)
16-16: Upgrade@pipedream/platformto ^3.1.0
This change is required for thegetFileStreamAndMetadatautility. Verify that all components build correctly against platform v3.1.0 and that no legacy code still references removed APIs.components/askyourpdf/package.json (1)
13-14: Dependencies bumped: platform → ^3.1.0, form-data → ^4.0.0
Aligns with the use ofgetFileStreamAndMetadataand multipart form-data improvements. Confirm that the asyncgetFormDataworkflow inaskyourpdf.app.mjsfunctions as expected with these versions.components/autodesk/package.json (1)
16-16: Upgrade@pipedream/platformdependency to ^3.1.0
Needed for the newgetFileStreamAndMetadataAPI in the upload-file action. Please validate that the action imports and error handling remain correct under v3.1.0.components/ilovepdf/package.json (1)
16-17: Dependencies updated: platform → ^3.1.0, form-data → ^4.0.0
Ensures compatibility withgetFileStreamAndMetadataand improved multipart handling. Run a full integration test to confirmprocess-filesaction behavior is unaffected.components/nyckel/actions/extract-text-from-image/extract-text-from-image.mjs (2)
9-9: Version bump looks correctMinor‐version increment from
0.0.1➜0.1.0matches the behavioural change (async file handling).
31-31: Good catch turninggetImageData()into an awaited callAwaiting the promise prevents a possible race condition where an unresolved promise would be spread into the request object. 👍
components/box/box.app.mjs (1)
16-18: Property description updated—check consistency with downstream validationDown-stream helpers (e.g.
common/utils.mjs -> getFileData) now accept both URLs and/tmppaths, so this UI text is aligned. Make sure any older validations that assumed a local path (e.g. regex checks) have been removed; otherwise users could hit confusing “invalid path” errors.components/_2markdown/_2markdown.app.mjs (1)
9-10: Label/description update LGTMWording is now consistent with the cross-component file-URL standard.
components/claid_ai/actions/remove-background/remove-background.mjs (1)
4-4: 👍 Switched togetFileStreamAndMetadataGreat move—eliminates manual
fshandling and guarantees metadata is present.components/openai/actions/upload-file/upload-file.mjs (1)
31-39: Good adoption of unified file helperTransition to
getFileStreamAndMetadataimproves reliability and preserves filename metadata—LGTM.components/nyckel/actions/classify-image/classify-image.mjs (1)
49-49: AwaitinggetImageDatafixes race conditionWaiting for the promise before spreading the object prevents undefined payloads—nice catch.
components/askyourpdf/actions/add-document-via-file-upload/add-document-via-file-upload.mjs (1)
9-16: Prop docs updated – looks goodLabel/description now matches the new cross-component convention (URL or
/tmppath). No further issues spotted.components/box/actions/upload-file/upload-file.mjs (1)
50-56: Nice catch makinggetFileUploadBodyawaitableThe added
awaitaligns with the now-async helper and prevents the request from being sent with an unresolvedPromise.
No additional concerns.components/google_drive/google_drive.app.mjs (1)
158-162: Change looks good – property description matches the new cross-component conventionOnly the label/description were updated; type and behaviour stay unchanged. No further action needed.
components/nyckel/nyckel.app.mjs (1)
30-34: Consistent wording – looks goodLabel/description now match the platform-wide “File Path or URL” wording.
components/google_drive/common/utils.mjs (1)
261-273: Let’s inspect the imports and usage ofgetFileStreamin the Google Drive update action:#!/bin/bash # Show import statements in update-file.mjs grep -n "import" components/google_drive/actions/update-file/update-file.mjs # Show lines around getFileStream usage grep -n -C3 "getFileStream" components/google_drive/actions/update-file/update-file.mjscomponents/imagga/actions/analyze-image/analyze-image.mjs (1)
9-18: Change looks good – single source of truth for the image inputSwitching to a single
imageFileprop simplifies the interface and aligns with the
new unified file-handling utility. No issues spotted here.components/_2markdown/actions/pdf-to-markdown/pdf-to-markdown.mjs (1)
2-2: Use ofgetFileStreamAndMetadatalooks goodSwitching from
fsto the platform helper keeps the action compatible with both URLs and/tmppaths and removes the sync I/O.components/cradl_ai/actions/common/common.mjs (2)
3-3: 👍 Replacing synchronousfs.readFileSyncwith streaming helperThe import aligns this module with the new file-handling convention and removes blocking I/O.
49-58: Confirm the API accepts a stream – it previously received a Buffer
getFileStream()returns aReadablestream whereas the prior implementation sent aBuffer.
Most axios back-ends handle both, but if the endpoint expectsContent-Lengthor random access it may break.Please re-run the integration tests for large (>5 MB) files and, if necessary, compute and set the
Content-Lengthheader:-const fileData = await getFileStream(this.filePath); +const { stream, metadata } = await getFileStream(this.filePath); + +// axios will set transfer-encoding: chunked; add length if the API demands it +const headers = { + "Content-Type": this.contentType, + ...(metadata.size && { "Content-Length": metadata.size }), +};components/_2markdown/actions/html-file-to-markdown/html-file-to-markdown.mjs (1)
2-2: Consistent helper import — good moveKeeps the two 2Markdown actions in sync with the new pattern.
components/ilovepdf/ilovepdf.app.mjs (1)
9-11: MakingfilePathsmandatory could be a breaking change
optional: truewas removed. Existing workflows that dynamically build the list may now fail validation.Recommend retaining optionality or bumping the major version of the app to signal a breaking contract.
- type: "string[]", - label: "File Paths or URLs", - description: "...", + type: "string[]", + label: "File Paths or URLs", + description: "...", + optional: true,components/box/common/common-file-upload.mjs (1)
30-33: Good addition ofcontentTypeand explicitfilenameThese headers help Box detect the MIME type and preserve the original name.
components/canva/canva.app.mjs (1)
42-44: Label & description update looks goodThe new wording clearly states that either a URL or
/tmppath is accepted. No further action needed.components/ocrspace/ocrspace.app.mjs (1)
14-16: Documentation tweak acknowledgedConsistent phrasing with the rest of the repo – thanks for the alignment.
components/google_drive/actions/upload-file/upload-file.mjs (1)
38-43: Prop name aligns but keep an eye on cross-file usages
filePathis now sourced from the app definition. Make sure all downstream actions/tests referring to the oldfileprop are migrated; otherwise callers will get a validation error at runtime.components/openai/actions/create-batch/create-batch.mjs (1)
54-61: Optional size headerSimilar to the Autodesk upload,
metadata.sizecan beundefined, yet it is passed asknownLength. While FormData toleratesundefined, some runtimes mis-handle it. You may want to conditionally supplyknownLengthonly when it is a finite number.components/monday/actions/update-column-values/update-column-values.mjs (1)
67-76: VerifyknownLengthreliabilityFor files coming from a URL without a
Content-Lengthheader,metadata.sizemay be undefined, yet it is forwarded to FormData asknownLength. Monday’s API can reject uploads without a valid size. Please confirm this edge case or conditionally omit the option.components/ilovepdf/actions/process-files/process-files.mjs (1)
78-83: Headers are correctly derived – nice!components/happy_scribe/actions/submit-file/submit-file.mjs (1)
24-29: Verify Happy Scribe API accepts a streamPreviously a
Bufferwas sent; switching to a stream may changeContent-Lengthhandling. Please confirm with the API or wrap the stream inFormDataif a multipart upload is required.components/askyourpdf/common/utils.mjs (1)
5-9: 👍 Async traversal fix looks goodSwitching from
forEachwith an async callback to an awaitedfor … ofloop guarantees sequential execution and proper error bubbling.components/openai/actions/chat/chat.mjs (1)
147-148: Awaiting_getChatArgsrequires it to be truly asyncGood catch changing the invocation to
await. Verify thatcommon._getChatArgsnow returns a Promise in every code path; otherwise this will wrap a non-promise value and may mask sync exceptions.components/openai/actions/common/common.mjs (1)
121-168: New async signature may break callers – double-check everyrun()that invokes_getChatArgs()
_getChatArgsis nowasyncand returns a Promise. Every consumer mustawaitit:-const chatArgs = this._getChatArgs(); +const chatArgs = await this._getChatArgs();Search all OpenAI actions in the repo to ensure none call it without
await; otherwise they will pass a Promise to the OpenAI SDK and fail at runtime.components/onlyoffice_docspace/actions/upload-file/upload-file.mjs (1)
43-50: Double-check multipart handling for thefilefieldYou pass a plain string to the
Filefield and setMULTIPART_FORM_DATA_HEADERSmanually.
Ifthis.app.post()no longer wraps the payload withFormData(since the utils were refactored to be async), the request will send JSON instead of multipart.Please verify that
app.post():
- Detects
data.Filevalues that are paths/URLs and converts them to a stream viagetFileStreamAndMetadata.- Re-creates the
Content-Type: multipart/form-dataheader with the correct boundary.
Manually providingContent-Typeoften leads to incorrect boundaries.If either of these assumptions is false, the upload will fail. Consider preparing the
FormDatahere instead of relying on hidden magic.components/google_drive/actions/update-file/update-file.mjs (1)
117-118: ```shell
#!/bin/bash
set -euxShow imports in the target file to identify where getFileStream is coming from
sed -n '1,50p' components/google_drive/actions/update-file/update-file.mjs
Search for getFileStream definition across the repo
rg -n "export (async )?function getFileStream" -C 3 --glob "*.{js,ts,mjs}"
Search for getFileStreamAndMetadata definition across the repo
rg -n "export (async )?function getFileStreamAndMetadata" -C 3 --glob "*.{js,ts,mjs}"
</details> <details> <summary>components/imagga/actions/process-batch/process-batch.mjs (1)</summary> `17-21`: **LGTM – clearer property label** The new label and description make the accepted input explicit. </details> <details> <summary>components/openai/openai.app.mjs (1)</summary> `257-260`: **Label/description update looks good** Consistent wording across components improves UX. </details> <details> <summary>components/llmwhisperer/actions/extract-text/extract-text.mjs (1)</summary> `30-35`: **Typos in prop names / docs leak into query-string** `pageSeperator` & `page_seperator` are misspelled (“Separator”). Unless the backend intentionally uses the typo, this will silently break the feature. Same for the description typo “stratergy”. Confirm the expected spelling on the API side; if it’s a typo, fix both the prop key and the query-param to avoid hard-to-trace errors. Also applies to: 41-41, 140-144 </details> <details> <summary>components/openai/actions/analyze-image-content/analyze-image-content.mjs (1)</summary> `80-88`: **Guard against unknown content-type** `getFileStreamAndMetadata` can return `metadata.contentType === undefined` when downloading from certain URLs. OpenAI rejects uploads without a valid `Content-Type`. ```diff - fileData.append("file", stream, { - contentType: metadata.contentType, + fileData.append("file", stream, { + contentType: metadata.contentType ?? "application/octet-stream",components/alttext_ai/actions/generate-alt-text/generate-alt-text.mjs (1)
16-21: Great usability touch with the alert propThe clarifying alert avoids common mis-configuration errors—nice!
| label: "File Path or URL", | ||
| description: "The file to upload. Provide either a file URL or a path to a file in the `/tmp` directory (for example, `/tmp/myFile.txt`)", | ||
| }, |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Name collision may confuse users (file vs filePath)
The app now exposes:
• file – “Source File Path” (path only)
• filePath – “File Path or URL” (path or URL)
Because both ultimately represent a file, consider unifying them (or clearly deprecating one) to avoid users picking the wrong prop when building flows.
🤖 Prompt for AI Agents
In components/docsbot_ai/docsbot_ai.app.mjs around lines 125 to 127, there is a
naming collision between the properties `file` and `filePath`, both representing
file locations but with different scopes. To fix this, unify these properties
into a single clear and descriptive name that covers both use cases or clearly
deprecate one of them with appropriate warnings and documentation updates to
prevent user confusion when selecting properties for flows.
| alertType: "info", | ||
| content: "Please provide exactly one of File URL or File Path.", | ||
| }, |
There was a problem hiding this comment.
“Info” alert is now misleading after URL/Path consolidation
Since the dedicated File URL prop was removed, users can no longer provide “exactly one of URL or Path.”
Keeping this alert will confuse them.
Either (a) drop the alert entirely, or (b) re-phrase to clarify that the single field accepts both formats.
- info: {
- type: "alert",
- alertType: "info",
- content: "Please provide exactly one of File URL or File Path.",
- },📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| alertType: "info", | |
| content: "Please provide exactly one of File URL or File Path.", | |
| }, |
🤖 Prompt for AI Agents
In components/flippingbook/flippingbook.app.mjs around lines 34 to 36, the alert
message still refers to providing exactly one of File URL or File Path, which is
misleading since the File URL prop was removed and the single field now accepts
both formats. Update the alert message to clarify that the single input field
accepts either a URL or a file path, or remove the alert entirely if it is no
longer necessary.
| const data = new FormData(); | ||
| const content = fs.createReadStream(file.includes("tmp/") | ||
| ? file | ||
| : `/tmp/${file}`); | ||
| const { | ||
| stream, metadata, | ||
| } = await getFileStreamAndMetadata(file); | ||
|
|
||
| data.append("file", content); | ||
| data.append("file", stream, { | ||
| contentType: metadata.contentType, | ||
| knownLength: metadata.size, | ||
| filename: metadata.name, | ||
| }); |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Guard against missing metadata & send MIME safely
metadata.size and metadata.contentType may be undefined for some URLs (e.g. when the server omits Content-Length or Content-Type). Passing undefined to FormData as knownLength/contentType triggers runtime errors in Node.
-const { stream, metadata } = await getFileStreamAndMetadata(file);
-
-data.append("file", stream, {
- contentType: metadata.contentType,
- knownLength: metadata.size,
- filename: metadata.name,
-});
+const { stream, metadata } = await getFileStreamAndMetadata(file);
+const options = {
+ filename: metadata.name,
+};
+if (metadata.contentType) options.contentType = metadata.contentType;
+if (Number.isFinite(metadata.size)) options.knownLength = metadata.size;
+
+data.append("file", stream, options);This avoids unexpected ERR_STREAM_PREMATURE_CLOSE / FormData length calculation errors.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const data = new FormData(); | |
| const content = fs.createReadStream(file.includes("tmp/") | |
| ? file | |
| : `/tmp/${file}`); | |
| const { | |
| stream, metadata, | |
| } = await getFileStreamAndMetadata(file); | |
| data.append("file", content); | |
| data.append("file", stream, { | |
| contentType: metadata.contentType, | |
| knownLength: metadata.size, | |
| filename: metadata.name, | |
| }); | |
| const data = new FormData(); | |
| const { stream, metadata } = await getFileStreamAndMetadata(file); | |
| const options = { | |
| filename: metadata.name, | |
| }; | |
| if (metadata.contentType) options.contentType = metadata.contentType; | |
| if (Number.isFinite(metadata.size)) options.knownLength = metadata.size; | |
| data.append("file", stream, options); |
🤖 Prompt for AI Agents
In components/openai/actions/create-transcription/create-transcription.mjs
around lines 92 to 101, the code uses metadata.size and metadata.contentType
directly when appending to FormData, which can be undefined and cause runtime
errors. To fix this, add checks to ensure metadata.size and metadata.contentType
are defined before passing them to FormData; if they are undefined, omit those
properties or provide safe defaults to prevent errors like
ERR_STREAM_PREMATURE_CLOSE.
| const { | ||
| stream, metadata, | ||
| } = await getFileStreamAndMetadata(this.image); | ||
| const formData = new FormData(); | ||
| formData.append("file", fs.createReadStream(checkTmp(this.image))); | ||
| formData.append("file", stream, { | ||
| contentType: metadata.contentType, | ||
| knownLength: metadata.size, | ||
| filename: metadata.name, | ||
| }); |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Header metadata could be incomplete & redundant local variable
metadata.contentType,metadata.size,metadata.nameare passed blindly.
•contentTypeorsizemay beundefinedwhen the source is a URL and the remote
server does not provide those headers – many do not.form-datawill then serialise
undefinedand the request may be rejected by the API.imageUrl(defined at line 41) is never mutated, so you can drop the alias and just
usethis.image.
- const {
- stream, metadata,
- } = await getFileStreamAndMetadata(this.image);
+ const { stream, metadata } =
+ await getFileStreamAndMetadata(this.image);
const formData = new FormData();
- formData.append("file", stream, {
- contentType: metadata.contentType,
- knownLength: metadata.size,
- filename: metadata.name,
- });
+ const fileOptions = {
+ filename: metadata?.name ?? "upload",
+ };
+ if (metadata?.contentType) fileOptions.contentType = metadata.contentType;
+ if (metadata?.size != null) fileOptions.knownLength = metadata.size;
+ formData.append("file", stream, fileOptions);This keeps the request robust when metadata is unavailable and eliminates the extra variable.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const { | |
| stream, metadata, | |
| } = await getFileStreamAndMetadata(this.image); | |
| const formData = new FormData(); | |
| formData.append("file", fs.createReadStream(checkTmp(this.image))); | |
| formData.append("file", stream, { | |
| contentType: metadata.contentType, | |
| knownLength: metadata.size, | |
| filename: metadata.name, | |
| }); | |
| const { stream, metadata } = | |
| await getFileStreamAndMetadata(this.image); | |
| const formData = new FormData(); | |
| const fileOptions = { | |
| filename: metadata?.name ?? "upload", | |
| }; | |
| if (metadata?.contentType) fileOptions.contentType = metadata.contentType; | |
| if (metadata?.size != null) fileOptions.knownLength = metadata.size; | |
| formData.append("file", stream, fileOptions); |
🤖 Prompt for AI Agents
In components/claid_ai/actions/upscale-image/upscale-image.mjs around lines 59
to 67, the metadata fields contentType, size, and name are used directly without
checking if they are undefined, which can cause issues if the source is a URL
lacking these headers. Update the code to conditionally include these fields
only if they are defined to prevent sending undefined values in the form data.
Also, remove the unused imageUrl alias declared at line 41 and use this.image
directly to simplify the code.
| const { | ||
| stream, metadata, | ||
| } = await getFileStreamAndMetadata(this.filePath); | ||
|
|
||
| let response = await this.canva.importDesign({ | ||
| $, | ||
| headers: { | ||
| "Import-Metadata": JSON.stringify({ | ||
| "title_base64": titleBase64, | ||
| }), | ||
| "Content-Length": fs.statSync(filePath).size, | ||
| "Content-Length": metadata.size, | ||
| "Content-Type": "application/octet-stream", | ||
| }, | ||
| data: fs.createReadStream(filePath), | ||
| data: stream, | ||
| }); |
There was a problem hiding this comment.
🛠️ Refactor suggestion
Content-Length may be undefined & wrong content-type hard-coded
metadata.size is only guaranteed for local /tmp files; for remote URLs it can be
undefined. Supplying an undefined Content-Length header breaks the request.
In addition, you already receive metadata.contentType, yet you always send
application/octet-stream. Let’s only include the header when present and use the
correct MIME type.
- "Content-Length": metadata.size,
- "Content-Type": "application/octet-stream",
+ ...(metadata.size != null && { "Content-Length": metadata.size }),
+ "Content-Type": metadata.contentType ?? "application/octet-stream",📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const { | |
| stream, metadata, | |
| } = await getFileStreamAndMetadata(this.filePath); | |
| let response = await this.canva.importDesign({ | |
| $, | |
| headers: { | |
| "Import-Metadata": JSON.stringify({ | |
| "title_base64": titleBase64, | |
| }), | |
| "Content-Length": fs.statSync(filePath).size, | |
| "Content-Length": metadata.size, | |
| "Content-Type": "application/octet-stream", | |
| }, | |
| data: fs.createReadStream(filePath), | |
| data: stream, | |
| }); | |
| const { | |
| stream, metadata, | |
| } = await getFileStreamAndMetadata(this.filePath); | |
| let response = await this.canva.importDesign({ | |
| $, | |
| headers: { | |
| "Import-Metadata": JSON.stringify({ | |
| "title_base64": titleBase64, | |
| }), | |
| ...(metadata.size != null && { "Content-Length": metadata.size }), | |
| "Content-Type": metadata.contentType ?? "application/octet-stream", | |
| }, | |
| data: stream, | |
| }); |
🤖 Prompt for AI Agents
In
components/canva/actions/create-design-import-job/create-design-import-job.mjs
around lines 33 to 47, the Content-Length header is set using metadata.size
which can be undefined for remote files, causing request failures. Also, the
Content-Type header is hardcoded to application/octet-stream instead of using
the actual metadata.contentType. Fix this by conditionally including the
Content-Length header only if metadata.size is defined, and set the Content-Type
header to metadata.contentType if available, otherwise omit it or use a sensible
default.
| async getFileData(filePath) { | ||
| const { | ||
| stream, metadata, | ||
| } = await getFileStreamAndMetadata(filePath); | ||
| return { | ||
| attributes: { | ||
| content_created_at: new Date(stats.ctimeMs).toISOString() | ||
| .split(".")[0] + "Z", | ||
| content_modified_at: new Date(stats.mtimeMs).toISOString() | ||
| .split(".")[0] + "Z", | ||
| name: path.basename(filePath), | ||
| parent: { | ||
| id: 0, | ||
| fileContent: stream, | ||
| fileMeta: { | ||
| attributes: { | ||
| content_created_at: new Date(metadata.lastModified).toISOString() | ||
| .split(".")[0] + "Z", | ||
| content_modified_at: new Date(metadata.lastModified).toISOString() | ||
| .split(".")[0] + "Z", | ||
| name: metadata.name, | ||
| parent: { | ||
| id: 0, | ||
| }, | ||
| }, | ||
| size: metadata.size, | ||
| contentType: metadata.contentType, | ||
| }, |
There was a problem hiding this comment.
metadata.lastModified can be undefined – guard to prevent Invalid Date exceptions
new Date(undefined).toISOString() throws, breaking every upload where the upstream server doesn’t send a Last-Modified header (common for signed URLs).
- content_created_at: new Date(metadata.lastModified).toISOString()
+ content_created_at: metadata.lastModified
+ ? new Date(metadata.lastModified).toISOString()
+ : new Date().toISOString()Apply the same guard to content_modified_at.
Committable suggestion skipped: line range outside the PR's diff.
🤖 Prompt for AI Agents
In components/box/common/utils.mjs between lines 5 and 24, the code uses new
Date(metadata.lastModified).toISOString() without checking if
metadata.lastModified is defined, which can cause Invalid Date exceptions if
lastModified is undefined. To fix this, add a guard to check if
metadata.lastModified exists before converting it to a date string for both
content_created_at and content_modified_at. If lastModified is undefined, set
these fields to null or an appropriate default value to prevent errors.
| const imageList = await imageFile.reduce(async (acc, cur) => { | ||
| await delay(1000); | ||
| const uploadId = await uploadFile({ | ||
| imageFile: cur, | ||
| imagga, | ||
| }); | ||
|
|
||
| return [ | ||
| ...await acc, | ||
| uploadId, | ||
| ]; | ||
| }, []); |
There was a problem hiding this comment.
Asynchronous Array.reduce causes race conditions and wastes the 1 s delay
Array.reduce does not await earlier iterations; all uploads start concurrently, defeating the sequential delay and potentially hitting rate limits.
Refactor to a simple for…of loop:
- const imageList = await imageFile.reduce(async (acc, cur) => {
- await delay(1000);
- const uploadId = await uploadFile({ imageFile: cur, imagga });
- return [
- ...await acc,
- uploadId,
- ];
- }, []);
+ const imageList = [];
+ for (const file of imageFile) {
+ await delay(1000);
+ const uploadId = await uploadFile({ imageFile: file, imagga });
+ imageList.push(uploadId);
+ }This guarantees one-second spacing between requests and preserves order.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
| const imageList = await imageFile.reduce(async (acc, cur) => { | |
| await delay(1000); | |
| const uploadId = await uploadFile({ | |
| imageFile: cur, | |
| imagga, | |
| }); | |
| return [ | |
| ...await acc, | |
| uploadId, | |
| ]; | |
| }, []); | |
| const imageList = []; | |
| for (const file of imageFile) { | |
| await delay(1000); | |
| const uploadId = await uploadFile({ imageFile: file, imagga }); | |
| imageList.push(uploadId); | |
| } |
🤖 Prompt for AI Agents
In components/imagga/common/utils.mjs between lines 7 and 18, the use of
asynchronous Array.reduce causes all uploads to start concurrently, ignoring the
intended 1-second delay between each. Replace the reduce with a for...of loop
that awaits each upload and delay sequentially to ensure uploads happen one
after another with a 1-second gap, preserving order and preventing rate limit
issues.
components/openai/actions/analyze-image-content/analyze-image-content.mjs
Show resolved
Hide resolved
|
Hi @GTFalcao, For OpenAI chat:
|
vunguyenhung
moved this from Ready for QA
to Changes Required
in Component (Source and Action) Backlog
|
Hello everyone, I have tested this PR and there're some test cases failed or needed improvement. Please check the test report below for more information |
|
For the record, base64 image input was already not supported in the code, despite the prop description saying so. I've changed the prop to standard file path/url input as well. |
GTFalcao
moved this from Changes Required
to Ready for QA
in Component (Source and Action) Backlog
vunguyenhung
moved this from Ready for QA
to In QA
in Component (Source and Action) Backlog
vunguyenhung
moved this from In QA
to Changes Required
in Component (Source and Action) Backlog
|
Hello everyone, I have tested this PR and there're some test cases failed or needed improvement. Please check the test report below for more information |
|
I've adjusted the image to always be an "image_url" instead of an "input_image", creating the URL from the base64 content. I also found and fixed the issue with the audio input: it was detecting the extension via a regexp which failed when there were query parameters (which was the case in your input, but not mine). I've adjusted this to use the extension from the file metadata instead. |
GTFalcao
moved this from Changes Required
to Ready for QA
in Component (Source and Action) Backlog
vunguyenhung
moved this from Ready for QA
to Ready for Release
in Component (Source and Action) Backlog
|
Hi everyone, all test cases are passed! Ready for release! |
|
/approve |
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
github-project-automation
bot
moved this from Ready for Release
to Done
in Component (Source and Action) Backlog
Partially addresses #16977
This branch was created off the first one (#16979 ), and will be merged only after that one is merged.
The 30 components changed in this branch are:
autodesk:
upload-filecradl_ai:
parse-documentparse-document-human-in-loopdocsbot_ai:
upload-source-fileegnyte:
upload-fileflippingbook:
create-flipbookupdate-flipbookgoogle_drive:
upload-fileupdate-filehappy_scribe:
submit-fileilovepdf:
process-filesimagga:
analyze-imageprocess-batchlokalise:
upload-filepodio:
upload-filezoho_crm:
upload-attachmentopenai:
analyze-image-contentchatcreate-batchcreate-transcriptionupload-fileairparser:
upload-document-parsealttext_ai:
generate-alt-textazure_storage:
upload-blobboldsign:
send-document-templatecanva:
create-design-import-jobupload-assetclaid_ai:
correct-color-lightingremove-backgroundupscale-imageSummary by CodeRabbit
New Features
/tmp) for file inputs, making file uploads and processing more flexible and convenient.Bug Fixes
Refactor
Chores