dify
dify copied to clipboard
Published
3 days ago
•
langgenius
langgenius
Workflow batch & api job requests fail when including a URL to a file
Self Checks
- [X] This is only for bug report, if you would like to ask a question, please head to Discussions.
- [X] I have searched for existing issues search for existing issues, including closed ones.
- [X] I confirm that I am using English to submit this report (我已阅读并同意 Language Policy).
- [X] [FOR CHINESE USERS] 请务必使用英文提交 Issue，否则会被关闭。谢谢！:）
- [X] Please do not modify this template :) and fill in all the required fields.
Dify version
0.11.1
Cloud or Self Hosted
Cloud
Steps to reproduce
Repro steps
- Setup a workflow using the configuration below (or any workflow which accepts images basically)
- Test that workflow works using a single job.
- Try a batch workflow or send requests to the workflow using API, you will get the below errors
Workflow configuration file (DSL)
description: Defines product image type and average image color as hex code value.
Also write ALT text for the image.
icon: 📸
icon_background: '#FFEAD5'
mode: workflow
name: Image ALT text, type and hex code definer (Version 2 - Joakim)
use_icon_as_answer_icon: false
kind: app
version: 0.1.3
workflow:
conversation_variables: []
environment_variables: []
features:
file_upload:
allowed_file_extensions:
- .JPG
- .JPEG
- .PNG
- .GIF
- .WEBP
- .SVG
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
enabled: false
fileUploadConfig:
audio_file_size_limit: 50
batch_count_limit: 5
file_size_limit: 15
image_file_size_limit: 10
video_file_size_limit: 100
workflow_file_upload_limit: 10
image:
enabled: false
number_limits: 3
transfer_methods:
- local_file
- remote_url
number_limits: 1
opening_statement: ''
retriever_resource:
enabled: false
sensitive_word_avoidance:
enabled: false
speech_to_text:
enabled: false
suggested_questions: []
suggested_questions_after_answer:
enabled: false
text_to_speech:
enabled: false
language: ''
voice: ''
graph:
edges:
- data:
isInIteration: false
sourceType: start
targetType: if-else
id: 1714920866750-source-1719391536361-target
selected: false
source: '1714920866750'
sourceHandle: source
target: '1719391536361'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: if-else
targetType: llm
id: 1719391536361-true-1728398212858-target
source: '1719391536361'
sourceHandle: 'true'
target: '1728398212858'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: if-else
targetType: llm
id: 1719391536361-8a4f1fc0-397b-424e-a744-5c4090f82295-1728398041688-target
source: '1719391536361'
sourceHandle: 8a4f1fc0-397b-424e-a744-5c4090f82295
target: '1728398041688'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: if-else
targetType: end
id: 1719391536361-false-1730886553319-target
source: '1719391536361'
sourceHandle: 'false'
target: '1730886553319'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: llm
targetType: end
id: 1728398041688-source-1714920954869-target
source: '1728398041688'
sourceHandle: source
target: '1714920954869'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: llm
targetType: end
id: 1728398212858-source-1719391856189-target
source: '1728398212858'
sourceHandle: source
target: '1719391856189'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: llm
targetType: end
id: 1731321244320-source-1731321721145-target
source: '1731321244320'
sourceHandle: source
target: '1731321721145'
targetHandle: target
type: custom
zIndex: 0
- data:
isInIteration: false
sourceType: if-else
targetType: llm
id: 1719391536361-2bd36163-37cb-407f-ad33-7c612bf2e791-1731321244320-target
source: '1719391536361'
sourceHandle: 2bd36163-37cb-407f-ad33-7c612bf2e791
target: '1731321244320'
targetHandle: target
type: custom
zIndex: 0
nodes:
- data:
desc: ''
selected: false
title: Start
type: start
variables:
- allowed_file_extensions: []
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
label: filetype
max_length: 256
options:
- Product Image
- Tech Logo
- Manual
- Energy Label
- Data Sheet
- Fiche
- EU Tariff File
- Battery Certificate
- Safety Data Sheet
required: true
type: select
variable: filetype
- allowed_file_extensions: []
allowed_file_types:
- image
allowed_file_upload_methods:
- local_file
- remote_url
label: file
max_length: 5
options: []
required: true
type: file
variable: file
- label: en_nameWeb
max_length: null
options: []
required: false
type: text-input
variable: en_nameWeb
- label: no_nameWeb
max_length: null
options: []
required: false
type: text-input
variable: no_nameWeb
- label: se_nameWeb
max_length: null
options: []
required: false
type: text-input
variable: se_nameWeb
- label: fi_nameWeb
max_length: null
options: []
required: false
type: text-input
variable: fi_nameWeb
- label: dk_nameWeb
max_length: null
options: []
required: false
type: text-input
variable: dk_nameWeb
height: 246
id: '1714920866750'
position:
x: -372.4019128777556
y: 302.8844452803611
positionAbsolute:
x: -372.4019128777556
y: 302.8844452803611
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
desc: ''
outputs:
- value_selector:
- '1728398041688'
- text
variable: output
selected: false
title: Logo analysis output
type: end
height: 90
id: '1714920954869'
position:
x: 875.2644922920333
y: 449.42070507111214
positionAbsolute:
x: 875.2644922920333
y: 449.42070507111214
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
cases:
- case_id: 'true'
conditions:
- comparison_operator: is
id: '1719391629749'
value: Product Image
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 850c6023-dfb0-4532-9da5-a19ec754d480
value: Feeling Image
varType: string
variable_selector:
- '1714920866750'
- filetype
id: 'true'
logical_operator: or
- case_id: 8a4f1fc0-397b-424e-a744-5c4090f82295
conditions:
- comparison_operator: is
id: df3688ee-ccf0-4726-9713-f39f9b537766
value: Tech Logo
varType: string
variable_selector:
- '1714920866750'
- filetype
id: 8a4f1fc0-397b-424e-a744-5c4090f82295
logical_operator: and
- case_id: 2bd36163-37cb-407f-ad33-7c612bf2e791
conditions:
- comparison_operator: is
id: 43fd393f-e959-4d2f-91a9-fc112623e595
value: Manual
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 7bc631ee-305a-4a2a-942b-ecab970db436
value: Energy Label
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: bfc78532-9f3d-4792-8580-3468d0b00f36
value: Data Sheet
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 3cc1475b-5540-4c62-891d-27d2567ff878
value: Fiche
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 9087f9cb-dc3b-4cda-974d-3f119626c8e8
value: EU Tariff File
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 55a98494-6556-4d73-a319-2793a97df162
value: Battery Certificate
varType: string
variable_selector:
- '1714920866750'
- filetype
- comparison_operator: is
id: 5c379fe1-dd23-4dd2-afcf-3692b56d94e2
value: Safety Data Sheet
varType: string
variable_selector:
- '1714920866750'
- filetype
id: 2bd36163-37cb-407f-ad33-7c612bf2e791
logical_operator: or
conditions:
- comparison_operator: is
id: '1719391629749'
value: productImage
variable_selector:
- '1714920866750'
- filetype
desc: ''
logical_operator: and
selected: false
title: FileType condition
type: if-else
height: 404
id: '1719391536361'
position:
x: -23.72524316829856
y: 302.8844452803611
positionAbsolute:
x: -23.72524316829856
y: 302.8844452803611
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
desc: ''
outputs:
- value_selector:
- '1728398212858'
- text
variable: output
selected: false
title: Image Analysis Output
type: end
height: 90
id: '1719391856189'
position:
x: 875.2644922920333
y: 302.8844452803611
positionAbsolute:
x: 875.2644922920333
y: 302.8844452803611
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.5
json_schema: "{\n \"name\": \"LogoAnalysisOutput\",\n \"schema\": {\n\
\ \"title\": \"LogoAnalysisOutput\",\n \"type\": \"object\",\n\
\ \"properties\": {\n \"logo_type\": {\n \"type\": \"\
string\",\n \"enum\": [\n \"Specification\",\n \
\ \"Processor\",\n \"Test winner\",\n \"Test product\"\
,\n \"Award\",\n \"Warranty\",\n \"Ecommerce\"\
\n ]\n },\n \"alt_texts\": {\n \"type\": \"\
object\",\n \"properties\": {\n \"English\": { \"type\"\
: \"string\" },\n \"Norwegian\": { \"type\": \"string\" },\n\
\ \"Danish\": { \"type\": \"string\" },\n \"Finnish\"\
: { \"type\": \"string\" },\n \"Swedish\": { \"type\": \"string\"\
\ }\n },\n \"required\": [\"English\", \"Norwegian\",\
\ \"Danish\", \"Finnish\", \"Swedish\"],\n \"additionalProperties\"\
: false\n },\n \"average_color_hex\": {\n \"type\"\
: \"string\",\n \"pattern\": \"^#([A-Fa-f0-9]{6})$\"\n }\n\
\ },\n \"required\": [\"logo_type\", \"alt_texts\", \"average_color_hex\"\
],\n \"additionalProperties\": false\n }\n}"
presence_penalty: 0.5
response_format: json_schema
temperature: 0
top_p: 0.75
mode: chat
name: gpt-4o-mini
provider: openai
prompt_template:
- id: f412d8ea-6ae8-4ba9-aebb-8932efb4c59d
role: system
text: You are an AI system specializing in analyzing e-commerce logo images
and generating SEO-optimized alt text descriptions in multiple languages.
Your task is to examine a logo, determine its type, create multilingual
alt text descriptions, and estimate the image's average color.
- edition_type: basic
id: 30d33c11-4583-417a-a28c-796af54626b1
role: user
text: 'Attached is the logo you need to analyze:
Please follow these steps:
1. Analyze the logo thoroughly, paying attention to visual elements, colors,
and any text present.
4. Determine the logo type based on the following categories:
- Specification: Features technical aspects of the product (e.g., 5G,
ink-saving feature)
- Processor: Intel or AMD processor logo
- Test winner: Indicates the product is best or best in test
- Test product: Shows product ranking in testing or recommendation by
a testing site
- Award: Highlights a design or sustainability award
- Warranty: Emphasizes the product''s warranty period
- Ecommerce: Relates to money-back guarantee, cashback, or return policy
5. Generate a JSON object with the determined logo type. For example:
{"logo_type": "Specification"}
6. Create SEO-optimized alt text descriptions for the logo in 5 languages:
English, Norwegian, Danish, Finnish, and Swedish. Follow these guidelines:
a) Accurately describe the product or feature represented by the logo.
b) Include relevant keywords naturally, without forcing them.
c) Keep the text concise, aiming for about 125 characters or less.
d) Avoid keyword stuffing.
e) Provide context when relevant, especially if the logo shows the product
in use.
f) Skip phrases like "Image of" or "Picture of".
g) Use proper spelling and grammar.
h) Focus on accessibility and providing an equivalent experience for all
users.
i) Be specific and detailed when necessary to differentiate between similar
items.
j) Include any text that is part of the image, translating it if necessary.
k) Front-load critical information to present key details first.
7. Estimate the average color of the provided image and output the color
as a HEX code.
Before providing your final output, wrap your analysis inside <detailed_logo_analysis>
tags. Include the following:
1. Detailed description of the logo''s visual elements, colors, and text.
2. Step-by-step reasoning for determining the logo type:
- List each potential logo type.
- Provide arguments for and against each type.
- Conclude with the most appropriate type and explain why.
3. Explanation of your SEO optimization strategy for each language:
- Brainstorm relevant keywords for each language.
- Explain how you''ll incorporate these keywords naturally.
4. Description of the method used for color estimation:
- Break down the process into steps.
- Explain how you''ll handle multiple colors if present.
Your final output should be structured as follows:
1. Logo type JSON object
2. Alt text descriptions in the specified languages
3. HEX code for the estimated average color
Example output structure (note: this is a generic example to illustrate
the format):
{"logo_type": "[Type]"}
English: [SEO-optimized alt text description]
Norwegian: [SEO-optimized alt text description in Norwegian]
Danish: [SEO-optimized alt text description in Danish]
Finnish: [SEO-optimized alt text description in Finnish]
Swedish: [SEO-optimized alt text description in Swedish]
#[HEX color code]
Please proceed with your analysis and provide the requested output.'
selected: false
title: Logo analysis
type: llm
variables: []
vision:
configs:
detail: high
variable_selector:
- '1714920866750'
- file
enabled: true
height: 98
id: '1728398041688'
position:
x: 453.63749851366333
y: 449.42070507111214
positionAbsolute:
x: 453.63749851366333
y: 449.42070507111214
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.5
json_schema: "{\n \"name\": \"ImageAnalysisOutput\",\n \"schema\": {\n\
\ \"title\": \"ImageAnalysisOutput\",\n \"type\": \"object\",\n\
\ \"properties\": {\n \"image_type\": {\n \"type\": \"\
string\",\n \"enum\": [\n \"Packshot Image\",\n \
\ \"Lifestyle Image\",\n \"Detail Image\",\n \
\ \"Infographic Image\",\n \"Comparison Image\",\n \
\ \"Set Image\"\n ]\n },\n \"alt_text\": {\n \
\ \"type\": \"object\",\n \"properties\": {\n \"English\"\
: { \"type\": \"string\" },\n \"norwegian\": { \"type\": \"\
string\" },\n \"danish\": { \"type\": \"string\" },\n \
\ \"finnish\": { \"type\": \"string\" },\n \"swedish\"\
: { \"type\": \"string\" }\n },\n \"required\": [\"English\"\
, \"norwegian\", \"danish\", \"finnish\", \"swedish\"],\n \"\
additionalProperties\": false\n },\n \"average_color\": {\n\
\ \"type\": \"string\",\n \"pattern\": \"^#([A-Fa-f0-9]{6})$\"\
\n }\n },\n \"required\": [\"image_type\", \"alt_text\",\
\ \"average_color\"],\n \"additionalProperties\": false\n }\n}"
presence_penalty: 0.5
response_format: json_schema
temperature: 0
top_p: 0.75
mode: chat
name: gpt-4o-mini
provider: openai
prompt_template:
- id: 6266ce57-ebea-4b84-b832-ab4a56e547a3
role: system
text: You are an AI system specializing in analyzing e-commerce product
images and generating SEO-optimized alt text descriptions in multiple
languages. Your task is to examine a product image, determine its type,
create multilingual alt text descriptions, and estimate the image's average
color.
- id: 56c5640e-560f-4fbd-a6a1-96ad5a99c5e3
role: user
text: 'Attached is the image you''ll be analyzing:
Now, here are the product names in different languages that you''ll be
working with:
<product_name_english>
{{#1714920866750.en_nameWeb#}}
</product_name_english>
<product_name_norwegian>
{{#1714920866750.no_nameWeb#}}
</product_name_norwegian>
<product_name_danish>
{{#1714920866750.dk_nameWeb#}}
</product_name_danish>
<product_name_finnish>
{{#1714920866750.fi_nameWeb#}}
</product_name_finnish>
<product_name_swedish>
{{#1714920866750.se_nameWeb#}}
</product_name_swedish>
Please follow these steps to complete your analysis:
1. Image Analysis:
Examine the provided image URL, focusing only on visible elements, colors,
and any text present in the image.
2. Image Type Determination:
Categorize the image into one of the following types:
- Packshot Image: Simple, clean images of the product against a plain
background (usually white).
- Lifestyle Image: Shows the product being used in a real-life context
or environment.
- Detail Image: Highlights specific features, textures, or intricate details
of the product.
- Infographic Image: Combines images with text or graphics to highlight
key features, benefits, or instructions.
- Comparison Image: Shows the product alongside similar items or alternatives.
- Set Image: Displays the product as part of a set or collection.
3. SEO-optimized Alt Text Creation:
Create alt text descriptions for the image in 5 languages: English, Norwegian,
Danish, Finnish, and Swedish. Use the provided product names in each respective
language. Follow these guidelines:
- Include the product name naturally within the alt text.
- Describe only specific visible features (color, texture, material, etc.)
actually present in the image.
- Use relevant keywords that potential customers might use to find this
product.
- Focus solely on clearly visible elements in the image, not general product
features.
- Include and translate any text that is part of the image, reproducing
it verbatim.
- Identify and describe the most impactful visible element first.
- Convey the overall tone and mood of the image without using superlatives.
- Use vivid, descriptive language, but avoid exaggeration or speculation.
- Present information in a logical order.
- Use appropriate language and punctuation to convey factual content.
- Omit irrelevant details and any information not directly visible in
the image.
- Aim for a balance between comprehensive description and conciseness.
- Describe the content directly without using phrases like "the picture
shows" or "the image contains."
4. Color Estimation:
Estimate the average color of the provided image and output the color
as a HEX code.
Before providing your final output, wrap your analysis inside <image_analysis>
tags. Include:
- A numbered list of all visible elements in the image, focusing on specific
details.
- Your reasoning for choosing the final image type classification.
- A list of potential SEO keywords for each language, considering only
the visible features of the product.
- Your strategy for adapting keywords and alt text for cultural differences
in each language.
- The method used for color estimation, including how you determined the
dominant colors.
- A detailed description of exactly what is visible in the image, focusing
on specific elements rather than general product features.
- A final review to ensure no speculative or non-visible details have
been included in the alt text.
Your final output should be a JSON object structured as follows:
```json
{
"image_type": "string",
"alt_text": {
"english": "string",
"norwegian": "string",
"danish": "string",
"finnish": "string",
"swedish": "string"
},
"average_color": "string"
}
```
Ensure that all alt text descriptions are properly escaped within the
JSON structure and focus solely on visible elements in the image.
Please proceed with your analysis and provide the requested output.'
selected: false
title: Image analysis output
type: llm
variables: []
vision:
configs:
detail: high
variable_selector:
- '1714920866750'
- file
enabled: true
height: 98
id: '1728398212858'
position:
x: 453.63749851366333
y: 302.8844452803611
positionAbsolute:
x: 453.63749851366333
y: 302.8844452803611
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
desc: ''
outputs:
- value_selector: []
variable: ''
selected: false
title: error
type: end
height: 54
id: '1730886553319'
position:
x: 453.63749851366333
y: 784.4741880970141
positionAbsolute:
x: 453.63749851366333
y: 784.4741880970141
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
context:
enabled: false
variable_selector: []
desc: ''
model:
completion_params:
frequency_penalty: 0.5
json_schema: "{\n \"name\": \"DocumentAnalysisOutput\",\n \"schema\"\
: {\n \"title\": \"DocumentAnalysisOutput\",\n \"type\": \"object\"\
,\n \"properties\": {\n \"DocumentContentsMarkdown\": {\n \
\ \"type\": \"string\"\n },\n \"DocumentLanguages\":\
\ {\n \"type\": \"array\",\n \"items\": {\n \"\
type\": \"string\"\n }\n }\n },\n \"required\": [\"\
DocumentContentsMarkdown\", \"DocumentLanguages\"],\n \"additionalProperties\"\
: false\n }\n}"
presence_penalty: 0.5
response_format: json_schema
temperature: 0
top_p: 0.75
mode: chat
name: gpt-4o-mini
provider: openai
prompt_template:
- id: 9512f95c-6542-4868-8afe-7bd09e1c2171
role: system
text: You are an advanced AI assistant specialized in document analysis
and transcription. Your task is to accurately replicate the contents of
a document page from an image file, while also providing context for visual
elements.
- id: edea1749-a0db-4ad0-927a-63b932dd6997
role: user
text: "Attached is the image file you need to analyze.\n\nYour goal is to\
\ perform detailed Optical Character Recognition (OCR), correct any errors,\
\ describe images inline, and present the information in a well-structured\
\ markdown format.\n\nBefore providing your final output, please analyze\
\ the document thoroughly. Wrap your analysis inside <document_examination>\
\ tags and address the following points:\n\t1.\tList all elements identified\
\ in the image (text blocks, images, tables, etc.), numbering each element\
\ as you go.\n\t2.\tProvide a step-by-step breakdown of the document’s\
\ visual layout, including the positioning of key elements.\n\t3.\tDescribe\
\ your OCR process and any challenges encountered.\n\t4.\tOutline your\
\ plan for recreating the document in markdown.\n\t5.\tExplain your strategy\
\ for creating comprehensive yet concise image descriptions.\n\t6.\tIdentify\
\ potential OCR challenges specific to this document and how you plan\
\ to address them. For each challenge, write down a potential solution.\n\
\t7.\tDiscuss how you will ensure that the language and terminology used\
\ in your output closely matches that of the original document. Provide\
\ examples if possible.\n\t8.\tPlan out the overall markdown structure\
\ for the document, including headers, subheaders, and other formatting\
\ elements.\n\t9.\tList any specific areas where you anticipate difficulties\
\ in transcription or formatting, and how you plan to overcome them.\n\
\t10.\tIdentify the primary language(s) of the document. If multiple languages\
\ are present, estimate the percentage of content in each language.\n\t\
11.\tCreate a concise checklist of the steps you’ll follow to ensure accuracy\
\ and completeness in your transcription.\n\nAfter completing your analysis,\
\ follow these steps to create your output:\n\t1.\tPerform OCR on all\
\ text elements, including those within images or graphics.\n\t2.\tCorrect\
\ any OCR errors you detect.\n\t3.\tFor each image, graphic, or figure\
\ in the document:\n\t•\tProvide a detailed description of its contents\
\ inline with the text.\n\t•\tUse markdown image syntax with elaborated\
\ alt text.\n\t•\tEnsure the description is concise, but comprehensive\
\ enough for accessibility.\n\t•\tTranscribe any text within graphics\
\ accurately.\n\t4.\tRecreate tables or other graphics using simple markdown\
\ syntax, ensuring that table elements remain readable and well-structured.\n\
\t5.\tUse markdown formatting to maintain the document’s structure and\
\ hierarchy, with a focus on simplicity to avoid complex characters or\
\ unsupported markdown syntax.\n\t6.\tAvoid overly large text blocks;\
\ summarize repetitive sections if necessary.\n\t7.\tOrganize all information\
\ into a well-structured markdown document that closely resembles the\
\ original layout.\n\t8.\tDo not include footers or page numbers in your\
\ output.\n\t9.\tDo not use emojis in your output.\n\t10.\tEnsure that\
\ language and terminology closely match the original document for consistency.\n\
\nYour final output must be in the following JSON format:\n\n{\n \"DocumentAnalysisOutput\"\
: {\n \"DocumentContentsMarkdown\": \"The full markdown representation\
\ of the document page, including all text, inline image descriptions,\
\ and recreated graphics or tables.\",\n \"DocumentLanguages\": [\n\
\ \"Detected languages in the document, listed in order of prominence.\"\
\n ]\n }\n}\n\nBegin your document examination now, followed by the\
\ document transcription as instructed."
selected: true
title: Document analysis
type: llm
variables: []
vision:
configs:
detail: high
variable_selector:
- '1714920866750'
- file
enabled: true
height: 98
id: '1731321244320'
position:
x: 453.63749851366333
y: 591.5672725566749
positionAbsolute:
x: 453.63749851366333
y: 591.5672725566749
selected: false
sourcePosition: right
targetPosition: left
type: custom
width: 244
- data:
desc: ''
outputs:
- value_selector:
- '1731321244320'
- text
variable: output
selected: false
title: Document Analysis output
type: end
height: 90
id: '1731321721145'
position:
x: 875.2644922920333
y: 591.5672725566749
positionAbsolute:
x: 875.2644922920333
y: 591.5672725566749
selected: true
sourcePosition: right
targetPosition: left
type: custom
width: 244
viewport:
x: 749.4435638438554
y: 177.77511584173362
zoom: 0.7764764961366506
✔️ Expected Behavior
- The workflow behaves the same way as single-run through the web interface and processes the image file
❌ Actual Behavior
- API throws error: File type not allowed
- Batch workflow web interface throws error: "file in input form must be a file"
I found some relevant information that might help with your issue:
-
File Type Not Allowed Error: This error can occur if the file type is not among the allowed types specified in your workflow configuration. Ensure that the file extensions and types you are trying to upload are included in the
allowed_file_extensionsandallowed_file_typessettings in your workflow configuration. The Dify API validates file types against a list of supported formats, including images like JPG, JPEG, PNG, GIF, WEBP, and SVG [1]. -
File in Input Form Must Be a File Error: This error is triggered when the input for a file variable is not a dictionary or an instance of the
Fileclass. Ensure that the input provided for the file variable in your batch workflow or API request is correctly formatted as a dictionary orFileobject [2].
Additionally, there have been recent changes in the Dify codebase related to file upload handling. The image upload feature has been integrated into a more comprehensive file upload functionality, which might affect how file uploads are processed in version 0.11.1 [3].
If these suggestions do not resolve the issue, you might want to check for any recent updates or changes in the Dify codebase that could affect file upload handling.
To continue talking to Dosu, mention @dosu.
![dosubot[bot] avatar](https://avatars.githubusercontent.com/in/324583?v=4 "Published by a pub.dev verified publisher"){kind=small}
I've checked both 1 & 2.
For reference I've tested with the testfiles below:
filetype,file,en_nameWeb,no_nameWeb,se_nameWeb,fi_nameWeb,dk_nameWeb Product Image,https://media.power-cdn.net/images/h-aiisawesome/products/3306919/3306919_1_1200x1200_w_g.jpg,"IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE"
And the change suggested by Dosu in 2:
filetype,file,en_nameWeb,no_nameWeb,se_nameWeb,fi_nameWeb,dk_nameWeb Product Image,"{""file_path"":""https://media.power-cdn.net/images/h-aiisawesome/products/3306919/3306919_1_1200x1200_w_g.jpg"",""file_type"":""image/jpeg"",""file_name"":""3306919_1_1200x1200_w_g.jpg"",""file_extension"":""jpg"",""upload_method"":""url""}","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE","IPAD AIR 11"" CL 2024 128GB BLUE"
Hi, @sigvardt. I'm Dosu, and I'm helping the Dify team manage their backlog and am marking this issue as stale.
Issue Summary:
- Reported bug in Dify version 0.11.1 with workflow batch and API job requests failing when a URL to a file is included.
- Single jobs work correctly, indicating a specific issue with batch/API requests.
- I suggested checking file type allowances and formatting, but you confirmed these were not the issue.
- The problem persists in version 0.12.1, suggesting a deeper issue with file upload handling.
Next Steps:
- Please confirm if this issue is still relevant to the latest version of the Dify repository by commenting here.
- If there is no further activity, this issue will be automatically closed in 15 days.
Thank you for your understanding and contribution!